CoreCodec/H.264 Codec "CoreAVC"

[Roco]

I have to wait for an alternative credit-card payment service too. I had more than one bad experiences from paypal in the past, so I am negatively biased to the point that I have them in my black list and will never use their service again. I hope an alternative will be available soon.

[popper]

Hey, you must do something with your website "http://coreavc.corecodec.org/"! It is too much LIGHT!!! It doesn't give the confidence that people want to take!!! If you want to sell more give it a better style!!!:sly:

Yes, it might be a good and fast decoder and yes you did a very good advertisment in the video forums, but if you want to have NEW customers who doesn't look the forums? What then???? A light website isn't always the best choice!!! Give it something! (Like: DivX, on2)!!!

id think they have to take one step at a time, after all
you seem to be forgetting that CORE doesnt just run on x86, there are many other platforms they could market too, i'd assume that if a webbrowser plug-in exists and they choose to make a seperate wiz-bang site/front page then we as their inital testers will point the less informed readers there.

its a shame doom9 has that registerd for 5 days rule before posting as i suspect there are many people wanting to comment on this thread, could that change some time soon DOOM9?.

[Selur]

http://forum.doom9.org/showthread.php?p=809261#post809261
=> seems like this only happens with the "BRAVIA Commercial: Extended version" from http://www.bravia-advert.com/commercial/braviaextcommhigh.html all other files play fine. (used Quicktime Pro 7 to save the file to hdd)

Cu Selur

[BetaBoy]

Hey, you must do something with your website "http://coreavc.corecodec.org/"! It is too much LIGHT!!! It doesn't give the confidence that people want to take!!! If you want to sell more give it a better style!!!:sly:

Yes, it might be a good and fast decoder and yes you did a very good advertisment in the video forums, but if you want to have NEW customers who doesn't look the forums? What then???? A light website isn't always the best choice!!! Give it something! (Like: DivX, on2)!!!

I did a small update about an hour ago... its a little more graphic. Our CMS was pushed out for a few more days so what you see now is a 'temp' site. Thanx for the feedback though.

[BetaBoy]

http://forum.doom9.org/showthread.php?p=809261#post809261
=> seems like this only happens with the "BRAVIA Commercial: Extended version" from http://www.bravia-advert.com/commercial/braviaextcommhigh.html all other files play fine. (used Quicktime Pro 7 to save the file to hdd)

Cu Selur
I've tried to duplicate this on three seperate PC's with no luck... all play fine. Anyone else?

[Ice =A=]

About that website: I really like it that way, it's refreshingly concise and clear.
I'm really fed up with those flamboyant sites that try to be stylish but only succede in making navigating much more complicated!!!

[BetaBoy]

id think they have to take one step at a time, after all
you seem to be forgetting that CORE doesnt just run on x86, there are many other platforms they could market too, i'd assume that if a webbrowser plug-in exists and they choose to make a seperate wiz-bang site/front page then we as their inital testers will point the less informed readers there.


You mean like this?
http://tcpmp.corecodec.org/tcpmpx/index.htm

btw... thats an older CoreAVC... and a build of CorePlayerX from over 2 months ago... but you see where we are going.

[CEC]

I am not saying that they must make it more complicated! The site funcionality is very good! I am saying that they must make it more beautiful!!!:)

We are speaking about a VIDEO DECODER here (people want to see something with style)!!! They have to use nicer colors and buttons!:D You can't just use 3-4 very basic colors :eek: for a website that sell a VIDEO DECODER!!!!! It is so....... Linux!!!:rolleyes:

Anyway, I hope to see something cool in it!!!:cool:

[Eretria-chan]

I can't view the site in Firefox. What plugin or whatever are you using there?

EDIT:

Selur: I tested your file by browsing to it, downloading via Save As, renaming the extension to mp4 and played it (by building a graph with graph builder). I built the graph with haali / coreavc / ffdshow audio and haali renderer. It plays just fine.

Also noticed:
The installer did not register the CoreAVC Decoder filter (using Pro); had to manually register it.
I did not encounter the aforementioned seeking issue (although I seeked only within graphedit).

Also regarding coreavc, it reduced the cpu % from about 70% to 50%. I got 70% when I did not use coreavc. I don't know WHAT I used, though, but I guess it was nero, because that is the only avc decoder installed and active. I tried this file BEFORE I registered the coreavc decoder.
I know this computer is utterly POC - the resolution was only 700 x 394.

EDIT2:
Now trying in WMP. I notice a small bug in seeking. It seems to either grab some frame before or simply stall a little before resuming playback the specified position. Cpu utilization does not peak to 100%, but stays normal, so I doubt it is the processor that is having trouble.
This laptop has an Amd Turion 64 Mobile ML-28 @ 800Mhz and 512 mb ram. As stated, I use Haali's Media Splitter included in the installation of CoreAVC. But I did notice that it didn't successfully install; some files might still be in use, so I shall try to investigate that, too, by reinstalling it tomorrow after having rebooted the computer.
I have the pro version.

[CruNcher]

@ CEC ehh especialy because it is a video decoder (something people never see work except they set something in the config dialog) it doesn't need to be flashy and so the website doesn't need to be if you want to market a Player thats a different thing but for a Video Codec it doesn't have to be and "less is more" here.


I've tried to duplicate this on three seperate PC's with no luck... all play fine. Anyone else?

Works here fine too no crash with Raw decoding and manipulateing Picture Properties

@ Eretria-chan
http://tcpmp.corecodec.org/tcpmpx/tcpmpAx.cab <- download extract and install but close firefox before then go to the webpage again :)

[Selur]

I've tried to duplicate this on three seperate PC's with no luck... all play fine. Anyone else?
Deinstalled ffdshow, Haali, CoreAVC, installed ffdshow from http://www.free-codecs.com/FFDShow_download.htm, reinstalled CoreAVC, now it works,.. (will report if I can recreate the problem somehow)

Cu Selur

[BetaBoy]

Not to get to OT... but we also created a new Codec we call 'COSP' (internal name). It is a Panorama based codec... have a look.

DEMO #2 - COSP Panorama Codec DEMO... grab the picture and move it...
COSP Demo 1: http://tcpmp.corecodec.org/demo/first/
COSP Demo 2: http://tcpmp.corecodec.org/demo/second/

This older version still has the FireFox 'crash' after you go to the webpage... this has been fixed in newer version. As well as a new right click menu.

[tomos]

Dan, is this spam blacklist still ongoing? could you get carlos to mail me the link again?

thanks :)

[Disabled]

If it could be integrated into XBMC I would instantly buy it - nah dang it, I'll buy it.
Ok got the professional version now.
But I'm still interested in the question about Mplayer/XBMC.

[popper]

You mean like this?
http://tcpmp.corecodec.org/tcpmpx/index.htm

btw... thats an older CoreAVC... and a build of CorePlayerX from over 2 months ago... but you see where we are going.

yea, thats the one i was thinking of, BTW anyone thinking
its not working after going to the above URL?.

its working for me with 'Advanced Browser'
http://www.advancedbrowser.com/ and IE.

try downloading the 'ice age 2' (Just Love that squirrel) 8)

from the wayback page4
http://forum.doom9.org/showthread.php?p=761387#post761387
and load that local file into the browser player, works a treat here, although im not sure if its the browser plug-in or the Version 0.0.0.4 Alpha (20060113) DS filter.

[Kostarum Rex Persia]

I am sorry for asking this, but from where I can download official version of CoreAVC decoder?

Thank you.

[BetaBoy]

I'll update the first post in this thread... its @ http://www.coreavc.com

[ChronoCross]

I am sorry for asking this, but from where I can download official version of CoreAVC decoder?

Thank you.

and before you ask no it's not free, they will not be making it free, and the preview time-limted version will not be out for another month.

[popper]

Not to get to OT... but we also created a new Codec we call 'COSP' (internal name). It is a Panorama based codec... have a look.

DEMO #2 - COSP Panorama Codec DEMO... grab the picture and move it...
COSP Demo 1: http://tcpmp.corecodec.org/demo/first/
COSP Demo 2: http://tcpmp.corecodec.org/demo/second/

This older version still has the FireFox 'crash' after you go to the webpage... this has been fixed in newer version. As well as a new right click menu.

dan , both of these url instantly crash both avanced browser
(see other post) and IE 6.0.2900.2180 with XP and SP2 installed, luckly, ABrowser remembers all the open urls
and reloaded all my pages (60+) and i just double clicked the bad url before it fully loaded so not crashing it.

[tomos]

odd, the last beta was ok, but the one i;ve been mailed goes back to the probs of beta 2 for me. the audio seems to be playing about 250ms too early. i have to correct this in mpc. tried this with several avi i've done myself and they're all the same. exact same audio delay needed in each one.

anyone else had the same problem?

[HookedOnTV]

Not to get to OT... but we also created a new Codec we call 'COSP' (internal name). It is a Panorama based codec... have a look.

DEMO #2 - COSP Panorama Codec DEMO... grab the picture and move it...
COSP Demo 1: http://tcpmp.corecodec.org/demo/first/
COSP Demo 2: http://tcpmp.corecodec.org/demo/second/

This older version still has the FireFox 'crash' after you go to the webpage... this has been fixed in newer version. As well as a new right click menu.

Javascript errors.

[BetaBoy]

yeah... they are basic demos... we have more complex and simplistic examples in the works.

[Sirber]

Are you calling a website test a "codec"?

[SeeMoreDigital]

yeah... they are basic demos... we have more complex and simplistic examples in the works.This is great fun....

Me thinks you need to start a new thread about it :D

[ChronoCross]

This looks very similar to the way plone looks. With the exception of the navbar being on the right instead of left.

[Abnormal1]

I just tried to use one of the early releases of the ds filter and sadly it will not work with mplayer on XBMC. XBMC tries to use the codec but errors.

05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll FlsAlloc
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll FlsGetValue
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll FlsSetValue
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll FlsFree
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll EncodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll EncodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll EncodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll EncodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll EncodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll EncodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll EncodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll EncodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll EncodePointer
05-04-2006 18:57:53 WARNING Unable to resolve: kernel32.dll DecodePointer
05-04-2006 18:57:53 WARNING kernel32.dll fake function DisableThreadLibraryCalls called
05-04-2006 18:57:53 ERROR cmplayer::openfile() smb://##.#.#.#/videos/test.mp4 failed
05-04-2006 18:57:54 ERROR Playlist Player: skipping unplayable item: 0, path [smb://##.#.#.#/videos/test.mp4]

[BetaBoy]

Are you calling a website test a "codec"?

No, we have developed a new panorama codec called 'COSP'. The website is just a demo to show it off.

[tomos]

hi Dan, any guesstimates for updates to the codec? logged my probs on your site bug reports section (audio sync probs)

[Disabled]

@Abnormal1 sad to hear it does not work out of the box. But I still have hope that the actual filter does work with it or may work in the future. Unfortunately I don't have a way to test this right now.
I Still would like to have an official statement here...

I did a quick and unfair speed comparison between VLC and CoreAAC.
252secs with AAC audio at 1Mbit/s (the video, not the audio) with every option Nero had to offer.
I just looked how many cpu-time was wasted by the two.
VLC used 1:48 min
MPC + CoreAAC + CoreAVC + Haali used 1:23 min
That results in a speed boost of 30% (or VLC is 25% slower, how you like it)
Quite impressive I think.
Does Nero ShowTime use its own filter? Just benched it and it took also 1:48 min to decode. So the clear winner is CoreAVC.

*edit*
DAMN! (Sorry) I'm able to watch this nice 720p Quicktime trailer. I first tried to play it in its .mov container, that sucked badly, but remuxed to .mkv it played flawlessly: 0 dropped frames, duration: 2:21 CPU-Time: 1:26 => average CPU load: 61% and that with an Athlon XP 2200+ 768MB Ram and sucky old GF2 TI 64MB

[Abnormal1]

Aparently mplayer's ability to use Directshow filters is limited so unless mplayer or CoreAVC are altered to work together it will never work.

[Eretria-chan]

If it works fine in other dshow apps, then it is mplayer's fault, yes? Then something should be done to mplayer.

[nm]

Yes, MPlayer has its own DShow interface implementation which is partly based on Wine code. That enables the DShow filters to work on other x86 platforms like Linux, *BSD and probably OS X x86, and not only Windows. Since the interface is not complete and implements only specific parts of the full Win32 API, new binary-only filters usually need some coding work in the MPlayer's side. However, CoreAVC developers are probably more willing to work with the community than the codec manufacturers generally, so getting the filter to work should not be too difficult given some time and effort.

[ravemax]

mplayer - coreavc: Since both mplayer and VLC are GPL, a easy way would be to write a stub, which loads CoreAVC using "dlopen". So at the end BetaBoy & Co only would have to release the source of the stub.

[nm]

That would be the way to do it, but getting that stub approved to the official MPlayer and VLC code might prove more difficult unless it is only an extension to an existing mechanism, like the DShow loader.

[welew]

A small bug of CoreAVC (0.0.0.4 beta version):
1, when CoreAVCDecoder.ax put and registered in the non-english directory, it can not be called by filter-based player (MPC, Graphedit, ...)

2, registry entry would be modified to fix the problem, and i write a simple program to do it.

3, do not know CoreAVC commercial version already fixed it.

[BetaBoy]

welew while I apprecite that report... 0.4 was a test codec. Please only report 1.xx bug reports. You can do so on our bug tracker here:
http://corecodec.org/tracker/?atid=313&group_id=64&func=browse

[breez]

Sorry if this is a stupid question or one already answered, but what about the TCPMP AVC plugin? Is the "1.0" version of it already available and is it a separate purchase from the DS filter?

[BetaBoy]

CoreAVC will be a included with the FULL version of CorePlayer v1.0. We are pushing the launch to be in mid-may... atm we will launch with versions for Palm, CE, Symbian, and Linux.

Also... BetaPlayer 1.0 will likely be out sooner... We will also offer a 'codec bundle' for all the pay codecs for it as well.

[Gnerma]

So is the plan for a "FFDSHOW like" all in one DirectShow decoding package? Or at least ASP, AVC, AAC, Vorbis and a few other formats?

[lexor]

CoreAVC will be a included with the FULL version of CorePlayer v1.0. We are pushing the launch to be in mid-may... atm we will launch with versions for Palm, CE, Symbian, and Linux.

no windows? :confused:

[BetaBoy]

no windows? :confused:
Our desktop strategy is very complex... not sure this is the right thread to be talking about it.

But basically by Xmas will will have alot to show you all. Alot has to do with our web integration, online store, metadata managment at the client(s)/server level and what we are internally calling 'CoreTunes'... which syncs the multiple DB's with all your CorePlayers on each device (PDA, Cell, Desktop, etc).

[Foreigner999]

@betaboy

Do you know if people who bought the CoreAVC DSfilter will get any kind of upgrade path for the official TCPMP release? Any discounts or will we have to plunk down another $XX to get that as well?:D

[tomos]

@betaboy, any info on this audio delay bug? it's odd that it was fixed in the beta i received but its back in the final release?

[BetaBoy]

@betaboy, any info on this audio delay bug? it's odd that it was fixed in the beta i received but its back in the final release?
tomos.... we have not been able to duplicate this at all. ;-(

[Sirber]

which syncs the multiple DB's with all your CorePlayers on each device (PDA, Cell, Desktop, etc).iPod too? :D

[tomos]

tomos.... we have not been able to duplicate this at all. ;-(

weird, someone i've spoken to on another forum has the same prob. can i send you a small section of an avi for you to test with?

[Sirber]

audio delay is related to the use of HE (and PS).

[DigitalDeviant]

iPod too? :D

Well, I guess since you can make your iPod run Linux (http://ipodlinux.sourceforge.net/index.shtml) it could work.

Somehow, I don't think this is what you meant.

[tomos]

sorry, i have no idea what HE/PS is :confused:

[Sirber]

sorry, i have no idea what HE/PS is :confused:
:search:

http://encyclopedia.tfd.com/aacPlus

[BetaBoy]

audio delay is related to the use of HE (and PS).

Thx Siber...

[BetaBoy]

iPod too? :D
Our 'CoreTheque' DB already syncs with iTunes perfectly. So yes... now playing back that content... thats another thing ;-/

[BetaBoy]

@betaboy

Do you know if people who bought the CoreAVC DSfilter will get any kind of upgrade path for the official TCPMP release? Any discounts or will we have to plunk down another $XX to get that as well?:D
No since, the DS filter is a totally different products. Technically CoreAVC will be a part of CorePlayer (mobile)... but you will not have the advanced filter properties that is available... mainly because they are not needed for a mobile device.

We have two clear cut avenues to use the codecs we have created.. one is threw the DS framework and the other is with CorePlayer/BetaPlayer. Our non-DS efforts for playback within our players will _always_ be faster then our DS efforts.

Also note that there are even features that are in the current TCPMP that even Haali and picard noted are faster then what even dxva could provide.

[Sirber]

Thx Siber...
Sirber ;)

[BetaBoy]

Sirber ;)
/me smashes Toff's 'frenglish' keyboard... sry :stupid:

[HookedOnTV]

I know it is still premature but... any indications of a prefference between the two camps for gpu acceleration?

[Gnerma]

BetaBoy isn't saying anything about GPU support yet. I suspect its because they're not all too sure how its going to turn out themselves. Heres to hoping that they'll manage to hammer out support for all SM 3.0 GPUs.

[CEC]

Do you have a release date of the trial version???? How limited is going to be????

[BetaBoy]

Do you have a release date of the trial version???? How limited is going to be????
No... and like I said the other day it will be at least a month away. its not a priority atm.

[BetaBoy]

BetaBoy isn't saying anything about GPU support yet. I suspect its because they're not all too sure how its going to turn out themselves. Heres to hoping that they'll manage to hammer out support for all SM 3.0 GPUs.
Just know that without saying specifics... we are working on it.

[lcx]

audio delay is related to the use of HE (and PS).
I am using the MainConcept/Elecard decoder. It has no audio delay with aacplus muxed videos.
With CoreAVC there is a significant audio delay.

[SeeMoreDigital]

I am using the MainConcept/Elecard decoder. It has no audio delay with aacplus muxed videos.
With CoreAVC there is a significant audio delay.That's odd!

I'm able to decode 1280x720 AVC video using CoreAVC. And decode 6Ch AAC-HE audio and transcode it to AC3 using FFdshow, with no delay..... Even on my 2.8GHz P4

[tomos]

I am using the MainConcept/Elecard decoder. It has no audio delay with aacplus muxed videos.
With CoreAVC there is a significant audio delay.

Icx, is it the same prob i have where the audio is always playing around 250ms too early?

[MaxL]

WHEN ppl without possibility to pay over PayPal will be ABLE to buy CoreAVC decoder. PayPal doesnt accept countries like Easter Europe and 2/3 of other countries worldwide. That really sux.

Why dont u have enabled alternative paying systems like WorldPay or similar?

MaxL

P.S. Sorry if this is not right place to ask this Q.

[ChronoCross]

The reason that paypal doesn't accept alot of forign accounts and CC's is because that (at my last look at stats) 60% of all purchases made using paypal outside of the US were fraudulent.

I'm sure they will come up with an alternative for international buyers.

[soresu]

Once purchased, will an updated version be sent to my e-mail automatically once it becomes available? Or will I have to download it from the website?

[BetaBoy]

Once purchased, will an updated version be sent to my e-mail automatically once it becomes available? Or will I have to download it from the website?
Our 'Core Account' system will be up soon. Once it is live... each purchaser will get an email asking them to log into it for the first time. In there you will be able to enable/disable update notification via email for every CoreCodec product you have purchased... that also includes update notifications via email for our Open Source projects as well.

[7zeal]

So are you going to change paypal for some other exchange site? I live in Romania and it's not stated there.. and I really want to buy the new coreavc. You'd better hurry, there are lots of customers in unsupported countries too. :) You don't want to lose them, do you?

[Eretria-chan]

When the core account system is online, hopefully there should be alternate payment methods available. Be a little more patient and you'll see a solution will show...

[7zeal]

cool. i doubt it, but .. cool :P

[BetaBoy]

So are you going to change paypal for some other exchange site? I live in Romania and it's not stated there.. and I really want to buy the new coreavc. You'd better hurry, there are lots of customers in unsupported countries too. :) You don't want to lose them, do you?
Of coarse we dont... we will have a more universal payment system in place and some additional payment options very soon. It will not be at first launch of the account system but soon afterwards.

[JoeBG]

@ Beta Boy

Is it allowed to make a download package with the old and free version 00004 of CoreAVC?

[BetaBoy]

@ Beta Boy

Is it allowed to make a download package with the old and free version 00004 of CoreAVC?
No, as a matter of fact I have sent 6 take down notices over the past few days.

[7zeal]

I would like to become a beta tester for CoreAVC.
If you consider it, please let me know. Btw I can do programming too in many languages, but .. just tell me how I could help. :sly:

[bob0r]

No, as a matter of fact I have sent 6 take down notices over the past few days.

I want one too :scared:

[BetaBoy]

I want one too :scared:
Like I told Jarod... we are not here to bust balls... I just need to state that we never distro'd 0.04 with 'any' license and as we are the copyright holders, that others should not state what 'we' actually did or what 'they' can do without the right to do so.

well... except for those .NL users ;-)

[BetaBoy]

On a devel note... Haali noted some minor changes for the next release and when Toff gets back today we will work on the new installer that merges Haali's splitter with the installer.

On a side note... Would everyone want our new CoreAAC v2.0 packaged with CoreAVC? . . and before ppl ask... the price would remain the same.

[maxleung]

I am finally allowed to post! Arrrghhh to the Doom9 forum for giving us CoreAVC users a 5 day delay!

Anyways, yes, I would love to see CoreAAC 2 bundled - I've been having a hell of a time finding a place to download a RECENT version of CoreAAC.

Also, what version of the Core Player should I be using? Core Media Player? The Core Media Player? The Core Player? Core Player? There are so many permutations of the name I have no idea which one would work with CoreAVC 1.0.

I am using Windows by the way...

[maxleung]

Another question, this time about benchmarking:

Is there a tool, similare to TimeCodec.exe, that will allow me to tell if my system is dropping frames because my CPU isn't powerful enough to decode in real time?

Timecodec.exe, while a nice idea, is useless at telling you if the decoding framerate dips below the source video framerate.

For example, Timecodec.exe tells me that the Marie Antoinette 1080p trailer plays back at 34 frames per second. However, when I play it back in realtime, I can see frame drops here and there - obviously in some spots my PC cannot keep up. I would like to know if there is a tool out there that can tell me what the MINIMUM framerate achieved was?

Thanks!

[ChronoCross]

i would love to have coreaac bundled. as long as it's fully featured and not different between basic pro and enterprise. I'm still using your old beta's of it and would appreciate greatly a new version =D . Since I am very satisfied with my purchase of pro coreavc and plan on upgrading to enterprise version in the future.

[niknik]

yes indeed, CoreAAC and AVC pack! :)

[Sirber]

Does CoreAAC v2 decode AAC HE+PS?

[Schrade]

Is the graphic on the CoreAVC website a condom with a face, ears and arms on it?

http://coreavc.corecodec.org/images/mlogo.gif (http://www.coreavc.com/)

[sillKotscha]

muuuaaaaahhhahahaaa,

yes, it looks like :D

[CiTay]

Is the graphic on the CoreAVC website a condom with a face, ears and arms on it?

Funny, yesterday it had a different symbol than that "copy controlled" one. But i also noticed the condom-shape.

[SeeMoreDigital]

On a side note... Would everyone want our new CoreAAC v2.0 packaged with CoreAVC?Sounds like an excellent plan ;)

[dk75]

Is there a tool, similare to TimeCodec.exe, that will allow me to tell if my system is dropping frames because my CPU isn't powerful enough to decode in real time?

Timecodec.exe, while a nice idea, is useless at telling you if the decoding framerate dips below the source video framerate.

For example, Timecodec.exe tells me that the Marie Antoinette 1080p trailer plays back at 34 frames per second. However, when I play it back in realtime, I can see frame drops here and there - obviously in some spots my PC cannot keep up. I would like to know if there is a tool out there that can tell me what the MINIMUM framerate achieved was?

Maybe you'll try to use actual renderer in it? (see "dfps" at picture below)

http://img416.imageshack.us/img416/357/timecodec2hk.jpg

[maxleung]

dk75, I did that - the dfps of 34 was with VMR9. It does NOT tell me that dropped frames would result, like it did when I watched it in a real directshow player.

[videomixer9]

VMR9 is too much depended on the player, also I think it'll have less cpu usage with smaller image. And if you try VMR9 in timeCodec and then use MPC renderless e.g. to play the video it's only natural that it varies, MPC VMR9 renderless is dependant on the players additional processing too, especially software scalers etc.

[dk75]

dk75, I did that - the dfps of 34 was with VMR9. It does NOT tell me that dropped frames would result, like it did when I watched it in a real directshow player.
Then use ZoomPlayer and during playback hit "I" - you will see actual FPS.

[maxleung]

dk75, so what you are saying is there is no way to determine if you get framedrops or drop below source framerate without actually monitoring the fps manually? Is that "actual FPS" an average or instantaneous?

IF the above is an average, what you propose is useless for the Marie Antoinette 1080p trailer on my rig - it takes the VMR9 renderer tens of seconds to settle the framerate, and by that time the critical high-bitrate section at the beginning has just been played - and then the framerate holds steady even when one or two frames gets dropped because it is an AVERAGE framerate, not an instantaneous one. Make sense?

So then there is no automated way of benchmarking video performance including minimum framerate achieved?

videomixer9, that's okay - I'm merely concerned with raw codec performance, not whole playback chain performance.

This is like those quick-and-dirty videogame benchmarks - "Oh look, Doom3 averages 60 frames per second on this rig!" - ignoring the fact that key scenes in the game dip below 10 frames per second! Playability? Who cares? :p

EDIT: Oops, removed a sentence that didn't make sense.

[CEC]

Is coreaac only going to be free???

[ChronoCross]

Is coreaac only going to be free???

Hopefully not. It should be paid and come with any purchase of the video decoder.

[starlabs]

I wouldn't mind CoreAAC being included, but please make sure it's an optional install as with Haali...

BTW it's a great piece of software you guys have. Very fair pricing. Thumbs up! :)

[BetaBoy]

CoreAAC v2.0 would be available for CoreAVC users who already paid for any version. It would also be in our 'Core Media!' playback bundle... and yes would be set as an optional install.

[Hans Ohlo]

dk75, so what you are saying is there is no way to determine if you get framedrops or drop below source framerate without actually monitoring the fps manually? Is that "actual FPS" an average or instantaneous?if you use media player classic (mpc) you can enable a statistic display in the view menu. there you can see current bitrate, average bitrate, dropped frames (accumulated), average fps and so on. maybe this is what you are looking for...

[maxleung]

Thanks Hans! I'll give that a shot!

[MrWizard]

Just as an FYI, you may have a slight problem with your current download system. In my registration email, it said "In the mean time, you should keep a copy of this file somewhere safe, as downloads are automatically deleted after 7 days." Just tried it again and it still works. Will likely be fixed with your new accounting system, yes? So, it may not matter to you, but I thought I'd mention it.

[maxleung]

This is tougher than I thought - I could not get Zoomplayer or MPC to report any dropped frames, even if I do something insane like run Cinema Craft Encoder at the same time (at a higher priority too!). It looks like the video will slow down and the audio will stutter - hence, no dropped frames!

So, your suggestion DK75, of using the Information dialog box in Zoomplayer is the best bet - it reports the frame rate that is updated quite frequently and isn't a longrunning average. Dips to 19 fps on this P4 3.0 GHz PC with a crappy ATI card in it.

Thanks dk75.

[BetaBoy]

Just as an FYI, you may have a slight problem with your current download system. In my registration email, it said "In the mean time, you should keep a copy of this file somewhere safe, as downloads are automatically deleted after 7 days." Just tried it again and it still works. Will likely be fixed with your new accounting system, yes? So, it may not matter to you, but I thought I'd mention it.
Correct... we also just added the delete script and have change it to 24 hours.

[dk75]

This is tougher than I thought - I could not get Zoomplayer or MPC to report any dropped frames, even if I do something insane like run Cinema Craft Encoder at the same time (at a higher priority too!). It looks like the video will slow down and the audio will stutter - hence, no dropped frames!
I think it's bycause H264 DS filter. I've seen already few times a dropped frames but never with H264. It's always slowmotion.
And as for slower playback than timeCodec - are You sure that You 're using CoreAVC to playback with Your player?

[maxleung]

dk75, sorry, the playback I did today was on my work PC which is a P4. I was running the Timecodec test on my home machine, which apparently is a lot faster. :)

[ACrowley]

Icx, is it the same prob i have where the audio is always playing around 250ms too early?

same problem here with DirectShow CoreAVC

But only with "some" h264 720p Files

No dropped Frames so far and stable framerate on A64 3400/2gig RAm/x800xt-pe

But theres the audioDelay-Audio async,how subscribed from the others here!

Nevermind what Container is used ,or what for a DirectShow Player is used!
Also nevermind if i use Haali Spliter or something else.

Problems with ALL CoreAVC Versions (0001-1.0)

And i know lot of People with CoreAVC and a Audio delay,async, Problem too

So its a bigger Problem...

With the CorePlayer ,tcpmpW32, and the AVC Plugin theres NO Audio async !
So ,theres only the Bug with the DirtectShow Versions of CoreAVC

BetaBoy...i hope we can get a bugfix for that !

As is say only with some h264-HD Files..

Heres a sample with async Audio on (only) CoreAVC :
http://rapidshare.de/files/17432320/sample_h264_mkv.mkv.html

Greetz

[tomos]

weird, happens with all of mine. whether 720, or 1080. if i add 250ms to my players delay then its all fine for every single file.

this didnt happen during older releases tho, just 1 beta.

[BetaBoy]

same problem here with DirectShow CoreAVC

But only with "some" h264 720p Files

No dropped Frames so far and stable framerate on A64 3400/2gig RAm/x800xt-pe

But theres the audioDelay-Audio async,how subscribed from the others here!

Nevermind what Container is used ,or what for a DirectShow Player is used!
Also nevermind if i use Haali Spliter or something else.

Problems with ALL CoreAVC Versions (0001-1.0)

And i know lot of People with CoreAVC and a Audio delay,async, Problem too

So its a bigger Problem...

With the CorePlayer ,tcpmpW32, and the AVC Plugin theres NO Audio async !
So ,theres only the Bug with the DirtectShow Versions of CoreAVC

BetaBoy...i hope we can get a bugfix for that !

As is say only with some h264-HD Files..

Heres a sample with async Audio on (only) CoreAVC :
http://rapidshare.de/files/17432320/sample_h264_mkv.mkv.html

Greetz

I am still trying to determine what the actual issue is... 4 systems I test on here do not have the issue. I have asked 3 beta testers to test the sample MKV so i'll report what they tell me.

I'll ping Haali and Toff on this.

[tomos]

do you want an example of an AVI that has this problem?

[HookedOnTV]

I played the above sample on two machines and they were both in sync (atleast as close as you could expect being a foreign language dub).

[net1999]

Had a quick test, the avi which is encoded with vfw mode x264 had the sync problem when I tried to do the fast searching. MKV which is done by MeGUI was acting perfect...

[jeffhuge]

Had a quick test, the avi which is encoded with vfw mode x264 had the sync problem when I tried to do the fast searching. MKV which is done by MeGUI was acting perfect...
momo net1999 sama
:)

[BlackSharkfr]

www.hardware.fr did a small review of Core AVC
They tested it with an Athon FX 55 and they compared CoreAVC with Cyberlink and Nero, with and without graphics card acceleration :
link to the article (french) (http://www.hardware.fr/news/8117/coreavc-plus-fort-qu-avivo-purevideo.html)
http://www.hardware.fr/medias/photos_news/00/16/IMG0016450.gif
http://www.hardware.fr/medias/photos_news/00/16/IMG0016453.gif

[maxleung]

Betaboy says:

I am still trying to determine what the actual issue is... 4 systems I test on here do not have the issue. I have asked 3 beta testers to test the sample MKV so i'll report what they tell me.

This one plays strange on my system - audio is garbage because Zoomplayer picked the DScaler Audio Filter and I get massive static. Hmmm...what is the audio format anyways?

EDIT: Well, I don't know why it picked DScaler, but when I told it to pass the MPEG audio through SPDIF it started to work. THe audio is out of sync.

[BetaBoy]

do you want an example of an AVI that has this problem?
yes... send it via yousendit.com and use betaboy@corecodec.com . .thx

[tomos]

uploading via that site now. 3 mins left at this point

while doing this i found something tho. i cut out a speech from a movie i have encoded. playing the movie straight away and it seems fine. the audio is perfectly in sync. now if i seek at all within the clip, start/middle/end, then the audio is 250ms out of sync.

same in mpc/zplayer etc. just playing the clip and not seeking at all leaves the audio percetly in sync.

hopefully you can confirm this on your end with this clip Dan :)

[anisota]

I thought someone on this forum might be interested in work I've done using the CoreAVCCodec in Linux. Specifically, I have hacked up the win32 filter loader in mplayer to handle CorAVCDecoder.ax. I initially got it working with 0.0.4, but that was only 2 days before 1.0 was released. After 1.0 was released, I had to do a lot more work to get it working again due to a lot more checking of data integrity inside the DLL itself. But I have now succeeded in doing so. Of course, I was informed by CoreCodec support that they'll be releasing a linux version in a couple of weeks, so this may not be useful to anyone, but my main goal is actually to use this filter with mythtv (which I have succeeded in as well), and I don't have any indication that their linux release will support that.

The code is still very crude, and I can't gaurantee it will work for anyone else, but it should. Also, it was built for a 32-bit pentium-4. I have no idea what will happen if it is compiled for 64bit, or on anything but gcc/linux. Specifically, I had to implement some nasty assembly stack manipulation to implement a Windows-style exception handler, and that is certainly very non portable. Also, the code probably won't play nicely with any other win32 filters. you've been warned.

Also, because the official codec does a key-validation, you need to supply your key to mplayer. I am not aware of any way to edit the registry file, so it is currently hardcoded (of course I didn't provide my key, you'll need to provide your own)

for mplayer:
apply the patch, then edit loader/dshow/DS_VideoDecoder.c
Search for 'Joe User' and replace the key/user-id from what is in your registry after installing the codec (luckily enough, the installer works great with wine, so no windows system is required)
run configure and make as usual
place the CoreAVCDecoder.ax file into /usr/lib/win32 (or wherever you configured mplayer to look for it)

append the following to the end of your codecs.conf file:

videocodec coreavc
info "CoreAVC DShow H264 decoder for x86 - http://corecodec.org/"
status untested
format 0x10000005
; format 0x31637661 ; avc1
fourcc H264,h264
fourcc X264,x264
fourcc avc1
fourcc davc,DAVC
driver dshow
dll "CoreAVCDecoder.ax"
guid 0x09571a4b, 0xf1fe, 0x4c60, 0x97, 0x60, 0xde, 0x6d, 0x31, 0x0c, 0x7c, 0x31
out YV12,IYUV,I420,YUY2


you should be ready to go.

for mythtv:
install the patch
cd libs/libdshowloader
make
cd dshow
make
cd ../../..

make and install as usual
copy CoreAVCDecoder.ax to /usr/lib/win32
copy the dshowcodecs file to ~/.mythtv/ (of the user you run mythfrontend as)
edit dshowcodecs and change the key and user id to the one from your registry (no need to edit the source for mythtv)
You should be good to go.

I have only tried the above with the 'Pro' version. It may or may-not work with any other version.
http://www.files.bz/files/9309/Mythtv/mplayer_coreavc.patch.20060413.bz2
http://www.files.bz/files/9309/Mythtv/directshow.patch.20060413.bz2
http://www.files.bz/files/9309/Mythtv/dshowcodecs

Here is some performance comparison I did (with 0.0.4) on linux:

for a 54 second 1280x720 video on a P4 2.8GHz (on linux)

using ffmpeg:
mplayer -vo null -nosound -benchmark h264.mpg
BENCHMARKs: VC: 47.711s VO: 0.038s A: 0.000s Sys: 0.732s = 48.481s
BENCHMARK%: VC: 98.4132% VO: 0.0779% A: 0.0000% Sys: 1.5089% = 100.0000%

using coreAVC:
mplayer -vo null -nosound -benchmark -vc coreavc h264.mpg
BENCHMARKs: VC: 35.498s VO: 0.040s A: 0.000s Sys: 0.750s = 36.288s
BENCHMARK%: VC: 97.8222% VO: 0.1109% A: 0.0000% Sys: 2.0669% = 100.0000%

[BetaBoy]

uploading via that site now. 3 mins left at this point

while doing this i found something tho. i cut out a speech from a movie i have encoded. playing the movie straight away and it seems fine. the audio is perfectly in sync. now if i seek at all within the clip, start/middle/end, then the audio is 250ms out of sync.

same in mpc/zplayer etc. just playing the clip and not seeking at all leaves the audio percetly in sync.

hopefully you can confirm this on your end with this clip Dan :)

This is a known bug and has already been fixed and will be in the next release.

[ACrowley]

@BetaBoy

Something new about the Audio async Bug ?

I think i will upload a new sample which is only with CoreAVC Async later.

[soresu]

Has anyone else here ordered CoreAVC and not been sent a download link email yet? Ive been waiting a week already.

[BetaBoy]

soresu... there are a handlful of ppl that seem to have not gotten their orders. You can email me betaboy@gmail.com or betaboy@corecodec.com to take care of it.

[BetaBoy]

@BetaBoy

Something new about the Audio async Bug ?

I think i will upload a new sample which is only with CoreAVC Async later.
Pls upload a clip... so far all of us that have tested, have not had any of the audio sync issues that have been reported. The demos ppl have sent us do not point to any issue with the decoder at all, but a local config issue.

[Shazam]

Hi guys !

I'm downloading HD trailers in 1080p (*.hdmov).

I'm using coreavc, MPC and zoomplayer with ffdshow.

My pb is that hdmov files, according to graphedit, are associated with coreavc AND ffdshow mpeg-4 video decoder, which jeopardizes the playback quality

http://img399.imageshack.us/img399/5725/graph1xs.jpg

When I remove ffdshow in graphedit , everything's fine !!

So how to get the right association to playback hasslefree in media player ?

http://img157.imageshack.us/img157/5171/graph10yf.jpg

Thanks for your help !

PS : h.264 has been deactivated in ffdshow...

[breez]

Deactivate "raw video" from ffdshow.

[Shazam]

Deactivate "raw video" from ffdshow.

Thanks for the advice : works !

...and why this has to be done ?

[[)370|\|470!2]

Because it has higher merit in dshow.

[Shazam]

Thanks !

[ravemax]

@anisota: Sweet.
- Is the hardcoded 1280x720 only a default-case, only used when the resolution was not specified?
- Still got the 0.04 .patch?
Thanks.

[ACrowley]

Pls upload a clip... so far all of us that have tested, have not had any of the audio sync issues that have been reported. The demos ppl have sent us do not point to any issue with the decoder at all, but a local config issue.

http://rapidshare.de/files/18167628/test.mkv.html

Here another sample where i (and others) have Problems with Audio async. only with CoreAVC DirectShow Version

TCPMPW32 CoreAVC .PLG plays it 100% in sync!

Greetz

[BetaBoy]

ok... 4 PC's here and this tests fine both via DS and CorePlayer Pro (win32)...... tell me specifically what audio codec you are using for DS AC-3 playback?

[sillKotscha]

ok... 4 PC's here and this tests fine [...] via DS

up to 5 ;)

[bob0r]

http://rapidshare.de/files/18167628/test.mkv.html

Here another sample where i (and others) have Problems with Audio async. only with CoreAVC DirectShow Version

TCPMPW32 CoreAVC .PLG plays it 100% in sync!

Greetz

It plays here very out of sync also.
Even tho media player classic is not dropping any frames, i think i have found MY problem: 1-AMD XP, 1530MHz, 256KB
What's yours?

[welew]

When coreavc (1.0.0.0 Pro) is installed in a non-english direcory (chinese directory for example), then CoreAVCDecoder.ax would not be called.

Solution: unrar the patch file to the same directory containing CoreAVCDecoder.ax, then run it. Done!

welew 2006-4-17
http://corecodec.org/tracker/download.php/64/313/727/32/CoreAVC_Patch.rar

[kirkers]

The latest FFDShow build (2006-04-13) by Milan Cutka posted on the x264.nl site has solved a huge problem with C++ assertion error crashes in Winamp streams and files using H/x264.

I don't know if this is exactly the right place, but I wanted to acknowledge the tremendous improvement in quality.

(Moderators: please post this in other thread(s) based on your best judgment)

[BetaBoy]

The latest FFDShow build (2006-04-13) by Milan Cutka posted on the x264.nl site has solved a huge problem with C++ assertion error crashes in Winamp streams and files using H/x264.

I don't know if this is exactly the right place, but I wanted to acknowledge the tremendous improvement in quality.

(Moderators: please post this in other thread(s) based on your best judgment)

Way OT ;-)

[BetaBoy]

When coreavc (1.0.0.0 Pro) is installed in a non-english direcory (chinese directory for example), then CoreAVCDecoder.ax would not be called.

Solution: unrar the patch file to the same directory containing CoreAVCDecoder.ax, then run it. Done!

welew 2006-4-17
http://corecodec.org/tracker/download.php/64/313/727/32/CoreAVC_Patch.rar
EDIT: NM... I see it was added to the tracker... thx

[ACrowley]

It plays here very out of sync also.
Even tho media player classic is not dropping any frames, i think i have found MY problem: 1-AMD XP, 1530MHz, 256KB
What's yours?

No...mine is strong enough for 720p h264!

A64 3400/2gb ram/x800xt-pe/audigy2zs/Wxp Pro

My DirectShow Interface is working fine!
CoreAVC ,Haali or all other Decoder/filter for each MediaFiles are used proberly in each Dshow player.
Graphedit shows me too..


As i say CorePlayer is playing theses problamtic Files 100% in sync!
Only Dshow CoreAVC not.. No dropped Frames, 21,50 average frames

All other HDTV Content like mpeg2(ts) wmv-9 ..no async


@betaBoy
I use AC3 FIlter for decoding AC3 or DTS !
Also test ffdshow or other AC3 decoder like cyberlink, nero...all the same!
And i test these Files without any Audio stream too...the same!

With the Samples i have a CPU load of 98-100% with CoreAVC ..
Whats you CPU load with the samples ??

[BetaBoy]

With the Samples i have a CPU load of 98-100% with CoreAVC ..
Whats you CPU load with the samples ??
32%... using cyberlink and it plays fine... can you specifically show me captures of graphedit? Can you uninstall FFDShow and see if this changes anything?

[ACrowley]

32% ....hardware accel. ??

i know absolutly no User who has only 32% with it on around 3,2gig CPU?? 100% not!
Are you sure ?
What for a CPU are you using ? DualCore?

And no..ffdshow unin. didnt change anything!

As i say only HaaliSplitter/CoreAVC/AC3Filter is used for those h264 Files!
In every DirectShowPlayer , also in Graphedit (with Video/AudioRenderer at the End)

So its 100% not a Problem with my Dshow Interface and with wrong used Decoders

Its CoreAVC himself ,because tcpmp or other Decoders play it in sync.
And only "some" h264- htv Files (other codes=no Problem) are not in sync..so it must be CoreAVC only

[BetaBoy]

32% ....hardware accel. ??

i know absolutly no User who has only 32% with it on around 3,2gig CPU?? 100% not!
Are you sure ?
What for a CPU are you using ? DualCore?

And no..ffdshow unin. didnt change anything!

As i say only HaaliSplitter/CoreAVC/AC3Filter is used for those h264 Files!
In every DirectShowPlayer , also in Graphedit (with Video/AudioRenderer at the End)

So its 100% not a Problem with my Dshow Interface and with wrong used Decoders

Its CoreAVC himself ,because tcpmp or other Decoders play it in sync.
And only "some" h264- htv Files (other codes=no Problem) are not in sync..so it must be CoreAVC only

Yes Dual Core T2600 Inspiron... we will continue to look into the issue.

[emmel]

No...mine is strong enough for 720p h264!

A64 3400/2gb ram/x800xt-pe/audigy2zs/Wxp Pro

With the Samples i have a CPU load of 98-100% with CoreAVC ..


Just tried your test.mkv with my A64 3200/512M/X600/XP Pro and HaaliSplitter/CoreAVC 1.0.0.0. I get 60% load with overlay renderer and 80% with vmr9. All n'sync.

[soresu]

Still no reply to my e-mail about missing CoreAVC pro download link. Gettin desperate now! :scared:

[ChronoCross]

Still no reply to my e-mail about missing CoreAVC pro download link. Gettin desperate now! :scared:'

check your spam folder....

[soresu]

Nope, ive got the purchase notification from service@paypal.co.uk - but no download link e-mail

[BetaBoy]

Nope, ive got the purchase notification from service@paypal.co.uk - but no download link e-mail

PM me your PP transaction # and i'll resend the email.

[KoD]

AC3Filter <- that's the problem usually when desync happens. How about uninstalling it ?

[BetaBoy]

Note ACrowley stated he was not using AC3Filter and got the same results... Maybe uninstalling it?

[emmel]

Hi,

I would like to ask a small question considering the timing of interlaced h264 when played with CoreAVC.

I have a couple of mp4-files containing interlaced h264. The content is obtained from a local satellite test transmission, demultiplexed into h264, and finally wrapped in a mp4 container with fps=25.

The h264 contains two kinds of 1-2 min sets following each other in a loop. According to the in picture timing sei message the sets consist of a) pict_type=1 and pict_type=2 and b) pict_type=3. In both sets, the clock_timestamp_flag is set to zero. In both cases the vui params num_units_in_tick=1 and timescale=50. Thats all timing info I can find from the data by using h264_parse.

Now, my problem is: The video plays smooth with correct frame rate only for picture of type a). For type b), the playback slows down by a factor of two. So, I have a video which is either perfect or in slow motion in any dshow player.

My question is, is there something I could take into account when trying to fix the case? I'm not sure if I have provided sufficiently information, or if this eventually has got anything to do with CoreAVC - it happens to be the only sufficiently powerful decorer I have capable of reveiling this timing issue anyways.

Thanks and br,
emmel

[BetaBoy]

Hi,

I would like to ask a small question considering the timing of interlaced h264 when played with CoreAVC.

I have a couple of mp4-files containing interlaced h264. The content is obtained from a local satellite test transmission, demultiplexed into h264, and finally wrapped in a mp4 container with fps=25.

The h264 contains two kinds of 1-2 min sets following each other in a loop. According to the in picture timing sei message the sets consist of a) pict_type=1 and pict_type=2 and b) pict_type=3. In both sets, the clock_timestamp_flag is set to zero. In both cases the vui params num_units_in_tick=1 and timescale=50. Thats all timing info I can find from the data by using h264_parse.

Now, my problem is: The video plays smooth with correct frame rate only for picture of type a). For type b), the playback slows down by a factor of two. So, I have a video which is either perfect or in slow motion in any dshow player.

My question is, is there something I could take into account when trying to fix the case? I'm not sure if I have provided sufficiently information, or if this eventually has got anything to do with CoreAVC - it happens to be the only sufficiently powerful decorer I have capable of reveiling this timing issue anyways.

Thanks and br,
emmel

Are you using the Haali splitter? or another third party one?

[emmel]

It is the Haali splitter (and no, "force fake 25 fps" in the splitter options does not fix the issue).

[ravemax]

@anisota: I build an XBMC mplayer.dll with your patch applied, but doesn't work on the Xbox. That's probably because I had to remove the "//" before "#ifdef WIN32_LOADER / #endif".
Guess I have to wait for a Linux version.

[Inventive Software]

Hey there! I'm really interested in this, as I downloaded one of the very early betas for TCPMP. However, my question is simple. Is there a free version of the DirectShow decoder? I read something about it earlier in the thread, and I don't have time to read 500 odd posts to see what! All I wanna do is play standard definition non-interlaced AVC content on my not-brilliant PC. CoreAVC does that (miraculously!) Can you gimme any info please?

[Eretria-chan]

No free version available...
But they aren't expensive. Check the homepage...

[emmel]

Are you using the Haali splitter? or another third party one?

I was able to do some verification with a more powerful computer and found out that some other filters seem to play the interlaced content corretly.

To me it seems that if only simple vui timing info is available, and if there is a sequence of fields followed by a sequence of frames (even if they are correctly marked in the sei), they all get played with the same number of ticks/field = ticks/frame. I'm not an expert here, this is just a guess, but it at least seems like that when watching the output.

Anyways, I was able to solve the problem by running the output through Avisynth with a fixed fps=25, or by adding an additional dshow frame rate converted between the CoreAVC and renderer. The output is then 100% perfect.

emmel

ps. I'll add for reference a more detailed description of the timing info (by h264_parse):

- For field based sequences::
ref 0 type 6 SEI
payload_type: 1 pic_timing
payload_size: 1 0x14
pict_struct: 1
clock_timestamp_flag[0]: 0

ref 0 type 6 SEI
payload_type: 1 pic_timing
payload_size: 1 0x24
string is "$"
pict_struct: 2
clock_timestamp_flag[0]: 0

-For frame based sequences:
ref 0 type 6 SEI
payload_type: 1 pic_timing
payload_size: 1 0x32
string is "2"
pict_struct: 3
clock_timestamp_flag[0]: 0
clock_timestamp_flag[1]: 0

-Vui timing info:
timing_info_present_flag: 1
num_units_in_tick: 1
time_scale: 50
fixed_frame_scale: 1

[Revgen]

@BetaBoy

I just bought your decoder and the speed is great. I was able to encode a 1280x720@60FPS (Basketball game) .ts file with X264 using just about all the b-frame settings,High Profile, CABAC, 0:0 deblocking, 6 subme, and everything else at defaults and play it back smoothly on my AMD 4600+ Dual Core.

Unfortunately I'm experiencing graphical errors when playing with Zoom Player. When I first start playing the video it plays upside down and the colorspace is screwed up. However, after the video is done playing and Zoom Player restarts it automaticaly again, it looks fine. Fortunately, the video is only 11 secs and it ends and restarts pretty quickly. This wouldn't be good for longer videos.


You can download the file at http://rapidshare.de/files/18504471/luke.mkv.html

[emmel]

Managed to craete a small sample for the interlacing problem:

http://rapidshare.de/files/18508978/hdf.mp4.html

The first 5 secs of the clip play badly, the last 5 secs play well.

[Revgen]

Don't you mean "decode" ;)

No I meant "encode". Just how I wrote it. I was describing the clip that I encoded for playback purposes. Read the paragraph again.

[BetaBoy]

@BetaBoy

I just bought your decoder and the speed is great. I was able to encode a 1280x720@60FPS (Basketball game) .ts file with X264 using just about all the b-frame settings,High Profile, CABAC, 0:0 deblocking, 6 subme, and everything else at defaults and play it back smoothly on my AMD 4600+ Dual Core.

Unfortunately I'm experiencing graphical errors when playing with Zoom Player. When I first start playing the video it plays upside down and the colorspace is screwed up. However, after the video is done playing and Zoom Player restarts it automaticaly again, it looks fine. Fortunately, the video is only 11 secs and it ends and restarts pretty quickly. This wouldn't be good for longer videos.


You can download the file at http://rapidshare.de/files/18504471/luke.mkv.html

Did you try to switch colorspace to see if this has any effect? in teh mean time... i'll download the clip and give it a test.

[Revgen]

I can't switch colorspace. X264 only supports YV12 encodes. The colorspace and the upsidedown picture is fine once Zoomplayer automatically replays it. The problem has to be either my Dual Core CPU, videocard drivers, or the codec itself.

Oh that reminds me.

Video Card: Nvidia GeForce 6800 Ultra w/71.89 drivers.

EDIT:

Okay, I found a solution. Apparently the native overlay colorspace convertor is screwy. So I enabled FFDShow's by choosing Raw Video--> "all supported" in the FFDShow video decoder page. Zoomplayer now uses FFDShows convertor instead. So, now the video no longer looks upside down and no longer has color issues.

[puffpio]

I havnet used CoreAVC since the beta, but has image quality improved? Back then the decoder seemed to be really efficient, but at the cost of image quality. Of particular note was blocking/banding of solid colors or gradients in dark areas that other decoders did not show (ie a flat wall w/ a gradiant of low light...)

[tomos]

from the 1st public alpha/beta that i tried i didnt notice a single problem with image quality versus ffdshow/nero showtime. truth be told, the only probs i have ever had with this decoder is the audio sync part.

can you show some shots of what you've noticed? with coreavc and some other codec to show? i'm curious if i am just not noticing these things :D

[Avish]

@Betaboy, sorry If I'm being rude, but What do I have to do to get a simple reply from u? I posted a simple question here http://forum.doom9.org/showthread.php?p=810623#post810623, but no reply for 2 weeks! So I sent PM to u 2-3 days back, & still havn't got any answer!! So pls tell me what is it that'll bring me your reply?:rolleyes:

Sorry again for the rudeness but I've waited long enough.

[videomixer9]

h264 decoders are supposed to output all exactly the same quality, however most decoders except for ffdshow maybe don't output YV12 and there's plenty of older crappy gfx card that have YV12 problems.

[puffpio]

I havnet used CoreAVC since the beta, but has image quality improved? Back then the decoder seemed to be really efficient, but at the cost of image quality. Of particular note was blocking/banding of solid colors or gradients in dark areas that other decoders did not show (ie a flat wall w/ a gradiant of low light...)
I dont have the sample with me...it was a 1280x720 @ 24fps (ie 24P) encode of an episode of 24 I did as a test using x264 w/ one of Sharktooth's high quality profiles...it was a dark room and the wall of the room was a flat color or had a slight gradiant...

when I played it in tcpmp w/ coreavc I got blocking on the wall...when I played it with ffdshow in media player classic, the wall was smooth....

so not sure...

[CEC]

I've done some tests with MPC+FFDSHOW, MPC+COREAVC, VLC-0.8.6-svn-20060421-0000 and here are some results:

MPC+FFDSHOW = 20~23 CPU :rolleyes:

MPC+COREAVC = 16~19 CPU :p

VLC = 10~13 CPU :cool:


Vlc seems to beat coreavc!!! Can someone else do some tests with vlc??

I have a 3 ghz pentium 4 HT

[MrWizard]

They must have made some serious improvements. Past builds of VLC were outperformed by CoreAVC, at least in my tests. Have you messed around with the various renderers? That can significantly impact the results you get with MPC.

[GmorG McRoth]

I'd be happy to test but I can't realy... I dont have coreavc and I can't find VLC-0.8.6-svn-20060421 best I found was vlc-0.8.5-test3

also what was your test video? was it an movie trailer I could download?

-edit-

Dear lord I just tested vlc-0.8.5-test3 against old Coreavc alpha and difrence was great! with VLC i had 0-1% usage and from mpc with coreavc alpha it was 9-30% WOW

I need to get some higher resolution test sample (heads to apple trailers page)

--edit---
on second test it was 6-9% for VLC I had to mess something up for first test (like watching wrong program cpu usage percentage)

[emmel]

Hmm.. actually yes. At least with 720p the performance is quite equal, maybe vlc is even better. The problem is, that vlc still dies with most of my clips, whereas CoreAVC does not. Well, they are mostly interlaced and tv-originated and less popular formats anyways.

[IgorC]

Guys, VLN uses open source libraries of H.264 decoder which is very slow right now.

Most probably your videocards support H.264 hardware aceleration. That's why VLN doesn't use CPU.

Without hard. ac. it's very slow
http://nightlies.videolan.org/build/win32/

[GmorG McRoth]

I have geforce Fx 5700 no hardware h.264 acceleration there.

[IgorC]

Sure? http://es.nvidia.com/object/winxp_2k_84.21_es.html

Just think. AVC 0-1% . Is it possbile? Even not a misc. doubt that is videocard.

[GmorG McRoth]

Support for high definition H.264 hardware decode acceleration on GeForce 6 and 7 series GPUs.

mine is generation 5... (sob)

resolution of video I tested was low (480x360 25.00fps). so it's possible that audio decoder had influence. I'm downloading high res trailer now.

[IgorC]

look . I have VLN 22d of april and it's slow as hell.

[breez]

IgorC: I don't think the open source decoders in VideoLAN support hardware acceleration at all. But anyway, 0-1% doesn't seem right either.

[GmorG McRoth]

I retested to duble check and it seems I just watched wrong program CPU usage indicator -_- sorry for missinforming. it was 6-9% this time

with cars-tlr_1_h1080p.mov (renamed to mp4 and played) trailer: VLC 26%-43% (with spikes to point of video freeze), MPC+CoreAVC 24%-40% (mostly under 30%)

CPU= prescott 3.0 GHz HT

not so Impressive, seems like I again jumped the gun.

[BetaBoy]

I've done some tests with MPC+FFDSHOW, MPC+COREAVC, VLC-0.8.6-svn-20060421-0000 and here are some results:

MPC+FFDSHOW = 20~23 CPU :rolleyes:

MPC+COREAVC = 16~19 CPU :p

VLC = 10~13 CPU :cool:


Vlc seems to beat coreavc!!! Can someone else do some tests with vlc??

I have a 3 ghz pentium 4 HT

Can you please enighten me with you testing methods, the clip that you used and the version of CoreAVC you tested with please?

[Revgen]

I've done some tests with MPC+FFDSHOW, MPC+COREAVC, VLC-0.8.6-svn-20060421-0000 and here are some results:

MPC+FFDSHOW = 20~23 CPU :rolleyes:

MPC+COREAVC = 16~19 CPU :p

VLC = 10~13 CPU :cool:


Vlc seems to beat coreavc!!! Can someone else do some tests with vlc??

I have a 3 ghz pentium 4 HT

These are strange results.

I have the CoreAVC professional version. I can say with certainty that both VLC and CoreAVC play Standard Definition (720x480, 24fps, 2000kbps, HP,CABAC, all b-frame options) H264 files at about the same CPU usage according to the taskbar.

High Definition (1280x720, 60fps, 6500kbps, HP, CABAC, all b-frame options) files can't hardly be played at all with VLC player but play smooth in CoreAVC by using my Dual-Core CPU.


All I can think of is that this problem may be related to HyperThreading. I remember reading articles talking about how HyperThreading sometimes slows down system performance.

[JoeBG]

I have the CoreAVC professional version.

I bought this version too. But I only got a e-mail with a download for a Test Version. Is this correct?

[sjchmura]

Suddenly FFDShow is crushing the old 0.4 build of coreavc. Did they change it or break it? I need to look carefully but it lookd very very very good

[soresu]

Something is seroiusly wrong here, I have religiously checked my e-mail account since wednesday when I was told that my download link e-mail was being resent, and yet still nothing!

Is there problems with the payment system or something? Inquiring minds NEED to know!

[CEC]

Can you please enighten me with you testing methods, the clip that you used and the version of CoreAVC you tested with please?

I used a movie I have at 688x288 25fps vorbis audio mkv container

The CoreAVC version was 0.0.0.4!!! You haven't released any trial of your latest to test with!!! :sly: But anyway, how much big differnce is going to have from your old alpha version in speed???

The version of vlc I tested can use both virtual cores (HT) of my CPU!!!

[videomixer9]

Something is seroiusly wrong here, I have religiously checked my e-mail account since wednesday when I was told that my download link e-mail was being resent, and yet still nothing!

Is there problems with the payment system or something? Inquiring minds NEED to know!

Sounds like a usual problem with crappy email providers. Btw. CoreAVC domain is still in many blocklists, not only blocklists like ipfilter.dat by several providers but also others :P Seems the IP range was assigned to some weird site before ...

[siddharthagandhi]

How come people with Dual Core processors are having trouble decoding HD content when I can play HD content and encode in HD at the same time using Nero Showtime on my Pentium 4 3.4 ghz processor with HT and my 1.5 GB of Ram. Its not a dual core processor that I have.

[BetaBoy]

I bought this version too. But I only got a e-mail with a download for a Test Version. Is this correct?
The test version is the PRO Edition that we had to manually send emails to.

[BetaBoy]

I used a movie I have at 688x288 25fps vorbis audio mkv container

The CoreAVC version was 0.0.0.4!!! You haven't released any trial of your latest to test with!!! :sly: But anyway, how much big differnce is going to have from your old alpha version in speed???

The version of vlc I tested can use both virtual cores (HT) of my CPU!!!
CEC... thats what I thought... please do NOT post any results for 0.0.4. You are confusing results with the current 1.0 version.

Results w/o HT are about a 10%... add HT and add another 10%. Add SMP and it drops significantly. Results will vary per system.

[BetaBoy]

Sounds like a usual problem with crappy email providers. Btw. CoreAVC domain is still in many blocklists, not only blocklists like ipfilter.dat by several providers but also others :P Seems the IP range was assigned to some weird site before ...
You are correct..... we are in the middle of moving to a DIFF subnet on our host. This will eliminate issues like this.

Anybody not receiving emails with their download can email me betaboy@corecodec.com with their email address and the PayPal transaction ID and i'll handle it.

[nm]

How come people with Dual Core processors are having trouble decoding HD content when I can play HD content and encode in HD at the same time using Nero Showtime on my Pentium 4 3.4 ghz processor with HT and my 1.5 GB of Ram. Its not a dual core processor that I have.
Encoding at the same time doesn't matter much since the encoding process is most likely executed at a much lower priority. Or at least it should be. What kind of HD content have you tested? Your processor should have a hard time with High profile 1080p24, at least with other decoders than CoreAVC. Depending on the player, this can be experienced as A/V sync problems or dropped frames.

Edit: Try this (at 720p and 1080p): http://trailer.nerodigital.com/eng/The_Greatest_Game_Ever_Played.html

[BetaBoy]

Encoding at the same time doesn't matter much since the encoding process is most likely executed at a much lower priority. Or at least it should be. What kind of HD content have you tested? Your processor should have a hard time with High profile 1080p24, at least with other decoders than CoreAVC. Depending on the player, this can be experienced as A/V sync problems or dropped frames.
Speaking of dropped frames.... in 1.1 we are adding a dropped frames option to DS.

[ravemax]

@BetaBoy:
Already an decision with regards to the Linux/Unix release format? E.g. universal, dynamic library with an FFMPEG like interface, or an application (e.g. mplayer) plugin?

[HookedOnTV]

So I tested the VLC vs CoreAVC. On my 1.6 Pentium M I can play the King Kong trailer with Core and it plays smooth at about 95% CPU. Using the latest VLC build dated 4/22 CPU is pegged at 100% and playback is less than real time.

[Revgen]

I used a movie I have at 688x288 25fps vorbis audio mkv container

The CoreAVC version was 0.0.0.4!!! You haven't released any trial of your latest to test with!!! :sly: But anyway, how much big differnce is going to have from your old alpha version in speed???

The version of vlc I tested can use both virtual cores (HT) of my CPU!!!

The speed difference is quite noticable. Especially at HD resoultions using using SMP with my AMD 4600+. The decoder clearly uses both cores when processing these clips. DVD resolution clips aren't going to show much difference though.

CoreAVC 0.0.0.4 doesn't even hold a candle to this new version.

[BetaBoy]

@BetaBoy:
Already an decision with regards to the Linux/Unix release format? E.g. universal, dynamic library with an FFMPEG like interface, or an application (e.g. mplayer) plugin?
Well... let me speak more of a general perspective on the player side, then of our plans for CoreAVC.

We are working on several versions of CorePlayer and BetaPlayer for Linux atm which will feature CoreAVC Professional Edition. Those being GTK, QT, QTopia, JUCE, and a native version for Linux based PMP's utilizing our CoreTheque DB (see http://www.coretheque.com ) and our upcoming SVG interface for CorePlayer 1.xx.

But more specific to your question... we are going to have a more 'Universal' approach for our codecs (CoreMP3, CoreASP, CoreAAC v2, CoreAVC, CoreAC3, CoreDTS, CoreAVS) on Linux with Libraries/Headers and API documentation... but this is more targeted to our much LARGER goal of a possible Open Source alternative to FFMPEG which i'll talk about more over the few months as the encoder side of our codecs matures.

[jblade]

Is it at all likely we will see a codec for the XBMC (Xbox Media Center, which is based off of mplayer) relatively soon?

[starg4ze]

Is it at all likely we will see a codec for the XBMC (Xbox Media Center, which is based off of mplayer) relatively soon?

We are alot of people DYING to have CoreAVC for XBMC...

Hey BetaBoy...Who do I pay? How much? Will you accept my wife as a downpayment?:D

[CobraX]

First of all I wanna say that I hate paying for a "decoder". But I'm considering buying CoreAVC Standard or Pro. I have never used PayPal before but I have a Visa card so this won't be a problem.

I still have a few questions though:

1) Do you get lifetime free updates for the version you buy (like with e.g. DVD Rebuilder Pro)?
2) Do you just download an installer and receive a serial? Or is there a key system like with DVD Rebuilder Pro? I'm asking this cause I want to use it on all of my PCs now and in the near future (for now I only have 2 desktops and 1 laptop).
3) I'm using CoreAVC Alpha 0.0.0.4 now and 1080p trailers from apple.com don't run decently on my Athlon64 3000+@3200+ with 1 GB RAM. Is there a big speed improvement in the 1.0 version?
4) I read something about CoreAVC users getting CoreAAC 2.0 for free? Can someone give me some more info about this?
5) I've also read something about a new account system (for ordering CoreAVC). Maybe I can better wait for this account system to get active before ordering CoreAVC?

Thanks in advance

CobraX

[CEC]

1)no

2)you Download The Installer! It Has Licenturion Protection!

3)yes

4)what More Info? You Get The Latest Version Of Coreaac!

5)there Is No Reason!

[tomos]

hmm, will have to check out this coreaac thing, havent heard much and dont know really what the big deal about aac is :)

Dan,

any guesstimates about the encoder? i am curious as to its potential and how far along you've gotten

[CobraX]

Pfff I've just read a bit about that Licenturion. Those extreme protections with activation stuff etc make it a big hassle and can demotivate possible buyers (like me). I just don't like the fact that you have to connect to some "strange" source (internet) to activate your software (feels like intrusion of my privacy). What if you upgrade your PC every year (or buy a new PC every year and sell your current one to stay up-to-date) and your software will only work with e.g. 3 hardware IDs? Then you're f*cked.

Also I don't like paying 19,99 USD every 6 months for a new version of CoreAVC Pro. I thought the developers didn't do it for the money (but to keep other companies from "stealing" their sourcecode). I guess I was a bit wrong here. :( Maybe I'd be better of buying a faster cpu then. :(

Over and out for now

A bit disappointed CobraX...

[jthunder]

We are alot of people DYING to have CoreAVC for XBMC...

Hey BetaBoy...Who do I pay? How much? Will you accept my wife as a downpayment?:D

I second this motion - this is one of the best uses of such a codec where the machine is cpu limited (I can't wait to find out how high a bitrate we can jam through xbmx)

Please update us on this!!!

Thanks.

[BetaBoy]

We are expecting our suite of encoders by the end of the summer to match our decoders (CoreMP3, CoreASP, CoreAVC).

As stated a few times in this thread... all 1.xx get 1.xx upgrades. Once the Core Account system is up you will be able to upgrade any of our non-Open Source products you have purchased.

2.0 is a way off but we have already been working on the framework for it even before 1.0 was finalized.

XBMC version of CoreAVC will not happen anytime soon as we spoke specifically with the devels of XBMC and were not able to work on a solution.

BTW... your 4 months too late for the 'wife' offer... as I am now on wife version 2.0 . . . thx anyway ;-)

[BetaBoy]

On CoreAAC v2.0.... All current CoreAVC users will gets this as a part of their CoreAVC 1.0 purchase...

We will have two releases for CoreAAC v2.0... 1st release will be a plain jane AAC decoder and be available in about 2 weeks.. 2nd release will be HE/SBR, however this version will not come out till after CorePlayer 1.0 is launched in June.

[remlap]

BBC have began thier HD tests on Astra 2 in DVBviewer of course theres an error needing MBAFF.

Is there away to get this working?

Cheers.

[xkodi]

BBC have began thier HD tests on Astra 2 in DVBviewer of course theres an error needing MBAFF.

Is there away to get this working?

Cheers.

there is also another problem with DVBViewer:

The CoreAVC relies on the resolution reported by the DVBViewer Filter when connecting to it. Obviously it doesn't evaluate the resolution info in the video headers (in contrast to all up-to-date MPEG2 video decoders)...

more here:

http://www.dvbviewer.com/forum/index.php?showtopic=9206&st=75#

and the picture looks strange (HDForum, 11240MHz, Hotbird 13.0E) :

http://img86.imageshack.us/img86/8318/coreavcdvbviwer5jh.th.jpg (http://img86.imageshack.us/my.php?image=coreavcdvbviwer5jh.jpg)

i hope soon there will be a solution to that problem too

[Avish]

Any news about trial version?? :D

[saratoga]

XBMC version of CoreAVC will not happen anytime soon as we spoke specifically with the devels of XBMC and were not able to work on a solution.



What was the issue? GPL licenseing or similar?

A fast decoder for XBMC would be amazing. I've been secretly hopeing it would happen since you guys annouced the beta test.

[starg4ze]

Regarding the XBMC version of CoreAVC and licencemodels.

The important thing is not how the codec is distributed. If there is problem with mplayers/xbmc licensemodel, don't distribute it with it then. Make a package that every user have to install on their xbox by them self. The important thin is that it should not require more work than copying the dll and changing a few rows in a script.

How hard can it be to compile the decoder with the XBOX SDK?:confused:

[Gnerma]

I doubt the problem is technical starg4ze. Reading between betaboy's lines money seems to be the sticking point between the two parties.

[BetaBoy]

I doubt the problem is technical starg4ze. Reading between betaboy's lines money seems to be the sticking point between the two parties.
Its nothing complicated... and it was never about money... and it was on the development side of things.... and to be fair thats all I can really say on it.

[MaxL]

hmm, still no other way to buy CoreAVC but F...... PayPal :(

[Gnerma]

Its nothing complicated... and it was never about money... and it was on the development side of things.... and to be fair thats all I can really say on it.Fair enough. I apologize if I offended you with that comment. So when can pro owners expect an update? Ok ok I'm sorry, you don't have to answer that :D

[ChronoCross]

well since no one has responded to my bugreport on your page hopefully here it will get some attention.

Lossless decoding in coreavc does not work. 2 different samples. one in mkv, one in mp4 different resolutions. they work fine using ffdshow, and nero showtime 3. Hopefully this will be fixed soon.

[BetaBoy]

well since no one has responded to my bugreport on your page hopefully here it will get some attention.

Lossless decoding in coreavc does not work. 2 different samples. one in mkv, one in mp4 different resolutions. they work fine using ffdshow, and nero showtime 3. Hopefully this will be fixed soon.
CC... I saw the bug report on the tracker.... i'll look into it... thx for the report.

[popper]

will someone with the tech skills and knowlage please go this
thead "MPEG2, X264, Xvid HD 1080i - The Lowdown "
http://forum.digital-digest.com/showthread.php?t=65336
and correct these guys posts with the right information and
insight please.

his main source of info seems to be http://en.wikipedia.org/wiki/ and even then makes reference to mpeg4 without explaining the parts2 and 10 and lumping them all together to the conclusion that mpeg2 is best for HD.

im just now becoming wiser with the help of this board and
some BBC tech peoples posts elseware as regards HD PAL and NTSC etc but this thread above just grates on my head that its so wong to to state something as conclusive without reading up on it and especially the fact that he's trying to pass it off as "the LOWDOWN" without what seems much clue
about even the correct names and how they fit together and the timeline, never mind the passing off xvid as a H.264, arrrr

some PAL 50hz HD posts and how they relate to good down conversion would be most welcome to there please.

[Revgen]

First of all, there isn't a "best" codec for HD material. MPEG2 and AVC have their distinct advantages.

Second of all, MPEG2 bit-for-bit video quality generally doesn't measure up to AVC. But, it also doesn't require as much CPU resources, and genrally would be better suited for those who don't have CoreAVC and a Dual-Core CPU. And finally, MPEG2 editing solutions are more widely available than AVC editing solutions, and at less cost. Whether MPEG2 or AVC is "best" is up to you.

[siddharthagandhi]

What about VC-1....you're leaving out what currently is #1 in HD encoding

[ChronoCross]

#1 for commercial solutions. It's certainly not the best codec.

[ryoohki360]

#1 for commercial solutions. It's certainly not the best codec.

Well we don't have acces to that codec. It's the 2005 revision of the original WMVHD Codec, i know that added B frames and 3 pass rendering , it's 'closer' than h264.

[ChronoCross]

I would imagine that x264 will be shown superior come time for testing.

[siddharthagandhi]

Hey BetaBoy I wasn't able to play the HD file you gave me the link to last week (greatest game ever played). Guess I might have to buy your decoder eventually. But....I'll wait a little while, because I have no immediate need to decode HD and there might be good freeware and open source solutions in the future.... :P

[IgorC]

please don't pollute the forum with open source and commercial projects wars.
Five minutes ago you were crazy about to probe VC-1 Microsoft . Commercial project N1 not only here but in the world.

[siddharthagandhi]

i'm not starting a war that was totally friendly comment

thats why i have the smiley and i said "might"

[BetaBoy]

funny... it was ot me who sent you that link.. it was SMD...

[siddharthagandhi]

Oh it was?... lol i forget.....my memory sucks....

[blubberbirne]

@BetaBoy

It the actually version of coreavc ready to work with dvb-s2 cards?
e.g. KNC-ONE DVB-S2?

[BetaBoy]

your referring to TS support and MBAFF interlaced... Haali has indicated his splitter should soon support it and MBAFF will be added in a few weeks to our Pro Edition.

[eLupus]

about the xbmc integration.. i think we and the core devs kinda stumbled on understanding what the other wanted :). As far as i know, there wasn't really any direct contact between devs, only "intermediaries".

Generally our problem with supporting the coreavc decoder, is that the only available one is directshow based, wich kinda is a pain to support under any other operating system than windows. There is a wrapper in the mplayer version that we use, but it's far from complete and doesn't work for CoreAVC directshow filter as it is now. The thought of writing our own emulator for this to use in our alternate player has come to mind, thou has not been done yet.

Registry dependancy is abit complicated too, however can be handled.

A though for the CoreAVC people would be to provide an alternate interface for decoding than using directshow.. Possibly even providing docs on how the coreplayer interfaces with the decoder as I kinda assume that is simpler than ds. On our part a windows precompiled dll is no problem, aslong as it doesn't have gui dependancies in use during init or decoding as those are painfull to emulate. I suppose just creating a new interface for use ith XBMC is more work than what you can expect it to return, however refactoring the code to support more operating system, could definatly increase the return.

Licencing issues shouldn't be any real problem aslong as the interface for the dll is available under some gpl compatible format ( see no reason for it not being free for all, as it only would increase support for your product and increase sales )

We are still very interested in supporting the decoder.

<edit> verified once more and it stumbled on the idea to do some mutual exchange of help. at that time none of our devs had any time or interest in what coreavc had in mind. so i guess we are somewhat to blame here </edit>

/Team-XBMC

[bkman]

I hope you don't mind me asking, Betaboy, but what kind of hardware requirement will the upcoming GPU acceleration have? Any PS2.0 card? Or higher?

[ACrowley]

I hope you don't mind me asking, Betaboy, but what kind of hardware requirement will the upcoming GPU acceleration have? Any PS2.0 card? Or higher?


yeah....and what about "GPU support added later" ?

3Month....6Month, 1 Year ??

[tomos]

right now, i havent seen anyone, anywhere who has this working on nvidia cards :(

[futurex]

you're going to need a card that actually handles H264 acceleration, eg geforce 7 series and ATI X1300 and up

[Revgen]

your referring to TS support and MBAFF interlaced... Haali has indicated his splitter should soon support it and MBAFF will be added in a few weeks to our Pro Edition.

CoreAVC can currently read TS files using the latest version of MPC which comes bundled with it's own TS reader.

[BetaBoy]

Sure.... but to be honest imho its still second to Haali's splitter.

[pankov]

that's good, but when will this splitter be available?
will it be free or it will come only bundled with CoreAVC?

btw
how is the work on the alternative payment methods going? Can we expect it any time soon?

[Shinigami-Sama]

halii has always been free to the best of my recollection

[pankov]

I know that the current version is free and I hope that even this version specially developed with/for Core developers will still be free ... but who knows

[Shinjite]

I am using CoreAVC 1.0.0, and I am using MPC to play my files, normally when I seek, it lags a bit, issit normal?

[GodofaGap]

Yes, this is normal.

Say you have a GOP of 150 frames long and you seek to the 100th frame. The decoder will first have to decode the 99 preceding frames and, of course, you can't expect it to do this instantaneously.

[Shinjite]

Oh okay thanks for explaining bro :)

Oh yeah another question, is the CoreAVC 1.0.0.0 anyhow faster in decoding than the beat version?

[starlabs]

Yes, this is normal.

Say you have a GOP of 150 frames long and you seek to the 100th frame. The decoder will first have to decode the 99 preceding frames and, of course, you can't expect it to do this instantaneously.


I get a lag (well, more of a speedup of playback followed by a possible playback repeat of the prior 2 seconds or so) if I use the Haali splitter.

If I use the Nero splitter, no such issues. This is with Nero-encoded h.264 content of course. I'm assuming Nero knows something about it's own format/streams to "prevent" this?

I'll like to use Haali as I feel it has more features, but this issue (and another regarding aspect ratio that appears on Haali but not on Nero) has me sticking to the Nero splitter...

[Dethis]

Oh okay thanks for explaining bro :)

Oh yeah another question, is the CoreAVC 1.0.0.0 anyhow faster in decoding than the beat version?

OK you could search this thread.. but it is huge

Depends on the case

Coreavc 1.0 is 1%-2% (Betaboy says) faster than previous directshow (2006-13-01) but if your CPU is HyperThread capable its 15% faster.
Coreavc 1.0 Pro can be about 2X or 4X faster on 2 or 4 CPU (cores) systems

Corecodecs TCPMP 0.71-0.72 with avc.plg was 15-20% faster than directshow version (in my case : P4 2400, no HT)

I am looking forward the upcoming CorePlayer to catch the ultimate speed with my current CPU.

[Scoty]

how can use CoreAVC pro with MPC ? ok i have re-register but MPC will only use Haali Video Renderer. when i disabled the intern decoder from MPC i use standart MPG2 Decoder, with enabled i use Haalis Renderer but not the CoreAVC.

[jblade]

about the xbmc integration.. i think we and .......
<edit> verified once more and it stumbled on the idea to do some mutual exchange of help. at that time none of our devs had any time or interest in what coreavc had in mind. so i guess we are somewhat to blame here </edit>

/Team-XBMC

Hopefully we will see chatt continue then between the two. Who knows maybe even eventually, we shall see a similiar quicktime alternative, that will rename the movies to .hdmov so that coreavc will be used on the xbox. It seems that the main problem is the lack of linux support in coreavc, but it seems that very soon that will not be a problem.

[Nick [D]vB]

Hello again,

Sorry to keep on about this but I’m going to give it one last shot. I appreciate you have set an internal development schedule according to your available resources and perceived demands of the marketplace but I just won’t to make sure you are fully aware of the current situation in the UK as I feel there is a potentially huge and, as of yet, untapped market for you here.

I’m sure you are aware the BBC have just started an HD trail across all platforms, it will show a wide range of programming but without question the biggest attraction will be the football World Cup which starts in just under a month. There has been a lot of media coverage of the trail and specifically the fact the major providers have massively underestimated demand for equipment that can receive it. Watch the short national News segments here for the background:

http://news.bbc.co.uk/1/hi/programmes/breakfast/4752017.stm

Basically the only cable provider offering the service, called TeleWest, is currently only available to a small number of homes in the UK. The other major player is $ky satellite, it has enough kit but is having trouble training enough installers in time, and now can not install any systems until after the World Cup has finished. :mad:

If I’m boarding you just bear with me a minute this next bit is crucially important. The BBC had also started a DVB-T trail which would let a large part of London receive the same HD programs through a normal TV aerial. BUT the BBC has chosen to run this as a closed trail [due to bandwidth limitations and other “political issues” with the regulator OfCom] and so has only supplied 500 set-top decoders! The service is FTA and unencrypted but there are no other commercially available AVC DVB-T decoders that can be used to watch the service. :confused:

Because of this a PC based solution seems to be the ONLY option for many people wanting to watch the World Cup in HD this summer, and there are plenty of them! In fact it would be hard to over-estimate the demand here at the moment, I know hundreds of people currently looking for a real-time MBAFF decoding solution in time for the World Cup. According to tests I’ve done with the Elecard and Cyberlink decoders they are just not able to provide acceptable performance even on ultra high-end dual core A64FX based system with ATI X1900’s [using GPU acceleration]. I don’t know about NVIDIA’s acceleration performance but most report it is not working with most products in the latest drivers.

I am sure if you are able to provide a solution in time for the Worlds Cup [even a fuctioning beta release] you would have a massive number of customers, but the timing is critical, with your current development schedule MBAFF support would not be available until well after the World Cup has stated, I don’t want to jinks the team but should England be knocked out of the tournament in the early stages a large part of this potential market would just evaporate over–night. Let’s just hope they are still on form by the time you’ve got something ready! Well I’ve said my piece. I am not trying to tell you how to run your business; I just wanted to make sure you were fully aware of this opportunity. Keep up the good work guys – just go a bit faster… :D

Thanks,
Nick

[Revgen]

Nick,

I believe Betaboy addressed this earlier, but I'll try to explain it.

Core is a very, very, very small company. Right now, all of the demand from the doom9 forum has forced them to have to upgrade their purchasing system. But even that isn't helping. Ask Scoty about it.

I doubt that they would be able to supply decoders for everyone in Britain who has a computer and wants to watch the games.

Then again I might be wrong, but I doubt it.

PS
I think Betaboy mentioned that MBAFF support would come out next month. Doesn't the world cup start next month? If so, then it should be right on time.

[Nick [D]vB]

I believe Betaboy addressed this earlier, but I'll try to explain it.
He has, I don't mean to be a nuisance about it but I really want to make sure Core appreciate the scale of the potential market.

Core is a very, very, very small company. Right now, all of the demand from the doom9 forum has forced them to have to upgrade their purchasing system. But even that isn't helping. Ask Scoty about it.

I doubt that they would be able to supply decoders for everyone in Britain who has a computer and wants to watch the games
If it was really necessary couldn't they find a third-party distributor that could cope with a large demand in the short term?

I think Betaboy mentioned that MBAFF support would come out next month. Doesn't the world cup start next month? If so, then it should be right on time.

I was told it would be over 6 weeks, I'm not holding anyone to that because I suspect it was just an estimate. The world cup starts in about three weeks, to make matters worse the England team just lost one if its star players so weather we are still in the tournament by the time the codec is available is anyone’s guess. Also I think domestic and business customers would both want to see a workable solution before investing in hardware. I really hope they can get something together for us in time, and I know I'm not alone...

[BetaBoy]

Try we will but no promises... We have delayed CorePlayer long enough mostly for codec development. Yes we are small but we are probally gonna grow 4x in the next 2-3 months with all the devel that is going on here.

For Scoty... his issue was the emails flagged as 'spam' but to fix this we will let the user manage his/her account and download any upgrades available and hence why we have opted to devel a 'Core Account' system to do this. We will have no less then 20 products within the next year coming out and this seemed like the most logical step.

No worries as our testing for DVB-H with CorePlayer Mobile will commence the beginning of summer... so MBAFF is a big priority for us.

[fields_g]

We will have no less then 20 products within the next year coming out and this seemed like the most logical step.


Small yet very ambitious. Great Work!

[Nick [D]vB]

Try we will but no promises... so MBAFF is a big priority for us.

Thanks for getting back to me, we are counting on you.

I don't know you have seen this before but it makes very interesting reading!

http://www.behardware.com/news/8117/coreavc-stronger-than-avivo-purevideo.html

:D

[Ice =A=]

@Nick [D]vB: Very interresting indeed, thanks a lot!

[chros]

My guess the ddraw overlay mode didn't kick in or it couldn't use yv12 format. The fallback mode is kind of useless (unoptimized) at the moment. Fullscreen slowing down, means the player uses (fallback) software scaling.
I have no picture in fullscreen-overlay mode with MPC ... (VMR9 renderless works OK.)
Any idea ?
Thanks

[seftel]

so MBAFF is a big priority for us.

London is counting on you :)

Even an alpha release by the ninth would be awesome.

[Revgen]

vB']Thanks for getting back to me, we are counting on you.

I don't know you have seen this before but it makes very interesting reading!

http://www.behardware.com/news/8117/coreavc-stronger-than-avivo-purevideo.html

:D

I believe that this was found out earlier in these forums awhile back. But it's still good to know that somebody else did a test.

[Nick [D]vB]

I realise you guys just might have looked into it before... :D

I was really talking about this part:

ATI indicated us that the decompression engine used by Cyberlink's codec is still the one that was released in December and that in the meantime (4 months), a more efficient version has been integrated in the drivers. As it has a different interface, it can't be exploited by Cyberlink's codec. The older acceleration engine disappeared from the recently released Catalyst 6.4 version (we used the 6.3 for this test), so it means that they will no longer accelerate H.264 decoding. We will have to wait for the Catalyst 6.5 or 6.6 for the new engine to be exploited and/or exploitable. Cyberlink apparently don't want to change their codec, so ATI will have to find a way to use the new engine when the codec calls the previous one.

It looks like Cyberlink have ditched ATI and gotten into bed with NVIDIA, especially when you look at the 7900GTX performance in that test.

Neros acceleration performance is pathetic by comparison so you'd think someone else would want to get in on the act by now?

.