QAAC: discussion, questions, feature requests, etc., [originally a thread for a feature request] Options

[lunkhead]

Duration:
qaac 1.47 q73: 3wk 0d 8:34:06.898 (81 375 388 218 samples)
FLAC and LAME: 3wk 0d 8:34:06.899 (81 375 388 229 samples)

I noticed foobar2000 incorrectly adds the total sample size of large number of items. It's always very slightly less. It's not noticeable when the list is small. Just an observation.

This post has been edited by lunkhead: Feb 19 2013, 06:00

[lvqcl]

Do you have HE-AAC files ?

[sluggy]

[qaac] release 2.16 (refalac 1.16)
posted 14 hours ago by nu 774
Read and handle multichannel layout of TAK files.
Write fact chunk when decoding into WAVEFORMATEXTENSIBLE format. As far as I can see, even WMP is not in honor of the fact chunk, so this would be pretty much useless. However, since it looks like RIFF/WAV spec require it in WAVEFORMATEXTENSIBLE, this was implemented to be more spec compliant. fact chunk is not written on piped output or WAVEFORMATEX format.
Automatically kill progress message when stderr is connected to nothing.

https://sites.google.com/site/qaacpage/cabinet

[hellside]

I just run into QACC, and I am wondering which is better, TVBR127 or cvbr256?
some say that cvbr sometimes reach 400 br

just need a quick word, thanks!

[eahm]

TVBR 127 = ~320kbps

You should ABX, you'll be surprise how low you can go with AAC.

I use -V73 (~150kbps) just because I want to use "transparent for me" +1. I like AAC @~128kbps and since -V63 gives ~135kbps (almost all the time around ~100kbps with my music) I use -V63 + 1 (just in case) = -V73.

This post has been edited by eahm: Mar 4 2013, 04:04

[LigH]

I know what you mean ... everything was fine with "oggenc2 -q 4" until I heard echoes in the break beats of "Prodigy: Your Love (Remix) — Experience". There are always extraordinary exceptions.

[o-l-a-v]

[qaac] release 2.17 (refalac 1.17)
posted 13 minutes ago by nu 774

Can't find the changelog yet

"
Fixed github issue 27 (regression on 1.26). --decode was writing invalid wav file.
Added --gapless-mode option (same as fdkaac).
Interestingly, iTunes seems to support both of iTunSMPB and ISO standard gapless mode. QuickTime supports only the latter. In the past, I thought QT silently assumes 2112 samples of delay. However, it turned out that QT actually looks elst media_time when sbgp and sgpd are present, so it can be used generally (as described in QTFF spec).

As far as I know, iTunes is the only music player that supports gapless playback in both way.
"

Download: https://sites.google.com/site/qaacpage/cabi...rects=0&d=1

This post has been edited by o-l-a-v: Mar 19 2013, 14:34

[ktf]

Fixed github issue 27 (regression on 1.26). --decode was writing invalid wav file.

I reported that issue a few hours ago and apparently it's already fixed? That's really, really fast!

[nu774]

Fixed github issue 27 (regression on 1.26). --decode was writing invalid wav file.

I reported that issue a few hours ago and apparently it's already fixed? That's really, really fast!

Thanks for reporting. It was simply due to missing initialization of one new boolean flag variable introduced on 1.26, and was quite easy to fix.
Maybe it could have been caught earlier by the compiler warning if I had killed tons of stupid C4244 or something (casting to smaller type).

[lvqcl]

1.26 or 2.16 ?

[nu774]

1.26 or 2.16 ?

Oh, sorry. I meant 2.16.

[me7]

...so, if I want gapless playback to work across most players/devices, what --gapless-mode should I choose? The default is "iTunSMPB only".

EDIT: To be clearer: Why is "iTunSMPB only" preferred over "both"?

This post has been edited by me7: Mar 19 2013, 19:48

[nu774]

...so, if I want gapless playback to work across most players/devices, what --gapless-mode should I choose? The default is "iTunSMPB only".

EDIT: To be clearer: Why is "iTunSMPB only" preferred over "both"?

It's default not because it's considered superior, but it's backward compatible and conservative.
Probably "both" will do no harm, but I cannot assure that. Other encoders (including iTunes) only write iTunSMPB.

[me7]

You've got a point. Even if the ISO specification says something else, it might be better to do "what all other encoders do"

[db1989]

I just removed blatant spam, and a full quote thereof by a regular member, from a user whose name began with “laptop” and whose post consisted of an 1.5 year–old paragraph by nu774 copied and pasted and with a link to an external website added. Please think about things such as usernames and links before replying!

[LigH]

Some ad bots use this behaviour (quoting a previous posting) to cover their activity by pretending to refer to it.

[robertcollier4]

Is it possible to use qaac as an audio encoder in GraphEdit? I would like to connect the ffdshow audio decoder output to qaac encoder. I am currently having to do ffdshow audio decoder -->WAV Dest --> File writer. Then convert the WAV file on hard dive with qaac.

Given that qaac can already accept piped input - is there any way to set this up in graphedit to pass its output to a command-line program? If not, then please add it as a feature request to have a qaac output filter for GraphEdit. Thanks.

[nu774]

Given that qaac can already accept piped input - is there any way to set this up in graphedit to pass its output to a command-line program?

It might be possible by creating special audio writer/renderer filter or something to stream audio to a pipe, but I've never heard of such things.

If not, then please add it as a feature request to have a qaac output filter for GraphEdit. Thanks.

Why don't you just pipe from ffmpeg instead of graphedit+ffdshow ?
I don't feel like writing dshow encoder filters.

[eahm]

CODE

-V63 --adts --no-delay -o %d -

gives me "Conversion failed: Unsupported format or corrupted file".

What am I missing?

This post has been edited by eahm: Apr 2 2013, 03:21

[nu774]

CODE

-V63 --adts --no-delay -o %d -

gives me "Conversion failed: Unsupported format or corrupted file".

What am I missing?

Try with older version of fb2k.

[eahm]

I went all back to 1.1.18 but nothing, same error.



This post has been edited by eahm: Apr 2 2013, 06:36

[nu774]

I went all back to 1.1.18 but nothing, same error.

It seems fb2k tries to copy tags on the resulting ADTS file after encoding, and fails.
You can try converting first to an intermediate file, then convert it to ADTS.
Use AIFF (for example) as the intermediate file format, since fb2k doesn't support tags on AIFF.

Otherwise, you can cheat fb2k by passing qaac "-o %d.aac" or something as output filename (an extra ".aac" extension is appended to the output filename).
fb2k will show error, but resulting file will not be removed by fb2k, since it's named different from what fb2k thinks.

[sohm]

nu774, firstly many thanks for qaac!

i've encountered a rather strange glitch trying to encode in HE-AAC:
output file's length(in seconds) is exactly 50% of the input.
meanwhile, qaac.exe command line output displays correct processed length in seconds, also no complaints in --verbose mode except that it runs suspiciously fast.

CODE

..\qaac_2.17\qaac>qaac --verbose --he test1.wav
qaac 2.17, CoreAudioToolbox 7.9.8.2

test2.m4a
Using default channel layout.
Output layout: Stereo
AAC-HE Encoder, CVBR 80kbps, Quality 96
[100.0%] 7:11.506/7:11.506 (28.8x), ETA 0:00.000
19029444/19029444 samples processed in 0:15.000
Overall bitrate: 88.272kbps
423/423 chunks written (optimizing)

..\qaac_2.17\qaac>qaac --verbose test2.wav
qaac 2.17, CoreAudioToolbox 7.9.8.2

test2.m4a
Using default channel layout.
Output layout: Stereo
AAC-LC Encoder, TVBR q91, Quality 96
[100.0%] 7:11.506/7:11.506 (25.1x), ETA 0:00.000
19029444/19029444 samples processed in 0:17.234
Overall bitrate: 213.081kbps
423/423 chunks written (optimizing)

test1.m4a
4.57MB (4 802 231 bytes)
3:35.753 (9514722 samples)

test2.m4a
11.0MB (11 571 769 bytes)
7:11.507 (19029444 samples)

9514722*2=19029444 )))

apart from this issue with --he, AAC-LC performs flawlessly in both --tvbr or --cvbr modes.
any suggestions much appreciated))

[Big_Berny]

Hi
I try to use qaac for transcoding in subsonic. I want to transcode my 256Kbit/s AAC to like 96 KBit/s AAC. I know that's evil , but I need iot for streaming if I have a bad mobile connection.

I tried this transcoding command:
Step 1: faad -w %s
Step 2: qaac -a %bk --adts - -

It seems to work but at the end of transcoding I always get a error message "Cannot seek back the input". Does someone has an idea why this happens?

CODE

...SHORTENED...
[5/9/13 5:57:11 PM CEST]    DEBUG    InputStreamReaderThread    (c:\subsonic\transcode\faad) 99% decoding D:\Users\Admin\Music\iTunes\iTunes Media\Music\Absolute Beginner\The Early Years 1992-1994\05 Planet 2000.m4a.
[5/9/13 5:57:11 PM CEST]    DEBUG    InputStreamReaderThread    (c:\subsonic\transcode\faad) Decoding D:\Users\Admin\Music\iTunes\iTunes Media\Music\Absolute Beginner\The Early Years 1992-1994\05 Planet 2000.m4a took: 6.40 sec. 45.19x real-time.
[5/9/13 5:57:11 PM CEST]    DEBUG    InputStreamReaderThread    (c:\subsonic\transcode\qaac) 4:46.511 (45.3x)
[5/9/13 5:57:11 PM CEST]    DEBUG    InputStreamReaderThread    (c:\subsonic\transcode\qaac) 4:49.297 (45.2x)
[5/9/13 5:57:11 PM CEST]    DEBUG    InputStreamReaderThread    (c:\subsonic\transcode\qaac) 12758016/-1 samples processed in 0:06.395
[5/9/13 5:57:11 PM CEST]    DEBUG    InputStreamReaderThread    (c:\subsonic\transcode\qaac) Overall bitrate: 110.335kbps
[5/9/13 5:57:11 PM CEST]    DEBUG    InputStreamReaderThread    (c:\subsonic\transcode\qaac)
[5/9/13 5:57:11 PM CEST]    DEBUG    InputStreamReaderThread    (c:\subsonic\transcode\qaac) stdin.aac
[5/9/13 5:57:11 PM CEST]    DEBUG    InputStreamReaderThread    (c:\subsonic\transcode\qaac) ERROR: Cannot seek back the input

[lvqcl]

Try "qaac -a %bk --adts - -o -"
(if you want qaac to output to stdout)

This post has been edited by lvqcl: May 9 2013, 17:33