ExtremeTech Wages War of the Codecs 356
prostoalex writes "ExtremeTech tested Windows Media, DivX, QuickTime/Sorenson and QuickTime/MPEG4 codecs. They encoded clips from Matrix Reloaded, Monsters, Inc., X2 and Spider-Man. QuickTime/Sorenson won the encoding speed contest, for the quality tests read the entire review, as each movie sample was encoded with 500KB and 1MB bitrates. Video samples provided on the site as well, so see for yourself."
Dont forget ffmpeg (Score:2, Informative)
No XVid? (Score:4, Informative)
Re:I love Slashdot! (Score:4, Informative)
Moderators, wake up!
If you do check the site you will see that Windows Media didn't win - it was a toss up.
Doom9's Comparison (Score:5, Informative)
doom9.net (Score:2, Informative)
In case you wanted to watch... (Score:5, Informative)
Re:No XVid? (Score:5, Informative)
Re:These encoding algorithms... (Score:5, Informative)
Re:I love Slashdot! (Score:5, Informative)
Terrible reporting - used wrong programs to encode (Score:2, Informative)
Extremetech REALLY blew it.... even in the apple world quicktime pro is known to be a poor ENCODER. The architecture is not the problem, it is the programs... Those beauitiful trailers that are highly compressed are Qicktime, but they are encode3d in Sorenson 3 using another program... It's called "Cleaner" by CreativeMac...
Extremetech REALLY REALLY blew it... I have never had such bad results when i used quicktime pro, (before i asked around how come I couldnt get the amazing detail of the trailers and was told that they're done in Cleaner)....
again, WMA is a codec, Quicktime is an architecture (thus, useing the Sorenson 3 codec)...man, I am firing off a letter to them for incompetence...
Doom9's Comparison (Score:4, Informative)
And yes, Doom9's comparison includes XViD.
Re:Made on a Mac? (Score:4, Informative)
Indeo? What the fuck? (Score:5, Informative)
STUPID! YOU'RE SO STUPID!!!
Re:What's up with MPEG4? (Score:4, Informative)
3ivx, Xvid and divx all postprocess, not unreasonably. The Apple codec makes itself look bad for no good reason.
Re:QTPro doesn't have the best encoders (Score:3, Informative)
Hell there *are* free MPEG-4 encoders that are better than Apple's encoder, e.g. ffmpeg.
Re:No XVid? (Score:2, Informative)
I hope you haven't been thinking you had to step in after 3 hours for too long.
That wasn't nipple (Score:2, Informative)
With nipple you get skin
With pokies you get shirt.
Re:Bah....... (Score:5, Informative)
this page [comicscontinuum.com] has quite a shot, though this is the one [moviecritic.ca] people usually think of, with the webslinger getting an upsidedown kiss.
Re:Well it might be licensing (Score:2, Informative)
The really cool thing about XviD though is that it can be decoded by a "standard" MPEG4 decoder, which means all the modern DVD players with MPEG4 decoders built in can play XviDs by default.
Suprised (Score:2, Informative)
Re:But no Xvid? (Score:3, Informative)
Re:stills vs. motion... (Score:5, Informative)
The article can really give people the wrong idea - it's not the MPEG4 codec, but maybe Apple's implementation that's to blame. Perhaps it just doesn't support all of MPEG4's features. Then again, perhaps the people doing the review just didn't know how to set up the encoder properly. Regardless of codec, there's quite an art to good encoding.
Quicktime != codec (Score:4, Informative)
This misunderstanding doesn't invalidate your argument, although I would disagree with you about MPEG-4. I've gotten good results with it, sometimes even great results.
Re:Made on a Mac? (Score:3, Informative)
There are simply too many video codecs out there for us to test them all -- and most of them wouldn't be useful anyway. We focused on four codecs, all of which are free and can be used with free tools. (Or very cheap ones - QuickTime requires a $30 Pro registration for full encoding capabilities.) You don't want to pay $500 for a professional video authoring program just to send grandma a video of baby's first steps, so we stuck with these four very popular codecs...
The article is testing encoders used by non-professionals, so they aren't going to test something that costs $300.
Inept methodology (Score:1, Informative)
1- Starting with 6Mbps MPEG is *BROKEN*. You are encoding artifacts, even if you can't see them. The only meaningful shootout is to use uncompressed SD which is 270 Mbps. (or HD if you want to get really serious). Codecs see artifacts as detail and waste bits trying to reproduce them. Starting with a source that's natively 270 Mbps and has been compressed to 6mbps (and VBR at that but that's another story) *will* introduce artifacts.
No professional would use a DVD as a source and an amateur would get worse results from even higher bitrate source materials that were shot/edited/color-corrected poorly. Garbage in/Garbage Out. The only reason the encoded hollywood DVD's look acceptable at all is the sources THEY were encoded from were expertly shot, edited, color corrected and tele-cine'd.
2- The Sorenson 3 codec as included in the free QuickTime download (yes free... you don't need QT Pro to export QuickTime, any QuickTime application can expose the export codec) is not capable of 2-pass VBR. The difference in quality using the pro version of the Sorenson codec with VBR (and having a clue) is enormous.
The shootout is not a shootout of codecs. It's a shootout of idiot presets using iffy source material. It's a shootout of tools and approaches appropriate for amateurs.
It completely ignores architectural differences between the formats which, again, for anyone with real content production goals. QuickTime is a vastly richer and more flexible architecture. (and no I'm not on Apple's payroll)
-A.C.
Re:No "fair-use" or editing comparisons are missin (Score:3, Informative)
The codec itself is neutral from any copy protection mechanism, or you just like to yell "DRM" for some cheap mod points.
Re:Not the best evidence. (Score:5, Informative)
Well, if you want a somewhat technical explanation, I would recommend reading This [fastvdo.com] (warning, PDF). Very well written, with enough technical details to satisfy the casually interested geek, while readible enough for non-geeks to get the general idea.
For just the quick-and-dirty... The MPEG4 AVC (aka MPEG4 part 10, aka H.26L aka H.264.10) includes quite a few new techniques at every step of the encoding, from preprocessing to interframe prediction to new frame types to new residual handling methods. These make encoding a lot more CPU intensive, but produce considerably better results (Oddly, most sources claim only 40-50% better than MPEG2, which I find absurd, since even ASP encoders manages to do better than that).
It may help some people to better appreciate the difference by seeing some side-by-side comparisons (not exactly the best possible test conditions, but they make their point)... Balooga [balooga.com] has a brief overview of the MPEG4 AVC vs the ASP and even MPEG2 available... Check out the screen shots, in particular.
Interstingly, on the topic of nomenclature, I think it would make people far less confused if we all called it H.264.10, rather than MPEG4 AVC. Up to and including what we normally think of as MPEG4 (the MPEG4-2 ASP), all the MPEG versions remained backward compatible. An MPEG1 stream counts as a valid MPEG2 stream, and an MPEG2 stream counts as a valid MPEG4-2 ASP stream. The AVC standard, however, departed from that backward compatibility. Not necessarily a bad idea, but by not picking a new name, everyone seem rather confused about exactly which names refer to which standards (similar to USB2, but worse, because each version has several sub-versions).
Re:But no Xvid? (Score:2, Informative)
Mac OS X server has Quicktime streaming built in, and it's damn easy. But you want it for windows or linux? That's ok, because it's free [apple.com].
Terrible, useless article (Score:5, Informative)
Some fundamental errors:
They're using MPEG-2 sources, which are already highly compressed (this has been amply covered by other posters).
They talk about converting to an "uncompressed" AVI, but never specify which flavor of uncompressed. They should have used a lossless codec that uses the native Y'CbCr color space of video, like Huffyuv. They way they just said "uncompressed" suggests they used the AVI "None" codec, which is uncompressed RGB. This causes two lossly color space conversions - one from the Y'CbCr of the source to RGB, and then back to Y'CbCr in the delivery codec.
They used Indeo 5.1 as their intermediate codec. This is terrible. Indeo uses what's called YUV-9 sampling. There is only one measurement of color per 4x4 block of pixels. This throws away 75% of the color information from the DVD (which uses 4:2:0 sampling, with 2x2 blocks), before it even touches a codec. And this results in very ugly blocks whenever there are highly saturated regions with sharp contrast. So, all the output is going to look highly compressed when rendered from these intermediates, even if further compression is lossless. Look at the Spider Man test frame for an example. Notice the red blooming around the shoulders of the vocalist. And the color everywhere is very muddled. Indeo can also be slow to decode, unless it was encoded with all keyframes. And how slow it is to decode will vary with the tool, which probably added measurable error to their encoding time measurements.
They don't know the difference between Sorenson Video 3, which comes free with QuickTime, and Sorenson Video 3.3 Professional, which you have to pay for and is what Apple uses for their movie trailers. With the Pro version, critical features like B-frames and 2-pass VBR are available.
Apple's MPEG-4 encoder isn't very good - 1-pass only, tuned for speed more than quality. A file with the exact same compatibility could be made with Squeeze, Compression Master, Envivio, etcetera with MUCH better quality. And the Divx MPEG-4 codec is, of course, also MPEG-4.
They didn't use 2-pass encoding! No quality-concious encoder would ever put content on spinning disc without using 2-pass. And they didn't mention most of the other encoding settings they used, which by context I'd guess were basic defaults.
That's from an initial skim. If I spent more time with the article.
In summary, these guys spent hours and hours analyzing the results of tests, where they would have been WAY better off spending an hour asking someone who knew anything about video compression how to administer this kind of test.
Oddly enough, their results are vaguely like you'd expect - WMV9 and DivX do well, Sorenson less so, and Apple MPEG-4 at the rear. Done properly, I imagine WMV9 would have had a slight lead, and Sorenson 3 Pro would have been a lot closer to DivX. And no one uses Apple's MPEG-4 codec for content distribution. QuickTime's decoder is fine, so folks would use a professional-grade MPEG-4 encoder instead.
Re:The opposite is true (Score:3, Informative)
Bitrate limited is great if you care about file size or bitrate more than quality.
Each has its place. Real-time streaming obviously has a hard requirement for the latter.
And VP6! (Score:3, Informative)
VP3 was the one that was open-souces, and is used as the basis of Ogg Theora.
The current On2 codec is VP6, which is free for personal use.
Re:Give it some time (Score:3, Informative)
Most MPEG-4 professionals would use something like Squeeze or Compression Master instead to make a
Taxonomy of MPEG-4 (Score:3, Informative)
QuickTime encodes and decodes Simple Profile MPEG-4
DivX did Simple in V4, and V5 added support for Advanced Simple.
Most of this will be moot soon, since the MPEG-4 Part 10/AVC/H.264 codec is way better than the old Simple or Advanced Simple, and will rapidly replace the old versions in the next couple of years.
what about ogg? (Score:3, Informative)
All hail ogg!
Slightly OT... (Score:2, Informative)
Re:what about ogg? (Score:3, Informative)
Are you talking about using Ogg Vorbis as the audio codec? Yes, it is very good. I wouldn't use anything else for either my CD rips or DVD rips.
Or are you talking about the bastardisation of the Ogg container format that is the OGM container format? Do some googling. From the mailing list postings I saw, the Ogg guys aren't too happy about this effort by one windows programmer to hack the Avi/VfW information into the Ogg container format. If that's what you're referring to, and using, I recommend you instead look at the Matroska [matroska.org] container format. It's much more flexible and is slightly more efficient space-wise than OGM. Mplayer [mplayerhq.hu] supports it, don't know about Xine [sf.net]. There's a Matroska splitter/demuxer thingy for windows [sourceforge.net], don't know about Mac OS/X support.