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: QAAC: discussion, questions, feature requests, etc. (Read 686456 times) previous topic - next topic
0 Members and 2 Guests are viewing this topic.

QAAC: discussion, questions, feature requests, etc.

Reply #800 –
Quote from: sluggy on
Sorry if this sounds like a silly question but i don't understand what difference the new -A option does? Anyone care to explain?

It was just a request so apps like foobar2000 wouldn't have to add different encoder parameters for both qaac and refalac, now the parameter is always -A and what changes is only the .exe.

See here another request, for Peter now: http://www.hydrogenaud.io/forums/index.php...=108274

QAAC: discussion, questions, feature requests, etc.

Reply #801 –
Quote from: eahm on
Quote from: sluggy on
Sorry if this sounds like a silly question but i don't understand what difference the new -A option does? Anyone care to explain?

It was just a request so apps like foobar2000 wouldn't have to add different encoder parameters for both qaac and refalac, now the parameter is always -A and what changes is only the .exe.

See here another request, for Peter now: http://www.hydrogenaud.io/forums/index.php...=108274


Ah so you can access the lossless encoder using qaac.exe with -A? I didnt realise that. I always just use refalac.

QAAC: discussion, questions, feature requests, etc.

Reply #802 –
My apologies if this has already been covered somewhere in the thread (yes, I searched; no, I didn't find it), but does the following parameter directly map to the "good," "better," and "best" options in QuickTime?

Quote
-q, --quality <n>      AAC encoding Quality [0-2]


I ask because the Quicktime 7 User's Guide includes the following note in the MPEG-4 Audio Export Options, at the top of page 51:

Quote
Encoding Quality: Available only with AAC audio. The Good setting is optimized for the highest-speed encoding, for higher-quality, choose Best for 16-bit audio, or Better if your audio source is 24-bit.


As a quick experiment, I encoded a 24-bit audio source via qaac/CoreAudioToolbox 7.9.9.4 to produce two separate files, one of which used --quality 2 (assuming that to be "Best"), and the other of which used --quality 1 (assuming that to be "Better"). I then converted each of those files back to WAV via foobar2000. Next I loaded the original into Audacity, and inverted the signal, prior to loading one of the decoded WAV files into Audacity, so that I could mix and render a residual signal.

From there, I loaded both residual signals in Audacity, and mapped both channels of the --quality 1 ("Better") output Left, with both channels of the --quality 2 ("Best") output mapped Right. Playing the synchronized pair of residual signals resulted in an experience that was significantly—as in, no subtlety about it, it was that easy to discern—noisier on the "Best" side, and that counter-intuitively seems to imply the "Better" setting preserved an audio signal closer to the original, 24-bit source than the "Best" setting managed... which, if correct, matches the advice in the Quicktime 7 User's Guide.

QAAC: discussion, questions, feature requests, etc.

Reply #803 –
Quote from: M on
does the following parameter directly map to the "good," "better," and "best" options in QuickTime?
Quote
-q, --quality <n>      AAC encoding Quality [0-2]


Yes, I think so.

Quote
Encoding Quality: Available only with AAC audio. The Good setting is optimized for the highest-speed encoding, for higher-quality, choose Best for 16-bit audio, or Better if your audio source is 24-bit.

I dunno. Only Apple dev should be able to explain it, but codec itself works on 32bit float. Then why the input bit depth matters? I'm somewhat skeptical.

Quote
As a quick experiment, I encoded a 24-bit audio source via qaac/CoreAudioToolbox 7.9.9.4 to produce two separate files, one of which used --quality 2 (assuming that to be "Best"), and the other of which used --quality 1 (assuming that to be "Better"). I then converted each of those files back to WAV via foobar2000. Next I loaded the original into Audacity, and inverted the signal, prior to loading one of the decoded WAV files into Audacity, so that I could mix and render a residual signal.

I don't think your test procedure is valid for comparing perceptual encoding.

QAAC: discussion, questions, feature requests, etc.

Reply #804 –
Quote from: nu774 on
I don't think your test procedure is valid for comparing perceptual encoding.


Nor did I suggest that it was. What I did suggest was that, based on Apple's own advice, the possibility merited technical examination... hence the reason I began with the "quick experiment" I described, rather than jumping straight to a double-blind test of perceptual output.

At any rate, the information seemed interesting enough to mention, so that any other HA members who might also be inclined to experiment with alternative methods of encoding 24-bit source material could do so. (And should any then do so, surely multiple individuals contributing results of their own double-blind tests would be more indicative than anecdotal information from a single individual?)

QAAC: discussion, questions, feature requests, etc.

Reply #805 –
Quote from: M on
My apologies if this has already been covered somewhere in the thread (yes, I searched; no, I didn't find it), but does the following parameter directly map to the "good," "better," and "best" options in QuickTime?

Quote
-q, --quality <n>      AAC encoding Quality [0-2]


I ask because the Quicktime 7 User's Guide includes the following note in the MPEG-4 Audio Export Options, at the top of page 51:

Quote
Encoding Quality: Available only with AAC audio. The Good setting is optimized for the highest-speed encoding, for higher-quality, choose Best for 16-bit audio, or Better if your audio source is 24-bit.


As a quick experiment, I encoded a 24-bit audio source via qaac/CoreAudioToolbox 7.9.9.4 to produce two separate files, one of which used --quality 2 (assuming that to be "Best"), and the other of which used --quality 1 (assuming that to be "Better"). I then converted each of those files back to WAV via foobar2000. Next I loaded the original into Audacity, and inverted the signal, prior to loading one of the decoded WAV files into Audacity, so that I could mix and render a residual signal.

From there, I loaded both residual signals in Audacity, and mapped both channels of the --quality 1 ("Better") output Left, with both channels of the --quality 2 ("Best") output mapped Right. Playing the synchronized pair of residual signals resulted in an experience that was significantly—as in, no subtlety about it, it was that easy to discern—noisier on the "Best" side, and that counter-intuitively seems to imply the "Better" setting preserved an audio signal closer to the original, 24-bit source than the "Best" setting managed... which, if correct, matches the advice in the Quicktime 7 User's Guide.


Interesting, over here https://developer.apple.com/library/mac/tec...237/_index.html

It says:
Encoding Quality: Good, Better or Best. The Good  setting is optimized for the highest-speed encoding, for higher-quality  choose Better or Best (optimal for 24-bit source). The tradeoff is  between encoding speed and audio quality

This seems to me that better and best could both apply to 16 and 24 bit files, with best resulting in the highest quality setting *which would make sense*. And it also has the distinction of mentioning 24 bit but not proclaiming it to be the highest source it accepts, simply optimal.

QAAC: discussion, questions, feature requests, etc.

Reply #806 –
Does anyone know what the default lowpass settings for QAAC's CVBR and TVBR modes in the higher bitrate range are (256 kbps and up for CVBR and Q 100 and up for TVBR)?

QAAC: discussion, questions, feature requests, etc.

Reply #807 –
Hello, nu774. Thank you so much for 64-bit support, for me it runs about 20% faster than x86.

But can you tell us why does qaac64 gives considerably different stream? I encoded the same file with x64 and x86 (both qaac 2.46/7.9.9.4) and have got their differential file with peaks up to -48 dBFS. Don't think it's normal (I wouldn't care if they was less than -96 dBFS, but actually they're much higher).

How could you explain it? Thanks.
🇺🇦 Glory to Ukraine!

QAAC: discussion, questions, feature requests, etc.

Reply #808 –
Quote from: Steve Forte Rio on
have got their differential file with peaks up to -48 dBFS. Don't think it's normal (I wouldn't care if they was less than -96 dBFS, but actually they're much higher).

Well, not being the developer of the codec, I have nothing to explain. However, to me it doesn't look as abnormal as you say.
Have you tried the same with other lossy encoders (LAME, Vorbis, Opus,...) ?

QAAC: discussion, questions, feature requests, etc.

Reply #809 –
Quote from: nu774 on
Have you tried the same with other lossy encoders (LAME, Vorbis, Opus,...) ?


For other codecs the differential signal generally has peaks much lower so it's even almost inaudible.

Then, you are right, it's rather a question for Apple developers 
And, anyway, I'm almost sure that even such differences, as for QAAC, will be non-abxable.
🇺🇦 Glory to Ukraine!

QAAC: discussion, questions, feature requests, etc.

Reply #810 –
The difference between SSE vs. non-SSE builds of "Oggenc2.87 using libVorbis v1.3.4" is comparable or even bigger than the difference between qaac vs. qaac64.

QAAC: discussion, questions, feature requests, etc.

Reply #811 –
remove please sorry.

QAAC: discussion, questions, feature requests, etc.

Reply #812 –
It's new QAAC 2.47:

- Large file (>= 4GB) output is now supported. Very long duration (beyond 32bit limit) is also supported, but the latter is not compatible with QuickTime 7.
- On very large files, container optimization can take several minutes. You can disable it by --no-optimize.

QAAC: discussion, questions, feature requests, etc.

Reply #813 –
Quote from: eahm on
Thanks for the -A switch too nu774.

Just noticed that Peter used --alac for foobar2000 and not -A, I don't know if he cares enough to change it and if he doesn't -A won't resolve anything for the normal user.

Here what he uses: "--ignorelength -s --no-optimize --alac -o %d -"

QAAC: discussion, questions, feature requests, etc.

Reply #814 –
while i was finding out what setting to encode my collection at i discovered a nasty artifact that i only seem to get on Apple AAC but not on other formats?. i can make sample if anyone whats hear or not.

QAAC: discussion, questions, feature requests, etc.

Reply #815 –
Quote from: lonelyroads on
while i was finding out what setting to encode my collection at i discovered a nasty artifact that i only seem to get on Apple AAC but not on other formats?. i can make sample if anyone whats hear or not.

Of course do please and thank you  Many say they prefer MP3 to AAC because of the different artifact habit I guess they have.

Open a new thread, this is about QAAC not AAC, the sample must be less than 30 secs.

QAAC: discussion, questions, feature requests, etc.

Reply #816 –
Quote from: eahm on
Quote from: lonelyroads on
while i was finding out what setting to encode my collection at i discovered a nasty artifact that i only seem to get on Apple AAC but not on other formats?. i can make sample if anyone whats hear or not.

Of course do please and thank you  Many say they prefer MP3 to AAC because of the different artifact habit I guess they have.

Open a new thread, this is about QAAC not AAC, the sample must be less than 30 secs.
will do and thanks.

QAAC: discussion, questions, feature requests, etc.

Reply #817 –
Quote from: eahm on
Quote from: eahm on
Thanks for the -A switch too nu774.

Just noticed that Peter used --alac for foobar2000 and not -A, I don't know if he cares enough to change it and if he doesn't -A won't resolve anything for the normal user.

Here what he uses: "--ignorelength -s --no-optimize --alac -o %d -"

nu774, sorry for always requesting, can you add --alac too to refalac, I guess? I don't even know if refalac is compatible with the other commands, let's see what Peter is going to do with the new foobar2000. Thanks.
 

QAAC: discussion, questions, feature requests, etc.

Reply #818 –
Quote from: eahm on
nu774, sorry for always requesting, can you add --alac too to refalac, I guess? I don't even know if refalac is compatible with the other commands, let's see what Peter is going to do with the new foobar2000. Thanks.

I think --alac is already allowed.

QAAC: discussion, questions, feature requests, etc.

Reply #819 –
Quote from: nu774 on
I think --alac is already allowed.

Perfect, sorry didn't have the file to check. Just checked.

QAAC: discussion, questions, feature requests, etc.

Reply #820 –
nu774, could you please add support for Sound Check?

QAAC: discussion, questions, feature requests, etc.

Reply #821 –
Not using iTunes, I did not even know what Sound Check means; apparently it is similar to ReplayGain.

QAAC: discussion, questions, feature requests, etc.

Reply #822 –
Quote from: LigH on
Not using iTunes, I did not even know what Sound Check means; apparently it is similar to ReplayGain.


Yes, it is similar to RG. I don't use iTunes either, but my iPod doesn't support RG, only SC.

QAAC: discussion, questions, feature requests, etc.

Reply #823 –
I will not implement replaygain scanner in qaac since:

1) A replaygain scanner shares no functionality with an encoder and it can be implemented as a separated, independent program (Scanner will use AAC decoder, not encoder).

2) Multiple choices of algorithm (R128 or something), multiple way of implementing it (aacgain way/metadata only).

3) qaac can be executed in many ways. What "Album" means is not always clear to qaac.

The same applies true for sound check, but it is even worse than replaygain.

QAAC: discussion, questions, feature requests, etc.

Reply #824 –
Hi,
I have a question about --ignorelength.
I use eac3to and get 3.85G wav file from video. Then I use qaac to convert it to the aac.
first I use --ignorelength and get a m4a file, then I try not use --ignorelength and I also get a m4a file.
And I listen these 2 files and didn't find any problem. But these 2 files are different(MD5). So, do I need to use --ignorelength ?
Thank~