Divx 5.1 found marginally better than WMP 9
andApple's MPEG-4 implementation is absolutely awful. It shows plenty of artifacts even in relatively easy, low-motion scenes, and turns into a mess of blocks when the action gets even the least bit intense.
+ Reply to Thread
Results 1 to 10 of 10
by encoding them to very high bitrate (around 8 megabit) files using Indeo 5.1 compression, as all our test applications could easily read this format. It was at this stage that we reduced the resolution (using a bicubic filter) to create the quarter-resolution master clips for the low resolution/bitrate encoding tests.
Indeo Video version 5.1 had bugs in it and was not fully XP compatable .. they should have at least used 5.2 ....
better yet they should have used a lossless compression .... because they have introduced artifacts for sure.....
better even yet would have been to test right from the mpeg file --
those guys don't come here i guess to learn a few things ..."Each problem that I solved became a rule which served afterwards to solve other problems." - Rene Descartes (1596-1650)
Originally Posted by BJ_MHope is the trap the world sets for you every night when you go to sleep and the only reason you have to get up in the morning is the hope that this day, things will get better... But they never do, do they?
BJ_M is right to question their methodology and conclusions..
I just posted so others could see what they did..
But BJ_M gets to the heart of the matter
"THEY DIDN'T COME HERE" to find out as they shoulda
I read that article and was wondering if they'd stacked the deck by using bad settings on the Quicktime mp4 test. I like mpeg4, seems like the wave of the future for video and I thought Quicktime had a good implementation in Quicktime 6.5 Pro.
So I made my own test clip from a good (not great) quality DVD and reviewd the uncompressed video for obvious artifacts. Didn't see much of anything, so I used all three mpeg4 compressors I have, QT 6.5 Pro, XviD (Koepi's latest build) and Nero Recode.
I set the bit rate to 1meg per sec on all. Used two pass where possible, highest quality. (QT didn't offer two pass, though it may have been hidden in the advanced settings). Kept the screen size the same too.
Well, I have to say the Nero Recode and XviD results were very good, almost undetectable artifacts that occured with a section of sky that had a small gradient from top to bottom during a crossfade. No blocking, just a sort of very faint contouring and scintillation for the second or so of the crossfade. If I had to pick a 'best' I'd say Nero Recode had a slight edge.
The Quicktime was another story. Took twice as long to compress and there was very obvious contouring and scintillation with some small blocking. Not nearly as bad as what ExtremeTech's shootout showed, though. But enough that I sent Apple's Quicktime people a rocket about the poor performance. I don't really expect them to pay attention to it, but at least it's something.
The clip was still very watchable and I had to really look for the problems. So I think MPEG4 is still a good compressor, and it's not proprietary like DivX, Micro$loth's WMV9 or Sorenson3, just Apple's implemenation needs work.
DivX, Micro$loth's WMV9 and xvid are all mpeg4 types ....
in reality -- wmv9 is extremly good ...
for low bit rate -- Sorenson3 pro (pro has many extra settings that make it really shine) is still one of the best .. certainly better than apples mpeg4 by a long shot ...
also -- apples dv codec isnt the best one out there either ... not the worse , but not one of the better ones"Each problem that I solved became a rule which served afterwards to solve other problems." - Rene Descartes (1596-1650)
Originally Posted by BJ_M
But, most of the choices are for converting DV (QT imported?) to various quality levels of MPEG2. But no mention, that I've seen, as to quality of import of the DV - and export back to tape.
It would make it seem the import/export process doesn't degrade the image quality.
Mike"Dare to be Stupid!" - Wierd Al Yankovic
Tomshardware re-does this test every so often. I trust them to keep it all even. XviD 'won' last month. But in reality, XviD, DivX 5.x, WMV9 and RM9 are all MPEG4. WMV9 can be incredible, just takes forever to encode :PTo Be, Or, Not To Be, That, Is The Gazorgan Plan
Originally Posted by CrustyCurmudgeon
I think the people at MPEG LA would beg to differ with you.
Just because it got approved as a "standard" doesn't mean there aren't licensing fees to be paid.
That is why the other versions of "MPEG4" are out there, to reduce the fees to be paid by using "home-grown" alternatives - and hopefully, charge their own licensing fees - like Divx now licensing its' codec for implementation in stand-alone DVD-players, and M$ trying to get WMV9 approved as a "standard" for future digital and HD-TV formats.
It's the reason Apple charges for the MPEG2 import/edit capabities, even after you've paid for the upgrade to QT Pro - fees to MPEG LA for each codec (included in the price of DVD SP and FCPro)
Edited: to remove my needless "history" of MPEG4 development - as if I have an insider's view of things!!"Dare to be Stupid!" - Wierd Al Yankovic
I would say divx/xvid are supported on hardware playback devices and wmv9 support will come when M$ stumps up the money, in siz to twelve months.Corned beef is now made to a higher standard than at any time in history.
The electronic components of the power part adopted a lot of Rubycons.