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 GT3 beta 1 (Read 60533 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Ogg Vorbis GT3 beta 1

Reply #75
Slightly off-topic, but I've just uploaded a version of the standard oggenc, but with the unicode patch applied. It's at Mirror 1. Would anyone interested kindly test and confirm if it's OK, please? Check the help file first for the option.

If this is OK, I'll repeat the exercise for the other versions of oggenc.

Ogg Vorbis GT3 beta 1

Reply #76
Quote
Well it does under win2k, but maybe not with other win versions.

Looked up what I wrote to Garf about setargv.obj. Something like this caused problems:
Code: [Select]
replaygain -a "my folder\*.ogg"

And that's on Windows 2000...

Ogg Vorbis GT3 beta 1

Reply #77
Quote
Slightly off-topic, but I've just uploaded a version of the standard oggenc, but with the unicode patch applied. It's at Mirror 1. Would anyone interested kindly test and confirm if it's OK, please? Check the help file first for the option.

If this is OK, I'll repeat the exercise for the other versions of oggenc.

I tried it with --utf8 flag (CDDB2 stores them in UTF-8 and this information is transferred through Player -> cdplayer.ini -> EAC, I assumed there will be no loss of UTF-8 info in these steps). It appears to be working fine. The encoder does not produce error messages related to the conversion as it used to and the tags appear correct so far...  Thanks a lot again.
The object of mankind lies in its highest individuals.
One must have chaos in oneself to be able to give birth to a dancing star.

Ogg Vorbis GT3 beta 1

Reply #78
Quote
I tried it with --utf8 flag (CDDB2 stores them in UTF-8 and this information is transferred through Player -> cdplayer.ini -> EAC, I assumed there will be no loss of UTF-8 info in these steps). It appears to be working fine. The encoder does not produce error messages related to the conversion as it used to and the tags appear correct so far...  :rolleyes: Thanks a lot again.

Good news!! I'll amend and recompile the other versions soon.

Ogg Vorbis GT3 beta 1

Reply #79
OggEncGT3b1 still fails badly at BadVilbel. I can easily ABX it up to -q 7. And that with a really crapy sound system. At -q 8 I don't hear any difference anymore, but it takes 404kbits/s! 
Is this due to limitations in the ogg vorbis codec or will the problem be eliminated (no abx possible at ~220kbits/s using regular pc-speakers) with further tuning? I can't ABX the same sample encoded with lame (3.90.2) --alt-preset fast standard, wich results in an acceptably sized (232kbits/s) mp3.
By the way, I don't know any sample that sounds that bad or requires nearly as much kbits per second using lame --alt-preset fast standard.

Ogg Vorbis GT3 beta 1

Reply #80
You can generally assume that when something encodes ok with mp3, Vorbis can do at least just as good a job in theory. That doesn't mean I'll actually be able to get the codec to perform like that though.

Quote
By the way, I don't know any sample that sounds that bad or requires nearly as much kbits per second using lame --alt-preset fast standard.


Of course not, MP3 is hard limited to 320kbps

Ogg Vorbis GT3 beta 1

Reply #81
Quote
You can generally assume that when something encodes ok with mp3, Vorbis can do at least just as good a job in theory. That doesn't mean I'll actually be able to get the codec to perform like that though.
I see. But I think it would be good if you could implement your quality switches so that they really produce sound of a constant quality at the given level. The mentioned sample shouldn't sound that horrible at say -q 6. Because of this sample I don't dare encoding my music with ogg vorbis as I never know whether it sounds ok or not.
Quote
Of course not, MP3 is hard limited to 320kbps
Well, I don't even know anything that sounds that bad at 320 kbit/s 

Ogg Vorbis GT3 beta 1

Reply #82
Quote
Quote
You can generally assume that when something encodes ok with mp3, Vorbis can do at least just as good a job in theory. That doesn't mean I'll actually be able to get the codec to perform like that though.
I see. But I think it would be good if you could implement your quality switches so that they really produce sound of a constant quality at the given level. The mentioned sample shouldn't sound that horrible at say -q 6.

If this were an easy task, Garf (or even Monty) would have already done it already, and quite awhile ago even.

Ogg Vorbis GT3 beta 1

Reply #83
Thanks for the quick reply. I think I learned something today...
Do You know any samples that break your alt-presets  that badly? I mean something that even I can hear ... even with my crappy pc-speakers.

Ogg Vorbis GT3 beta 1

Reply #84
Quote
I see. But I think it would be good if you could implement your quality switches so that they really produce sound of a constant quality at the given level. The mentioned sample shouldn't sound that horrible at say -q 6.


That is the goal of course, but as I said, this will take a while, and...

Quote
Because of this sample I don't dare encoding my music with ogg vorbis as I never know whether it sounds ok or not.


...things will never be perfect with ANY lossy codec. From what you just said I would seriously recommend FLAC over Vorbis, MP3 or MPC to you.

Ogg Vorbis GT3 beta 1

Reply #85
@garf

why is the nominal bitrate at q6 212kbps and not 192kbps?

Ogg Vorbis GT3 beta 1

Reply #86
Quote
@garf

why is the nominal bitrate at q6 212kbps and not 192kbps?

First read thread, then ask questions.

Ogg Vorbis GT3 beta 1

Reply #87
Quote
Quote
@garf

why is the nominal bitrate at q6 212kbps and not 192kbps?

First read thread, then ask questions.

"Because at a given quality level, GT3 will generate larger files." © 2003 by garf

Ogg Vorbis GT3 beta 1

Reply #88
I would like to use it under Linux, frankly I don't know how to handle this .x86... never seen this as a file ending... won't execute in Konqueror nor in shell.

Ogg Vorbis GT3 beta 1

Reply #89
Just do
Code: [Select]
chmod +x

on the file and execute it as normal

Ogg Vorbis GT3 beta 1

Reply #90
I just noticed that the Linux binary still reports a nominal bitrate of 160 Kbps at -q 5, and even though it doesn't really matter (it is just a number), I thought I'd point it out.

Ogg Vorbis GT3 beta 1

Reply #91
I've been working a bit on it with not much success so far. The brahms sample is interesting, I have no idea what goes wrong there, except that it's not a stereo issue this time. Looks almost like a bug or as a regression, I don't remember seeing this with GT2.

If I don't find anything, I'll skip it and tackle a new preecho method I have in my head, but I really want the noise issue fixed because it is indeed annoying. Would be nice if monty at least answered his email, grmpf.

Ogg Vorbis GT3 beta 1

Reply #92
Quote
The brahms sample is interesting, I have no idea what goes wrong there, except that it's not a stereo issue this time. Looks almost like a bug or as a regression, I don't remember seeing this with GT2.

The RC2, the GT2 and the RC3 hadn't this problem. I noticed it with final version. Something (wrong ?) has changed between RC3 and final. You should maybe take a look on late RC3 version (or pre-RC4). I haven't the binary ; I can't help you.

The noise generation at high bitrate is concomitant with the apparition of noise at lower bitrate (final version, or post-RC3). This might help you to find the guilty element .
Wavpack Hybrid: one encoder for all scenarios
WavPack -c4.5hx6 (44100Hz & 48000Hz) ≈ 390 kbps + correction file
WavPack -c4hx6 (96000Hz) ≈ 768 kbps + correction file
WavPack -h (SACD & DSD) ≈ 2400 kbps at 2.8224 MHz

Ogg Vorbis GT3 beta 1

Reply #93
Thanks a lot, this is very usefull info. If I can't get Monty's help on this, I will back up from 1.0 for new GT versions. (Since most of RC3->1.0 was low bitrate tuning, this shouldn't matter for quality/bugs)

CVS should have all versions between RC3 and 1.0, that should make it possible to pinpoint the guilty change as well.

 

Ogg Vorbis GT3 beta 1

Reply #94
Quote
Thanks a lot, this is very usefull info. If I can't get Monty's help on this, I will back up from 1.0 for new GT versions. (Since most of RC3->1.0 was low bitrate tuning, this shouldn't matter for quality/bugs)

CVS should have all versions between RC3 and 1.0, that should make it possible to pinpoint the guilty change as well.

what or who is CVS?

if the certain CVS doesn't have the needed versions, ask me . I kept each version I downloaded on my hd.
I love the moderators.

Ogg Vorbis GT3 beta 1

Reply #95
CVS: http://www.cvshome.org/
A riddle is a short sword attached to the next 2000 years.

Ogg Vorbis GT3 beta 1

Reply #96
Quote
...bitrates will tend to be a bit higher for a given quality level.

Emboldened by participation in another thread (which went quiet, I hope I didn't kill it  ) I'll take part in this one too.

I'm not sure of the logic in having different quality outputs for the same -q value. If GT3 q6 sounds as good as 1.0 q5.4 or whatever (I haven't done listening tests) ought GT3 not be rescaled to match? The benefit of GT3 is then reduced bitrate. Otherwise there's a risk of confusion when v1.1 is released with new folk discussing their experiences and replies coming back asking which particular encoder they used. Presenting new users with unexpected technicalities is not a perfect way to gain market share.

IMHO

Ogg Vorbis GT3 beta 1

Reply #97
Quote
Quote
Now we see true power of OGG Vorbis, keep tuning Garf...
Is there any chance to beat MPC through tuning? How much You can tune?

Beat MPC? Not going to happen. I'll be happy if it has no more big issues at around 170-190kbps.

Never, ever?  Intrinsically impossible? ... or are you being modest?

Ogg Vorbis GT3 beta 1

Reply #98
@SwiftBiscuit:
Quote
Quote
ok 
Will this be the case in the final version as well? Are you hoping to bring the bitrates into line with vorbis 1.0?

Yes. Actually, ideally I'd want to correct the quality levels, so -q5 with GT3 corresponds roughly with -q5 of 1.0 in quality, but GT3 does it at a lower nominal bitrate.


Ogg Vorbis GT3 beta 1

Reply #99
 Okay, where the hell was that? I've been reading this thread avidly I swear!....
....bugger. I need more sleep I guess.

God I feel stupid.