.Ogg Vorbis aotuv, Question for Aoyumi Options

[AshenTech]

just want to say thanks to Aoyumi for all your hard work and report a bug in your(and almost every other ogg encoder I have found) with LONG audio file conversion.

on files 6+hours long the encoder just errors out (not sure where the cut off is, 3hrs is fine, 6 errors), the only vorbis encoder I have found that dosnt have this issue no matter the file length/size is Spoon's encoder (dbpoweramp).

to explain I listen to alot of audio books, And size for quality I prefer vorbis to mp3, I just happened to run into this bug A while back and hope somebody can address the issue so i dont have to use spoons SLOW encoder despite its great quality its just SLOW.

[Aoyumi]

just want to say thanks to Aoyumi for all your hard work and report a bug in your(and almost every other ogg encoder I have found) with LONG audio file conversion.

on files 6+hours long the encoder just errors out (not sure where the cut off is, 3hrs is fine, 6 errors), the only vorbis encoder I have found that dosnt have this issue no matter the file length/size is Spoon's encoder (dbpoweramp).

The encoding is possible via standard input in the current VENC.
I don't know about other encoders. Sorry.

[Nikaki]

Aoyumi, is your patch available somewhere so I can patch my libvorbis 1.3.1 in Linux with it?

This post has been edited by Nikaki: Jul 21 2010, 05:46

[Aoyumi]

Aoyumi, is your patch available somewhere so I can patch my libvorbis 1.3.1 in Linux with it?

There is not the patch for the moment.
If you use Wine on Linux, you can use the test version.

[Nikaki]

Aoyumi, is your patch available somewhere so I can patch my libvorbis 1.3.1 in Linux with it?

There is not the patch for the moment.
If you use Wine on Linux, you can use the test version.

libvorbis is used by other software I use: Native Linux software. Wine doesn't help. Wine only runs Windows programs.

Anyway, it would be nice if you could post your patch, not just the binary. Please don't forget that you are required (I think) to do this in the first place because of the libvorbis license terms.

This post has been edited by Nikaki: Jul 21 2010, 20:41

[lvqcl]

Anyway, it would be nice if you could post your patch, not just the binary. Please don't forget that you are required to do this in the first place because of the libvorbis license terms.

libvorbis has BSD-style license.

libvorbis - Copyright © 2002-2008 Xiph.org Foundation

Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

- Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.

- Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.

- Neither the name of the Xiph.org Foundation nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

That's all.

This post has been edited by lvqcl: Jul 21 2010, 20:46

[Aoyumi]

libvorbis is used by other software I use: Native Linux software. Wine doesn't help. Wine only runs Windows programs.

That it is used by various programs was not assumed.
I am concerned about a test copy being used in a wide range.

Anyway, it would be nice if you could post your patch, not just the binary. Please don't forget that you are required (I think) to do this in the first place because of the libvorbis license terms.

I respect software with developers.
Therefore I think that the license issue should be read carefully.

By the way, did you confirm a license terms?

[>> lvqcl]
Thank you for quotation.

[Nikaki]

I wasn't sure about the license. I assumed Vorbis was open source, which means we need to have the source. At least I'm sure it was open source in the past at some point. But I guess other people know this stuff better than me.

Anyway, it's sad that you don't provide the patch for this That way you only support Windows users and windows programs, which I don't use. Why not provide the patch? You provide the binary, so why not the patch? It would help many people. I'm still stuck to libvorbis 1.2.1, which has been dropped by my Linux distribution because it has security issues, and also it makes using some new versions of various programs impossible because they only work with new versions of libvorbis.

Please, consider providing a patch so that we can use new libvorbis + aotuv.

This post has been edited by Nikaki: Jul 22 2010, 17:42

[john33]

I wasn't sure about the license. I assumed Vorbis was open source, which means we need to have the source. At least I'm sure it was open source in the past at some point. But I guess other people know this stuff better than me.

Anyway, it's sad that you don't provide the patch for this That way you only support Windows users and windows programs, which I don't use. Why not provide the patch? You provide the binary, so why not the patch? It would help many people. I'm still stuck to libvorbis 1.2.1, which has been dropped by my Linux distribution because it has security issues, and also it makes using some new versions of various programs impossible because they only work with new versions of libvorbis.

Please, consider providing a patch so that we can use new libvorbis + aotuv.

If you have observed the history of Aoyumi's work, you should know that he always tests in this way and then once he is satisfied with his tuning efforts, full source is provided. The testing currently tasking place should probably be regarded as alpha testing of the incremental tuning that's taking place. Please be patient, you'll have access to the source once this next version is released. It's not usual to release source while testing is taking place.

BTW, 'open source' is a loose term that doesn't begin to define the terms of the licence under which it may be made available.

[AshenTech]

I love people who are rude(sorry but the attitude Nikaki puts forth is quite rude in my opinion) and expect to get what they demand.

I see this more in "open source" situations then in any others, I dont know why it is other then the fact that many FOSS people feel entitled to get what they want when they want it.....

I want to thank you again for your hard work Aoyumi, You are the one person who is actually trying to improve vorbis, and I for one respect that, dont let Hikaki's attitude bother you, and keep up the excellent work!!

[Aoyumi]

so why not the patch?

Preparations are necessary to open it. It takes far time than I show test build. For example, the source cord overflows with personal memo, and it is necessary for it to rewrite it beforehand.
Of course the work is not only it.

It is very likely that I don't show the test binary at present if there was the premise to open the source cord.

[Nikaki]

@AshenTech
I'm not rude. I only wanted to use this with non-Windows operating systems. One has to be pretty emo to be offended by anything I said, since it was not rude by any stretch of the imagination.

@Aoyumi
We don't need the source code because we want to look at it We only need it so we can compile it. It doesn't matter if the source is pretty or not.

This post has been edited by Nikaki: Jul 24 2010, 14:09

[Tommy Carrot]

We don't need the source code because we want to look at it We only need it so we can compile it. It doesn't matter if the source is pretty or not.

Well, you wont get the source code, no matter how much you whine, because test version is only for testing (and there are enough testers on windows), not for people using it. You need to develop some patience, until Aoyumi considers the new aotuv ready you can always use the previous version.

This post has been edited by Tommy Carrot: Jul 24 2010, 14:26

[Destroid]

@AshenTech
I'm not rude. I only wanted to use this with non-Windows operating systems. One has to be pretty emo to be offended by anything I said, since it was not rude by any stretch of the imagination.

Actually, I interpreted it as rude. A polite request and the graciousness to accept the answer you don't want would be open-minded.

So... (resume topic) I am always enthusiastic about cutting-edge tunings of Vorbis in the low bitrate range, I am rather fond of the results of -q-2 as an alternative for my radio show online archive to the more obscure Speex format (not knocking Speex, it just Vorbis has the advantage of having more supported players).

[AshenTech]

just a quick note, the sansa clip/clip+ and fuze all now have rockbox so speex is OK on its support by affordable players, but you are correct vorbis would be more supported.

I personally find q-0 to be the best choice for audio books and radio shows as when the audio gets more complex it scales up, but normally it keeps the bitrate very close to what speex gives me at its top quality settings.
q--2 (-2) can sound horrible on some headphones due to artifacts(dont really think anybody has optimized the - numbers to much even for voice)

and Im glad Im not the only one who saw his comments/attitude as rude, it came off as very demanding, as if hes entitled to get what he wants when he wants it........reminds me of my buddies teen age daughter who lived with her mother for a few years, spoiled little brat......

[AlexDDR]

@Aoyumi:

Hello, I found that vorbis 1.3.1 for q4 or more has better stereo separation (or stereo fidelity) than the aotuv b5.7.

I look in the source of 1.3.1 and they put a lot of rewrite code in psy.c

please consider take attention in stereo sound stage and intensity of the music in your tunnings, they do a very good job in the harshness of some songs IMHO

sorry i sound rude, my inglish is poor



thanks

[Nikaki]

We don't need the source code because we want to look at it We only need it so we can compile it. It doesn't matter if the source is pretty or not.

Well, you wont get the source code, no matter how much you whine

Now you're the one being rude, accusing me of whining.

you can always use the previous version.

Well, that's the problem. I can't. It's been blacklisted for security reasons (that older libvorbis version has security bugs). I know that Windows doesn't care about stuff like this, but other OSes do.

This post has been edited by Nikaki: Jul 27 2010, 00:59

[Synthetic Soul]

Enough with the open source, Windows vs *nix yawnathon. Any further posts on the subject will be permanently deleted, and users will receive a TOS #5 warning.

[Aoyumi]

We don't need the source code because we want to look at it We only need it so we can compile it. It doesn't matter if the source is pretty or not.

It is the same thing for me.
I can understand your thought, but cannot go along.

Hello, I found that vorbis 1.3.1 for q4 or more has better stereo separation (or stereo fidelity) than the aotuv b5.7.
I look in the source of 1.3.1 and they put a lot of rewrite code in psy.c
please consider take attention in stereo sound stage and intensity of the music in your tunnings, they do a very good job in the harshness of some songs IMHO

I can say nothing as far as there is not the sample.
But the stereo processing is changed with the existing test version.

[AlexDDR]

I can say nothing as far as there is not the sample.
But the stereo processing is changed with the existing test version.

I test venc,exe version aoTuV post-beta5.7 [20100718] at -q2

with de aotuv b5.7 the song Wild Wild West of will smith, and in the chorus at the 2nd voice that have a high pitch i hear a little loss of power and stereo, is like a loss of high frecuency fault like blur or something.


but with de testing version post b5.7(above) i can't hear a big difference, in fact i hear differences (i guess for the quality level) but the energy of the music is almost the same

b5.7 = 99kbps (q2)
testing= 107 kbps (q2)

CODE

K:\>venc -q2 www.wav www.ogg
aoTuV Post Beta 5.7 - OggVorbis Encoder (c) 2003-2010 Aoyumi

Output file : "www.ogg"
Stream info.: 2ch/44.1kHz/16bit(int) -> 2ch/44.1kHz/32bit(float)

Quality 2 (nominal bitrate : 96kbps)
Processing...Done.

[Report of the encoded file]
   play time    : 245.533 sec.
  encode time   : 11.692 sec. (x21.000)
stream bitrate : 107.34 kbps
bit reduction  : 1 / 13.1

i think tha is most visible the stereo at high quality (q4 or q5) when the music is more detailed and at this level of q i cant hear (q2) any artifacts but i hear the loss of power or wide stereo image or blur or roughness .


in conclusion, i'm waiting the new aotuv version based in the code of stereo procesing of 1.3 series

thanks mr. aoyumi

PD: i hear the difference in the stereo in the songs with 360º stereo sound like the chorus of "s club 7- friday night" but others with flat stereo i can't hear nothing

This post has been edited by AlexDDR: Jul 28 2010, 21:44

[IgorC]

AlexDDR,

It will be helpful if you will upload up to 30 seconds (allowed limit) of mentioned sample somewhere. It could be Upload section or somewhere else like rapidshare.
The words has a little meaning without sample.

Also, read the rules. Especially N8.

This post has been edited by IgorC: Jul 28 2010, 21:54

[AlexDDR]

AlexDDR,

It will be helpful if you will upload up to 30 seconds (allowed limit) of mentioned sample somewhere. It could be Upload section or somewhere else like rapidshare.
The words has a little meaning without sample.

Also, read the rules. Especially N8.

Sorry my mistake for not having posted results of ABX to confirm my findings, until then I beg you not to take into account my long post until I have the ABX support we'll see how it goes with foobar and ABX.

thakns for the warning, I was a little disoriented

[Aoyumi]

I released the new test version. (for tester)

CHANGES (from post-beta5.7 [20100718])
* The one of the issue of noise to rarely occur at the low bit rate (eg -q-2) was improved.
* New dynamic stereo processing is added.

If you found the problem that is peculiar to this version, please inform me.

This post has been edited by Aoyumi: Aug 1 2010, 12:43

[lithoc]

I released the new test version. (for tester)

CHANGES (from post-beta5.7 [20100718])
* The one of the issue of noise to rarely occur at the low bit rate (eg -q-2) was improved.
* New dynamic stereo processing is added.

If you found the problem that is peculiar to this version, please inform me.

Any feedback on this build it's been a month without any reply or bugs found.

Keep up the good work Aoyumi, I've been following your progress for few years.

Thank you for your great work!!!

[akapuma]

Hello,

I'm looking forward to the next version.

Thank's for your excellent work, Aoyumi.

Best regards

akapuma