Slashdot Mirror
Quicktime 5 vs. Everybody?
Dean Siren asks: "Now that Apple has released the Quicktime 5 Public Preview, they've upgraded their video codec to Sorenson 3 from Sorenson 2 which Jeremy Neish says pales in comparison to MPEG4 and its derivatives. So, how well does Sorenson 3 compete against MPEG4 as far as quality per bitrate, processor demand, compression times, and streaming are concerned? Can Apple now stay ahead of DivX, Windows Media Player 7 and Real Player 8?" Neish notes that none of the latest codecs are not compatible with Unix so only Mac and Windows users get to take advantage of the latest in digital video formats. When someone gets around to fixing this, then I can again get excited about Quicktime, until then it's just another format to me (and yet another reason why I have to have a Windows box lying around).
quicktime is an *architecture* far more complex than just codecs and audio and video apple is very active in mpeg4 development even if apple can't get it out the door for a while, there are plenty of 3rd parties that can do a codec that will plug in - with the new autodownloading components in qt5, this makes installing third party components (skins, codecs, etc) automatic, secure, and painless. there will be linux someday - not this year probably next (understandably, osx is steve's primary new os porting target right now - after that, they will do linux - it's just a matter of time and priority)
"None of the latest formats are not compatible with UNIX so..."
:-).
Sorry, I didn't catch that, is it "None of the latest formats are compatible," or "The latest formats are not compatible," or is it REALLY supposed to be "None of the latest formats are not compatible with UNIX?"
See, that would come as a surprise to me since MS's format sure as hell isn't UNIX friendly, Real's new format sure as hell isn't UNIX friendly (unless I missed something), and the only one that is (somewhat) is DivX
(actually, I obviously love to keep harping on this, or I wouldn't.)
Apple is not now, nor have they ever been, a friend of the open source movement. Were they a friend, they would at least release a binary only, closed source version of Quicktime for Linux. Better still, a binary-only plug-in for Xanim. Best, they would release the Sorenson decoder source (the real magic of these things is not in the decoder, but in the ENCODER: figuring out what data to eliminate is the hard work, reconstructing it is relatively easy.)
What has Apple done for the open source or free software movements? They've released yet another kernel for PPC. Nice, but we already have several OSS and free software kernels out there.
The next time Apple tries to milk the movements for free publicity, stand up and call them on it.
www.eFax.com are spammers
... you can play DivX;) on a linux box. IA32 only, but hey. It's called "Avifile" and "Xtheater". Go check it out, it rocks the house...
Xtheater.sourceforge.net
-----
I am under the impression that Apple can't release source to their Sorenson codec. I think they license it from somewhere.
Which of course begs the question, if it isn't as good as MPEG4, then why don't they add MPEG4?
I'll see your senator, and I'll raise you two judges.
Apple have an exclusive license from Sorenson to use the codec. This means that when you ask Sorenson about it, they say "Sorry, we'd love to sell (not give) you a license, but our hands are tied - Apple have an exclusive license. Go bug them about it."
While I don't know the exact details of the license, I'm pretty sure it prevents Apple from releasing any source. Even if it doesn't prevent them, Apple have made it quite clear that they have no intentions of ever releasing the source. The license almost certainly forbids Apple from relicensing the technology, so it's unlikely they could allow any third-party (such as the Xanim author) to develop an implementation of the codec.
Hell, they're not even interested in a totally non-free binary-only port. There's definately nothing (legally) stopping them from doing this, it'd just be another platform which QT5 supports in addition to the usual suspects. It'd make good business sense, too, since Apple are (essentially) fighting a codec war with MPEG4 and RealVideo, and having the free software movement on their side would help them immeasurably. Instead, people using free software platforms can't even view Sorenson encoded stuff, which means they'll just pass it over as an option.
Quidquid latine dictum sit, altum viditur.
Given that MPEG4 isn't finalized as a standard, there's really no reason for Apple to release it. Since the Quicktime framework allows for Apple to easily add new codecs to it (and MPEG4 uses Quicktime as a model) it will be possible (and very likely) that when the standard is finalized, Quicktime will support it.
Quicktime on Mac already handles MPEG1, MPEG2 support is included in QT 5. If it offers significant improvements over the Sorenson Codecs MPEG4 could be used as the encoding instead of Sorenson and made usable on all platforms that QT is supported on (QT5 might run on Linux by the time the standard is finalized).
Just a personal opinion, but I have my reasons. I'm still running a K6-200 (with 64 megs), and Quicktime under Win95 is practically unusable. The player takes inordinately long to load, and all but the lowest resolutions are extremely choppy.
Quicktime was ahead of it's time, and normally I'd give it the benefit of the doubt and just say it's time I bought a new system. What's unacceptable, however, is the fact that MediaPlayer plays Quicktime files (the ones it can play at all) with better quality and half the overhead of the Quicktime player. There's just no excuse for that sort of bloat (and we're comparing it to a Microsoft product here).
As I said, Apple is not a friend of Open Source/Free Software.
www.eFax.com are spammers
Regardless of the grammatical correctness discussed elsewhere, I'm curious as to how a codec can be incompatible with Unix. Not currently implemented for Unix, yes, but incompatible, no. It's just a stream of bytes at the end of the day -- and that's *very* compatible with Unix.
"The invisible and the non-existent look very much alike." -- Delos B. McKown
my point however is that QT also works fully on Mac OS X (I know I work with it) which is based on BSD and the Mach kernel. With the libraries available and working, I really think (and hope) that they will work with Linux some day. The format doesn't even need to be published as long as the libs are around. you make QT do all the work of loading movies and playing them. An app just needs to make calls to the library and it just works.
I speak as someone who does some development with QT on the Mac (as well as tyring to get a Linux project going). Here's hoping!
-- DuckWing
Because last I heard, QuickTime was the starting point for MPEG4. Check it out:
//w ww.info.apple.com/pr/press.releases/1998/q2/980211 .pr.rel.iso.html
http:
And...
http://www.cselt.it/ mpe g/standards/mpeg-4/mpeg-4.htm#E11E10
----
----
Am I the only one who thinks Microsoft is a misnomer? Perhaps Macrosoft would be a better fit?
Hell, they're not even interested in a totally non-free binary-only port.
... having the free software movement on their side would help them immeasurably.
Sure they are. They just don't see that the benefits match the costs.
The equation will balance a bit more once the OS X port is finished, but with no demand from the authoring side and just a couple million non-revenue-producing whiners on the playback side, it's reeeeeeeeeaaaaallllly hard for anyone with a clue to see how an argument convincing to a shareholder could be made that it would be worthwhile even then.
It'd make good business sense, too,
If you actually have a quantifiable argument of sufficient quality to convince the Board of Directors, I can have it on Frank Casanova's desk within the day, and he can have it in front of the Board within a week.
So put up or shut up.
I think there is a little confusion on what a codec is. MPEG-4 is NOT a codec, it is an architecture (like QuickTime and to a much lesser degree Windows Media). An architecture controls everything from file formats to blitting algorithms to sound resampling to rich media support. A video codec turns uncompressed bitmaps into a binary stream and back again. It's an important part of the whole web video experience, but certainly less than half the work.
There are PLENTY of MPEG-4 codecs, including several in development for QuickTime, and several different implementations for Windows Media. However, the MPEG-4 codec is only a small part of the full MPEG-4 spec and file format, which also includes audio codecs, client/server negotiation, packet recovery and retransmission, rich media support like text tracks and clickable hotspots, and lots of other groovy stuff.
XAnim may someday have MPEG-4 codec support, but that doesn't mean it'd be able to play any but the simplest MPEG-4 files, and then only via local playback or TCP. MPEG-4 file playback is probably an order of magnitude more complex than the entierly of XAnim today, with its primative support for lossless binary streams of linear media.
This is HARD stuff people, which is why teams of experts spend years designing these things. We're talking many engineer years to do even a basic implementation or port.
I'm not sure I see how releasing a binary-only closed-source version of anything would make someone more a friend of the open-source movement. They have released a good deal of OSS, and it's a little disingenuous to carp that they're not a friend of the movement because they haven't released the particular bit that *you'd* like to see. By your argument SGI isn't an open-source contributor because they've only released XFS and I'd really like to see CXFS. Get it straight - that concerns me, or you, but not the open-source movement.
Releasing a binary-only version for (presumably) x86 Linux would only help x86 Linux users, like (presumably) you. It wouldn't help Linux users on other platforms, or users of other x86 platforms, and would generally fly in the face of one of the main points of the open-source movement - platform independence. It would make you happier as a whiny seeker of free-beer software for the platform you happen to be using, but I personally couldn't care less about that, and neither would "the open-source movement" (whomever that is).
-- Life is short. Forgive quickly. Kiss slowly. ~ Robert Doisneau
The biggest problem I've seen with how the MPEG group works is that it allows patented technologies into the standard without requiring (as the JPEG group requires) a royalty-free license from the patent owner to use the technology for implement the standard.
<O
( \
XPlay Tetris On Drugs!
Will I retire or break 10K?
Regardless of the grammatical correctness discussed elsewhere, I'm curious as to how a codec can be incompatible with Unix. Not currently implemented for Unix, yes, but incompatible, no. It's just a stream of bytes at the end of the day -- and that's *very* compatible with Unix.
Any stream of bytes is compatible with the UNIX® components that correspond to GNU Fileutils (`chgrp', `chmod', `chown', `cp', `dd', `df', `dir', `dircolors', `du', `install', `ln', `ls', `mkdir', `mkfifo', `mknod', `mv', `rm', `rmdir', `sync', `touch', and `vdir'). On the other hand, GNU Binutils (`ar', `c++filt', `demangle', `gas', `gprof', `ld', `nlmconv', `nm', `objcopy', `objdump', `ranlib', `size', `strings', and `strip') and GNU libc are the packages that really matter for building and running programs on a GNU/Linux system, and they can't handle Win32 very well ;-)
<O
( \
XPlay Tetris On Drugs!
Will I retire or break 10K?
Now, whether Apple will do the legwork to port the libs
The libs in question are mostly the Carbon libraries. The Carbon API is a reimplementation of the Toolbox API of MacOS 1 through 9; it's one of the things you pay $2,000 for when you buy a Macintosh® computer.
Fat chance Apple will port it.
<O
( \
XPlay Tetris On Drugs!
Will I retire or break 10K?