CoreCodec/H.264 Codec "CoreAVC"

[nm]

1) Where did you get the 90% info from? From our demographic studies you are way off... that is unless your talking about open source and consumer, but even then your % is not even close. In any case pls posts facts not speculations.
2) You are wrong again.... If you have not followed this thread or the last release (which at times is alot too follow) we support out of range MV. There was one post from one persons saying he had an issue with non-standard BV with MV, but never a) sent info on the sample b) never stated a title c) never stated what was used to encode it with. So without more info...
I think that with "out-specs" Dark Eiri was referring to the Blu-ray/HD DVD levels that most x264 encodes don't follow, not to the out of range MV problem. If you read his post again, you may notice that he was arguing against HW decoding and in favor of CoreAVC ;)

[Sasovics]

BetaBoy,

Any ETA for the new release fixing our AviSynth problem ??

[Shinigami-Sama]

I think that with "out-specs" Dark Eiri was referring to the Blu-ray/HD DVD levels that most x264 encodes don't follow, not to the out of range MV problem. If you read his post again, you may notice that he was arguing against HW decoding and in favor of CoreAVC ;)

yeah I think he was talking about the level 4.1 profile error, when the buffer(VBV?) gets out of range

[BetaBoy]

BetaBoy,

Any ETA for the new release fixing our AviSynth problem ??

Next release.

[BetaBoy]

I think that with "out-specs" Dark Eiri was referring to the Blu-ray/HD DVD levels that most x264 encodes don't follow, not to the out of range MV problem. If you read his post again, you may notice that he was arguing against HW decoding and in favor of CoreAVC ;)

Re-read it... gotcha... thx ;-)

[Dark Eiri]

1) Where did you get the 90% info from? From our demographic studies you are way off... that is unless your talking about open source and consumer, but even then your % is not even close. In any case pls posts facts not speculations.
2) You are wrong again.... If you have not followed this thread or the last release (which at times is alot too follow) we support out of range MV. There was one post from one persons saying he had an issue with non-standard BV with MV, but never a) sent info on the sample b) never stated a title c) never stated what was used to encode it with. So without more info...

I was talking about the HD3870 card.

[ADude]

Any ETA for the new release fixing our AviSynth problem ??
Several weeks after the last AviSynth user gives up on reading this thread. ;)

[Digiface]

Should i use any deinterlacing with CoreAVC? I watch avc/h264 encoded videos with it. I mean is there any need to use deinterlacing?

[Sharktooth]

sorry but your question is a bit of a nonsense... what you mean?

[Digiface]

Ok, let me ask this way: Why i should use DI? Is there any advantages to use it? What if i don't use it?

[nm]

You should deinterlace only if the video is interlaced, which is not very likely unless you have a TV tuner card and receive H.264 channels.

[Digiface]

Ok. Thanks.

[molitar]

Just got a brand new 512MB Radeon 3870 video card.. Now I just want to know when can we expect GPU support? Currently I am not even using CoreAVC but MPC Home Cinema version with H264 DXVA.

[HowlerX]

You mean you are using a FREE media player with WORKING hardware-accelerated GPU decoding of h.264 content? I use MPC Home Cinema all the time. You mean to tell me that a few MONTHS since the last release of this great update to Media Player Classic, that the dude behind it has added GPU-decoding of h.264? Still free? All in a few months of work?!

Gee, maybe CoreAVC should hire the guy...

[Rectal Prolapse]

No please don't! Then MPC-HC DXVA will no longer be supported! :P :)

Besides, I like the idea of CoreAVC being software-only - h/w accel still has lots of issues with non-compliant streams.

[Sharktooth]

MPC HC doesnt do any h.264 decoding. it supports EVR and DXVA 2.0 interface.
in case you have a decoder which supports HW acceleration, MPC HC can "talk" to it. Just that.

[MatMaul]

MPC HC doesnt do any h.264 decoding. it supports EVR and DXVA 2.0 interface.
in case you have a decoder which supports HW acceleration, MPC HC can "talk" to it. Just that.
not anymore, casimir also wrote a DXVA decoder ;)

the last beta version can use hardware acceleration with its own internal filters if your card support full hardware decoding (VC1 and H264 for ATI cards, H264 only for nvidia cards)

[Sharktooth]

link?

EDIT: Nevermind, found it. it uses ffmpeg.

[BetaBoy]

Related to CoreAVC... we have just released CorePlayer Pro for Windows and OS X. It has CoreAVC Professional integrated into it as well as all our other decoders (CoreASP, CoreAAC v2.0, etc.).

This initial release is NOT for everyone... its not MPC, it not WPM, Zoom, PowerDVD and the like... no way even close. However like us, we think that you might find that CorePlayer Pro is likely the most efficient desktop media player out there and will continue to grow as the entire mobile and in this case desktop platform grows. So with that there are a few things to mention to for those that might want to give it a spin...

- Our new CoreAC3 filter will be added in next Thursdays version 3/6.
- We will add a filter properties access page in 2-3 releases
- CoreAVC Pro does not feature deinterlacing with this initial release
- CoreASP post processing is not included at this time
- The CoreTheque Media Library interface is changing by v1.2.5 to make it less 'mobile' like.
- No mouseovers support
- No subtitle support till a v1.3.x release
- We will add an external directshow interface in a later 1.2.x version
- We will add YouTube support (mobile) into CE/Windows Mobile/Win32/OS X in next Thursdays version 3/6... sure its mobile and the quality of many of those videos are not that great but hey, its not Flash video either!

We have however included our advanced networking stack udp/rtp/rtsp/sdp/unicast/multicast/http tunneling/rtcp (keep alive only) in the current release (in 2-3 releases we will also include our new UPNP additions with DLNA to follow soon afterwards). We have also tested our new streaming component against, Darwin, Live555, VLC, MS, Real, Shoutcast, and Icecast servers. However since this is our initial release we still have a few things that need to be added including proxy authentication.

I know I will be asked... Do we plan on releasing a Linux version? Short answer is yes. Our OEM version of CorePlayer Pro is already available on Linux using QT, Qtopia, GTK, CoreUI, and X11 just around the corner. (The big project for us on Linux atm is our new native CorePlayerAPI (similar to GStreamer but we include all the codecs). So how does this relate to our consumer version on Linux? Alot actually... as we will use the current Windows and OS X versions to integrate more features before choosing which UI to use for the public (although CoreUI and x11 are looking to be the winners there).

More info here:
http://www.corecodec.com/forums/index.php?topic=758.0

[lexor]

We have however included our advanced networking stack udp/rtp/rtsp/rtcp (keep alive only) in the current release and in 2-3 releases we will also include our new UPNP additions with DLNA to follow soon afterwards.

Is that DLNA two way like WMP11, which can serve content to the network or is it play back only?

[BetaBoy]

Yes, DLNA will be as well as our current UPNP stack in CP is both client and server.

[popper]

Related to CoreAVC... we have just released CorePlayer Pro for Windows and OS X. It has CoreAVC Professional integrated into it as well as all our other decoders (CoreASP, CoreAAC v2.0, etc.).

This initial release is NOT for everyone... its not MPC, it not WPM, Zoom, PowerDVD and the like... no way even close. However like us, we think that you might find that CorePlayer Pro is likely the most efficient desktop media player out there and will continue to grow as the entire mobile and in this case desktop platform grows. So with that there are a few things to mention to for those that might want to give it a spin...

- Our new CoreAC3 filter will be added in next Thursdays version 3/6.
- We will add a filter properties access page in 2-3 releases
- CoreAVC Pro does not feature deinterlacing with this initial release
- CoreASP post processing is not included at this time
- The CoreTheque Media Library interface is changing by v1.2.5 to make it less 'mobile' like.
- No mouseovers support
- No subtitle support till a v1.3.x release
- We will add an external directshow interface in a later 1.2.x version
- We will add YouTube support (mobile) into CE/Windows Mobile/Win32/OS X in next Thursdays version 3/6... sure its mobile and the quality of many of those videos are not that great but hey, its not Flash video either!

We have however included our advanced networking stack udp/rtp/rtsp/sdp/unicast/multicast/http tunneling/rtcp (keep alive only) in the current release (in 2-3 releases we will also include our new UPNP additions with DLNA to follow soon afterwards). We have also tested our new streaming component against, Darwin, Live555, VLC, MS, Real, Shoutcast, and Icecast servers. However since this is our initial release we still have a few things that need to be added including proxy authentication.

I know I will be asked... Do we plan on releasing a Linux version? Short answer is yes. Our OEM version of CorePlayer Pro is already available on Linux using QT, Qtopia, GTK, CoreUI, and X11 just around the corner. (The big project for us on Linux atm is our new native CorePlayerAPI (similar to GStreamer but we include all the codecs). So how does this relate to our consumer version on Linux? Alot actually... as we will use the current Windows and OS X versions to integrate more features before choosing which UI to use for the public (although CoreUI and x11 are looking to be the winners there).

More info here:
http://www.corecodec.com/forums/index.php?topic=758.0

very nice BetaBoy.

but you still dont compile for the PowerPC Linux do you, and so cant use it on the likes of the PPC based PS3 while running for instance Gentoo PPC linux (with Lu-Zero VLC etc Altivec optimisations) can we?

regarding the UI, id like to use rebol/View http://www.rebol.net/ to interface with any HTML/TCP:IP/UDP based IO


ideally, id like to see a combined Multicast tunnel and streaming video app in one unit (so as to bypass the worlds ISPs refusal to re-activate the upload bandwidth saving IP Multicast/IP broadcast on all their kit to and from the end users) with Multicast P2p DHT (bambooDHT) doing the hard interconnecting work, and rebol view scripting interfacing to it all and making your End user prefered GUI on what ever OS you like.

[BetaBoy]

Man popper... you went 20 levels passed what I thought anyone would reply to on this thread ;-)

Rebol.... nice concept I had studied that a few years ago and on how info is stored... not sure it would work with our current framework however.

PPC Linux? Not gonna happen ;-( Noting however that CPP on OS X is optimized for both Intel and Altivec.

Not to get to OT again... but We are working on a P2P component in CP that will obfuscate past ISP's blocks that works in conjunction with our 'Seek and Share' technology. This links individual CorePlayer/CoreAccount's... but this is not planned to be released till later this year.

bambooDHT is intresting especially when handling the hashes... and perms... but from a legal perspective I have read that it is flawed on liabilities. Plus its written is Java :-P

[popper]

LOL, only 20 levels , must try harder :)

well you must remember BB, iv been advocating PPC (thats PowerPC remember) Multicasting and tunnels to bypass the inept ISPs deactivation of generic Multicast for a long time (remember the old MBONE network).

but from a company and long term profits/Market share POV , it seems someone will take my points seriously one day (as the Unicast streaming grows and kills the net throughput)and make a finantial killing in these markets.

lets take the PPC based PS3, its now the defacto HD platform in the worlds homes, now that HD-DVD is being sidelined.

so theres a MASSIVE potential for a Codec/Video app vendor to make some good profits in that ever growing HD market.

you obviously have your reasons not to compile coreAVC and the other codecs and apps for that near untouched PPC market.

i hope that you do take my point about the PS3 as the HD platform of choice today for many people, as its cheap, can be software (firmware)re-programed with any Codec should the need arise, and many people want and are trying PPC Linux on it every day to expand its abilitys even more ?

next: the Multicast part and the most important and long needed ability in my mind at least.

and as for 'obfuscate past ISP's blocks' then why mess around from a commercial POV, just build for the long term future, and make an Encypted Muticasting tunnel and throw all the content through that client/server or P2p model from day one, sidestepping in one leap all your competitors.

its clear when you look at the likes of the UK BBC Unicast Iplayer reports coming out that the ISPs are not liking its bandwidth growth, not to mention your partners again, Unicast JOOST contibution to the masive one to one Unicast traffic.

its clear with the new Docsis3 (bronze) Multicast, and later this year, the (silver) Qos Multicast in the compulsary spec ,plus the BT 21 ST network also mandating Multicast as a generic option that even the worlds ISP will sometime soon ( but alas not soon enough)need to finally take advantage of Multicast capabilitys.

its always been my stance that some company or Open source vendor will one day make a very good name and profits for themselves if they are the first to bring the worlds end users this required Multicasting streaming video.

they need something thats easy to use, yet saves them lots of their precious upload bandwidth and still make it effective for the their sharing of their data streams, Multicast and (tempory, while the ISPs get their acts together, if ever)tunnels are the clear way forward ,all the better if its P2p Multicast DHT in some form.

Unicasting video is so last decade, Multicasting and tunnels to push it through if needs be over the interweb will save vast amounts of data throughput when more than a single user per stream is connected.

we are there on AVC/H264 codec, we are nearly there with the old Unicast P2p , we need the next step and thats multicast swarms.

as for Rebol View , as iv told you before ,go ask carl the creator of Rebol (and Amiga OS) if he's interested in working on a simple small experimental (commercial ?)collaberation, the new rebol R3 is in the process of being built, see the API, and you might find it to your commercial advantage if you simply ask him and his team.

we dont mind you making a profit betaBoy, as long as we get some real innovation and the things we want to pay for, its a real shame your not thinking of PPC compiling though, all those PS3s going begging for a good and effective codec base and related PPC apps.

guess its up to the likes of Lu_zero and the other Alivec coders, to continue the work of optimising the FFMpeg and other open code bases , AND thanks you very much lads for your dedication and commitment.

[madshi]

Our new CoreAC3 filter will be added in next Thursdays version 3/6.
Maybe slightly out of topic. But can you tell us a bit more about this CoreAC3 filter?

(1) Will it be available as a standalone DirectShow filter usable in e.g. MPC/ZoomPlayer etc?
(2) Will it support E-AC3? If yes is 7.1 decoding supported?
(3) Does it output the same data as Dolby's reference decoder?

Same questions about your CoreDTS filter:

(1) Will it be available as a standalone DirectShow filter usable in e.g. MPC/ZoomPlayer etc?
(2) Will it support DTS-HD High Resolution and DTS-HD Master Audio? If yes is 7.1 decoding supported?
(3) Does it output the same data as DTS' reference decoder?

Thanks! :)

[Selur]

Note that today we...
- Fixed: AVCHD M2TS blockiness
- Fixed: DS deinterlacing
- Fixed: INI for missing crop1088 and the value
Any ideas when a new version with this fixes will be released?
(a quick release for cebit would be nice)

[BetaBoy]

Any ideas when a new version with this fixes will be released?
(a quick release for cebit would be nice)

We'll know more on Sunday night as we are planning working on 1.7 some more over the weekend.

[BetaBoy]

(1) Will it be available as a standalone DirectShow filter usable in e.g. MPC/ZoomPlayer etc?
A: Later on yes

(2) Will it support E-AC3? If yes is 7.1 decoding supported?
A: Next phase adds support for "E".

(3) Does it output the same data as Dolby's reference decoder?
A: It has too ;-) but as with all our decoders we coded this frm scratch. Our initial testing shows it better in many ways compared liba52 as well as weak is some areas like volume levels, but there is still plenty of room for ASM optmizations and speed improvements.

I cannot comment on CoreDTS atm....

Also... we have listened to the feedback and it apparent that releasing AC3 without digital output support for suck for alot of ppl.... so we are going to push out adding AC3 now in Pro to support this feature as well as add our filter properties now rather then later.

[dimzon]

Who need CoreAC3/CoreDTS since we have free AC3Filter and ffdshow? I think speed is not so critical for AC3/DTS like for AVC.
So, I really doesn't understand why to waste your time for such development, sorry.

[BetaBoy]

Our decoders are internal towards our CorePlayer efforts atm.... but as I stated last year ago we do plan on seling 'all' our decoders as a directshow bundle. I agree however, that for many liba52/AC3Filter will be good enough.

I would disagree with the performance as a waste of time even as it relates to AC3/DTS (you ever see the spec for E-AC3?). Bundle the performance of Audio with Video and Player Platform and its the reason why CorePlayer Pro is at times 30% faster then VLC. Not caring about performance for even 'one' codec or component would not match our overall objectives.

[dimzon]

I would disagree with the performance as a waste of time even as it relates to AC3/DTS
I think AC3/DTS encoding doesn't eat more then 1-2 % comparing to 1080p AVC decoding. And I doesn't think You can improve AC3/DTS decoding performance more than 30%. In this case overall decoding speed gain (for 1080p AVC + AC3) will be (0.3 - 0.5 %). I believe such optimization is waste of time at all (sorry).

you ever see the spec for E-AC3?
By the way, I really don't understand current E-AC3 isteria. I believe AC3 @ 640kbps provide transparent sound...

Bundle the performance of Audio with Video and Player Platform and its the reason why CorePlayer Pro is at times 30% faster then VLC.
Dont' take me wrong, I prefer to spent my 50$ for better HW (by better CPU) and achieve same performance (or even better) than by your player. Cheap CoreAVC Basic/Pro is good compromise, but complete player for 50$ is not. Just think about it.

100$ - CPU AMD Athlon 64 X2 5000+ (ADA/ADO5000*) Socket AM2 OEM dual core
150$ - CPU AMD Athlon 64 X2 6000+ (ADA/ADX6000IAA6*) Socket AM2 OEM dual core

[Inventive Software]

Bear in mind, AFAIK CorePlayer is aimed at the portable devices market, PDAs and the like. So for them, a fast, clean player is essential. :)

[dimzon]

Bear in mind, AFAIK CorePlayer is aimed at the portable devices market, PDAs and the like. So for them, a fast, clean player is essential. :)
I'm talking about DirectShow bundle and Core Player for windows

[CruNcher]

dimzon how about users that want to use this on their Laptops that can't be updated neither GFX wise nor CPU i think Core Products are a nice solution and the price of every solution i saw today is fair but granted alot of work has still to go into CorePlayers current UI, currently it's the pure horror usability wise for a Desktop Application and also what they did with the Icons (giving icons that are used for years in other Windows UIs another meaning) is wrong. But else it is a nice coded App with a startup memory usage behaviour alot of Windows Players nowdays can dream of (including MPC), and there will also be a OpenSource version soon www.betaplayer.com wich for sure will be enhanced also in different directions like MPC :).

[madshi]

I would disagree with the performance as a waste of time even as it relates to AC3/DTS (you ever see the spec for E-AC3?). Bundle the performance of Audio with Video and Player Platform and its the reason why CorePlayer Pro is at times 30% faster then VLC. Not caring about performance for even 'one' codec or component would not match our overall objectives.
One good target for you might be DTS-HD Master Audio decoding. The Sonic DTS-HD decoder is *PAINFULLY* slow with decoding DTS-HD Master Audio. There you could really make a difference.

I'm talking about DirectShow bundle and Core Player for windows
But if they need to write the decoders for their portable platform, anyway, why not making them available on the PC platform, too? Shouldn't cost them much time and some people might value such a decoder collection. If you don't like the idea, nobody forces you to use/buy the DirectShow bundle... ;)

By the way, I really don't understand current E-AC3 isteria. I believe AC3 @ 640kbps provide transparent sound...
If you have a source which has an E-AC3 audio track, how are you planning to play it? It's not about whether E-AC3 or AC3 is better. It's just about being able to play the audio tracks in their original format. Of course you can transcode E-AC3 tracks to AC3. But you wouldn't have any benefit from doing that. Directly decoding the original audio format is simpler and better.

[nm]

But i don't see that use DXVA and i also interesting in full deinterlace support , maybe when these features are included , it would be more interesting.
Nowadays it's best to leave deinterlacing for the graphics card on Windows. DirectShow decoders just need to ask for it and I think CoreAVC has supported that for a long time already.

MPC Home Cinema should have working DXVA support for the latest GPU generations, so you don't necessarily need to buy filters for that. However, CoreAVC is still quite useful for H.264 streams that can't be properly decoded with current hardware solutions.

[Yo]

and there will also be a OpenSource version soon www.betaplayer.com

Is that a joke or what? What is your definition of "soon"? How many decades? Centuries?

When Coreplayer first came out (I forget now exactly when, but more than a year ago), the open-source Betaplayer was supposed to follow very soon. It has never been released. Go to the URL above, and it says nothing, other than the egotistical ad about "never be the same". No impending release date.

Also, when Coreplayer Mobile first came out, it was said, on the Corecodec forums, by Corecodec management, that a trial version would be released shortly. Also, more than a year later, no trial version has been released. (How difficult could it be, to slap a 30 day restriction on an app?) Even though, just about every other commercial mobile app (look at sites like handango.com, for instance) has a freely downloadable trial version, "try before you buy".

Why is it taking so long for a trial version of Coreplayer to be released? Could it be that they are afraid, that if people try it, they will decide NOT to buy it? I have read many reports on forums, of people who have bought Coreplayer, had problems with it, not liked it, and gone back to TCPMP (definitely an excellent product), or gone to a different commercial mobile media player. If Core has confidence in their products, why still no trial versions, of any of their products?

I'll stick with TCPMP.

[molitar]

My surprise is they have no GPU support yet someone can write a DXVA for MPC and make it work. Why can't Core get GPU support out now?

[BetaBoy]

Yo... keep it OT... If you goto the CoreCodec Forums you know about our plans for BP and why there is no trial available atm for CP but will be once we add all the mainstream features into it (we have said this since day 1). Same with CoreAVC... as soon as we get the next release out we finally consider it 'feature complete' (as we move towards our 2.0 release) for the mainstream and will then release a trial version... and 'yes' you can hold me to that.

[CiNcH]

My surprise is they have no GPU support yet someone can write a DXVA for MPC and make it work. Why can't Core get GPU support out now?

A standalone DirectShow filter is provided too. So use this one instead of CoreAVC...

[Disabled]

My surprise is they have no GPU support yet someone can write a DXVA for MPC and make it work. Why can't Core get GPU support out now?

MPC cant do DXVA with any card besides the very last generation of Nvidia (>8000 series, but NOT 8800 GTS(320/640)/GTX) and ATI cards (HD series). (Correct me if I'm missing something).

With those cards its pretty simple to do DXVA, as you just have to pass the bitstream to the GPU and thats it. And it only works with streams encoded to Level 4.1.

Even if Core would add GPU like MPC does, there would be no added reason to buy Core, as you would get exactly the same acceleration as with MPC.

(I hope this post made sense)

[Yo]

Yo... keep it OT... If you goto the CoreCodec Forums you know about our plans for BP and why there is no trial available atm for CP but will be once we add all the mainstream features into it (we have said this since day 1). Same with CoreAVC... as soon as we get the next release out we finally consider it 'feature complete' (as we move towards our 2.0 release) for the mainstream and will then release a trial version... and 'yes' you can hold me to that.

It's been a while since I've looked at the Corecodec forums, but why don't you give the reasons here--for the absence of the new Betaplayer, and the lack of a trial version for Coreplayer, both of which were promised to be available soon, well over a year ago.

I don't know why you would wait until you have "added all the mainstream features" to Coreplayer, before you have a trial version. If Coreplayer is not really ready yet, then shouldn't it be beta, available for free? However, since you are selling it commercially (and have done so for some time), shouldn't a trial version (time-limited, not crippled) be available now, of the current version, so that users can "try before they buy", as is available with EVERY OTHER COMMERCIAL MOBILE APP?

Why would you need to wait until "all features are added into CP", before releasing a trial version of the currently available commercial version? Software is always under development, it is never perfect, new features are always being added. But people should be able to try out your product RIGHT NOW, with its current features, to decide if they want to purchase it now or not.

Therefore, although I have been a longtime user of TCPMP, and the original Betaplayer before that, and highly admire and am appreciative of Picard's work on those projects (is he still with you?), and would have therefore been interested in your newer development of Coreplayer, I have not purchased it, because I would not want to do so without trying it first. I'm sure I am not alone in that, you have probably lost many potential sales, due to not having a trial version available. And apparently, from comments of people who did go ahead and purchase Coreplayer on blind faith, many regret that they did, some even regressing to TCPMP.

Perhaps, if Coreplayer really isn't ready yet, it should still be in beta, rather than selling it, while not providing a trial version (fearing that people who try it won't want to buy it).

[molitar]

MPC cant do DXVA with any card besides the very last generation of Nvidia (>8000 series, but NOT 8800 GTS(320/640)/GTX) and ATI cards (HD series). (Correct me if I'm missing something).

With those cards its pretty simple to do DXVA, as you just have to pass the bitstream to the GPU and thats it. And it only works with streams encoded to Level 4.1.

Even if Core would add GPU like MPC does, there would be no added reason to buy Core, as you would get exactly the same acceleration as with MPC.

(I hope this post made sense)

No GPU support does work with 8800 just fine too. Problem is most use ffdshow and you can't have any other filter working for DXVA. Now their would still be a reason to use CoreCodec if it had GPU support and that is because it has some leniancy built in so even if the format is not 100% proper it will still play the file.

[BetaBoy]

Yo... I'm not gonna waste my time on your continued OT rants and attacks... Bring some CoreAVC talk or move it over to the CC forums... Moderator anyone?

[Inventive Software]

If memory serves, there's a CoreAVC rant thread I started up a couple years ago....

Found it! About 18 months ago, I started this:

http://forum.doom9.org/showthread.php?t=118775

About a year since the last post. Not bad considering I started the thread and the issue at the time was activation! :D

If a mod wants to split non-topical / ranting / unhelpful posts to that thread, feel free, I'm sure CC employees would rather their tech support thread was populated with tech issues, and not moaning about no GPU support. :)

[grumpy]

Yo... keep it OT... If you goto the CoreCodec Forums you know about our plans for BP and why there is no trial available atm for CP but will be once we add all the mainstream features into it (we have said this since day 1). Same with CoreAVC... as soon as we get the next release out we finally consider it 'feature complete' (as we move towards our 2.0 release) for the mainstream and will then release a trial version... and 'yes' you can hold me to that.

could you point to that thread. I've been searching CC forum for 40 minutes and haven't come across the discussion your talking about.

[bond]

guys, be nice to each other and use the ranting thread!!! ;)

[BetaBoy]

Thx bond.... grumpy, on the CoreAVC discussions... the trial discussions are spread over the past few dozen pages. ATM... we just fixed some 'core' bug fixes for AVCHD and some .M2TS related ones... but as I stated the last big issue is the registry connectivity for pin output before we release the trail.

As far as anything else... go to CC and we can talk over there.

[madshi]

With those cards its pretty simple to do DXVA, as you just have to pass the bitstream to the GPU and thats it. And it only works with streams encoded to Level 4.1.
And the image quality is not as good as CoreAVC's software decoding, I've been told (haven't really compared myself yet, though). I hope that CoreAVC's hardware decoding support will be implemented in a way which doesn't change image quality compared to software decoding. I guess that maybe the hardware decoding of the latest ATI/NVidia cards have cut a few corners to make things easier? Don't know...

@BetaBoy, do you have an idea why image quality with ATI/NVidia's hardware decoding looks slightly different (worse?) than your software decoder? Do you think that you will be able to add hardware acceleration without losing any image quality?

Thanks!

[Shinigami-Sama]

And the image quality is not as good as CoreAVC's software decoding
lol whut?

a decoder will output the same image as the next
its just DXVA doesn't allow for post processing