PDA

View Full Version : CoreCodec/H.264 Codec "CoreAVC"


Pages : 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 [22] 23 24 25 26 27 28 29

roytam1
27th November 2009, 06:44
/me was heard that coreavc doesn't support x264 weightp parameter. will it be fixed in 2.0?

Mixer73
27th November 2009, 06:54
Mixer73... Depending on which method used for DXVA, it will always use less CPU over CUDA. RAM usage in our testing is negligible outside of the GPU, but the wave in recent months has been an increasing number of apps going 64bit. Windows 7 (and cheap ram) imho is driving this and that's a good thing, although technically there are not many advantages for decoding.

Clarify for me, CUDA will use less CPU than outright DXVA or the other way around.

I'm a fan of your product, seat of the pants tells me it worked better with CoreAVC than Win7's default codec, but I feel that I want to install as few third party things as possible on the machine.

CiNcH
27th November 2009, 08:13
Clarify for me, CUDA will use less CPU than outright DXVA or the other way around.
AFAIK, CUDA copies back the decompressed video into main memory, which takes some CPU time. The pro is that you can post process video in "software".

Shii
27th November 2009, 09:21
As stated, CoreAVC 2.0 w/Haali splitter now makes MKV the default for directshow

What about mp4? Will WMC and WMP use CoreAVC 2.0 for decoding mp4 h264 files?

hydra3333
27th November 2009, 10:04
... and you can only go back and blame MS for creating MediaFoundation which is really not needed (in our opinion) unless you want to push EVR/DXVA, DRM, or finite Codec/Splitter control (although some would argue this point). Does anyone see a compelling reason to switch after seeing this comp chart? http://msdn.microsoft.com/en-us/library/aa468614.aspx#appendix_feature_comparisons__zftb

Nope, M$ are only in it for the almighty US $ that they and their rich mates pray to... text from that M$ link says it all
The key feature of Media Foundation that supports protected content is the Protected Media Path (PMP), which provides a protected environment for running audio and video processing pipelines.
i.e. they want to suck the life out of your PC and stop you from doing things you want to on your own PC ... and you pay for the privilege of them sticking it to you.

cheeky so and so's, they've lost the plot about who their customer really is... unless big content is bribing them to do it.

G_M_C
27th November 2009, 10:39
Nope, M$ are only in it for the almighty US $ that they and their rich mates pray to... text from that M$ link says it all

i.e. they want to suck the life out of your PC and stop you from doing things you want to on your own PC ... and you pay for the privilege of them sticking it to you.

cheeky so and so's, they've lost the plot about who their customer really is... unless big content is bribing them to do it.

Hmm, your opinion of DRM is quite clear. But I think you overreact somewhat.

Most PC's dont even have a BD-drive or another component that requires PMP/PAP. Most people just want to use youtube or what not. And realistically speaking; Most people play MKV's or AVI (i.e the wellknown "x264-somegroup" type videos) too. And MS codecs makes it easy for clueless-n00bs (*) to do so without having to mess with codec-packs or worm-toolbar-spyware-infected-shit .

The bad thing is that they made it impossible for people that are NOT clueless n00bs (*) to choose their own software. Personally, i'd even consider buying a Win7-N (without windows media-player), cause i dont like MS mediaplayer. With my current Win XP I don't use it ever, and i've blocked it with my firewall atm. But I fear that i dont have freedom anymore in Win7, hence my idea of buying Win7-N. They should also have made a Win7/Stripped-edition, without mediaplayer or MS codecs; I would have bought that.

Anyway; Just wait a few months and there will probably a plethora of tools circumventing this Mediafoundation hinderence ;)

((*) clueless n00b => (C) Doom9 ;) )

hydra3333
27th November 2009, 10:48
Thanks for the label :p If you don't care what people do with your PC whilst making you pay for it, then that's up to you.

MS codecs and the stuff that goes with it don't work as well as say this codec here nor ffdshow... and witness media foundation whos primary purpose is to lock stuff down... agh, you're entitled to your view.

I too await tools and techniques built for the users to use as they choose not as enforced by the mates of big content.

dimitrik
27th November 2009, 22:50
Well.... I disagree.

CoreAVC already works on x64 just fine, my desktop rig is such. Its not a 64bit component but then you have to have every filter in the chain being 64bit which complicates matters, even if you wanted to use 64bit MPC-HC with CoreAVC 64bit when its out, you can do that but I just don't see any real benefit. 64bit seems to me to be the revolution that never was.

If it wasn't for the large amounts of cheap RAM I'm not sure anybody would be rushing out to get x64 OS.

And CoreAVC isn't as critical to 7 as it was to Vista IMO, with 7 having a working default h.264 codec built in. Vista was much harder to get to play DXVA in MCE. CoreAVC 'may' be better on 7 but even as an owner of CoreAVC I'm not sure if I will install it on my 7 rig.

Well, it's your right to disagree and I appreciate your points. However I don't think you are the whole of the market.

While x64 users might not be the biggest share of the market, they're growing fast and for valid reasons which makes us an important segment for CoreAVC to address - 32-bit will eventually disappear, just like 16-bit did. To take your points one by one:

1) It's good that you are happy with the 32-bit codecs, and so am I - on my desktop. But on my HTPC I want to run windows media center, which is only 64-bit on Win7 x64. So if I want to use CoreAVC, I'm stuck with 32-bit windows - not cool.

2) Why not cool? Well, for one thing x64 windows is significantly more robust, responsive and faster on newer systems. It can access all the physical memory, has better threading on multi-core CPUs, and fully utilizes the 64-bit CPU capabilities. In most tests, x64 systems beat x86 systems hands down. True it does little for decoding, but it does a lot for everything else.

3) About a year ago windows x64 adoption was tripling every 3 months. It has possibly accelerated since especially with win 7 coming out and MS encouraging adoption of the x64 platform.

4) As for your choice to stick with the MS codecs, that's fine if DXVA is all you care about. Speaking for both me and many other HTPC users, we want subtitles (not available with MS codecs and WMF, multiple-audio streams (also not possible - you need haali), more audio decoding and processing options (forget it, you're stuck with the MS audio decoder only - at least in MC) and lots of other things which the default codecs don't support or allow you to configure.

As for DXVA - seriously who cares? Most CPU's today handle 1080p in software just fine. I heat is a problem, you can buy a 2-core energy efficient CPU for less than $100. DXVA might be useful for a couple of years but just ask yourself who cares about hardware accelerated DVD decoding today?

Virtual_ManPL
28th November 2009, 17:42
I find 2 bugs in CoreAVC Professional 1.9.5.0

1 bug
I see blocks in CUDA mode, soft mode decode this properly
x264 - core 55 - H.264/MPEG-4 AVC codec - Copyleft 2005 - http://www.videolan.org/x264.html - options: cabac=1 ref=5 deblock=1:-6:-6 analyse=0x3:0x133 me=umh subme=7 brdo=1 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=2 deadzone=21,11 chroma_qp_offset=0 threads=3 nr=0 decimate=1 mbaff=0 bframes=2 b_pyramid=1 b_adapt=1 b_bias=0 direct=3 wpredb=1 bime=1 keyint=250 keyint_min=25 scenecut=40(pre) rc=2pass bitrate=5299 ratetol=1.0 rceq='blurCplx^(1-qComp)' qcomp=0.60 qpmin=10 qpmax=51 qpstep=4 cplxblur=20.0 qblur=0.5 ip_ratio=1.40 pb_ratio=1.30 zones aq=1:0.3:15.0
file:
CoreAVC-large-motion-vectors-x264-b0rk.mkv (http://www.cccp-project.net/beta/test_files/CoreAVC-large-motion-vectors-x264-b0rk.mkv)
http://img101.imageshack.us/img101/9430/coreavc1.th.png (http://img101.imageshack.us/i/coreavc1.png/)



2 bug
frames are wrongly decoded in CUDA and soft mode, internal MPC-HC decoder decode this properly
OLS05lossless--avisource_82_92.mp4
x264 - core 55 svn-663C - H.264/MPEG-4 AVC codec - Copyleft 2005 - http://www.videolan.org/x264.html - options: cabac=1 ref=7 deblock=1:-1:-1 analyse=0x3:0x133 me=umh subme=7 brdo=1 mixed_ref=1 me_range=32 chroma_me=1 trellis=1 8x8dct=1 cqm=2 deadzone=21,11 chroma_qp_offset=0 threads=1 nr=0 decimate=1 mbaff=0 bframes=2 b_pyramid=1 b_adapt=1 b_bias=0 direct=2 wpredb=1 bime=1 keyint=240 keyint_min=25 scenecut=40 rc=2pass bitrate=1717 ratetol=1.0 rceq='blurCplx^(1-qComp)' qcomp=0.60 qpmin=10 qpmax=51 qpstep=4 cplxblur=20.0 qblur=0.5 ip_ratio=1.40 pb_ratio=1.30
file:
OLS05lossless--avisource_82_92.mp4 (http://www.cccp-project.net/beta/test_files/OLS05lossless--avisource_82_92.mp4)

CoreAVC
http://img192.imageshack.us/img192/4695/coreavc2.th.png (http://img192.imageshack.us/i/coreavc2.png/)
internal MPC-HC decoder
http://img192.imageshack.us/img192/6159/mpchc.th.png (http://img192.imageshack.us/i/mpchc.png/)



and you can also try other files on this FTP (http://www.cccp-project.net/beta/test_files/) to check that everything working properly with CoreAVC 2.0 before official release ;)

hydra3333
29th November 2009, 00:23
Latest new nvidia driver 195.62 says
"Adds support for OpenCL 1.0 (Open Computing Language) for all GeForce 8-series and later GPUs."
Does that change anything ?

LoRd_MuldeR
29th November 2009, 00:41
No, it doesn't. And it has been explained a dozen times why OpenCL isn't relevant for CoreAVC.

:search:

roozhou
29th November 2009, 09:52
No, it doesn't. And it has been explained a dozen times why OpenCL isn't relevant for CoreAVC.

:search:
It is not easy to explain to those who knows little about computer architecture. People see those words in NV's new drivers and believe GPU+OpenCL may help speed up everything.

LoRd_MuldeR
29th November 2009, 14:07
It is not easy to explain to those who knows little about computer architecture.

That's why it has been explained in this thread. More than one time ;)

TheShadowRunner
30th November 2009, 01:00
Please take as much time as you want to release 2.0, but by all means correct the "live resolution switching issue (http://forum.doom9.org/showthread.php?p=1282057#post1282057)" before release!

mandarinka
30th November 2009, 03:30
Oh, I also have an issue with resolution switching, this time with both software and hardware mode. Actually, I don't use coreAVC myself (although I admire how it's optimized for K7/K8!), a friend pointed out the error for me.

What happens: there are video corruptions or crashes if you encounter/seek through a resolution change in h.264 video - in my case they are two matroska segments linked together into one merged timeline. I used different pps/sps id for both segments when encoding to x264 (I also found out that I have to use raw output in x264, for those who are interested), which causes filter chain reconnection (?) and enables decoder to switch the resolution.

This works perfectly with most if not all renderes - if I use divx decoder or ffdshow-tryouts (in its case, it works with revision 3008 and lower, there is a regression after that, coincidentally). Obviously this is with haali splitter, else there would be no linking.
Now with coreavc (both software and hardware decoding), one gets video corruption when the resolution switches, and the decoding isn't corrected in further gops either, it even corrupts when you seek elsewhere - forward, or back to the previous segment. I checked if the decoder switches the resolution, and it apparently does so (else the video would be totally beyond recognition from my experience), mybe the parsing somehow seems to get glitched, resulting in the corruptions.

Because the affected video files (movie with changing AR) were too cumbersome, I made a couple of small samples with the same workflow, to replicate the issue.

------------------------------------------------------------
http://www.mediafire.com/?2z1ziktwlyj
http://www.mediafire.com/?tmvmjn2nemz

They are for demostration purposes only (pardon for disclaiming :eek:). Video was borrowed from import ntsc dvd of Osamu Tezuka's Cleopatra movie, audio is a random mp3 with some amateur performance (used just to simulate audio track, note that it's intentionally the same for both parts) and subtitle script muxed in is just a dummy (added for the same reason)...

x264 commandline:
--crf 19 --no-mbtree --tune grain --b-pyramid --keyint 48 --preset slow --sps-id 1 --output "cleoA.264"
--crf 19 --no-mbtree --tune grain --b-pyramid --keyint 48 --preset slow --sps-id 2 --output "cleoB.264"


sample mkvmerge 2.4.2 commandline (the same is used for segment A too, simple blocks not enabled):
-o cleoB.mkv --priority lower -s 0 -D -A sample.ssa -a 0 -D -S BABICKA.MP3 --default-duration 0:24000/1001fps -d 0 -A -S cleoB.264 --track-order 2:0,1:0,0:0
------------------------------------------------------------
So much for the report, thanks for your attention!

PS: Hello to forum members :)

[Edit:] PPS: I tried this (http://forum.doom9.org/showpost.php?p=1348169&postcount=211) version of Diavc; it doesn't show the corruptions I was experiencing with coreavc and links fine if width of video changes, while for some reason crashing on height change (so not a complete pass).

Mixer73
30th November 2009, 10:44
2) Why not cool? Well, for one thing x64 windows is significantly more robust, responsive and faster on newer systems. It can access all the physical memory, has better threading on multi-core CPUs, and fully utilizes the 64-bit CPU capabilities. In most tests, x64 systems beat x86 systems hands down. True it does little for decoding, but it does a lot for everything else.

Personally I've yet to see a single benchmark that actually shows this.

Potentially the adoption of a new 'lowest common denominator' instruction set optimisation could provide a speed boost but I'm yet to see any applications that can back this up.

You're right, unlike most here I don't use subs or any such thing so I don't have those requirements, but I still had CoreAVC on my Vista system because it was the simplest and most compatible solution. However, I see no point at all in running 64bit for a MCE platform - I fail to see any benefits at all for this usage, so I'm running 32bit.

I have no argument that its the right way to go, but the way I see it, right now, it costs a lot of development time and money, without providing any additional benefits to the consumer.

My POV on this is influenced by my employment position, which is in the video editing & video compression fields. Personally there have been better ways for our staff to spend their development time to give more features for our customers than to work on 64bit code which at this point in time doesn't deliver a tangible benefit across all relevant areas.

Indeed as you've stated its something that will need to be done, but there's more productive areas IMO.

Guest
30th November 2009, 14:15
Debate about x64 Windows is OT here. Please start a new thread in an appropriate forum.

squid_80
2nd December 2009, 10:56
I find 2 bugs in CoreAVC Professional 1.9.5.0

1 bug
I see blocks in CUDA mode, soft mode decode this properly

file:
CoreAVC-large-motion-vectors-x264-b0rk.mkv (http://www.cccp-project.net/beta/test_files/CoreAVC-large-motion-vectors-x264-b0rk.mkv)
http://img101.imageshack.us/img101/9430/coreavc1.th.png (http://img101.imageshack.us/i/coreavc1.png/)
Looks like it was encoded with an old build of x264, which didn't restrict vertical motion vectors properly.


2 bug
frames are wrongly decoded in CUDA and soft mode, internal MPC-HC decoder decode this properly

file:
OLS05lossless--avisource_82_92.mp4 (http://www.cccp-project.net/beta/test_files/OLS05lossless--avisource_82_92.mp4)

CoreAVC
http://img192.imageshack.us/img192/4695/coreavc2.th.png (http://img192.imageshack.us/i/coreavc2.png/)
internal MPC-HC decoder
http://img192.imageshack.us/img192/6159/mpchc.th.png (http://img192.imageshack.us/i/mpchc.png/)

I tried the following decoders and none of them decoded the clip properly:
- DivX
- ArcSoft
- ffmpeg
- DiAVC
- MPC-HC DXVA

Virtual_ManPL
2nd December 2009, 16:02
Looks like it was encoded with an old build of x264, which didn't restrict vertical motion vectors properly.
dunno, soft mode works good, CUDA not ;)


I tried the following decoders and none of them decoded the clip properly:
- DivX
- ArcSoft
- ffmpeg
- DiAVC
- MPC-HC DXVA
but MPC-HC can

Astrophizz
3rd December 2009, 01:17
MPC-HC (internal) is ffmpeg...

sneaker_ger
3rd December 2009, 17:32
Can someone enlighten me about the reference frame limit of CoreAVC in CUDA mode? On the CoreCodec site it says that 16 frames is the limit but on many pages it says that the limit is actually 15 frames. Or does this depend on the version of CoreAVC, the resolution, graphic card and frame rate?
And how does CoreAVC decide if it will decode a video with CUDA or in pure software mode? Does it only read the level or does it inspect the x264 info (if x264 was used and the info not stripped)?

honai
3rd December 2009, 18:06
squid_80 answered this a while back:

http://forum.doom9.org/showthread.php?p=1262588#post1262588
http://forum.doom9.org/showthread.php?p=1192953#post1192953
http://forum.doom9.org/showthread.php?p=1262558#post1262558

sneaker_ger
3rd December 2009, 18:23
That answers my first question, thank you! The formulation on the CoreAVC page (http://forum.corecodec.com/viewtopic.php?f=3&t=69) can lead to misunderstanding.
Still leaves me wondering about those "out-of-order frames" though. How does CoreAVC make sure those frames won't get corrupted e.g. by using pure software mode?

schweinsz
5th December 2009, 16:19
I find 2 bugs in CoreAVC Professional 1.9.5.0

1 bug
I see blocks in CUDA mode, soft mode decode this properly
The.Illusionist.2006.HDTV.720p.x264-iLL.mkv
x264 - core 55 - H.264/MPEG-4 AVC codec - Copyleft 2005 - http://www.videolan.org/x264.html - options: cabac=1 ref=5 deblock=1:-6:-6 analyse=0x3:0x133 me=umh subme=7 brdo=1 mixed_ref=1 me_range=16 chroma_me=1 trellis=1 8x8dct=1 cqm=2 deadzone=21,11 chroma_qp_offset=0 threads=3 nr=0 decimate=1 mbaff=0 bframes=2 b_pyramid=1 b_adapt=1 b_bias=0 direct=3 wpredb=1 bime=1 keyint=250 keyint_min=25 scenecut=40(pre) rc=2pass bitrate=5299 ratetol=1.0 rceq='blurCplx^(1-qComp)' qcomp=0.60 qpmin=10 qpmax=51 qpstep=4 cplxblur=20.0 qblur=0.5 ip_ratio=1.40 pb_ratio=1.30 zones aq=1:0.3:15.0
file:
CoreAVC-large-motion-vectors-x264-b0rk.mkv (http://www.cccp-project.net/beta/test_files/CoreAVC-large-motion-vectors-x264-b0rk.mkv)
http://img101.imageshack.us/img101/9430/coreavc1.th.png (http://img101.imageshack.us/i/coreavc1.png/)



2 bug
frames are wrongly decoded in CUDA and soft mode, internal MPC-HC decoder decode this properly
OLS05lossless--avisource_82_92.mp4
x264 - core 55 svn-663C - H.264/MPEG-4 AVC codec - Copyleft 2005 - http://www.videolan.org/x264.html - options: cabac=1 ref=7 deblock=1:-1:-1 analyse=0x3:0x133 me=umh subme=7 brdo=1 mixed_ref=1 me_range=32 chroma_me=1 trellis=1 8x8dct=1 cqm=2 deadzone=21,11 chroma_qp_offset=0 threads=1 nr=0 decimate=1 mbaff=0 bframes=2 b_pyramid=1 b_adapt=1 b_bias=0 direct=2 wpredb=1 bime=1 keyint=240 keyint_min=25 scenecut=40 rc=2pass bitrate=1717 ratetol=1.0 rceq='blurCplx^(1-qComp)' qcomp=0.60 qpmin=10 qpmax=51 qpstep=4 cplxblur=20.0 qblur=0.5 ip_ratio=1.40 pb_ratio=1.30
file:
OLS05lossless--avisource_82_92.mp4 (http://www.cccp-project.net/beta/test_files/OLS05lossless--avisource_82_92.mp4)

CoreAVC
http://img192.imageshack.us/img192/4695/coreavc2.th.png (http://img192.imageshack.us/i/coreavc2.png/)
internal MPC-HC decoder
http://img192.imageshack.us/img192/6159/mpchc.th.png (http://img192.imageshack.us/i/mpchc.png/)



and you can also try other files on this FTP (http://www.cccp-project.net/beta/test_files/) to check that everything working properly with CoreAVC 2.0 before official release ;)

I extracted the raw 264 bitstream in OLS05lossless--avisource_82_92.mp4 and decode it to locate the error and I found it is a corrupted bitstream. When MB with (mbx=26, mby=1)is decoded, the end_of_slice_flag=1 in the first P picture with POC=4, so I believe no decoder can get the "correct" result, the needed operation is error concealment.

Virtual_ManPL
6th December 2009, 01:02
@ schweinsz - yep, it's probably like you said, but look how decode this newest MPC-HC in soft mode vs your decoder, CoreAVC (CUDA & soft) and DivX (+ MPC DxVA decoding)
only with MPC decoder in soft mode I get not pixelated frame like you can see on attached SS, odd ;p

Keiyakusha
6th December 2009, 01:17
Maybe it works in MPC because of some bug ^_^

Anyway may I ask about status of this multi-something bug in CoreAVC 2.0? Its not that I'm asking about releasing it now, I just thought that bug was not so serious...

schweinsz
6th December 2009, 08:54
@ schweinsz - yep, it's probably like you said, but look how decode this newest MPC-HC in soft mode vs your decoder, CoreAVC (CUDA & soft) and DivX (+ MPC DxVA decoding)
only with MPC decoder in soft mode I get not pixelated frame like you can see on attached SS, odd ;p

The first P picture with poc=4 is almostly same with the preceding I picture with poc=0, if error concealment is done, the P picture will be recovered perfectly. the current DiAVC is without the error concealment and I will add it after beta version is released.

bluebebe
7th December 2009, 15:49
hi there, i thought it was another error with the encoder, but its coreavc, it produces in a few videos of mine errors like this:

http://i46.tinypic.com/23j692o.jpg

http://i46.tinypic.com/4kzq6w.jpg


It is a bug or what? when i try decoding a x264 thru ffdshow i dont have that artifacts. hope anyone can tell me what the problem is.

JEEB
7th December 2009, 16:01
You might want to read some of the last pages of this thread. I think it's pretty sure that it's just the decoding bug/feature that's present in the 1.X series of CoreAVC and that has been fixed/added for quite some time already, just that CoreCodec hasn't gotten to releasing their 2.X version yet (which contains the support for weightp). nvidia's VP2 decoding doesn't have problems with decoding such content, just to note it one more time (CoreAVC's CUDA feature uses the VP2 hardware on the GPU).

You might want to use another decoder until CoreAVC 2.X gets released, and then check if it contains enough reasons for you to upgrade :)

squid_80
7th December 2009, 16:38
It's virtually impossible to confirm the cause of a problem from screenshots, stream samples are needed.

bluebebe
7th December 2009, 19:26
no samples needed, with cuda the video is working fine. thanks for help :)

pankov
8th December 2009, 00:40
Well, I too have problems (very rare to be fair) with CoreAVC's decoding.
Here is a sample showing the problem captured in the picture bellow. The same file plays fine both with FFDShow and MPC-HC decoders.
[link removed]

http://img44.imageshack.us/img44/3894/samplemkv00000.th.jpg (http://img44.imageshack.us/i/samplemkv00000.jpg/)

I'm using ATI so I don't have a chance to test it with CUDA. Can someone confirm/deny that this is the same weightp problem?

pankov
8th December 2009, 01:24
no, it's a sample I've found on the net with complaints about bad quality and just wanted to check if the problem was present on my machine too.
As you can see - it does and I'm still not sure which is the problem the source or CoreAVC.

I read that it's not against the rules to post samples, but if I misread/misunderstood the rules I'll remove it.

Stephen R. Savage
8th December 2009, 03:35
If it was encoded with x264, you can look at the headers in a hex editor (search for ASCII string "x264"). Streams encoded with weight-p on will display either "wpredp=1" or "wpredp=2". If it was not encoded with x264, you probably have to use a stream analyser like Elecard StreamEye or something.

LoRd_MuldeR
8th December 2009, 14:25
If it was encoded with x264, you can look at the headers in a hex editor (search for ASCII string "x264").

Or more user-friendly, look at the x264 encoding parameters with a tool like MediaInfo or AVInaptic :)

weasel_
9th December 2009, 20:33
Well, I too have problems (very rare to be fair) with CoreAVC's decoding.
Here is a sample showing the problem captured in the picture bellow. The same file plays fine both with FFDShow and MPC-HC decoders.
[link removed]

http://img44.imageshack.us/img44/3894/samplemkv00000.th.jpg (http://img44.imageshack.us/i/samplemkv00000.jpg/)

I'm using ATI so I don't have a chance to test it with CUDA. Can someone confirm/deny that this is the same weightp problem?
put in mediainfo and see if weightp=2

pankov
10th December 2009, 01:39
Well, I did try this (with the latest version from sourceforge - GUI 0.7.25) and here is the result
Writing library : x264 core 79 r1342 e8501ef
Encoding settings : cabac=1 / ref=5 / deblock=1:-3:-3 / analyse=0x3:0x133 / me=esa / subme=10 / psy=1 / psy_rd=1.0:0.3 / mixed_ref=1 / me_range=24 / chroma_me=1 / trellis=2 / 8x8dct=1 / cqm=0 / deadzone=21,11 / chroma_qp_offset=-4 / threads=6 / nr=0 / decimate=0 / mbaff=0 / constrained_intra=0 / bframes=5 / b_pyramid=1 / b_adapt=2 / b_bias=0 / direct=3 / wpredb=1 / wpredp=2 / keyint=250 / keyint_min=25 / scenecut=40 / rc=2pass / mbtree=0 / bitrate=14648 / ratetol=1.0 / qcomp=0.60 / qpmin=10 / qpmax=30 / qpstep=4 / cplxblur=20.0 / qblur=0.5 / ip_ratio=1.40 / pb_ratio=1.30 / aq=1:0.70

I have to admit I don't understand 90% of these parameters but I do see that there is no wieghtp or alike.

kemuri-_9
10th December 2009, 01:47
I have to admit I don't understand 90% of these parameters but I do see that there is no wieghtp or alike.

weightp is written as wpredp in the SEI,
see the wpredp=2 in your options there which indicates weightp 2 was used.

pankov
10th December 2009, 01:51
10x for the help
Now I'm officially a "victim" of the weightp bug/feature of CoreAVC
;)

BetaBoy
10th December 2009, 20:08
Well... a good thing now is that we are reworking how we handle MV's now (we now support non-spec compliant streams), which in-turn had a direct effect on our current weightp support.

ajp_anton
10th December 2009, 21:15
Will the 64-bit version be any faster? And if so, how much?

BetaBoy
10th December 2009, 21:38
Will the 64-bit version be any faster? And if so, how much?

We've discussed this several times here in this thread, but... you'll likely see negligible speed improvements for 64bit. I'll leave it for you all to tell us that after its released just how 'true' that to be, but I think many of you will be surprised.

Snowknight26
10th December 2009, 21:45
Any updates on that one bug you guys were aware of that hadn't been fixed at the time you mentioned it... whatever that bug was/is?

BetaBoy
10th December 2009, 21:51
iirc I did state its a weightp bug related to our support for the updated x264 and is a part of what we are working on now.

hajj_3
10th December 2009, 22:06
don't suppose you have a new eta?

dead_screem
10th December 2009, 22:44
Well... a good thing now is that we are reworking how we handle MV's now (we now support non-spec compliant streams), which in-turn had a direct effect on our current weightp support.

How is it supported? With two internal decoders? One with spec compliant MV limits and the other unlimited? (as mentioned a few times in this thread,). Or just one decoder with unlimited MV support?

Hans Ohlo
10th December 2009, 22:44
it is really disappointing as always with core. we get an eta for the avc decoder and haali with x64 support weeks ago and still nothing. really great as always. we should have known better.

Shakey_Jake33
10th December 2009, 23:11
In fairness, people have been fussing for support for out of spec MVs and the like for ages, and people critisised CoreCodec a few pages ago for not supporting it. I don't think it's necessarily fair to critisise them for doing what people have asked them to do.

Snowknight26
10th December 2009, 23:16
iirc I did state its a weightp bug related to our support for the updated x264 and is a part of what we are working on now.

My mistake - I thought there was only one bug in CoreAVC and HMS, not simply CoreAVC. You mentioned in one of your previous posts you were only aware of one bug in the 2.0 codebase, but I skipped over the part where you mentioned that you don't work in tandem with Haali. Bummer.

hydra3333
11th December 2009, 01:29
it is really disappointing as always with core. we get an eta for the avc decoder and haali with x64 support weeks ago and still nothing. really great as always. we should have known better.
Crikey, you're a bit on the down side ! Cheer up mate, when it arrives think how you'll enjoy it then.

DigitalDeviant
11th December 2009, 01:35
it is really disappointing as always with core. we get an eta for the avc decoder and haali with x64 support weeks ago and still nothing. really great as always. we should have known better.

LOL, when has Core ever met a release date they set? This has been a tradition going back to 1.0 back in April 2006, why break it now?

squid_80
11th December 2009, 04:39
LOL, when has Core ever met a release date they set? This has been a tradition going back to 1.0 back in April 2006, why break it now?

When was a definite release date given?

BetaBoy
11th December 2009, 05:12
My mistake - I thought there was only one bug in CoreAVC and HMS, not simply CoreAVC. You mentioned in one of your previous posts you were only aware of one bug in the 2.0 codebase, but I skipped over the part where you mentioned that you don't work in tandem with Haali. Bummer.
Well we found out that the Motion Vector limit we had has a direct effect on WeightP, so its a 2-for-1 bug fix and requires we rewrite a nice chunk of code. The good thing about this though is as I stated, a) It removes the MV limitation, and now b) Will take into account some future x264 WeightP changes, which we already know are coming sometime next year.

While its obvious we have a vested interest in Haali's work with the splitter, he works at his own pace but has always listened to what needs to be fixed/changed... and if you ask many of the testers (*cough*Jarod) the latest versions have fixed almost every known issue (outside of VC1), but I'm sure more needs to be done.

betaking
11th December 2009, 07:09
Well we found out that the Motion Vector limit we had has a direct effect on WeightP, so its a 2-for-1 bug fix and requires we rewrite a nice chunk of code. The good thing about this though is as I stated, a) It removes the MV limitation, and now b) Will take into account some future x264 WeightP changes, which we already know are coming sometime next year.

While its obvious we have a vested interest in Haali's work with the splitter, he works at his own pace but has always listened to what needs to be fixed/changed... and if you ask many of the testers (*cough*Jarod) the latest versions have fixed almost every known issue (outside of VC1), but I'm sure more needs to be done.

:)today i test Haali Media Splitter 1.9.355.21! on windows7 wmp12 can use haali play avi files but test ts/tp file not work! use mpc-hc use haali play some ts file encoder by vc1 not see video but sound is ok! and some vob files can see video but no sound! sound encoder by lpcm! can repot it to haali ?thanks!

BetaBoy
15th December 2009, 01:48
A big hurdle has been overcome. CoreAVC 2.0 now supports Non-Compliant Motion Vector's.... and it gets even better, WeightP is now fully supported ;-) (this also includes support for future x264 weightp changes coming in 2010). We are going through QA now and tweaking the code as we go. No timeline on the release, let's see how QA goes.... but current results puts us on top again, but there is still room for improvement ;-)

Thanx all for your patience.

Mixer73
15th December 2009, 05:36
Bring it on Dan!

hydra3333
15th December 2009, 08:27
Yep, do a good job in QA and then please release it as soon as you reasonably can !

link2009
15th December 2009, 22:59
When 2.0 gets released and I see a decent improvement for playback, I will re-purchase it to support the project.

Romario
16th December 2009, 01:26
BetaBoy,please, I have one question for you. :)

Can you, please, tell me from which site I can download latest Development version of Halli media splitter, which you will use in CoreAVC 2.0 ?

thanks. I want to test it on Windows 7.

BetaBoy
16th December 2009, 03:56
Technically its the same that's on Haali's site right now.... all we do is add our special sauce (non-hacking) to make it work on Windows 7.

honai
16th December 2009, 04:38
While its obvious we have a vested interest in Haali's work with the splitter, he works at his own pace but has always listened to what needs to be fixed/changed... and if you ask many of the testers (*cough*Jarod) the latest versions have fixed almost every known issue (outside of VC1), but I'm sure more needs to be done.

Can you, please, tell me from which site I can download latest Development version of Halli media splitter, which you will use in CoreAVC 2.0 ?

thanks. I want to test it on Windows 7.

Technically its the same that's on Haali's site right now.... all we do is add our special sauce (non-hacking) to make it work on Windows 7.

Now you got me really confused. So does the latest version at your hands realize bug-fixes not implemented in the latest downloadable splitter (dated 11/01/2009), or not?

BetaBoy
16th December 2009, 05:21
Now you got me really confused. So does the latest version at your hands realize bug-fixes not implemented in the latest downloadable splitter (dated 11/01/2009), or not?
Have you tested the new version Haali posted on his site for any bugs?

ChronoCross
16th December 2009, 05:43
Have you tested the new version Haali posted on his site for any bugs?

The one on Haali's site is from January.

BetaBoy
16th December 2009, 05:53
I thought Haali had posted the new version already. Pinging him now.

ajp_anton
16th December 2009, 06:24
A little offt, but a "11/01/2009" date can be either Nov 1st or Jan 11th, depending on which one of the weird formats you use.
2009/01/11 = no confusion.

LoRd_MuldeR
16th December 2009, 13:47
I'd prefer the ISO 8601 format. For example:
2009-12-16

And the latest version on Haali's site definitely dates back to 2009-01-11 ;)

THX-UltraII
16th December 2009, 16:40
great news that the new Haali Renderer will be up soon!

Carpo
16th December 2009, 16:45
http://haali.net/mkv/mkx.y.9.exe - thought that was the latest unofficial version about

Keiyakusha
16th December 2009, 16:47
THX-UltraII
Renderer? I guess you mean splitter? Or there is some changes to the renderer too?

THX-UltraII
16th December 2009, 16:50
THX-UltraII
Renderer? I guess you mean splitter? Or there is some changes to the renderer too?

sorry, mistyped

hydra3333
17th December 2009, 03:54
Technically its the same that's on Haali's site right now.... all we do is add our special sauce (non-hacking) to make it work on Windows 7.

Hmm, is this http://www.videohelp.com/tools/Preferred_Filter_Tweaker your work in getting codecs to work under Win7 ?

Jeff Flowerday
17th December 2009, 04:30
Hmm, is this http://www.videohelp.com/tools/Preferred_Filter_Tweaker your work in getting codecs to work under Win7 ?

:confused:

http://forum.doom9.org/showthread.php?t=146910

hydra3333
17th December 2009, 05:18
Ah, it's something different. Thanks.

honai
17th December 2009, 08:14
I thought Haali had posted the new version already. Pinging him now.

I understand you are on a busy schedule, but are there any followups on this?

STaRGaZeR
17th December 2009, 10:26
You can get the (lastest?) Haali beta with CCCP beta. Deselect everything but Haali and you're done. It supports TrueHD in MKV and detects Blu-ray stereo LPCM tracks ;)

BetaBoy
17th December 2009, 10:57
I understand you are on a busy schedule, but are there any followups on this?

Haali will post it on his site when we launch 2.0.

tal.aloni
17th December 2009, 11:37
You can get the (lastest?) Haali beta with CCCP beta.

the revision included is Haali 1.9.355.21 (2009-11-14)
you might as well grab it from his site (sorry):
-removed-

nice progress haali, thanks!
however, this release crash with the m2ts samples I posted earlier (TrueHD / EAC3), and also crash with another Blu-Ray I have (TrueHD).

Note:
I have not yet investigated the origin of the crash, there's a slight chance it's unrelated to the splitter.

Edit:
The crash is related to ffdshow, later build resolve the crash, but the playback is still very choppy with TrueHD, I'm not sure if it's a bug with the splitter or the decoder.

BetaBoy
17th December 2009, 11:53
tal.aloni.... Excuse me if I ask but, has Haali given you the ok to post that? As far as we are concerned we have respected Haali's work until we have the OK to distribute it or he posts a public link to it (which I do not think has occurred).

tal.aloni
17th December 2009, 13:11
I have removed the link, I don't want to offend anybody,
however, it's already embedded into many codec packs.
the file name was published already by a few codec-pack authors.

BetaBoy
17th December 2009, 18:10
tal.aloni... Thx it's appreciated. It seems some of the testers decided it was ok to give out the links ;-/ . . . there is a reason its only posted on Haali's site when its ready.

Keiyakusha
17th December 2009, 18:15
Not necessarily. It's not hard to guess that link using capabilities of any download manager with support downloading by mask.

BetaBoy
17th December 2009, 18:19
True.... but in this case I already have confirmation that the testers in fact did give out download links. No biggie now.... we just don't want reports coming back on versions that maybe old and fixed (or not ;-).

Carpo
17th December 2009, 23:34
so is there anything wrong with that build ;)

BetaBoy
18th December 2009, 00:16
It dosen't really matter at this point as Haali will be posting the update soon enough.

BetaBoy
18th December 2009, 23:37
Well the time has come... We are about to release CoreAVC 2.0 Professional Edition for Windows. Current customers will be notified first... Purchasers in the past 60 days get the upgrade included.... post 60 will get a unique code to upgrade. Emails are going out now for; 1) The new portal login account (all users get this), and 2) The upgrade email.

I'll post more details next.

BetaBoy
18th December 2009, 23:47
After 9 months of work, CoreCodec is proud to release CoreAVC 2.0 Professional Edition.

More details below.... but lets get to what's new:

CoreAVC H.264 Video Codec - Version 2.0.0.0 (20091218)
- ADD: Initial support for Windows 7
- ADD: Support for Windows Media Center (in .MKV)
- ADD: Custom fourcc to match the Haali splitter
- ADD: Support for Non-Compliant Motion Vectors (MV)
- ADD: mmx optimizations
- ADD: mmx2 optimizations
- ADD: sse optimizations
- ADD: sse2 optimizations
- ADD: sse3 optimizations
- ADD: ssse3 optimizations (almost 70)
- ADD: sse4 optimizations
- ADD: faster handling of pure-DC chroma blocks
- ADD: new x86 cpu caps
- ADD: Multi-dupe weightp (+Future x264 changes)
- ADD: 64bit support
- FIX: Fix potential failure in CABAC MVD decoding.
- FIX: Proper alignment intrinsics for MSVC and GCC.
- FIX: AVC SEI+userdata fix
- FIX: Focus bug in Windows Media Center that disabled CUDA
- FIX: Better frame re-ordering on difficult streams
- CHG: Supports 16 CPU Cores (was 4 cores max)
- CHG: Support for QuadHD resolutions(4096Ũ4096)
- CHG: New dc_add
- CHG: Faster CAVLC
- CHG: Faster CABAC
- CHG: Faster Deblocking (Massive deblocking overhaul)
- CHG: SSSE3: Faster Motion Compensation (20% faster on Core 2 Conroe and more for Penryn)
- CHG: New CoreNumber for 2.0
- CHG: Initial support for CoreAccount. Linking purchase+account
- CHG: Integrated the Haali splitter into the installer
- CHG: New installer supports both 32/64 bit Operating Systems
- CHG: Filter compiled against ICC (2% speed increase for modern Intel Processors)
- CHG: Updated IDCT to support ARM NEON Cortex A8 Support
- CHG: Updated Blitter YUV/RGB for ARM NEON Cortex A8 Support
- OEM: Initial ARM NEON Cortex A8 Support (iPhone, Touch, Linux, Windows Mobile)
- OEM: SDK NVIDIA CUDA 2.3 support
- OEM: Removed the windows direct3d requirement for CUDA (it can now be native)
- OEM: Updated GStreamer plug-in


Haali Media Splitter (20091112)
- ADD: Official 64bit support
- ADD: Official support for Windows 7
- ADD: Custom fourcc option for windows 7
- ADD: Silent install flags for each install option
- CHG: Disabled explorer thumbnail support (off by default)

Haali Media Splitter supports the following command line options:
/S - silent install without any UI
/MKVONLY - register only Matroska components
/AVI=[yes|no] - register AVI support
/MP4=[yes|no] - register MP4 support
/OGG=[yes|no] - register OGG/OGM support
/TS=[yes|no] - register MPEG TS support
/PS=[yes|no] - register MPEG PS support
/WMP=[yes|no] - register WMP to play in Windows 7


Bulletpoints on CoreAVC 2.0
-------------------------------
- First on Windows 7..... Both CoreAVC and Haali's splitter now use a custom fourcc so that when you play an MKV video it uses Directshow instead of MediaFoundation in WMP and MC. No hacks and its a simple solution to gets ppl out of the pitfalls of MF. Note that this is ONLY for .mkv and does not affect any other container.
- You will also see CoreAVC now supports both 32/64bit, but like Haali we have opted to keep both .AX's in the same folder for simplicity's sake.
- We would also like to thank Clsid for his help with some of the changes we made to the installer for all OS's but mostly Win7.

clsid
19th December 2009, 00:26
Haali Media Splitter (20091112)Based on the date I assume this is version 1.9.355.21. What will happen with the bugs that have been reported for that version? Will there be another release in the near future?

BetaBoy
19th December 2009, 00:33
If I am not mistaken the date in the readme is wrong. The build # is: 1.9.355.21 that's what counts (as you mention). But otherwise yes, Haali said when he gets time (motivation) he will work on the additional bugs reported.

Cyber-Mav
19th December 2009, 01:19
criky thats a big changelog listing a lot of speed boosts, hopefully someone will post some test results here showing how coreavc 2.0 fairs against divx, diavc and against coreavc 1.9.5

the_corona
19th December 2009, 01:28
Awesome, lots of speed optimizations!

Any benchmarks you can share? Which proc generation saw the biggest speed bump (probably the latest which ironically needs it the least, lol)

Disabled
19th December 2009, 02:06
One Question: Do you sell a full featured High Profile* Decoder this time? Or is it again possible that in 2 years x264 will use a feature previously not used and everyone has to buy a new licence?
(This of course excludes new features that are added to the specs after the purchase, as you don't need to promise on those.)

*normal High profile, not High10 or with 4:4:4 support

Dark Shikari
19th December 2009, 02:08
Awesome, lots of speed optimizations!

Any benchmarks you can share? Which proc generation saw the biggest speed bump (probably the latest which ironically needs it the least, lol)
The speed bump is pretty big, though note that the following test was done on a Core i7 with HT, and CoreAVC 1.9.5 was capped to 4 threads max, so the real boost is not as large as the graph suggests. Obviously, CoreAVC 1.9.5 was still faster than ffdshow-mt on a per-core basis; it's just that 8 threads of ffdshow-mt beat out 4 threads of CoreAVC 1.9.5.

http://i50.tinypic.com/b7ett3.png

CoreAVC is now slightly faster than DivX on 1080p HD content. The weightp example is somewhat invalid, as DivX does not decode weightp bit-exact. The results of the 2160p test were a bit odd though: I'm going to guess that DivX is faster on really low bitrate content (the 2160p test case was 4 megabit).

Note "CoreAVC 2.0 Intel" is using ICC, so you can see the benefit of that.

Cyber-Mav
19th December 2009, 02:20
yikes looks like divx is still damn quick. would be good to see how diavc compares with the above. i guess coreavc is only good for those who want cuda support, cant see people paying good money for a decoder thats hardly any quicker and in half the cases above slower than a free alternative. ??

Dark Shikari
19th December 2009, 02:22
yikes looks like divx is still damn quick. would be good to see how diavc compares with the above. i guess coreavc is only good for those who want cuda support, cant see people paying good money for a decoder thats hardly any quicker and in half the cases above slower than a free alternative. ??Well DivX still doesn't handle weightp correctly, albeit the artifacts are generally not too noticeable. Also note that CoreAVC likely has a larger advantage on older CPUs, which is where you most need the speed of a fast decoder to begin with. I don't have an older CPU to test it on though; I'd be curious about the results.

Hopefully in 2.1 the speed will be even higher ;)

Cyber-Mav
19th December 2009, 02:24
need some tests done on older hardware such as athlon xp. wonder if there is some inprovements for atom cpu's too.

shikari any idea if divx will sort out the weight-p issues you mentioned in future divx updates?

Dark Shikari
19th December 2009, 02:26
shikari any idea if divx will sort out the weight-p issues you mentioned in future divx updates?Dunno. Slap them until they do. It's a trivial deblocking bug that's very common among decoders, and fortunately one whose effects are probably not noticeable in real situations. It still should get fixed though.

Do note that CoreAVC had to sacrifice a small amount of speed (not much, but measurable) to fix the MV range and weightp issues.

the_corona
19th December 2009, 02:26
yikes looks like divx is still damn quick. would be good to see how diavc compares with the above. i guess coreavc is only good for those who want cuda support, cant see people paying good money for a decoder thats hardly any quicker and in half the cases above slower than a free alternative. ??

I agree, that's kinda dissapointing.....hmmm

But from what I remember DivX did comparitably good on high core/thread counts (probably better multithreading), so I'm guessing the graphs would look different on a 2 core machine. At least I'm hoping.....

Dark Shikari
19th December 2009, 02:28
I agree, that's kinda dissapointing.....hmmm

But from what I remember DivX did comparitably good on high core/thread counts (probably better multithreading), so I'm guessing the graphs would look different on a 2 core machine. At least I'm hoping.....Likely true as well. We're probably going to look into some multithreading tweaking with lots of cores, though to be honest one hardly needs CoreAVC on a system as beefy as the one I benchmarked on.

the_corona
19th December 2009, 02:33
Likely true as well. We're probably going to look into some multithreading tweaking with lots of cores, though to be honest one hardly needs CoreAVC on a system as beefy as the one I benchmarked on.

Wait...."we're" as in "we" as in you and CoreAVC? Are you on their team now?

Does that mean CoreAVC got their optimizations from yours truly?

Edit: I guess you are since the ICC version indicates you have access to the source code.

Dark Shikari
19th December 2009, 02:35
Wait...."we're" as in "we" as in you and CoreAVC? Are you on their team now?

Does that mean CoreAVC got their optimizations from yours truly?Yes, all the x86 asm since 1.9.5 has been written by me, including some from x264 that I wrote as well ;)

A bit has gone in the other direction too; x264's cacheline-split chroma MC was originally written for CoreAVC.

Snowknight26
19th December 2009, 02:45
Any word on when we can expect HMS as a standalone download?

Astrophizz
19th December 2009, 03:48
Looking forward to trying out a trial version to compare with DiAVC which is even faster than DivX on my system. If Dark's chart is any indicator then I'd have to agree with Cyber-Mav that it doesn't look good. Maybe I'll see CoreAVC's "larger advantage on older cpus" since I have a Core Duo (not Core2). Exciting stuff either way.

ranpha
19th December 2009, 04:01
Nice new website, the same as my website provider support system. Is that affiliate feature really work? 20% off purchase price is really generous. :p

dead_screem
19th December 2009, 05:49
so what's the deal? the website still has 1.9.5 and I havn't gotten an email. for the record the last update email I got was for 1.6. 1.6.5 and up I had to use the resend form every time, comcast seems to be blocking your bulk emails. So how do I get my 2.0 update 20% off email now?

Edit: I found the new core account system, (unfortunately it doesn't use our old core account email/passwords). It used my paypal email as the account email. (I had originally used a different email for the old core account then the one I payed with paypal) I tried logging in to the new system with my paypal email and my old core acount pasword and that didn't work. I tried the forgot password link and it said validation email sent but I got nothing. My paypal email was with aol. can't submit a support ticket because I would need to access my account to even do that... so now what?

lnatan25
19th December 2009, 09:46
A small question, will the h264 decoding bugs also find their way into the CorePlayer Movile platform?

dvy
19th December 2009, 10:45
CoreAVC is definite fastest H.264 decoder in planet earth.



Note "CoreAVC 2.0 Intel" is using ICC, so you can see the benefit of that.

Dark Shikari,why don't compile CoreAVC of windows version in icl,icl is wondows version intel C++ compiler.I compile x264 in icl is faster than x264 compile in gcc.

hippoth
19th December 2009, 10:56
What about a CoreAVC version for Mac OS X ... is there anything in the pipe?

blubberbirne
19th December 2009, 11:46
When buyers get the email with the download link???

EDIT: I can't login in customer area. My eMail don't work!

zn
19th December 2009, 12:18
any news on linux version? or android market?

BetaBoy
19th December 2009, 12:54
When buyers get the email with the download link???

EDIT: I can't login in customer area. My eMail don't work!

We are doing exactly as I stated emailing current customers with the account login first, then releasing it to them. So users will start to see those emails over the next few hours as they go continue to go out. Then we will open purchasing to the public.

BetaBoy
19th December 2009, 12:59
Nice new website, the same as my website provider support system. Is that affiliate feature really work? 20% off purchase price is really generous. :p
Correct, we have an affiliate program in place. However the rate will be 15% for 2.0.

Also on price. Upgrades are $4.95 and the full version is $9.95 (discounted from $12.95 till xmas).

BetaBoy
19th December 2009, 13:05
Looking forward to trying out a trial version to compare with DiAVC which is even faster than DivX on my system. If Dark's chart is any indicator then I'd have to agree with Cyber-Mav that it doesn't look good. Maybe I'll see CoreAVC's "larger advantage on older cpus" since I have a Core Duo (not Core2). Exciting stuff either way.
I did state this a while back in this thread that we will not offer a trial version of 2.0 for the first couple of weeks. The trial is scheduled to be released January 7th as CES begins as it is expected to be included on Media and some press junket's (NVIDIA, etc.)

BetaBoy
19th December 2009, 13:08
A small question, will the h264 decoding bugs also find their way into the CorePlayer Movile platform?Its the same thing really... so all the current fixes are already in CorePlayer 2.0. That also includes CUDA which we added a few months ago, but CorePlayer 2.0 will not be released in the next couple of weeks as we are still working on UI elements and trying to integrate our 'Corenect' UPNP/DLNA client/server into it for CES.

BetaBoy
19th December 2009, 13:11
so what's the deal? the website still has 1.9.5 and I havn't gotten an email. for the record the last update email I got was for 1.6. 1.6.5 and up I had to use the resend form every time, comcast seems to be blocking your bulk emails. So how do I get my 2.0 update 20% off email now?

Edit: I found the new core account system, (unfortunately it doesn't use our old core account email/passwords). It used my paypal email as the account email. (I had originally used a different email for the old core account then the one I payed with paypal) I tried logging in to the new system with my paypal email and my old core acount pasword and that didn't work. I tried the forgot password link and it said validation email sent but I got nothing. My paypal email was with aol. can't submit a support ticket because I would need to access my account to even do that... so now what?
Accounts get added into the system as the emails go out. So, no email=no account. Account emails are still going out.

schweinsz
19th December 2009, 13:41
yikes looks like divx is still damn quick. would be good to see how diavc compares with the above. i guess coreavc is only good for those who want cuda support, cant see people paying good money for a decoder thats hardly any quicker and in half the cases above slower than a free alternative. ??
From the results given by Dark Shikari, I believe the DiAVC alpha is still faster than the DiVX and coreavc2.0. I believe it is especially true when the DiAVC beta is out.

BetaBoy
19th December 2009, 13:42
schweinsz... no offense, but pls post results and not speculations and keep that talk in your thread, not this one.

schweinsz
19th December 2009, 13:54
schweinsz... no offense but pls post results and not speculations.
There are some results about coreavc1.9.5, divx and DiAVC.
http://forum.doom9.org/showthread.php?p=1331966#post1331966
http://forum.doom9.org/showthread.php?p=1334702#post1334702
http://forum.doom9.org/showthread.php?p=1344770#post1344770
http://forum.doom9.org/showthread.php?p=1347162#post1347162
http://forum.doom9.org/showthread.php?p=1347187#post1347187

I have not a coreavc 2.0, but allrough Dark Shikari gives that coreavc2.0 is comparible to divx, I believe it is slower than DiAVC.

I will give more results after you release the coreavc 2.0 to public between coreavc2.0 and the upcoming DiAVC beta.

STaRGaZeR
19th December 2009, 14:07
Yes, all the x86 asm since 1.9.5 has been written by me, including some from x264 that I wrote as well ;)

A bit has gone in the other direction too; x264's cacheline-split chroma MC was originally written for CoreAVC.

So we can forget about you RE'ing CoreAVC and improving libavcodec/ffmpeg-mt with that, right?

squid_80
19th December 2009, 14:14
I will give more results after you release the coreavc 2.0 to public between coreavc2.0 and the upcoming DiAVC beta.

How about you wait until your filter can output to multiple colorspaces, perform image levels adjustment, handle anamorphic widths properly and deinterlace. Also don't forget to provide the source files used for the benchmarks so they can be verified.

edison
19th December 2009, 14:18
there is a sync glitch when playing TrueHD clip with the 09-12-19 version MatroskaSplitter .

Cyber-Mav
19th December 2009, 14:41
From the results given by Dark Shikari, I believe the DiAVC alpha is still faster than the DiVX and coreavc2.0. I believe it is especially true when the DiAVC beta is out.

this is why im hesitant on making the purchase for coreavc 2.0
i hate to say it but its just too little way too late.

diavc beta will be the decider for me.

shikari could you rerun those speed test by locking affinity to 1 or 2 cores and see how performance racks up?

tal.aloni
19th December 2009, 14:47
there is a sync glitch when playing TrueHD clip with the 09-12-19 version MatroskaSplitter .

I can confirm that,
1. TrueHD have issues with both m2ts and mkv files, including the sample I posted @ page 256,
I can post more samples if needed.
Haali, is this part of your "core parser library"? I'm willing to help. (sent you an email)

2. I have a Blu-Ray demo disc that have two audio tracks: 5.1 DTS-HD and 2-channel LPCM,
I enjoyed 5.1 with the previous version,
but with the new version, I can only select the 2-channel LPCM audio track.

3. DD+ now works great. (starting with the new release).

Thanks,
Tal

JohnnyFu
19th December 2009, 15:18
Correct, we have an affiliate program in place. However the rate will be 15% for 2.0.

Also on price. Upgrades are $4.95 and the full version is $9.95 (discounted from $12.95 till after xmas).

Betaboy, is this the xmas edition you announced in 2008 ? Sorry, couldn't resist :D

Btw, I'd like to upgrade my lincense but I can't seem to find an upgrade link, huh?

http://corecodec.com/products/coreavc2

LoRd_MuldeR
19th December 2009, 15:27
It was already told that CoreAVC 1.x customers will receive an email with a unique upgrade code...

JohnnyFu
19th December 2009, 15:31
Aight', didn't read the forum lately, I just stumbled upon the Coreavc2 link while looking for a solution on how to play MKV's in Win7 MCE.

BetaBoy
19th December 2009, 15:56
Betaboy, is this the xmas edition you announced in 2008 ? Sorry, couldn't resist :D

Btw, I'd like to upgrade my lincense but I can't seem to find an upgrade link, huh?

http://corecodec.com/products/coreavc2

That link is not live yet.... it will be in about an hour as we send the emails out to current customers and take down CoreAVC.com to redirect them to CoreCodec.com

BetaBoy
19th December 2009, 15:58
There are some results about coreavc1.9.5, divx and DiAVC.

I will give more results after you release the coreavc 2.0 to public between coreavc2.0 and the upcoming DiAVC beta.

I'm really not gonna say it again seeing that you are insisting to taking your Off-Topic discussion to new levels. IE; Stop sh*tting on the release, before its even out, got it?

Take 'your' comparison out of this thread. Mods notified.

weasel_
19th December 2009, 16:04
Problem with cuda nad high number of ref frame is fixed ?

BetaBoy
19th December 2009, 16:10
In case you all did not notice..... Haali has posted the new splitter now that is included in CoreAVC 2.0: http://haali.su/mkv/

edison
19th December 2009, 16:25
In case you all did not notice..... Haali has posted the new splitter now that is included in CoreAVC 2.0: http://haali.su/mkv/

http://forum.doom9.org/showpost.php?p=1354688&postcount=5370 :)

sneaker_ger
19th December 2009, 16:37
One Question: Do you sell a full featured High Profile* Decoder this time? Or is it again possible that in 2 years x264 will use a feature previously not used and everyone has to buy a new licence?
(This of course excludes new features that are added to the specs after the purchase, as you don't need to promise on those.)

*normal High profile, not High10 or with 4:4:4 support

The 1.x was already sold as featuring "High profile (http://web.archive.org/web/20071222072516/www.coreavc.com/index.php?option=com_content&task=view&id=30&Itemid=1)".

lexor
19th December 2009, 16:56
@http://corecodec.com/products/coreavc2 I see this (emphasis mine):
Get Matroska MKV and MKA support from the ppl here at CoreCodec that helped create Matroska in the first place!

Was that intentional? Seems oddly out of place. Also I think it would be a good idea to expand MF into its full name (possibly with a link to details). I doubt many people who do not visit this forum would know what MF is and why bypassing it is important.

BetaBoy
19th December 2009, 17:07
Intentional yes, but only to my auto-correct.

Disabled
19th December 2009, 17:11
The 1.x was already sold as featuring "High profile (http://web.archive.org/web/20071222072516/www.coreavc.com/index.php?option=com_content&task=view&id=30&Itemid=1)".

Thats why I asked for full High Profile support, so they can't give me the finger once again. Now that DarkShikari teamed up with them, theyre gonna win every speed comparison and I have to buy the upgrade... ;-)

BetaBoy
19th December 2009, 17:15
The 1.x was already sold as featuring "High profile (http://web.archive.org/web/20071222072516/www.coreavc.com/index.php?option=com_content&task=view&id=30&Itemid=1)".
With features subject to change. We do plan on releasing 4:2:2 and 4:4:4 at some point... but as I stated, we have aligned our efforts with that of x264's capabilities.

sneaker_ger
19th December 2009, 17:27
With features subject to change. We do plan on releasing 4:2:2 and 4:4:4 at some point... but as I stated, we have aligned our efforts with that of x264's capabilities.

There's no mention of that on the website I linked. We all know that "smart" weighted p prediction was already supported by the standard standard at that time.

P.S.: Your website still charges 12.95 $ in case you haven't noticed yet. (Yes, I know your site isn't online yet)

Thats why I asked for full High Profile support, so they can't give me the finger once again. Now that DarkShikari teamed up with them, theyre gonna win every speed comparison and I have to buy the upgrade... ;-)

A decoder that isn't able to correctly decode all compliant High profile videos can't claim to feature High profile support IMHO. But of course I'd like to hear the answer to your question as well. Just wanted to make sure that when they claim "High profile" support it doesn't necessarily answer your question.

Dark Shikari
19th December 2009, 17:29
There's no mention of that on the website I linked. We all know that "smart" weighted p prediction was already supported by the standard standard at that time.

A decoder that isn't able to correctly decode all compliant High profile videos can't claim to support High profile IMHO. But of course I'd like to hear the answer to your question as well. Just wanted to make sure that when they claim "High profile" support it doesn't probably won't answer your question.libavcodec, DivX, Mainconcept, Elecard, and DiAVC all have known bugs or limitations in decoding some valid High Profile streams. It's not as if CoreAVC is the only one; hell, the only decoder I know that has no known issues at all is JM. And that's probably just because I don't use it enough to notice the bugs.

FYI, a list:

libavcodec: There is at least one weird conformance vector it fails on, relating to MMCO + long term refs.
DivX: incorrect deblocking with duplicate refs
Mainconcept: incorrect motion compensation with duplicate refs
Elecard: broken direct mode with duplicate refs, incorrect handling of delta quants > 25
DiAVC: quite a few (but it's alpha, so it never claimed to be perfect yet, of course)

Disabled
19th December 2009, 17:45
A decoder that isn't able to correctly decode all compliant High profile videos can't claim to feature High profile support IMHO.
I totally agree with you, but they obviously don't. But what can you do, try to sue them about it? Because of 9.99$?

libavcodec, DivX, Mainconcept, Elecard, and DiAVC all have known bugs or limitations in decoding some valid High Profile streams. It's not as if CoreAVC is the only one;
Youre probably right, but CoreAVC are the only ones who got money from me, so they are the only ones I can complain to. And they are the ones with the most obvious artifacts (ie unplayable).
I wouldn't care if the not supported features werent in use, but they are now, so now I do care. I just don't want to shell out money again to a company that promises things, they don't want to keep.

sneaker_ger
19th December 2009, 17:46
Don't get me wrong. I'm no angry customer complaining about not getting fixed weight p. But I don't like the marketing which tries to make it look like it's x264 fault that CoreAVC is buggy. On the recent CoreAVC page there was a URL listing its limitations (good thing!) - but it wasn't even linked (bad thing!). If I sell a decoder and claim that it features High profile support I have to be prepared to be taken up on that. (But to be realistically: CoreAVC is no 2000$ product - it only costs as much as few packs of cigarettes so customers should not be too demanding)
In this case I just wanted to make clear that giving a "marketing friendly" answer, like "CoreAVC 2.0 supports High profile" would not be sufficient, as they already had claimed that in the past.

Cyber-Mav
19th December 2009, 17:54
we have aligned our efforts with that of x264's capabilities.

so playing videos that are not encoded with x264 will result in a negative impact on speed?

this is very insteresting indeed.

sneaker_ger
19th December 2009, 18:00
so playing videos that are not encoded with x264 will result in a negative impact on speed?

this is very insteresting indeed.

We all know that impact on decoding speed depends on the settings and not the software used for encoding.

Dark Shikari
19th December 2009, 18:05
We all know that impact on decoding speed depends on the settings and not the software used for encoding.It's quite possible for a decoder to take speed shortcuts based on things it knows about the encoder. I don't know any that do, but it could be done.

sneaker_ger
19th December 2009, 18:08
It's quite possible for a decoder to take speed shortcuts based on things it knows about the encoder. I don't know any that do, but it could be done.

Thanks for the clarification. I didn't think in that direction at all. I love how you make at least 10 people a day look stupid. :(

BetaBoy
19th December 2009, 18:10
sneaker_ger... your points had been discussed 100+ pages back and we have taken the steps in notify what is and is not. In fact we have done this since the first day it was sold as well as changed when we were notified on issues.

BetaBoy
19th December 2009, 18:19
We are sending out emails now for users who purchased it within the past 60 days for the free update. Then come the update emails with the $4.95 upgrade discount for current customers.

BetaBoy
19th December 2009, 18:23
For the $9.95 launch use the coupon code: 20LAUNCH or just click this link: https://customers.corecodec.com/cart.php?a=add&pid=1&promocode=20launch

the_corona
19th December 2009, 18:29
That's great, if only we had a trial to see if it was worth it......or at least a few benchmarks by people with different proc architectures. I'm guessing nehalem owners arent exactly impressed so far (not that they should care much in the first place). What about Atoms/Athlon 64/Core Duo/Core 2 Duos?

I will not make a blind purchase.

BetaBoy
19th December 2009, 18:29
Since its finally live... I would like to thank the staff here for their hard work on the release, Haali for his simple solution for bypassing MediaFoundation in MKV's, Dark Shikari for his input, Clsid for his help with the installer... and even though you don't see it because its in our CoreAVC SDK... Thx to Mans for his kick ass work on ARM NEON Cortex A8 which makes CorePlayer rock on the iPhone 3GS.

Disabled
19th December 2009, 18:29
https://customers.corecodec.com/cart.php?a=add&pid=1&promocode=20launch
That link still states "- SMP (supports 4 CPU Cores)". Wasn't 2.0 to support more cores like DS made me think (http://forum.doom9.org/showthread.php?p=1354537#post1354537)?

BetaBoy
19th December 2009, 18:34
That's great, if only we had a trial to see if it was worth it......or at least a few benchmarks by people with different proc architectures. I'm guessing nehalem owners arent exactly impressed so far (not that they should care much in the first place). What about Atoms/Athlon 64/Core Duo/Core 2 Duos?

I will not make a blind purchase.

Based on feedback you will see across the board improvements in overall speed with all processors. Noting however that the Motion Vector and WeightP changes did slowdown performance about 1%... this is why we switched to ICC to get that 2% gain in overall speed.

BetaBoy
19th December 2009, 18:37
That link still states "- SMP (supports 4 CPU Cores)". Wasn't 2.0 to support more cores like DS made me think (http://forum.doom9.org/showthread.php?p=1354537#post1354537)?
We are still in the middle of changes.... that one is next.

Yoshiyuki Blade
19th December 2009, 18:38
Thanks. Just ordered and installed the client. Everything appears to be working fine.

BetaBoy
19th December 2009, 18:56
Correct, we have an affiliate program in place. However the rate will be 15% for 2.0.

The Affiliate Program is now live at a 15% commission rate.
https://customers.corecodec.com

We have added the ability to directly link to the purchase page to maximize conversion. If you log in to your affiliate panel, at the bottom, you will now see a "Affiliate links" section. This will include a link you can use to directly add the purchase to cart. You can activate your account by logging into your account and clicking the top 'Affiliates' tab and then agree to the notification.

If you have any questions, or any specific issues, feel free to contact me.

BetaBoy
19th December 2009, 18:56
We are still in the middle of changes.... that one is next.

Done.... thx for the report.

blubberbirne
19th December 2009, 19:04
When will the eMail send to people who buy coreavc for more than one year???

BetaBoy
19th December 2009, 19:08
When will the eMail send to people who buy coreavc for more than one year???
For all users, yes.

sneaker_ger
19th December 2009, 19:12
sneaker_ger... your points had been discussed 100+ pages back and we have taken the steps in notify what is and is not. In fact we have done this since the first day it was sold.

I have linked a past version of your website above listing CoreAVCs features - and there's no asterisk or any mention of any limitations on High profile support. The same goes for your start post in this very thread.

Then we have this post by you:
1.x is also fully spec compliant against the features that are included.... there are many features not supported in both CoreAVC and x264 that are being added,

Yet your feature limitation page only lists "FMO" and "multi-dupe WeightP". (+ the profiles not supported)

That leads me to the conclusion that you just lied to me. Don't do that because you're now making me angry. I don't know if you are willing to clarify and/or apologize but I suggest you to at least answer Disabled's question.

zn
19th December 2009, 19:46
In case you all did not notice..... Haali has posted the new splitter now that is included in CoreAVC 2.0: http://haali.su/mkv/

1.9.355.21 correct? file is dated as 14th November, but website tells us 19th December

clsid
19th December 2009, 20:11
14th November is the production date. 19th December the official public release date.

JohnnyFu
19th December 2009, 22:05
Betaboy, are you guys still pushing the mails out? I have no mail until yet.

hajj_3
19th December 2009, 22:55
is there going to be a trial released of 2.0? Wouldn't mind trying it on my sisters older laptop to see if hers can handle 720p 4000bitrate x264 .mkv. Nice to see this out, a nice christmas suprise:)

BetaBoy
19th December 2009, 23:08
Betaboy, are you guys still pushing the mails out? I have no mail until yet.

Yes, they are trickling out... and I suspect they will be for the next 6-7 hours.

BetaBoy
19th December 2009, 23:09
is there going to be a trial released of 2.0? Wouldn't mind trying it on my sisters older laptop to see if hers can handle 720p 4000bitrate x264 .mkv. Nice to see this out, a nice christmas suprise:)

No trial till January 7th.

hd1080
20th December 2009, 00:57
.... Thx to Mans for his kick ass work on ARM NEON Cortex A8 which makes CorePlayer rock on the iPhone 3GS.
means this i can play 720p/1080p movies with my iPhone?

blubberbirne
20th December 2009, 01:15
I'm still waiting for the email with the download link :(

JEEB
20th December 2009, 01:26
I'm still waiting for the email with the download link :(
You should be able to download the app from the customer area after your payment has been checked. No need to wait for an e-mail :)

BetaBoy
20th December 2009, 01:26
means this i can play 720p/1080p movies with my iPhone?
Its not 'true' GPU hardware acceleration, but it does help (especially for mobile)

JohnnyFu
20th December 2009, 01:34
I'm still waiting for the email with the download link :(

http://corecodec.com/products/coreavc/retrieve

404

BetaBoy
20th December 2009, 01:43
You should be able to download the app from the customer area after your payment has been checked. No need to wait for an e-mail :)

You can now for both 2.0 and past customers for 1.9.5. Its just the upgrade emails and free email links that are still going out. As they go out they create/change the account to reflect the upgrade/discount.

JEEB
20th December 2009, 01:46
You can now for both 2.0 and past customers for 1.9.5. Its just the upgrade emails and free email links that are going out. As they go out they create/change the account to reflect the change.
Oh yes, forgot completely about the 60-day offer and so on.

blubberbirne
20th December 2009, 01:59
You should be able to download the app from the customer area after your payment has been checked. No need to wait for an e-mail :)

i buy coreavc in mid 2008 (i think) and can't do nothink at the moment. i wait for email, nothing happens. this really .....

BetaBoy
20th December 2009, 02:01
http://corecodec.com/products/coreavc/retrieve

404

No longer is valid... the customer portal now handles all of that: https://customers.corecodec.com

JohnnyFu
20th December 2009, 02:16
No longer is valid... the customer portal now handles all of that: https://customers.corecodec.com

"No client account was found with the email address you entered"

What the frak?

ranpha
20th December 2009, 02:42
timecodec tests with a couple of files.

A Blu-ray rip snippet, 1080p, no reencode, straight out from the disc:-


Divx 7.2 (latest available at their site)
User: 7s, kernel: 0s, total: 7s, real: 29s, fps: 295.4, dfps: 74.7
User: 8s, kernel: 0s, total: 8s, real: 30s, fps: 267.7, dfps: 72.1
User: 8s, kernel: 0s, total: 9s, real: 30s, fps: 241.4, dfps: 70.5

ffdshow-mt build 3099
User: 11s, kernel: 0s, total: 11s, real: 37s, fps: 189.2, dfps: 58.4
User: 11s, kernel: 0s, total: 11s, real: 36s, fps: 191.5, dfps: 59.1
User: 11s, kernel: 0s, total: 11s, real: 36s, fps: 191.0, dfps: 59.8

MPC-HC (DXVA) build 1249
User: 99s, kernel: 0s, total: 99s, real: 99s, fps: 22.0, dfps: 21.9
User: 99s, kernel: 0s, total: 99s, real: 100s, fps: 22.0, dfps: 21.8
User: 98s, kernel: 0s, total: 98s, real: 99s, fps: 22.1, dfps: 22.0

MPC-HC (no DXVA) build 1249
User: 10s, kernel: 0s, total: 11s, real: 94s, fps: 196.4, dfps: 23.0
User: 10s, kernel: 0s, total: 11s, real: 96s, fps: 189.4, dfps: 22.6
User: 10s, kernel: 0s, total: 10s, real: 96s, fps: 200.0, dfps: 22.6

CoreAVC 1.9.5 (no CUDA)
User: 6s, kernel: 0s, total: 6s, real: 27s, fps: 342.3, dfps: 78.9
User: 5s, kernel: 0s, total: 5s, real: 26s, fps: 368.4, dfps: 81.2
User: 6s, kernel: 0s, total: 6s, real: 27s, fps: 353.5, dfps: 80.7

CoreAVC 1.9.5 (CUDA)
User: 9s, kernel: 1s, total: 11s, real: 30s, fps: 191.5, dfps: 70.6
User: 9s, kernel: 2s, total: 11s, real: 30s, fps: 188.9, dfps: 70.6
User: 9s, kernel: 2s, total: 11s, real: 30s, fps: 188.4, dfps: 70.6

CoreAVC 2 (no CUDA)
User: 6s, kernel: 0s, total: 6s, real: 26s, fps: 344.8, dfps: 81.7
User: 5s, kernel: 0s, total: 5s, real: 26s, fps: 370.4, dfps: 81.3
User: 6s, kernel: 0s, total: 6s, real: 26s, fps: 328.6, dfps: 83.8

CoreAVC 2 (CUDA)
User: 7s, kernel: 1s, total: 8s, real: 29s, fps: 264.6, dfps: 74.2
User: 7s, kernel: 0s, total: 7s, real: 29s, fps: 277.2, dfps: 74.1
User: 7s, kernel: 0s, total: 8s, real: 29s, fps: 269.7, dfps: 74.1

Microsoft DTV-DVD Video Decoder (no DXVA)
User: 0s, kernel: 0s, total: 0s, real: 37s, fps: 13999.9, dfps: 57.9
User: 0s, kernel: 0s, total: 0s, real: 36s, fps: 15555.5, dfps: 59.1
User: 0s, kernel: 0s, total: 0s, real: 36s, fps: 17499.9, dfps: 59.1

DiAVC (latest version from website)
User: 7s, kernel: 0s, total: 7s, real: 25s, fps: 275.0, dfps: 86.6
User: 6s, kernel: 0s, total: 7s, real: 25s, fps: 309.7, dfps: 87.3
User: 7s, kernel: 0s, total: 7s, real: 25s, fps: 281.7, dfps: 86.5



A Blu-ray snippet reencode, 1080p, very high bitrate, Blu-ray DXVA-compatible:-


Divx 7.2 (latest available at their site)
User: 2s, kernel: 0s, total: 3s, real: 16s, fps: 212.3, dfps: 40.6
User: 2s, kernel: 0s, total: 2s, real: 16s, fps: 231.8, dfps: 41.5
User: 2s, kernel: 0s, total: 2s, real: 16s, fps: 230.6, dfps: 40.5

ffdshow-mt build 3099
User: 3s, kernel: 0s, total: 3s, real: 17s, fps: 178.7, dfps: 38.2
User: 3s, kernel: 0s, total: 3s, real: 17s, fps: 168.2, dfps: 38.4
User: 3s, kernel: 0s, total: 3s, real: 17s, fps: 182.5, dfps: 38.3

MPC-HC (DXVA) build 1249
User: 62s, kernel: 0s, total: 62s, real: 62s, fps: 10.7, dfps: 10.7
User: 62s, kernel: 0s, total: 62s, real: 62s, fps: 10.7, dfps: 10.6
User: 62s, kernel: 0s, total: 62s, real: 63s, fps: 10.7, dfps: 10.6

MPC-HC (no DXVA) build 1249
User: 3s, kernel: 0s, total: 3s, real: 53s, fps: 174.3, dfps: 12.6
User: 3s, kernel: 0s, total: 3s, real: 53s, fps: 177.9, dfps: 12.5
User: 3s, kernel: 0s, total: 3s, real: 53s, fps: 172.9, dfps: 12.6

CoreAVC 1.9.5 (no CUDA)
User: 2s, kernel: 0s, total: 2s, real: 16s, fps: 250.8, dfps: 40.6
User: 2s, kernel: 0s, total: 2s, real: 16s, fps: 250.8, dfps: 40.6
User: 2s, kernel: 0s, total: 2s, real: 16s, fps: 230.6, dfps: 40.0

CoreAVC 1.9.5 (CUDA)
User: 4s, kernel: 0s, total: 5s, real: 22s, fps: 131.5, dfps: 29.4
User: 3s, kernel: 0s, total: 4s, real: 22s, fps: 147.9, dfps: 29.4
User: 3s, kernel: 0s, total: 4s, real: 22s, fps: 142.9, dfps: 29.4

CoreAVC 2 (no CUDA)
User: 2s, kernel: 0s, total: 2s, real: 17s, fps: 268.0, dfps: 39.2
User: 2s, kernel: 0s, total: 2s, real: 16s, fps: 225.7, dfps: 40.1
User: 2s, kernel: 0s, total: 2s, real: 17s, fps: 238.2, dfps: 38.8

CoreAVC 2 (CUDA)
User: 2s, kernel: 0s, total: 3s, real: 22s, fps: 197.6, dfps: 29.6
User: 3s, kernel: 0s, total: 3s, real: 22s, fps: 195.8, dfps: 29.6
User: 2s, kernel: 0s, total: 3s, real: 22s, fps: 214.4, dfps: 29.6

Microsoft DTV-DVD Video Decoder (no DXVA)
User: 0s, kernel: 0s, total: 0s, real: 20s, fps: 10721.1, dfps: 32.8
User: 0s, kernel: 0s, total: 0s, real: 21s, fps: 7147.4, dfps: 31.7
User: 0s, kernel: 0s, total: 0s, real: 20s, fps: 5360.5, dfps: 33.3

DiAVC (latest version from website)
User: 2s, kernel: 0s, total: 2s, real: 14s, fps: 236.9, dfps: 46.9
User: 2s, kernel: 0s, total: 2s, real: 14s, fps: 268.0, dfps: 46.7
User: 2s, kernel: 0s, total: 2s, real: 14s, fps: 261.5, dfps: 46.7


System: Windows 7 x64, Phenom X4 9500, 4GB RAM, nVidia GT220 1GB with 191.07 drivers.

Mixer73
20th December 2009, 02:52
- First on Windows 7..... Both CoreAVC and Haali's splitter now use a custom fourcc so that when you play an MKV video it uses Directshow instead of MediaFoundation in WMP and MC. No hacks and its a simple solution to gets ppl out of the pitfalls of MF. Note that this is ONLY for .mkv and does not affect any other container.

Awesome release, thanks for all the hard work.

Can I please clarify the above. I'm using Windows 7MCE and I archive all my content in PS3/DXVA compatible h.264, MP4 container, does the above mean I won't be able to use CoreAVC for my system at all? Or does it just mean I will have to do trickery to get the merit sorted?

Apologies for the newb question, as yet I have not changed anything on the 7MCE system.

BetaBoy
20th December 2009, 03:37
IF when you want... you can simply rename the .mp4 videos to .mkv and it will work with CoreAVC/Haali.

JohnnyFu
20th December 2009, 03:42
Betaboy, I have not received an email link and I can't logon to your customer system to access my download/serial.

"No client account was found with the email address you entered"

blubberbirne
20th December 2009, 03:55
"No client account was found with the email address you entered"

What the frak?

same here :devil:

BetaBoy
20th December 2009, 04:46
Emails are still going out.... this is why I mentioned earlier that your account is modified/activated when the email is sent. But we think we can go even one better by adding an 'addon' to each of those accounts. We are looking into it now.

BetaBoy
20th December 2009, 04:58
All.... I know I have a bunch of PM's in my account... I'll get to them once we get through tonight/tomorrow AM. Thanx for being patient.

lych_necross
20th December 2009, 08:13
Thanks BetaBoy and the rest of CoreCodec! I just bought my copy and it's working fine on my system. Ignore the haters and keep up the good work. :)

puffpio
20th December 2009, 09:06
i just bought a copy and installed it woohoo...
notices some slowdowns playing 1080p...and i think it may be my system is not fast enough?
mobile core2duo @ 2ghz
mpc-hc 64bit build 1424, haali media splitter, coreavc 2, haali renderer

thanks for any info

Sharc
20th December 2009, 09:44
With version 2.0 the former --weightp issues (blocks) of 1.9.5 are solved, according to my tests.

Mixer73
20th December 2009, 10:04
IF when you want... you can simply rename the .mp4 videos to .mkv and it will work with CoreAVC/Haali.

With perhaps thousands of files this isn't practical, is there a way I could do the same through the registry?

DigitalDeviant
20th December 2009, 10:04
I'm still waiting for my upgrade link.

hippoth
20th December 2009, 11:57
Same here. I have bought CoreAVC april 2006 and did not get any mail and I canīt login --> "No client account was found with the email address you entered"

blubberbirne
20th December 2009, 13:48
This is sad. People who buy CoreAVC at the beginning are still waiting. New buyer are still happy

the_corona
20th December 2009, 14:00
Hmm, so seems DiAVC > CoreAVC > Divx (and sometimes DivX/CoreAVC are reversed or very close).

Whats interesting is that DiAVC supposedly has alot more optimizations left which will come in the beta.

Can someone test on a dual core machine (or simply only give timecodec 2 cores)? My HTPC is a Yonah so personally such results would help me the most (hint CoreAVC just release the trial and people can test themselves....I don't get it)

clsid
20th December 2009, 14:37
With perhaps thousands of files this isn't practical, is there a way I could do the same through the registry?Next version of my Win7DSFilterTweaker tool should be able to assist you with bypassing the evil MS stuff ;)

BetaBoy
20th December 2009, 15:12
Hmm, so seems DiAVC > CoreAVC > Divx (and sometimes DivX/CoreAVC are reversed or very close).

Whats interesting is that DiAVC supposedly has alot more optimizations left which will come in the beta.

Please keep it on-topic... but to your point... remember this is our 'first' optimizations release. While DiAVC has been doing opts for months now... you can only tweak so much till as DS says you code to match a 'specific' encoder.

BetaBoy
20th December 2009, 15:16
Next version of my Win7DSFilterTweaker tool should be able to assist you with bypassing the evil MS stuff ;)

Yeah... we have been thinking about an option to treat other containers as .mkv but I think its out of the scope of what 'we' should do.

the_corona
20th December 2009, 15:28
Please keep it on-topic... but to your point... remember this is our 'first' optimizations release. While DiAVC has been doing opts for months now... you can only tweak so much till as DS says you code to match a 'specific' encoder.

Is simply everything thats not praising CoreAVC "off-topic"?

Seriously, how is it not on topic to talk about CoreAVC performance in the CoreAVC thread. I mean come on...

We only have very few benchmark results, and these are my impressions. Are they wrong?

<stupid rant>Yeah yeah, I'm sure your next "optimization release" will blow us all away....in 2014 ..... after buying yet another license.</stupid rant>

Cyber-Mav
20th December 2009, 16:53
DiAVC has been doing opts for months now...

if your not happy with that then maybe you can get a court injuction placed on diavc to stop them from doing these optimisations if its hurting your business?

seriously betaboy i was expecting you to be more mature about this. i can see now why dark shikari left out the diavc tests in his performance benchmarks since it would have a negative impact on coreavc sales, not surprising since dark shikari and corecodec have "teamed up" so to speak.

also its pointless saying that this is your first release with speed optimisations, by the time the second release is done everyone will be using quad core cpu;s even on laptops.

all this talk is on topic since it is about coreavc and its performance. no one is crapping on the release here, they are just voicing thier opinions and first hand experience of the new codec you have just released.

for me the key reason i purchased coreavc 1.9.0 is because of the CUDA support and that was money well spent in my eyes.

Dark Shikari
20th December 2009, 17:38
seriously betaboy i was expecting you to be more mature about this. i can see now why dark shikari left out the diavc tests in his performance benchmarks since it would have a negative impact on coreavc sales, not surprising since dark shikari and corecodec have "teamed up" so to speak.How am I supposed to bench a codec that doesn't work on my computer? I use Windows 7 64-bit and DiAVC doesn't support it yet. :rolleyes:

I explicitly went out of my way to try to include DiAVC but it didn't work, so obviously I couldn't include it in the results.

honai
20th December 2009, 17:45
I hope we can all agree that competition is a good thing. Without the pressure from past and prospective customers I don't think CoreAVC 2.0 would include that much bug fixes and improvements as it does know.

Having said that, without a freely downloadable trial I'm going to refrain from purchasing upgrades for my 5 licenses for the time being.

BetaBoy
20th December 2009, 18:25
I hope we can all agree that competition is a good thing. Without the pressure from past and prospective customers I don't think CoreAVC 2.0 would include that much bug fixes and improvements as it does know.
Without a doubt competition is great... always will be. There is tons of space in H.264 space and it's just a small part of what we are doing here overall, and we are very grateful for our customers and commercial partners.

LoRd_MuldeR
20th December 2009, 18:43
How am I supposed to bench a codec that doesn't work on my computer? I use Windows 7 64-bit and DiAVC doesn't support it yet. :rolleyes:

Does work here. Windows 7, 64-Bit, latest DiAVC Alpha. Also works with TimeCodec :confused:

Here are my results:
http://forum.doom9.org/showpost.php?p=1347162&postcount=179

(CoreAVC 2.0 is missing in that comparison, as I don't have access to the 2.0 release of CoreAVC)

lnatan25
20th December 2009, 19:04
Just to say, DiAVC works here too, on 64-bit Windows 7.

LoRd_MuldeR
20th December 2009, 19:09
Just to say, DiAVC works here too, on 64-bit Windows 7.

As older Alpha versions did not work and the developer has updated the Alpha several times without renaming the download, Dark Shikari may be using an old/broken version.

(It also may help to unregister and re-register the filter manually from an Admin console. That did the trick for me at least...)

the_corona
20th December 2009, 19:14
DiAVC works for me on Win7 x64 too with MPC-HC 32bit.

Though I have to say the output looks different from other decoders, I'll have to investigate this when I find some time.

lnatan25
20th December 2009, 19:17
As older Alpha versions did not work and the developer has updated the Alpha several times without renaming the download, Dark Shikari may be using an old/broken version.

(It also may help to unregister and re-register the filter manually from an Admin console. That did the trick for me at least...)
I just added it as an external codec in KMP. Works very good. It's an alpha, so some bugs here and there, but I was surprised how well it works for such an early release.

To go "off-topic" for a moment (:rolleyes:), I think everyone should wait and see how DiAVC progresses. Beta should be out soon. No point in purchasing one product and then having to purchase a better one later. Just wait a little longer to see where things are headed.

LoRd_MuldeR
20th December 2009, 19:17
DiAVC works for me on Win7 x64 too with MPC-HC 32bit.

Though I have to say the output looks different from other decoders, I'll have to investigate this when I find some time.

You are 100% sure that you aren't misinterpreting different Luma Levels as "different output" ???

the_corona
20th December 2009, 19:52
You are 100% sure that you aren't misinterpreting different Luma Levels as "different output" ???

No, I think actually 32bit MPC had different renderer settings than 64bit which I was used to seeing things as. I didn't really bother, just a quick install to see if it runs on Win7 x64. Disregard that remark for now.

BetaBoy
20th December 2009, 20:00
To go "off-topic" for a moment (:rolleyes:), .
No.. what I suggest is simple, take your OT discussion out of this thread.

wayland
20th December 2009, 20:26
maybe i misunderstand this "Current customer from the past 60 days are getting updates emailed to them now. Post 60 days users get the upgrade for $4.95 and those emails are going out now as well." but longterm customers have to pay $15 more for 2.0? (was originally $20 iirc not $10 plus $5 upgrade fee)

LoRd_MuldeR
20th December 2009, 20:31
maybe i misunderstand this "Current customer from the past 60 days are getting updates emailed to them now. Post 60 days users get the upgrade for $4.95 and those emails are going out now as well." but longterm customers have to pay $15 more for 2.0? (was originally $20 iirc not $10 plus $5 upgrade fee)

If you bought CoreAVC in the last 60 days, you get the 2.0 upgrade for free. If you bought it earlier, you can upgrade for $5. New users will have to pay $10 (after xmas it will be $13).

Please correct me, if I'm wrong...

laserfan
20th December 2009, 20:42
If you bought CoreAVC in the last 60 days, you get the 2.0 upgrade for free. If you bought it earlier, you can upgrade for $5. New users will have to pay $10 (after xmas it will be $13).

Please correct me, if I'm wrong...Where'd the $5 upgrade come from? I have 1.9.5 and just paid $9.95 to upgrade... :confused:

wayland
20th December 2009, 20:48
i think they are supposed to email you a code that gives a discount, i havnt had any email about it yet though. i dont really see a good reason for paying the upgrade fee, might as well just use a decoder without mv and weightp problems and save $5

LoRd_MuldeR
20th December 2009, 20:50
Where'd the $5 upgrade come from? I have 1.9.5 and just paid $9.95 to upgrade... :confused:

From this post:

Correct, we have an affiliate program in place. However the rate will be 15% for 2.0.

Also on price. Upgrades are $4.95 and the full version is $9.95 (discounted from $12.95 till xmas).

I think as a CoreAVC 1.9.x owner, you should have received an email with a unique code for reduced-price upgrade...

laserfan
20th December 2009, 21:27
I think as a CoreAVC 1.9.x owner, you should have received an email with a unique code for reduced-price upgrade...
Maybe I "jumped the gun" then. I did receive an email Subject: CoreAVC 1.9.5 Purchase Update
Dear <my registered name>,

Thank you for your purchasing CoreCodec products. We have recently upgraded our order processing system.

An account has automatically been generated for you in our new system.
To log on, use your email address (myregisteredemail), and your temporary
password of "secretpassword". This password is stored encrypted and can be reset, but not recovered, if lost.

In addition, we have imported your existing CoreAVC 1.9.5 purchases into this account.

The following CoreAVC 1.9.5 serial numbers have been associated with your account:
- XXXXXXXXXXXX -

You can log in to view your serial numbers and downloads at any time by visiting https://customers.corecodec.com/ .
So I went to the link, logged-in to my account, didn't see any "upgrade" link anywhere but clicked on "order" and completed the order w/20LAUNCH to get to $9.95.

Got this email 2 days ago and just bought the new version this a.m. (no further emails received).

So has anyone here gotten theirs for $5bucks? How? :confused:

JEEB
20th December 2009, 21:44
Maybe I "jumped the gun" then. I did receive an email Subject: CoreAVC 1.9.5 Purchase Update

So I went to the link, logged-in to my account, didn't see any "upgrade" link anywhere but clicked on "order" and completed the order w/20LAUNCH to get to $9.95.

Got this email 2 days ago and just bought the new version this a.m. (no further emails received).

So has anyone here gotten theirs for $5bucks? How? :confused:

I had the same thing, only mail I got was that purchase update one. Oh well...

BetaBoy
20th December 2009, 21:58
We are looking into the reports of users not getting the upgrade emails.

Sharc
20th December 2009, 22:12
Maybe I "jumped the gun" then. I did receive an email Subject: CoreAVC 1.9.5 Purchase Update

So I went to the link, logged-in to my account, didn't see any "upgrade" link anywhere but clicked on "order" and completed the order w/20LAUNCH to get to $9.95.

Got this email 2 days ago and just bought the new version ..
Same here....

blubberbirne
20th December 2009, 22:17
We are looking into the reports of users not getting the upgrade emails.

:thanks: i hope i get my email this year :mad:

DigitalDeviant
20th December 2009, 22:29
No upgrade link here either.

BetaBoy
20th December 2009, 22:42
We are looking into the reports of users not getting the upgrade emails.

For any people that did not get the discount code you can email me directly: info@corecodec.com and I will issue you a unique coupon. Please provide your original purchase email address as this will need to match your current CoreAVC order.

ranpha
20th December 2009, 22:50
For any people that did not get the discount code you can email me directly: info@corecodec.com and I will issue you a unique coupon. Please provide your original purchase email address as this will need to match your current CoreAVC order.

I haven't got the upgrade e-mail either, so I have sent you an e-mail for the coupon so that I can grab myself a second copy.

BetaBoy
21st December 2009, 00:18
Coupons sent to all those you sent in their purchase info to me so far.

RiCON
21st December 2009, 00:32
So I tried to install CoreAVC through Wine, to use it on Mplayer, like I used with 1.9.5. But something changed that broke that possibility.

As the development of coreavc-for-linux seems to be kinda dead, is there any way CoreCodec could explain to get it to work again?

It's essential that I get it to work as ffmpeg-mt is not fast enough to run on the netbook for which I especifically bought CoreAVC.

Everything from coreavc-for-linux compiled just fine. It's just a problem with the .ax not detecting the serial registered through registercodec. I tried registering the new serial and even the user, but it doesn't seem to work.

Any ideas?

Mixer73
21st December 2009, 00:45
Next version of my Win7DSFilterTweaker tool should be able to assist you with bypassing the evil MS stuff ;)

Awesome CLSID. With some time off work soon I feel I might be able to do "risky" stuff with my 7 box ;)

Btw I gotta say you're involved in all my favourite projects, :thanks:

Yeah... we have been thinking about an option to treat other containers as .mkv but I think its out of the scope of what 'we' should do.

I can understand this 100%, there's only so much you should be doing as part of a commercial installer.

Coupons sent to all those you sent in their purchase info to me so far.

Well done - whether its $10 or $5 CoreAVC is a great product and very affordable. Thanks for all your hard work.

lnatan25
21st December 2009, 00:48
No.. what I suggest is simple, take your OT discussion out of this thread.
I guess you didn't catch the sarcasm. :rolleyes:

Or you tuck your head in the sand on purpose?

yesgrey
21st December 2009, 01:34
Here are my test results using an Intel E2160@2.7GHz with 1080p material:
ffdshow
User: 0s, kernel: 0s, total: 0s, real: 4s, fps: 479.6, dfps: 41.2
User: 0s, kernel: 0s, total: 0s, real: 3s, fps: 479.6, dfps: 54.2
User: 0s, kernel: 0s, total: 0s, real: 3s, fps: 647.4, dfps: 52.4

ffdshow-mt
User: 0s, kernel: 0s, total: 0s, real: 3s, fps: 719.4, dfps: 66.8
User: 0s, kernel: 0s, total: 0s, real: 2s, fps: 563.0, dfps: 67.4
User: 0s, kernel: 0s, total: 0s, real: 3s, fps: 588.6, dfps: 64.4

CoreAVC 1.9.5
User: 0s, kernel: 0s, total: 0s, real: 2s, fps: 719.4, dfps: 71.2
User: 0s, kernel: 0s, total: 0s, real: 2s, fps: 588.6, dfps: 73.2
User: 0s, kernel: 0s, total: 0s, real: 2s, fps: 616.6, dfps: 72.7

DiAVC alpha
User: 0s, kernel: 0s, total: 0s, real: 2s, fps: 588.6, dfps: 81.9
User: 0s, kernel: 0s, total: 0s, real: 2s, fps: 681.5, dfps: 80.9
User: 0s, kernel: 0s, total: 0s, real: 2s, fps: 479.6, dfps: 82.5

CoreAVC 2.0.0
User: 0s, kernel: 0s, total: 0s, real: 2s, fps: 498.0, dfps: 81.4
User: 0s, kernel: 0s, total: 0s, real: 2s, fps: 404.6, dfps: 81.9
User: 0s, kernel: 0s, total: 0s, real: 2s, fps: 462.5, dfps: 80.9

BetaBoy
21st December 2009, 04:11
We keep seeing Timecodec reports again and again on Doom9 with multi-core systems. Everyone needs to understand that timecodec does not produce accurate results for multi-core systems (single core however does produce accurate results), and as such any reports are not correct. Both Haali and I both have stated this a few times here. I have pinged Haali to see if he can mod it to accurately represent true 'dfps', although I suspect that would require a lot of work to make it accurate against all the AMD and Intel CPU's.

Keiyakusha
21st December 2009, 04:50
But even if result is not accurate, it should be not accurate in the same way for all decoders. So we still should be able to say "this one is faster than that one".
Anyway maybe yesgrey3 or someone can do another comparison using graphstudio?

BetaBoy
21st December 2009, 05:01
But even if result is not accurate, it should be not accurate in the same way for all decoders. So we still should be able to say "this one is faster than that one".
Anyway maybe yesgrey3 or someone can do another comparison using graphstudio?
Not at all, as each decoder will likely handle threading priorities differently and throw off the 'dfps' count.

laserfan
21st December 2009, 05:28
Maybe I "jumped the gun" then. I did receive an email Subject: CoreAVC 1.9.5 Purchase Update

So I went to the link, logged-in to my account, didn't see any "upgrade" link anywhere but clicked on "order" and completed the order w/20LAUNCH to get to $9.95.

Got this email 2 days ago (Dec 18th)and just bought the new version this a.m. (no further emails received).
Subject: CoreAVC 1.9.5 Purchase Update
Date: Fri, 18 Dec 2009 17:06:11 -0800 (PST)
From: CoreCodec Inc. <noreply@corecodec.com>
To: <my email>
Dear <my registered name>,

Thank you for your purchasing CoreCodec products. We have recently upgraded our order processing system.

An account has automatically been generated for you in our new system.
To log on, use your email address (myregisteredemail), and your temporary
password of "secretpassword". This password is stored encrypted and can be reset, but not recovered, if lost.

In addition, we have imported your existing CoreAVC 1.9.5 purchases into this account.

The following CoreAVC 1.9.5 serial numbers have been associated with your account:
- XXXXXXXXXXXX -

You can log in to view your serial numbers and downloads at any time by visiting https://customers.corecodec.com/ .

So has anyone here gotten theirs for $5bucks? How? :confused:
What bonehead(s) in your company are responsible for this: after getting the above two days ago, I tonight received the following:
Subject: Upgrade to CoreAVC 2.0 for $4.95!
Date: Sun, 20 Dec 2009 18:56:26 -0800
From: CoreCodec, Inc. <support@corecodec.com>
To: <yours truly, removed to protect my innocence>

CoreCodec, Inc.

Dear <my name>,

You are receiving this email because you have purchased CoreAVC 1.X.

We are pleased to announce the immediate availablity of a discounted upgrade to CoreAVC 2.0.

To purchase, simply log into your client portal and access your licenses at

https://customers.corecodec.com/clientarea.php?action=products

then click the license you wish to upgrade, and click upgrade. This will generate an invoice for the upgrade.

Upon payment, your license will be automatically upgraded to CoreAVC 2.0.

Thank you.

A little late CoreCodec, Inc.! I know it's only $5 bucks, but what a way to make your customers feel like jerks for responding to your first "come on"! How'd you manage to mess this up??? :devil:

BetaBoy
21st December 2009, 06:13
Not boneheads at all, but we apologize for your inconvenience. We sent another round of emails for those that had already been sent, but bounced back to us. If you already paid for it send me an email: info@corecodec.com and we will make up the difference.

G_M_C
21st December 2009, 07:10
I've got one question; How is the 32-64 bit / new windows migration arranged with you guys ?

I'm on 32-bits Win XP right now, but plan to upgrade to Win 7 in Q1 2010. When I buy CoreAVC now, wich would be Win XP32, will my license give me the abillity to also migrate CoreAVC to Win7/64 bits when i upgrade ?

tormento
21st December 2009, 08:26
I have just bought coreavc2 and I have 2 questions:

1) Why is not possible to install it in a custom directory?

2) How (on hell) is possible to change codec priority? I have tried every single codec modding program out there and CoreAVC does take the decoding process.

Oh, yeah, on my PC even tagging the preferences does not show the tray icon..

BetaBoy
21st December 2009, 08:38
1 - will be an option in a future release.
2 - Uncheck the preference for 'codec' priority in the decoder properties page.

On the tray... what OS are you running?

dimitrik
21st December 2009, 11:32
I just got my upgrade email - can't wait to install and test the new decoder. It helps me run MC7 x64 on my HTPC, I will be one happy guy.

By the way for those interested in speed comparisons, there is a major thread on this, which is the ideal place for those questions.

MPEG-4 AVC/H.264 decoder comparison (http://forum.doom9.org/showthread.php?t=99402)

I think its more relevant to discuss the specific features of this release here, good or bad, which will ultimately help both old and new customers.

JohnnyFu
21st December 2009, 11:49
I can confirm MC7 recognizes MKV files without any modification with CoreAVC 2.0 installed. Thanks!

Looking forward for ATi GPU support!

tormento
21st December 2009, 12:21
2 - Uncheck the preference for 'codec' priority in the decoder properties page.

On the tray... what OS are you running?

2) Tried, not working.

Windows ult 7 x64.

Disabled
21st December 2009, 13:41
We keep seeing Timecodec reports again and again on Doom9 with multi-core systems. Everyone needs to understand that timecodec does not produce accurate results for multi-core systems...

I wonder, how do you benchmark your decoder? Does the graphstudio benchmark provide correct results? I tested all decoders (CoreAVC 1.9.5) with graphstudio and timecodec and while the absolute numbers varied a little, the results were absolutely comparable...

LoRd_MuldeR
21st December 2009, 13:44
Side-note: In all my tests the "dps" value seems to correlate perfectly with the "real" decoding time and also with the subjective decoding experience.

And I assume TimeCodec does measure the (real) decoding time correctly. That really shouldn't be too hard to do ;)

tommy_vercetti
21st December 2009, 14:48
I can confirm MC7 recognizes MKV files without any modification with CoreAVC 2.0 installed. Thanks!

Looking forward for ATi GPU support!

You can try with just the Haali Splitter, it works the same, just it gets rendered by the default codecs in Windows 7 x64

BetaBoy
21st December 2009, 15:05
I wonder, how do you benchmark your decoder? Does the graphstudio benchmark provide correct results? I tested all decoders (CoreAVC 1.9.5) with graphstudio and timecodec and while the absolute numbers varied a little, the results were absolutely comparable...
Timecodec is good for benchmarking devel diff's of before Vs. after to see if it has a positive/negative impact on performance. Also from what I was told graphstudio also has the same threading problem, which would explain why they numbers are close for both.

I'm trying to chat with Haali more on it to get more details on what is and is not, and what we can do to get more accurate benchmarks.

BetaBoy
21st December 2009, 15:07
2) Tried, not working.

Windows ult 7 x64.

This seems to be a bug with Shark's Codec tool based on the reports coming in. Removing it seems to resolve the issue till he releases a fix.

Shark007
21st December 2009, 16:18
This seems to be a bug with Shark's Codec tool based on the reports coming in. Removing it seems to resolve the issue till he releases a fix.

Since i have not had a chance to test 2.0 (until now) my x64 components use x64 ffdshow as an H264 default. You will need to open the x64 settings application on the Config TAB, use the center dropdown to choose 'FFDshow Video' and set H264/AVC to disabled. I will add provisions to my x64 release to accomodate x64 CoreAVC a.s.a.p.

EDIT: version 2.2.8 of the x64 Components (http://www.majorgeeks.com/VistaCodecs_x64Components_d5535.html) accommodates the use of CoreAVC x64.

BetaBoy
21st December 2009, 20:08
Shark007... thank you!

Cyber-Mav
21st December 2009, 23:10
ok i bit the bullet and upgraded to coreavc 2.0 here are my results:

q9650@3.6ghz vista ultimate x64


720x304 1mbit

coreavc 1.9.5
User: 1s, kernel: 0s, total: 1s, real: 4s, fps: 4918.2, dfps: 1073.8

coreavc 2.0
User: 0s, kernel: 0s, total: 0s, real: 3s, fps: 6170.1, dfps: 1396.5
User: 0s, kernel: 0s, total: 0s, real: 3s, fps: 6284.4, dfps: 1396.5
User: 0s, kernel: 0s, total: 0s, real: 3s, fps: 5751.8, dfps: 1408.0

divx
User: 0s, kernel: 0s, total: 0s, real: 3s, fps: 7892.0, dfps: 1705.0
User: 0s, kernel: 0s, total: 0s, real: 3s, fps: 7069.9, dfps: 1705.0
User: 0s, kernel: 0s, total: 0s, real: 3s, fps: 8277.0, dfps: 1696.8

diavc alpha
User: 1s, kernel: 0s, total: 1s, real: 2s, fps: 3393.6, dfps: 1814.9
User: 1s, kernel: 0s, total: 1s, real: 2s, fps: 3201.5, dfps: 1814.9
User: 1s, kernel: 0s, total: 1s, real: 2s, fps: 3232.0, dfps: 1805.0
---------------------------------------------------------------------

720p 2.5mbit

coreavc 1.9.5
User: 2s, kernel: 0s, total: 2s, real: 11s, fps: 2177.1, dfps: 542.8

coreavc 2.0
User: 3s, kernel: 0s, total: 3s, real: 10s, fps: 1878.9, dfps: 606.9
User: 3s, kernel: 0s, total: 3s, real: 10s, fps: 1820.7, dfps: 605.1
User: 3s, kernel: 0s, total: 3s, real: 10s, fps: 2088.7, dfps: 604.2

divx
User: 4s, kernel: 0s, total: 4s, real: 11s, fps: 1480.1, dfps: 579.5
User: 4s, kernel: 0s, total: 4s, real: 11s, fps: 1480.1, dfps: 581.2
User: 4s, kernel: 0s, total: 4s, real: 11s, fps: 1418.9, dfps: 578.7

diavc alpha
User: 5s, kernel: 0s, total: 5s, real: 10s, fps: 1139.8, dfps: 604.2
User: 5s, kernel: 0s, total: 5s, real: 10s, fps: 1213.8, dfps: 606.9
User: 4s, kernel: 0s, total: 5s, real: 10s, fps: 1250.7, dfps: 606.0
---------------------------------------------------------------------

720p 5.7mbit

coreavc 1.9.5
User: 3s, kernel: 0s, total: 3s, real: 14s, fps: 1861.7, dfps: 473.0

coreavc 2.0
User: 3s, kernel: 0s, total: 3s, real: 14s, fps: 1815.3, dfps: 485.6
User: 4s, kernel: 0s, total: 4s, real: 13s, fps: 1602.6, dfps: 487.7
User: 3s, kernel: 0s, total: 3s, real: 14s, fps: 1822.9, dfps: 485.6

divx
User: 4s, kernel: 0s, total: 4s, real: 14s, fps: 1439.1, dfps: 480.8
User: 4s, kernel: 0s, total: 4s, real: 14s, fps: 1406.7, dfps: 481.3
User: 4s, kernel: 0s, total: 4s, real: 14s, fps: 1371.5, dfps: 481.8

diavc alpha
User: 4s, kernel: 0s, total: 5s, real: 13s, fps: 1271.8, dfps: 504.0
User: 4s, kernel: 0s, total: 5s, real: 13s, fps: 1317.8, dfps: 505.8
User: 4s, kernel: 0s, total: 5s, real: 13s, fps: 1337.9, dfps: 502.9
---------------------------------------------------------------------

1080p 4.5mbit

coreavc 1.9.5
User: 10s, kernel: 0s, total: 11s, real: 27s, fps: 741.5, dfps: 297.1

coreavc 2.0
User: 11s, kernel: 0s, total: 11s, real: 26s, fps: 685.2, dfps: 303.1
User: 11s, kernel: 0s, total: 11s, real: 26s, fps: 699.9, dfps: 303.1
User: 12s, kernel: 0s, total: 12s, real: 26s, fps: 631.5, dfps: 303.1

divx
User: 13s, kernel: 0s, total: 13s, real: 28s, fps: 588.2, dfps: 288.4
User: 14s, kernel: 0s, total: 14s, real: 28s, fps: 565.4, dfps: 287.7
User: 14s, kernel: 0s, total: 14s, real: 28s, fps: 572.2, dfps: 288.9

diavc alpha
User: 12s, kernel: 0s, total: 13s, real: 26s, fps: 605.9, dfps: 302.8
User: 14s, kernel: 0s, total: 15s, real: 27s, fps: 535.8, dfps: 301.7
User: 13s, kernel: 0s, total: 14s, real: 27s, fps: 574.7, dfps: 301.4

new coreavc 2.0 is deffo faster than 1.9.5 was. however on my laptop the pentium m740, coreavc 2.0 is around 1fps SLOWER than 1.9.5 is in all the timecodec benches i ran.

however im currently running benches on my opteron cpu and the results so far are very very interesting indeed. i will post results in few mins when testing is done.

//also how do i get my results embedded in a code box like others do?

yesgrey
21st December 2009, 23:35
//also how do i get my results embedded in a code box like others do?
select the results and click the "#" icon when creating/editing a message.

Cyber-Mav
21st December 2009, 23:37
select the results and click the "#" icon when creating/editing a message.

thankyou my good man.

the_corona
21st December 2009, 23:40
Well, it sure seems all those optimizations only made it faster than 1.9.5, but not really faster than the competition.

BetaBoy, if timecodec is as you claim not accurate at all and shouldn't be used to judge a codec's performance, then how can you claim CoreAVC to be the "world's fastest H264 software decoder"? How do you measure that? Or don't you and just put that up there for good marketing? Or was timecodec "good enough" while it showed CoreAVC to be the fastest in the past?

In either case, a bit questionable buiseness practice if I may say so.

Cyber-Mav
21st December 2009, 23:43
Well, it sure seems all those optimizations only made it faster than 1.9.5, but not really faster than the competition.

BetaBoy, if timecodec is as you claim not accurate at all and shouldn't be used to judge a codec's performance, then how can you claim CoreAVC to be the "world's fastest H264 software decoder"? How do you measure that? Or don't you and just put that up there for good marketing? Or was timecodec "good enough" while it showed CoreAVC to be the fastest in the past?

In either case, a bit questionable buiseness practice if I may say so.

wait till i get my opteron pc results up. its very interesting indeed.

BetaBoy
21st December 2009, 23:51
AMD opteron users have seen the great decrease in overall speed with CoreAVC 2.0 from the reports so far.

the_corona... take your continued attacks/flames elsewhere, its really getting old.

JohnnyFu
21st December 2009, 23:52
Opteron or AMD CPU's in general?

Cyber-Mav
22nd December 2009, 00:00
opteron is just more cache memory than regular athlon64, so i would assume all athlon 64 users would get similar speed to opteron. i almost finished benching the codecs on my opteron pc, just a few more runs to go then i install coreavc 2.0 and bench that and i will post up the results. but as of now coreavc 1.9.5 is showing amazing results for me.

BetaBoy
22nd December 2009, 00:28
Opteron or AMD CPU's in general?

typo.. decrease. But we still have a bunch of ppl testing and are getting mixed results. So I'll hold off on more reports till later this week.

Dark Shikari
22nd December 2009, 00:35
I'm guessing that a good bit of the new SSE2 assembly code is slower on A64 and should probably be disabled there.