Elecard Converter Studio 2.1

[IgorC]

Elecard 2.1
http://www.elecard.com/products/products-pc/consumer/converter-studio/

Changelog:
http://www.elecard.com/pdf/converter_studio/EConverterStudio_RN.pdf

Elecard Converter 2.0
+Film grain optimization

[DeathTheSheep]

Could this have the same core as the mainconcept reference AVC?

[Sharktooth]

Q: does it support CQMs?

[DeathTheSheep]

Oh my goodness, they finally got constant quantizer mode!! And fully adjustable quantizers... and "insane" modes, and HADAMARD transform, and no less than 3 AQ modes, 3 quant modes... Wow, what a treat. (If only it was free :D).

Doesn't look like there's CQM support yet. Oh well.

...and it looks like you can't specify GOP sizes either. That's the deal breaker for me... lots of flickery flickery quality-jumping in random places... :rolleyes:

[IgorC]

...and it looks like you can't specify GOP sizes either.
It's possible to adjust max gop size. The rest is a job of encoder. And the results are really good.

[DeathTheSheep]

Oh it is possible to set max gop? I didn't see it then. Never mind, I must not have looked hard enough then. :) Where is this option located?

[Trahald]

'advanced' button while you are in the 3rd tab

[DeathTheSheep]

Actually that's where I had been looking at first to no success.

I found it though--above the advanced button above the B-frames slider. XD

Gee I wish there was no logo, then SSIM/PSNR would take on a whole new meaning. As would bitrate comparisons...the extra bitrate in that thing is really a shame, lol.

[Jay Bee]

Great piece of software, love the interface and encoding quality. Now if only the price were more realistic for private users. And the single core limitation seems a bit silly for such a hardware intensive task as AVC encoding.

Two more things. MBAFF still doesn't deinterlace properly and 4:3 uses 16:15 PAR which is not compatible to HD Optical. (12:11 is).

[Dark Shikari]

Actually that's where I had been looking at first to no success.

I found it though--above the advanced button above the B-frames slider. XD

Gee I wish there was no logo, then SSIM/PSNR would take on a whole new meaning. As would bitrate comparisons...the extra bitrate in that thing is really a shame, lol.Letterbox your source video, that makes SSIM/PSNR possible because the logo will be on the black area of the video (and its bitrate use is negligable).

[DeathTheSheep]

Or I could just buy it, right? It's just pocket change, after all! :D

[Sergey A. Sablin]

Two more things. MBAFF still doesn't deinterlace properly and 4:3 uses 16:15 PAR which is not compatible to HD Optical. (12:11 is).

I guess MBAFF deinterlacing isn't the Converter Studio issue.
And could you please describe a bit about PAR/SAR problem? Resolution? HDDVD or BD?

Sergey.

[Jay Bee]

I guess MBAFF deinterlacing isn't the Converter Studio issue.
And could you please describe a bit about PAR/SAR problem? Resolution? HDDVD or BD?

Sergey.

MBAFF problem is described here (http://forum.doom9.org/showpost.php?p=972299&postcount=2637).

And PAR problem from here (http://forum.doom9.org/showthread.php?p=1047890#post1047890) onwards.

[SealTooGreat]

What is x264's cmd parameters for HSS rate (bps), CPB size (bits) and initial delays?

--vbv-maxrate, --vbv-bufsize and --vbv-init, respectivelly???

[iwod]

How does this compare to x264 in terms of quality?

[IgorC]

How does this compare to x264 in terms of quality?
At fast or medium settings MC is better. For insane settings x264 (subme 7 trellis umh and other rdo sets) is better but the results are enough close.

[Paddy97]

I bought the update and I am quite found of the new version.

The only problem I have had so far is that it do not work well when reading the video source from a DVD. It starts to encode and then quits rather quickly leaving a small fragment of the encoded file ( as far as it got in the encode ? ).

Have anybody else been able to use a DVD as source (ofcourse non CSS ) ?

[Sergey A. Sablin]

I bought the update and I am quite found of the new version.

The only problem I have had so far is that it do not work well when reading the video source from a DVD. It starts to encode and then quits rather quickly leaving a small fragment of the encoded file ( as far as it got in the encode ? ).

Have anybody else been able to use a DVD as source (ofcourse non CSS ) ?

Paddy97, you can contact techsupport with description of the issue, if you didn't it yet. We're working on the new update with some features, so we'd be glad to fix any problems current version have before next release.

[Paddy97]

Thanks Sergey.

I have now sent in a report through the web based form on the homepage. I also saw that there was a new version released. I will contact Elecard Sales as well to get the new download link.

[bond]

+Film grain optimizationis this fgm or some backwards compatible "optimization"?

if its fgm, can someone please post a sample. :thanks:

[Sergey A. Sablin]

is this fgm or some backwards compatible "optimization"?

if its fgm, can someone please post a sample. :thanks:

no, it's not fgm. it is hvs optimizations.

[RealTelstar]

Could this have the same core as the mainconcept reference AVC?

Looking at the versions yes, it seems so. It has the same limitations:
http://www.elecard.com/products/compare_table.php?product_id=converter-studio

So the "studio" seems to use the mainconcept h.264/avc while the "studio pro" the pro codec.

[bond]

no, it's not fgm. it is hvs optimizations.thanks :)

[CruNcher]

Yep overall something similiar is existing now for Avisynth as a Pre Processing Filter and should work for every codec called Grainoptimizer by Dark_Shikari, at least it should help x264 alot with it's partition decission to not drop every bit of grain away and preserve a reduced but visible version of the source grain layer even @ very low bitrates :).

[Sagittaire]

Yep overall something similiar is existing now for Avisynth as a Pre Processing Filter and should work for every codec called Grainoptimizer by Dark_Shikari, at least it should help x264 alot with it's partition decission to not drop every bit of grain away and preserve a reduced but visible version of the source grain layer even @ very low bitrates :).

Film grain optimization for Elecard/Mainconcept core encoder is not pre-process filter ... it's just particular way encoding to retain grain.

[CruNcher]

yes sure but it makes no difference Dark Shikaris Grainoptimizer could also be implemented as it is this way in x264 directly as a psy option so basicly it's the same thing a HVS optimization only Mainconcepts may work different allocating more bits maybe into the background directly, but how it works isn't really important @ all more then how it looks and feels in the final result :).

[Sagittaire]

yes sure but it makes no difference Dark Shikaris Grainoptimizer could also be implemented as it is this way in x264 directly as a psy option so basicly it's the same thing a HVS optimization only Mainconcepts may work different allocating more bits maybe into the background directly, but how it works isn't really important @ all more then how it looks and feels in the final result :).

1) Well in fact Dark filter is just temporal filter for better grain retention. it's not a pure encoding optimisation for better grain retention. This filter like all the other filter can potentialy produce major source modification and IMO should not be implemented in x264 encoder.

2) Film grain optimisation for Elecard is purely a coding optimisation (particular motion search mode, partition search mode, rdo search mode ... ???) and don't affect directly the source.

[CruNcher]

i agree that it's better to modify the motion estimation directly also the partition distribution for that very special case as grain layer preservation (HVS) but x264 isn't that far yet and Grainoptimizer seems to be a nice workarround for now (from what you hear), and indeed in my last tests it doesn't showed the results i hoped for but that was with Atemes Encoder not x264.

[Paddy97]

My problem with it having problems reading DVD as source has been fixed.

I recieved very promptly a new version when contacting tech support and I have now ran several test runs and it works flawlessly.

Thanks Elecard for fast action.

[IgorC]

Update.
Elecard 2.1
http://www.elecard.com/products/products-pc/consumer/converter-studio/

Changelog:
http://www.elecard.com/pdf/converter_studio/EConverterStudio_RN.pdf

[CruNcher]

Did they fixed the Extreme Memory allocation or made the memory usage a little more sane or even user choseable?
Im not sure if it's really efficient to allocate so much memory as converter studio does @ Encoding especialy with .AVS scripts that allready need alot of memory themselves. Other Gui based Encoders and Even Mainconcepts Reference Encoder are more sane with that :)

[NickP]

I try to encode avs script in converter stuido and in GraphEdit. Converter Studio allocate 25Mb more then graph edit, it is not so much.
CruNcher, could you tell me, how much memory converter allocates in your test?

It is my avs script:
Source=Mpeg2Source("lord1.d2v")
source=Crop(source,0,64,-0,-64)
source=LanczosResize(source,720,304)
Return(source)

[MuTeK]

Elecard Converter Studio 2.2 : Call For Testers (http://www.elecard.com/cs-tester.php)

[Sharktooth]

nice. just applied :)

[MuTeK]

Elecard Releases Fourth Generation Codecs and SDK

New Codebase to Surpass Digital Video Experts’ Expectations

TOMSK, RUSSIA – April 2, 2008 – Elecard, a leading developer of codecs and products for digital video, today announced the release of the fourth generation of its highly acclaimed video and audio compression codecs and codec SDK for MicrosoftTM WindowsTM.

Established in 1988, Elecard released the first set of codecs under its brand name. Later, the company developed the second generation known as Moonlight codecs, with IP rights retained by Moonlight Cordless Ltd. and its successors. The third generation was developed with MainConcept AG, which retained IP rights to most of the codecs.

“The three previous generations were among the best available, and are still in use by industry leaders in a variety of digital video products,” said Andrey Posdnyakov, president of Elecard. “In-house tests prove that the fourth generation is even better, which is no wonder as it is powered by almost 20 years of experience in codec development.”

The Elecard 4G codec line includes:

H.264 AVC, MPEG-2, MPEG-4 Part 2, DV50 and DV100 video encoders
AAC, MPEG-2 and MP3 audio encoders
H.264 AVC, MPEG-4 Part 2, DV50 and DV100 video decoders AAC audio decoder
MPEG-2 PS/TS multiplexer and demultiplexer
All of the above codecs are based on completely new engines. The Elecard product line also includes MPEG-2 video, MPEG-2 and MP3 audio decoders based on the third generation engines.

http://elecard.ru/news/news_arch.php?nid=161

ps Converter Studio 2.2 based on 4G codecs.

[Sharktooth]

yup, im beta-testing CS2.2 and i have to say the quality and speed are pretty damn good (even if the 2pass and CQ modes are not available yet). however i think CS2.2 does not expose all the encoder options...

[DeathTheSheep]

I share your reasoning, Sharktooth, about the quality and the speed, as far as I could see. Though I have used CQ mode, which works fine. You just have to override the quantizers in the advanced tab. Oh, and don't encode to AVI container in this build (not that you would anyway, but it's quite a bug).

However, the Elecard logo is still present on the encoded video, making it a bit harder to test the low-resolution handheld-type content that way, especially using metrics (directly).

[Sharktooth]

oh, this is CS2.2 beta testing... not the elecard AVC codec beta testing...
so i guess they're interested more in the application... raw metrics numbers are for the codec developers :)

[DeathTheSheep]

One could say it's really both:

ps Converter Studio 2.2 based on 4G codecs.

And metrics numbers are for everybody--They can't test all source material, so when anomalies present themselves, it's a good reporting tool to have at one's disposal.

[Dark Shikari]

However, the Elecard logo is still present on the encoded video, making it a bit harder to test the low-resolution handheld-type content that way, especially using metrics (directly).Solution: Pad the top of the video by 80 pixels of black, then crop on playback. Bit cost is nearly zero for the encoder to do this, since its a multiple of macroblock size.

[DeathTheSheep]

Haha, I remember you mentioned that before. Maybe even in the same words. :o

It's true, that works, but it does require more work, li'l more processing, and a synthetic playback step (the extra crop). It'd be nice to know I'm dealing with the codec and playing the files as they were meant to function, rather than beating around the bush. Perhaps most importantly, the mobile devices I'd like to test on don't have a crop-on-playback, and to test quality properly on them, I want all the bits where they should be, not in a flickering logo.

[rica]

Hi,
I'm the customer of
ECS 1*** version
ECS2****version
XMuxerPro
Elecard AVC Plug-in

Some notes for Elecard guys:
I'm happy with ECS since it is the fastest AVC encoder.
But Elecard still insists not to understand the HA revolution .
HA capable interlaced ready profiles are still missing.
MC h264 decoder is not ready for HD HA for months. (still it is April 07 version)
I encode with Elecard's AVC plug-in's decoder but watch with Cyberlink's decoder. It is funny.
I told this before at Elecard's forum months ago but nothing has changed yet.

[kuka2]

But Elecard still insists not to understand the HA revolution .
HA capable interlaced ready profiles are still missing.

Just interested what does it mean HD HA?

[Sharktooth]

Hardware Assisted (decoding)?

[kuka2]

Hardware Assisted (decoding)?

Maybe, but I don't think so. From other posts by rica - it is something concerning interlace/deinterlace. Maybe field based encoding and 50/60 fps progressive playback?
There is only one meaning for HD HA I found in internet - camera makers use it for HD wide angle cameras.

[nm]

This post from him suggests Hardware Assisted (or accelerated):
http://forum.doom9.org/member.php?u=146589

So, he seems to be complaining about the interlaced profiles not working properly on NVIDIA/AMD hardware and that the DxVA support in MainConcept's decoder does not work either.

[rica]

Hi guys,
I meant PureVideo HD (or AVIVO HD) /( HD HA = High Definition Hardware Acceleration)
HW based decoding and deinterlacing don't work for HD with
Elecard's MC h264 AVC decoder (coming with Elecard AVC Plug-in)