Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: .Ogg Vorbis aotuv (Read 493755 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

.Ogg Vorbis aotuv

Reply #300 –
Quote from: Aoyumi on
Supplement:
The aoTuV beta 5.7 DLL in my page does not need additional DLLs.
This depends on the difference of the compiler (setting) simply.

Oops. I should mention explicitly that my post was about rarewares compile.

.Ogg Vorbis aotuv

Reply #301 –
What does "BS1" mean? And does it mean next aotuv version will be a "Aotuv BS1" not a "beta 6"?

.Ogg Vorbis aotuv

Reply #304 –
If it really fixes the Block Switching issues I think Beta 6 is more appropriated than beta 5.x, because if it really achieve this IMHO it will be the biggest quality improvement for vorbis since aoTuV Beta 2. I didn't test BS1, so I have no clue at all actually. When you will be ready to release a "RC" for Beta 6, ring my bell, I will test it. I don't really want to waste my time, testing all the possible in between versions actually. I am not a low bitrate fan, so IMHO the biggest thing to fix is to make q6 transparent to compete with Nero AAC. From what I heard fixing Block Switching will affect all bitrates anyway, so if it happens, it will be a major update. The biggest unknow for me is to known if you are doing a quick hack like old GT version or if you have really decided to fix it once for good (because in the beginning you said you will not fix it now, so I was surprised that BS1 poped up so quickly), that's why I didn't test BS1, I never like GT Tuned version, so I don't really want the same kind of quick patch (cheating with bitrate) for block switching.

.Ogg Vorbis aotuv

Reply #305 –
Quote from: sauvage78 on
so IMHO the biggest thing to fix is to make q6 transparent to compete with Nero AAC. From what I heard fixing Block Switching will affect all bitrates anyway, so if it happens, it will be a major update. The biggest unknow for me is to known if you are doing a quick hack like old GT version or if you have really decided to fix it once for good (because in the beginning you said you will not fix it now, so I was surprised that BS1 poped up so quickly), that's why I didn't test BS1, I never like GT Tuned version, so I don't really want the same kind of quick patch (cheating with bitrate) for block switching.

The current test version does not solve the problem of all types(you showed it). It solves the problem of the specific pattern mainly by expanding the search range of the block switching (different from GT version).

The better solution that I think of now bumps in an existing patent surely.  It is not preferable for Vorbis.

.Ogg Vorbis aotuv

Reply #306 –
What concerns myself, Vorbis failure in the test became a bolt from the blue. I am thinking of fallback to mp3 

.Ogg Vorbis aotuv

Reply #307 –
The problem I highlighted is not new at all, vorbis has shined in some Roberto's listening test despite the block switching issue. So on average music, vorbis is still competitive, specially at 128Kbps & below, where every codec has holes anyway. The problem is that it is a codec that has some holes even at high bitrate, think of it as a colander. MP3 is not better qualitywise, I could add several samples where it fails if I had any interest in MP3, but it has the advantage that it is a big standard, with universal support. Personnaly 2 weeks after my test, I moved back to lossy|flac. I was quite happy with vorbis when I didn't knew anything about block switching  If I would be sure vorbis had a developper for high bitrates maybe I wouldn't have switched, but I know Monty doesn't care much anymore for vorbis, he only fixes some bugs here & there, but nothing that really improves audio quality & I know Aoyumi is mainly interested in low bitrate. In the beginning I had some hope that I would help aoyumi fixing this problem, but when I realised that this was the same issue that Garf tried to fix several years ago. I realised it would most likely never be fixed. Vorbis is very suited for low/mid bitrates streaming with icecast or for low/mid bitrates audio coupled with theora, so if you are a webmaster then vorbis is a great option. There it really shines as this is what it is built for. But for CD backup it is not very suited, it fails to bring transparency at high bitrate on electronic & live recordings which is not acceptable for an audiophile.

I think that many people who don't do listening test for themselves, mixed pre-echo & block switching (I did) & thought that aotuv beta 2 had solved all vorbis problems. That's why aotuv becomed quite popular when at the same time mpc was dying. Sadly if aotuv beta 2 has improved pre-echo a lot, it didn't affect the block switching issue much. I only realized it lately. Happyly enough I had already switched to lossless & deleted all my vorbis files several years ago personnaly. It was a bit of noltalgia to retry it, but I am not a romantic, I left

.Ogg Vorbis aotuv

Reply #308 –
Little quiet in the ogg vorbis forum

I was wondering if Aoyumi is planning to release any new version of his code or if he is working on something...
 

.Ogg Vorbis aotuv

Reply #309 –
A bit OT: And I wonder if Ogg Vorbis will get a boost ever again, especially now Firefox will support it natively. It's so sad to see industry are adding to hardware aac/mp4 support instead of, or adding too ogg/vorbis. I suppose it's due to the absence of a collective focussing forces.

.Ogg Vorbis aotuv

Reply #310 –
Could somebody compile a win64 copy of the latest aoTuV, I would like to see if there are any gains in speed as have been seen in LAME and MPC's 64bit editions thanks to newer compiler optimizations.

and Alexxander, theres a good reasion (good in the eyes of the riaa/mpaa) for not supporting ogg, unlike AAC/MP4 ogg isnt really built to have any kind of DRM used on it, and we all know how the riaa/mpaa/exct love their DRM.......

.Ogg Vorbis aotuv

Reply #311 –
What is block switching, forgive my ignorance?

.Ogg Vorbis aotuv

Reply #312 –
Quote from: Alexxander on
A bit OT: And I wonder if Ogg Vorbis will get a boost ever again, especially now Firefox will support it natively. It's so sad to see industry are adding to hardware aac/mp4 support instead of, or adding too ogg/vorbis. I suppose it's due to the absence of a collective focussing forces.


Well, despite the industry's seeming move to AAC/MP4, I see Ogg Vorbis gaining ground with game developers.

I didn't know that Firefox support Ogg Vorbis natively??

.Ogg Vorbis aotuv

Reply #313 –
Quote
Sadly if aotuv beta 2 has improved pre-echo a lot, it didn't affect the block switching issue much


If there are block switching issue problems and I haven't encountered many of them. Some cannot be fixed for one reason alone. There are some things you cannot adjust in the source code without running into patents. This happens to be one of them. A customized window function was also designed in order to avoid any patents in regard to that area.


Quote
But for CD backup it is not very suited, it fails to bring transparency at high bitrate on electronic & live recordings which is not acceptable for an audiophile


Do you have any listening tests to backup these claims? I use FLAC to archive my CD's, but I use Vorbis still for HD and streaming purposes. It seems adequate enough in the listening tests that I have done around -q 8, which is overkill if you ask some people.


Quote
What is block switching, forgive my ignorance?


It's a mechanism that the new lossy codecs AAC and Vorbis have a reliance upon heavily, because in the engineering world they are called frequency domain coders. The problem with frequency domain coders is sometime they fail to provide adequate time resolution on transients (short impulsive audio signals) and that's why you need to switch between long blocks (better frequency resolution) and short blocks (better time resolution). A block is nothing, but a set of PCM audio samples that are a multiple of 2.


Quote
Well, despite the industry's seeming move to AAC/MP4, I see Ogg Vorbis gaining ground with game developers.

I didn't know that Firefox support Ogg Vorbis natively?



Vorbis and Theora will be supported natively within Firefox 3.5 the newest release. This has a lot to do with the new <video> and <audio> tags in HTML 5 spec that are part of Firefox 3.5. The only problem is that Theora needs some more work to get it somewhat on par with H.264 that's what the "Thusnelda" branch is for. Dailymotion the popular video site has already started releasing their stuff in both Vorbis and Theora. Monty got a nice grant from the Firefox and Wikimedia people of $100,000 to work on Theora. He job right now is with Red Hat doing just that.
budding I.T professional

.Ogg Vorbis aotuv

Reply #314 –
Quote from: HotshotGG on
Vorbis and Theora will be supported natively within Firefox 3.5 the newest release. This has a lot to do with the new <video> and <audio> tags in HTML 5 spec that are part of Firefox 3.5. The only problem is that Theora needs some more work to get it somewhat on par with H.264 that's what the "Thusnelda" branch is for. Dailymotion the popular video site has already started releasing their stuff in both Vorbis and Theora. Monty got a nice grant from the Firefox and Wikimedia people of $100,000 to work on Theora. He job right now is with Red Hat doing just that.

Oh, you're talking about Fx3.5 ... I'm still staying away from that because it's not yet RC

As for Theora... why aren't they using Dirac?

.Ogg Vorbis aotuv

Reply #315 –
Quote from: HotshotGG on
There are some things you cannot adjust in the source code without running into patents. This happens to be one of them. A customized window function was also designed in order to avoid any patents in regard to that area.

OK, it's not appliable in the official Vorbis encoder, then. But what about an "underground" version ?
Does it could break the compatibility with the official decoder ?

For lancer resurrection (from mpg123 changelog):
Quote from: Canar on
Change log:
New and improved SSE optimizations! For x86-64, too! Also AltiVec! Fast float output! Faster stereo!
Finally, this should put mpg123 into the efficiency-leading position on current hardware!
Thanks go out to Taihei Monma for pushing lots of new assembler code.

Uhmm...

.Ogg Vorbis aotuv

Reply #316 –
Quote from: pepoluan on
Quote from: Alexxander on
A bit OT: And I wonder if Ogg Vorbis will get a boost ever again, especially now Firefox will support it natively. It's so sad to see industry are adding to hardware aac/mp4 support instead of, or adding too ogg/vorbis. I suppose it's due to the absence of a collective focussing forces.


Well, despite the industry's seeming move to AAC/MP4, I see Ogg Vorbis gaining ground with game developers.

The more documentation & standardization the easier (or not so impossible) & more worthwhile it is to implement a technology, I suspect that is the reason you see few hardware devices with OGG Vorbis support.

For game developers all they have to do is encode their files in the free format and make that little extra bit of profit.

.Ogg Vorbis aotuv

Reply #317 –
Quote from: PatchWorKs on
Quote from: HotshotGG on
There are some things you cannot adjust in the source code without running into patents. This happens to be one of them. A customized window function was also designed in order to avoid any patents in regard to that area.

OK, it's not appliable in the official Vorbis encoder, then. But what about an "underground" version ?
Does it could break the compatibility with the official decoder ?

Compatibility does not fail.
It is completely a problem of the encoder side. It is not a problem of a format.

.Ogg Vorbis aotuv

Reply #318 –
Quote
As for Theora... why aren't they using Dirac?


You know that's good question and I am not really sure to be honest with you. Maybe there aren't enough tools avaliable to start using it. Firefox 3.5 has an implementation of FFMPEG in it to do on the fly transcoding if I am not mistaken also.
budding I.T professional

.Ogg Vorbis aotuv

Reply #319 –
Quote from: Aoyumi on
Quote from: PatchWorKs on
Quote from: HotshotGG on
There are some things you cannot adjust in the source code without running into patents. This happens to be one of them. A customized window function was also designed in order to avoid any patents in regard to that area.

OK, it's not appliable in the official Vorbis encoder, then. But what about an "underground" version ?
Does it could break the compatibility with the official decoder ?

Compatibility does not fail.
It is completely a problem of the encoder side. It is not a problem of a format.

What about releasing an "underground"/secret version of the encoder which does not avoid the patents?

.Ogg Vorbis aotuv

Reply #320 –
Quote
What about releasing an "underground"/secret version of the encoder which does not avoid the patents?


... and risk getting sued six ways to Sunday by the MPEG consortium and loss all credibility surrounding the project?. The specific problems that arise from this are rare. The AoTuV tunings will eventually get merged into the official brance after the next release. It's probably not a wise to do that also.
budding I.T professional

.Ogg Vorbis aotuv

Reply #321 –
Quote from: HotshotGG on
Quote
What about releasing an "underground"/secret version of the encoder which does not avoid the patents?


... and risk getting sued six ways to Sunday by the MPEG consortium and loss all credibility surrounding the project?. The specific problems that arise from this are rare. The AoTuV tunings will eventually get merged into the official brance after the next release. It's probably not a wise to do that also.

AFAIK the LAME encoder uses patented methods as well, and they get away with it by only distributing the source code. Apparently, the source code is only a description of the patented method, and the description is publicly available anyway. Of course, merging the change into the official version would probably be a bad idea.

.Ogg Vorbis aotuv

Reply #322 –
Quote
AFAIK the LAME encoder uses patented methods as well, and they get away with it by only distributing the source code. Apparently, the source code is only a description of the patented method, and the description is publicly available anyway. Of course, merging the change into the official version would probably be a bad idea.


Well that's what's planned. AoTuV stand alone is great softare. It would be even better see it merged into the official Vorbis branch. That's what's going to happen with the next official release of libvorbis, because the demand is so high. It coincides with the beta release of the Thusnelda encoder as well.
budding I.T professional

.Ogg Vorbis aotuv

Reply #323 –
Quote from: MedO on
Quote from: HotshotGG on
Quote
What about releasing an "underground"/secret version of the encoder which does not avoid the patents?


... and risk getting sued six ways to Sunday by the MPEG consortium and loss all credibility surrounding the project?. The specific problems that arise from this are rare. The AoTuV tunings will eventually get merged into the official brance after the next release. It's probably not a wise to do that also.

AFAIK the LAME encoder uses patented methods as well, and they get away with it by only distributing the source code. Apparently, the source code is only a description of the patented method, and the description is publicly available anyway. Of course, merging the change into the official version would probably be a bad idea.


Exactly. I have in mind something like FLAKE, an unofficial implementation on an open source project.

This "Barelly legal Vorbis" encoder could be hosted in a Russian (or DVD Jon) website and NOT integrated in the official sources (of course).
Then we can involve "compiles freaks" (as LAME or XviD does) for binaries.

EDIT
OK, an idea for the name: Greebo, the Discworld's Nanny OGG's murderous cat.

Quote
He is a foul-tempered one-eyed grey tomcat whose owner, Nanny Ogg, insists against all the evidence that he is a sweet, harmless kitten. In the course of the books, he has killed two vampires, eating at least one of them in the novel Witches Abroad:

The bat squirmed under his claw. It seemed to Greebo's small cat brain that it was trying to change its shape, and he wasn't having any of that from a mouse with wings on.

And in Maskerade, Magrat recalls when Greebo once killed an elk.

In Lords and Ladies, Greebo's overall attitude is best described in an allusion to Schrödinger's cat:

Greebo had spent an irritating two minutes in that box. Technically, a cat locked in a box may be alive or it may be dead. You never know until you look. In fact, the mere act of opening the box will determine the state of the cat, although in this case there were three determinate states the cat could be in: these being Alive, Dead, and Bloody Furious.

Shawn dived sideways as Greebo went off like a Claymore mine.

"Don't worry about him," said Magrat dreamily, as the elf flailed at the maddened cat. "He's just a big softy."


This could have many meanings beside: OGG, cat, vampires killer (aka Vorbis, Hydrogenaudio, patents killer), etc..





.Ogg Vorbis aotuv

Reply #324 –
Technical question: could LAME bs code be used into "Greebo" ?

Another idea: is it possible to separate the bs algorithm from the encoder ?
If so, it could be en external library that may be exchangable between various codec's encoder engines...