HCenc vs QuEnc

[vinib]

Hello,
I'm using FAVC (by the way, nice stuff...) and I was wondering which encoder would you recommend and also how are different these 2 encoders?

many thanks
vini

[blutach]

Do a test using both and let your eyes decide. Refer rule 12.

Regards

[gizzin]

I say HC probably yields the best visual results, make sure you use AQ or luminace.

[video_magic]

At tending towards low-medium bitrates (around 3000) i.e. putting 3 hours on a DVD with a little menu, I preferred the visual quality of HC Encoder, and I think it's superior as a program as well as the DVD compliance of its' output - but do your own tests for visual preference.

I would shy away from the 'Luminance Gain' option in HC, standalone compatibility.

[vinib]

Do a test using both and let your eyes decide. Refer rule 12.

Regards That's what I intended to do, but also I wanted to get users comments/advices, but I will do of course tests. By the way, should I do it on small part of the movie (just to shorten the process) or on the whole movie? thanks

[blutach]

Just find a bit of video that you consider challenging and encode it at an average bitrate.

Regards

[Dark Shikari]

I've found HCEnc to, in general, be surprisingly good. I even have a test case where it outperforms a ~2003-generation H.264-like codec quality-wise...

[IgorC]

If you mean VP6 as ~2003-generation H.264-like codec. Its efficiency is good at low bitrate only. VP6 aplies very strong smooth filter to gain compressibility even with the most weak deblock settings.

However things could changed since then as there were rumors about 'optimizied 2008' version of VP6.

[Dark Shikari]

If you mean VP6 as ~2003-generation H.264-like codec.I'm referring to VP7, actually, in a low-bitrate setting. Also, VP6 is not H.264-like, it is ASP-like.

[IgorC]

Well, VP7 was released during end of 2004 and early 2005.
The VP6's specifications are very close to H.263+/++. Something between ASP and AVC.

[Dark Shikari]

Well, VP7 was released during end of 2004 and early 2005.
The VP6's specifications are very close to H.263+/++. Something between ASP and AVC.I don't really consider anything that doesn't support arithmetic coding to be AVC-generation.

(But if a company claims their format beats H.264, that automatically lumps it into AVC generation, which they probably won't like when the codec comparisons come out ;) )

[IgorC]

I'm triyng to say when people said that one codec is better (quality wise) than other it's also good to know if codec was designed to compete with other at given bitrate.

Let's do some parallels with audio compression. It's clear example. HE-AAC. It was designed to be a good codec up to 80 kbits stereo. Ok, it's right to say that mp3 (MPEG-1 layer 3) is bettet than HE-AAC at 128 kbit/s. It's true.
Would be it correct conclusion? "Look, MP3 is better than bleeding edge HE-AAC at 128 kbits"

Of course On2 will never say that their codecs disminuish efficiency as bitrate gets hgiher.

[Dark Shikari]

I'm triyng to say when people said that one codec is better (quality wise) than other it's also good to know if codec was designed to compete with other at given bitrate.That was my point as well. That's why I put Xvid, DivX, WMV3, and VP6 in the same category, just as I put VP7, H.264, RV40, and SVQ3 in the same category.

[IgorC]

I don't really consider anything that doesn't support arithmetic coding to be AVC-generation.

Well, On2 isn't MPEG organization. It's company wich has their own technology. People just compare VP7 with AVC as their were developed at the same period. Of course VP7 isn't AVC.
But VP7 and H.264 was on par in quality terms during some short period of 1 half of 2005.

[Dark Shikari]

Well, On2 isn't MPEG organization. It's company wich has their own technology. People just compare VP7 with AVC as their were developed at the same period. Of course VP7 isn't AVC.I'm rather curious how much or little they ripped off from H.264, given almost all similar proprietary formats of that generation (RV40, SVQ3, etc) were H.264 ripoffs.
But VP7 and H.264 was on par in quality terms during some short period of 1 half of 2005.To get this back on topic and explain why I find that doubtful:

(note: VERY unscientific test. VP7 and HCenc on best settings, though I didn't go tweaking HCenc's matrices or anything like that. VP7 used 600-frame gop, HCenc used 18-frame gop, since HCenc doesn't allow ridiculously large gops like 600. Both 3 megabits per second, 2pass, on 60fps input video.)

VP7
http://i33.tinypic.com/w4l7d.png

HCenc
http://i38.tinypic.com/34phs1y.png

AQ is almost surely HCenc's saving grace in this case, which is IMO why it is such a good MPEG-2 encoder, and why I'd happily recommend it over anything else on the market, at least that I've seen.

It'd be nice if it was open source though... I don't doubt x264 could probably borrow a thing or two from it.

[IgorC]

That was my point as well. That's why I put Xvid, DivX, WMV3, and VP6 in the same category, just as I put VP7, H.264, RV40, and SVQ3 in the same category.
It's absolutely right. But my point was bitrate category, not generation category. Vp6/7 aren't good at high bitrate. That's why even MPEG2 at enough high bitrate can be on par or even better. On2 will never admit that but they designed their codec for 'enough high quality for limited bandwith'.

[IgorC]

Looking at these screenshots I can only say that both codecs have serious distortions.
VP7 has serious colour issues and blurs too much. MPEG-2 has separate monotone scale blocks and it's even hard to call it artifacts. Just blocks.