GUIDE: Stereo to 5-Channel Surround

Surcode DTS 1.021
If any programmers are interested, I have pulled apart the Registry settings and the .SSF files. If interested I can give the info if wanted. May be easier to change the registry settings then call surcode. As do not have to load in .SSF file. But surcode keeps track of the last used .SSF file in the registry also. Does it load it back in on open?.(Better check this one out). I also found that if you change the .SSF file you have to restart Surcode, as it goes through the motions of loading again, but ignores the changes - unless restarted-reloaded.

I was going to program something but have got the Delphi version of the VST templates and am investigating some interesting options. So sidetracked.
Have made great strides in a DTS decompiler, but nothing to give yet. Actually I don't know if I will release it as it uses the routines and ideas from the Patent and ETSI tech specs. So copyright issues???? (I only attack this project when in the mood as it does not seem to be as easy as first thought )

[DSPguru]

what's going on with your brackets ?
reminder : every section should start with an openning bracket and end with a closing bracket.

it better look like this :
BeSweet.exe -core( -input track01.wav -output "C:\BeSweetv1.5b20\mono\temp-" -logfile c:\dts.txt ) -ota( -g peak ) -surdvd( -b 1536 -output "C:\DTS Rips\Besweet\track01.dts" -path "C:\Program Files\Minnetonka Audio Software\SurCode DVD DTS" ) -split( -start 1 )

@EoH
assuming you created a 32bit float wave with Bidule, -ota( -g peak ) will normalize that wave just like -ota( -g max ), but will do it in a single pass instead of two-pass (=faster).
for confirmation, i guess we'll have to wait to kempfand's logfiles.
btw, multichannel, 32bit floating-point waves bigger than 2gb are supported as well, so i don't see any reason why not to create 32bit waves. (unless you're short in H.D. space)

Originally posted by Eye of Horus
I had a spike today of 6 seconds... !!! ???

interesting.
can you attach the first 10kbytes in here ?


@specise_8472
BeSweet doesn't use a .ssf file, it sets the registry keys and call surcode.

[Eye of Horus]

I doubted a long time if I should post this, but I do.....

Kempfand and I developed an other method to make surround sound from stereo material with the use of Ambisonics. We use several tools to get to that with the help of the Plogue Bidule package. We wrote the guide from 2 channels to 5 or 6. I made a small (very limited) additional chapter about ways to convert these 5 or 6 channels to AC3 or DTS. But that's only there for completion !
What we see happen now, is that this is becoming the xxth thread about commandlines in Besweet. That's not where we wrote this guide for and not what we would like to see in this thread. IMHO it's off topic , because there are already several threads about that topic. So why start another one here ? (gee, feel like a "mod" now )

It's not my intention to sound pedant or to insult anyone, because I admire the works of DSPGuru a lot and I am a regular user of his tools. But this is IMHO not the right thread.

This current thread is for discussing the pros / cons of the method itself, and how to use & apply the tools (B-Proc, P-Pan, etc).

I hope to bring some structuring into this, helping the readers to keep the overview and get quick answers

For BeSweet usage, there are the DSPGuru / Doom9 guides in addition to plenty of threads.

And I think that for BeSeet development, a new one should be opened with a meaningful header.

I know there are a lot connections between all kind of soundprocessing-methods and one or other "Besweet"-family member, but I hope we can bring this back a bit to the start..... :

How to get the ultimate method to convert stereo to surround with software !?

with kind regards,

Eye of Horus

[Eye of Horus]

interesting.
can you attach the first 10kbytes in here ?

I can't..... I don't keep the failures :-) I wish I had, but it was between 5 and 6 seconds one large block that was not equal on all channels. 4 were the same, the other two were a fraction shorter.
The file I was converting was a stereo WAV with a length of 7 minutes.
In future I will keep this kind of faults in a separate folder....

Anyway I discovered more strange things today.
I converted a complete ripped stereo album to a 6CH file.
Shoot....... forgot to reset to 44.1 Khz in Bidule.
I changed the setting to 44.1 Khz, ended and restarted the program as suggested. Looked again at the preferences : indeed 44100 and started the 50+ minutes file for the second time. And....... ended again with a huge 6CH file in........48 Khz !!
I completely reset the machine (OFF/ON !), started Bidule again, checked the preferences and did the file again.
This time it was fine. (150 minutes !) Something's not quite right :-)

About the real-time versus speed : wouldn't it be simple to implement : no output module loaded (soundcard, midi, ms-mapper) then : SPEED. Output module loaded : real time. (I'm not a programmer, but this came into my mind.....).


grtz,

EoH

[DSPguru]

Originally posted by Eye of Horus
I made a small (very limited) additional chapter about ways to convert these 5 or 6 channels to AC3 or DTS. But that's only there for completion !

yes, you are right.
the main issue to deal with is the 2ch to 5ch/6ch process.
i myself don't enjoy those commandline-discussions threads, mainly, Because it's either documented or supported by DD's GUI.
note that my only motivation was to enhance my tool to complete your needs, as far as i understand them. (+have the time+have the interest).

anyway, thread splitted.

Cheers ,
Dg.

[bitsnbytes]

The readme was short n sweet so I dug up A page on this kind of filtering... I believe the best choice for music is the lowpass filtering
which would set it @ 80hz or lower some may like to test 100 and 125 but I think would be too bassy. I set mine to 60hz.125 would proly be suitable for explosive movie.

so in HNM filter first slider "filter" set to desired hz range so 100 hz would let 100 or less
pass.

the "cutoff filter" I believe sets too small of A slope. set to "0".

"resonance filter" Another steep cutoff filter
again I recommend against it. set to "0"

"slope" I have no idea but I presume its A no. set to "0"

"InAmp" I presume A pre Gain I leave at .500
"OutAmp" I presume a post Gain I set it to .380


I suggest some more research on these options
here's one http://home.cdsnet.net/~roberth/bandpass.htm


For those on A MAc I found some free VST :-(
http://www.arboretum.com/download/do..._freeware.html


A filter info page from above site:

http://www.arboretum.com/support/man...c_filters.html

[ilmanu]

when i try to convert wav file in a 6 mono wav besweet crash....on win2k....
why?
help me please


ps: i use
C:\audio\beswe\BeSweet.exe -core( -input f:\6ch.wav -output f:\wav\final- -type wav -6ch ) -ota( -G max )

[Eye of Horus]

Originally posted by ilmanu
when i try to convert wav file in a 6 mono wav besweet crash....on win2k....
why?
help me please


ps: i use
C:\audio\beswe\BeSweet.exe -core( -input f:\6ch.wav -output f:\wav\final- -type wav -6ch ) -ota( -G max )

I have no idea, but perhaps DSPGuru does.
In the meanwhile, you can try this program to get the 6 wav's.
Only you don't have the 'gain' option.... but that can be done with other programs too.
http://akson.sgh.waw.pl/~wj23277/wav2wav6.zip

good luck,

EoH

[ilmanu]

well, now there is a new problem plogue generate a corrupt file, the original is 45:34:700 but the new 6wav file is 74:56:650 why is too long then original?

[bitsnbytes]

A little note when playing music back.

Ambisonics is not limited to creating a reverberant rear sound field, and requires a different arrangement of speakers. Also, with Ambisonics all speakers cooperate to localise sounds, so the front-rear time delay is unnecessary (and would be detrimental).

[Eye of Horus]

Originally posted by ilmanu
well, now there is a new problem plogue generate a corrupt file, the original is 45:34:700 but the new 6wav file is 74:56:650 why is too long then original?

Is it +74 minutes after you made the 6 mono files ?

EoH

[ilmanu]

yes after....but i have found the problem.....i made a 16bit wav 2ch and try to convert in 5.1 24bit...

[Eye of Horus]

Originally posted by ilmanu
yes after....but i have found the problem.....i made a 16bit wav 2ch and try to convert in 5.1 24bit...

OKAY

EoH

[ilmanu]

well, my english is good like your italian

[Umma]

Eye of Horus:

Might it be possible to have your original thread on stereo-to-dts conversion archived somethere on this site? I find myself referring back to it for the absolutely fascinating information that you all put in there in the course of that year that it was posted.

All the posts here have been extremely informative and I've been pleased with my own ambisonics results. I'm still playing with the plogue bidule...I guess I'm just used to the 'old' way...I'll prolly feel better when I get the batch mode to work in Besweet. Drivin' me nuts right now.

Thanks everyone for your contributions! I've been lurking here for quite some time, now, and have learned quite a bit (though I'm no pro...)!

[Eye of Horus]

Originally posted by Umma
Eye of Horus:

Might it be possible to have your original thread on stereo-to-dts conversion archived somethere on this site? I find myself referring back to it for the absolutely fascinating information that you all put in there in the course of that year that it was posted.

All the posts here have been extremely informative and I've been pleased with my own ambisonics results. I'm still playing with the plogue bidule...I guess I'm just used to the 'old' way...I'll prolly feel better when I get the batch mode to work in Besweet. Drivin' me nuts right now.

Thanks everyone for your contributions! I've been lurking here for quite some time, now, and have learned quite a bit (though I'm no pro...)!

Thanks for the kind words !
That playing in Plogue Bidule is something I still do everyday. It's so nice to hear every change you make in real time ! I only wish some of these VST's were better documented !

Here is the link to the old one :
http://forum.doom9.org/showthread.php?s=&threadid=29277

(BTW a simple search on "ambisonics" would have brought you there too )

kind regards,

EoH

[TRILIGHT]

Thanks. I finally got it running with this command line...

BeSweet.exe -core( -input E:\TEST.wav -output E:\temp- -6chfloat ) -ota( -g peak )

I encoded using SoftEncode but I'm not overly impressed with what I'm hearing. Sure, it's clean and comes out all of the speakers (I was expecting something a bit more intelligent but vocals are in the rear also), but it is not as clear as the original. I can compare the two and the extracted WAV obviously has more clarity than what I'm hearing in the AC3 that was created. Any idea what could be causing this loss of clarity?

Just to clarify some info, what I did to run my particular test was to extract Track 01 on my Norah Jones CD. I resampled the WAV in SoundForge to 48kHz (if I didn't do this, my final result was pitch shifted up). After, I processed the WAV in Bidule as per the guide, then I split with the above BeSweet commmand line, and then encoded in SoftEncode. Any advice? Thanks!!

[kempfand]

@ TRILIGHT. Glad the technical side now worked out.

To the method: Depending on what you expect from Ambisonics, you will be delighted or disappointed. This method does not attempt to seperate isolated instruments (or voice) to discrete channels. It's more about producing a 'sound-image' in a defined space (with dimensions, walls, reflections etc) and then 'recording' that in what is called an Ambisonic B-file (1st order in this case).

I suggest you take some various samples (only 1 treack each) and see what it does for you. For my personal taste, I had some excellent results with Kate Bush, Pink Floyd, Massiv Attack, Hooverphonic, just to name a few). I had bad results on recordings that already were heavily 'manipulated' during a/o after recording (echo etc).

Regarding the method, some additional reading can be found at:

Richard Elen's page

Martin Leese's FAQ page

Ambiophonics

Technical Papers

Kind regards,

Andreas

[bitsnbytes]

Bit of An odd question.
I have A Mono .wav file... what do you suggest I do to get it mixed in surround?


Ok got it. Made a copy of the mono wav selected 2 1 channel players linked playing/playing.
first 1 also linked to recorder playing/recording.

so my subject :Though it was recorded on two-track tape, only the mono master of 'Love Me Do' was kept, so the song has never been available in true stereo.

as I just listened to the finished piece not bad not bad, one thing I've noticed Ambisonic seems a little bassier .. like I make A 2ch +LFE from the same wav.. not so much bass hmmm
maybe LFE should not run through ambisonic
but straight to the recorder what u think?

Hah beat you. Yes Straight to the recorder For LFE
Otherwise over bassy response is made.
IE That soft rock gonna sound like a Mad Chatterhouse or whatever the bloke said that cracked me up

[Eye of Horus]

@bitsnbytes

Any new bidules ???

please post them......


TIA.....


EoH