Help - Search - Members - Calendar
Full Version: Aotuv b3 bug
Hydrogenaudio Forums > Lossy Audio Compression > Ogg Vorbis > Ogg Vorbis - General
cyborg
I got odd results when I was testing the new "aoTuV b3" encoder (vorbis_vendor tag=AO; aoTuV b3 [20041120] (based on Xiph.Org's libVorbis), oggenc2.3aoTuVb3.zip from rarewares.org). I used this command line to encode a file:
CODE
-m 191 -M 193 --advanced-encode-option "bitrate_hard_max=193" --advanced-encode-option "bitrate_hard_min=191"

If I have understood correctly, the resulting file should be ~192 kbps VBR file, but encoded file's bitrate was 3 kbps and file size was 70.61 kB. The original file size was 33.1 MB. The original WAV file's length was 3:16.826 and the encoded vorbis file had the same length. The music lasts about 2 seconds in the beginning of the vorbis file, rest of the file is silent. If I remove the advanced encoder options and use this command line:

CODE
-m 191 -M 193


The file is encoded normally, even though the average bitrate is over 193 kbps, it's 194 kbps. If I use the "-M" switch only, the file is encoded normally. Sometimes the average bitrate is lower than what is specified in "-M" switch, sometimes higher.
DreamTactix291
Tried the top commandline and experience the same bug. I ended up with a 2kbps file. Same issue with the music starting and then stopping completely.
john33
This has actually come up before, IIRC. It is either an error in the code, or an error in understanding of the use of those 2 advanced encode option switches. The values attached to those options need to be expressed in bps, not kbps as supposed. If you rerun using 193000 and 191000, you will find you get the result you're looking for. BTW, the values in the -m and -M switches should be in kbps, as they are, which is no doubt where the confusion arises.

As mentioned, I've never been sure whether the code should be amended to take the kbps figures, which would bring a measure of consistency, or whether it is intended to be the way it is!! unsure.gif
john33
BTW, this applies to ALL versions of oggenc, not just the one referred to above. wink.gif
cyborg
@john33

QUOTE
This has actually come up before, IIRC. It is either an error in the code, or an error in understanding of the use of those 2 advanced encode option switches. The values attached to those options need to be expressed in bps, not kbps as supposed. If you rerun using 193000 and 191000, you will find you get the result you're looking for. BTW, the values in the -m and -M switches should be in kbps, as they are, which is no doubt where the confusion arises.

As mentioned, I've never been sure whether the code should be amended to take the kbps figures, which would bring a measure of consistency, or whether it is intended to be the way it is!!


thanks john33, I'll give it a try. smile.gif
DreamTactix291
That makes sense, but the inconsistency is kinda annoying. Not that I see myself using those switches a lot anytime soon wink.gif
cyborg
I tried those "advanced-encode-option" settings with 191000 bps and 193000 bps, they seem to work but the bitrate still goes over the maximum bitrate (193 kbps). Is this supposed to happen? Aren't those switches supposed to limit the maximum and minimum bitrate? I tried those settings with "aotuv b3" and "1.1" encoders and the results were same, bitrate went over the specified maximum bitrate. I would also like to know how can I encode true CBR files with Ogg Vorbis? I have tried the --advanced-encode-option "bitrate_hard_max=xxxxxx", --advanced-encode-option "bitrate_hard_min=xxxxxx" and --advanced-encode-option "bitrate_average=xxxxxx" settings combined with -b, -m and -M settings, but they don't seem to work. sad.gif
john33
The truth is there is NO true CBR mode in vorbis, it is only an emulation. By design, it is a VBR encoder and any other mode is merely emulated by playing with parameters to the encoder and forcing behaviour which was not really originally intended. This is why the recommendations are always to use the 'quality' settings.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2008 Invision Power Services, Inc.