Additions and comments on "GUIDE LIST: Stereo-to-surround Conversion.."

[tebasuna51]

The average bits per sample is:

BitRate / (SampleRate x NumChannels)

An ac3 5.1 48 KHz 6 channels 448 Kb/s have:
448000 / (48000 x 6) = 1.555

An ac3 5.1 44.1 KHz 6 channels 448 Kb/s have:
448000 / (44100 x 6) = 1.693

You can't say "i have always the ac3 in 32 bit", the ac3 isn't 32 bit, you use a decoder than decode the ac3 in samples of 32 bit.

You can instruct the decoder to output 16 bits samples for the same ac3, is your decision (in the decoder) not a ac3 feature.

[raquete]

...you use a decoder than decode the ac3 in samples of 32 bit.You can instruct the decoder to output 16 bits samples ...in the pc i understand but how my standalones players and DD decoder manage the files as i can't instruct both?
how the decoders will "know" "what is what" and "who is who" and how much bit to play? (16,32 etc)

[foxyshadis]

Standalones will probably all use 24-bit for everything, unless they're really cheap and limited to 16-bit.

It would be interesting to renormalize a 24-bit track to 32-bit -140dB and find out whether the codec really is capable of 32-bit, in the absense of other factors.

So I tried this. :p The input wav can be renormalized to sound just like the original, but all I get is a buzzing out of the ac3. At -120dB, it gives something resembling the sound effects to Atari games like Pole Position. :D Oh well. :p

[jordisound]

Well, I've done a test, with 32-bit waves, using the script to merge 6 wav and using aften like i told you in my last post. the new ac3 was only 30', 1/3 the length of the audio track (1h 30' aprox.).
I've tried again without global OPT_AllowFloatAudio=True in the script (it seems aften works 16bits). The ac3 was ok, the original lenght

[tebasuna51]

Well, I've done a test, with 32-bit waves, using the script to merge 6 wav and using aften like i told you in my last post. the new ac3 was only 30', 1/3 the length of the audio track (1h 30' aprox.).
I've tried again without global OPT_AllowFloatAudio=True in the script (it seems aften works 16bits). The ac3 was ok, the original lenght

You need Aften rev449 or new and a use the aften parameter:
-readtoeof 1

With this parameter Aften ignore the wav header field wrong than say is only 30' and continue encoding until the end of file 1h 30'.

When the AviSynth output is converted from 32 float to 16 bit (without global OPT_AllowFloatAudio=True) the size is half and the new problem limit is at 124 min.

[foxyshadis]

That parameter should probably be default for any 6 channel or 32-bit input. Or, if you want to get tricky, just calculate whether the input header is pointing to a length of right around 2GB or 4GB, and if so assume the header is broken. That way Aften will work more simply for most people.

[tebasuna51]

That parameter should probably be default for any 6 channel or 32-bit input. Or, if you want to get tricky, just calculate whether the input header is pointing to a length of right around 2GB or 4GB, and if so assume the header is broken. That way Aften will work more simply for most people.
I agree, but is not the opinion of aften creator.

You can see this post (http://forum.doom9.org/showthread.php?p=962080#post962080) and related.

[jordisound]

You need Aften rev449 or new and a use the aften parameter:
-readtoeof 1
Thank you tebasuna, it works.

[omega6666]

Try here (http://rapidshare.com/files/21180963/LeeAudBi.zip.html).

Tebasuna, I'm looking for your LeeAudiBi. Can I still download that one somewhere?

Thx...

[tebasuna51]

Tebasuna, I'm looking for your LeeAudiBi. Can I still download that one somewhere?

Not yet finished, and without guaranty, try this new version (http://www.mytempdir.com/2048795) with GUI. Say me if work for you.

Edit: New link and version LeeAudBi03b.7z (http://www.sendspace.com/file/7izmtk)

[omega6666]

Wow thanks! That showed me more info than I've seen so far! Unfortunately I'm still not sure if this solves my problem described here (http://forum.doom9.org/showthread.php?t=131120). I got this info from the first frame of the problematic dts file;

First frame data:
----------------------------------------------------------
Bytes before header 120867 ; Frame 1 in position 120868
CRC present 0 : Not
Number of PCM Sample Blocks 15 : 512 samples/frame
Primary Frame Byte Size 2012 : 2013 bytes/frame
Audio Channel Arrangement 9 : 5 C + L + R + SL + SR
Core Audio Sampling Frequency 13 : 48 kHz
Transmission Bit Rate 24 : 1536 Kb/s
Embedded Down Mix Enabled 0 : Not
Embedded Dynamic Range Flag 0 : Not
Embedded Time Stamp Flag 0 : Not
Auxiliary Data Flag 0 : Not
Mastered in HDCD format 0 : Not
Extension Audio Descr. Flag 0 : Channel Extension (XCh)
Extended Coding Flag 0 : Not
Audio Sync Word Insert. Flag 1 : Sub-sub-frame
Low Frequency Effects Flag 2 : Present, interpolation factor 64
Predictor History Flag Switch 1 : Yes
Multirate Interpolator Switch 0 : Non-perfect Reconstruction
Encoder Software Revision 7 : Current
Copy History 1 : Definition deliberately omitted.
Source PCM Resolution 6 : 24 bits
Front Sum/Difference Flag 0 : Not
Surrounds Sum/Difference Flag 0 : Not
Dialog Normalization Param. - 0 dB

The original where this was demuxed from was 96/24 for sure. I would like to know if the extension data for the 96 khz is still there, but is maybe flagged wrongly so that my receiver simply won't use the available extension data, or if it is just missing? Here (http://rapidshare.com/files/67478370/dts_samples.rar.html) are 2 dts files (short samples) where there's one that works fine as 96/24 and the other one just plays as 48 khz, even though the original was 96 khz. Can you maybe tell me if there's still extension data there to let my receiver play this one as 96/24?
Oh, and thanks again for your very informative program! I think I'm gonna use that one very often! :thanks:
[edit:] The proper dts file shows this flag; Extension Audio Descr. Flag 2 : Frequency Extension (X96k), and the faulty one shows this; Extension Audio Descr. Flag 0 : Channel Extension (XCh). When I mux the faulty dts track with some black bmp to a VOB file, and change the domain stream attributes of the dts to 24 bits and 96 khz with PgcEdit, and then demux the dts file again, it seems it hasn't changed the Channel Extension flag! How can I do this?

[tebasuna51]

@omega6666
See my answer in the thread 96/24 data in dts extension field (http://forum.doom9.org/showthread.php?t=131120) to keep clean this thread.