GUIDE: Stereo to 5-Channel Surround

[Covak]

pack2x2to4 isn't entire command line based (it asks you questions), which makes it easy to run from anywhere (including Explorer). But for ambidec you have to specify everything on the command line.

Let's say you have everything in F:\Temp...

F:\Temp>dir
...
01/10/2000 13:51 258,048 ambidec.exe
26/06/2003 12:03 896,463 sample.wxyz
...

Then you'd go:

F:\Temp>ambidec -rSurround1 sample.wxyz

If you wanna chat me: ICQ# 18969614, AIM: GaryCartel, MSN: covak@hotmail.com, it might be easier to help that way if you still need it :)

[matrix]

Has anybody tryed ambidec with a large file?
I have this movie soundtrack, 135min long, and I see some other people are thinking to do this.
I run it through the whole process, and I ended up with a 135 min wxyz file. But then, when trying to run it in ambidec,(I had some problems, but with the great assistance from Covak, I was able to do that too)... well, I got an 81 min long wav file.
Now I don't know if I did something wrong in the process, but I don't think so since the wxyz file was the right length.
Somehow ambidec did only 81 min out of it. And I run it through ambidec a few times. The command was the one provided in the guide.
ambidec -r Pentagon stereo.wxyz and once ambidec -r Surround1 stereo.wxyz
Every time I ended with the short file.

So it mekes me belive ambidec has some problems with large files.?
Can someone confirm this or have I really done something stupid.
Anyway, is there a way to split the wxyz file?
If I still had the wx and yz files, I could do it, but I deleted them.
I'm curious though, I did something wrong or it can't process large files.

Regards

[Eye of Horus]

Originally posted by matrix
Has anybody tryed ambidec with a large file?
I have this movie soundtrack, 135min long, and I see some other people are thinking to do this.
I run it through the whole process, and I ended up with a 135 min wxyz file. But then, when trying to run it in ambidec,(I had some problems, but with the great assistance from Covak, I was able to do that too)... well, I got an 81 min long wav file.
Now I don't know if I did something wrong in the process, but I don't think so since the wxyz file was the right length.
Somehow ambidec did only 81 min out of it. And I run it through ambidec a few times. The command was the one provided in the guide.
ambidec -r Pentagon stereo.wxyz and once ambidec -r Surround1 stereo.wxyz
Every time I ended with the short file.

So it mekes me belive ambidec has some problems with large files.?
Can someone confirm this or have I really done something stupid.
Anyway, is there a way to split the wxyz file?
If I still had the wx and yz files, I could do it, but I deleted them.
I'm curious though, I did something wrong or it can't process large files.

Regards

Never tried such a big file, because Aurora and Cooledit won't accept such lengthy ones !

I didn't try it yet on my new PC, but my Pentium 3 500 with 384 MB couldn't handle files longer than 9-10 minutes.

EoH

[matrix]

Never tried such a big file, because Aurora and Cooledit won't accept such lengthy ones !
Oh, but they do. Both of them. That's what I used to get as far as ambidec.
Well, it takes forever even with a P4 2.66Ghz, let alone the random aurora crashes, but I can do the job.
I'm doing it again right now, but I'm just going to split it before.
Hopefully I'll see soon what I spent all this time for.

Matirx

[matrix]

Originally posted by Bert Schlichter
Thanks to the guide from EoH we can convert stereo files into 5 channel Ambisonic DTS or DD5.1...Great EoH ;-)
For those of you who want to use Ambisonic in their music-clip DVD's. Here is a small addendum to this guide to keep the audio files 100% in sync with the video. As i was told the Aurora plug-in produces so called 'room pulses' and therefore increase's the filesize with two times 0.3715 along the whole 'EoH guide' route.
So we have to correct this to keep the video in sync.

Here are the steps:

1: Use Smartripper to rip the various chapters.
2: Use TMPG to demux the .vob to a .m2v and and a 48 Khz .wav file.
3: Use Cooledit to down-sample the 48 Khz file to a 44.1 Khz .wav file.
4: Follow the 'EoH guide' exactly until you have got 5 mono files produced by Bsplit.
5: Cooledit: Pulldown menu Timebar - Define custom frames - Custom Time Code Display - Put: 10.000 frames/sec. 6: Find for each file the exact start-pulse and delete 0:3715 towards the middle from that point.
7: Find for each file the exact end-pulse and delete 0:3715 towards the middle from that point.
8: Use Cooledit to up-sample each 44.1 Khz file to a 48 Khz .wav file.

Use SurCode (DTS) or SoftEncode (DD 5.1) and your favourite Authoring program to make a Ambisonic DVD!

Have Phun,

Bert Schlichter :-)

In step 6: Find for each file the exact start-pulse
What start-pulse are you\is he refering to, and how do you find it?

[Malow]

is the same prob in the first others encodes,
the volume level is not procise, is just a "medium" too you imagine the "peak level" of the audio channel.

i do another on encode again, and just on convolve the channels with the wav´s files provided by ambitool, the level of audio go down.

and im doing exactly as the guide. after use ambidec the channels packed is the same right? only one channel is added and others remain intact right? im seeing now, is right on convolve the levels is modified... maybe use another settings on convolve with aurora?

Thank you for the help.

Malow

[sazanon]

First attempt to script Eyes of Hours Ambisonic Guide

1.- Soft I use
--------------
1.1 - Cool Edit Pro 2.1
1.2.- Aurora's Convolve with Clipboard Module (Convolve.xfm vers.1.0.0.1)
1.3.- Ambisonic Impulse files UHJ_W_4416, UHJ_X_4416 and UHJ_Y_4416 placed in c:\Cool Edit Pro\Imps. Modify the script according to your needs.

2.- Procedure
-------------
2.1 - Load your original stereo 16 bits 44.100 Hz wav file in CEP
2.2 - Options --> Scripts & Batch Processing --> Open New Collection --> Ambisonic.scp --> WXYZ --> Run Script
2.3 - At the end of the process you'll have
2.3.1 - Ambisonic W file in Clipboard 1
2.3.2 - Ambisonic X file in Clipboard 2
2.3.3 - Ambisonic Y file in Clipboard 3
2.3.4 - Ambisonic WX file in Clipboard 4
2.3.5 - Ambisonic YZ file in Clipboard 5

So you just have to copy WX & YZ to your H.D.
Set Current Clipboard 4/5 --> Paste to New --> File save as WX/YZ.wav

3.- Questions I ask myself
--------------------------
3.1 - Why is it that in the script I have to reverse channels in WX to get the same wav file as when I do it manually ?
3.2 - Why YZ has a lower volume as when I do it manually ?

4.- Remarks
-----------
4.1 - Sorry for my english, it's not my mother tongue
4.2 - What ring should we use for a AC3/DTS system ?, Pentagon ?, Surround1 ?
4.3 - Any enhancement or CORRECTION to this script are welcome

5.- Well, here it goes :

Collection: Ambisonic

Title: WXYZ
Description: Ambisonic
Description: ----------------
Description: Load your original Stereo 16 bits 44,100 Hz file first
Description:
Description: At the end of the script, you'll have the 5 processed files as follows :
Description:
Description: Wav W in Clipboard 1
Description: Wav X in Clipboard 2
Description: Wav Y in Clipboard 3
Description: Wav WX in Clipboard 4, Copy to Hard Disk as WX.wav
Description: Wav YZ in Clipboard 5, Copy to Hard Disk as YZ.wav
Description: *************************************************************************
Description: Debe cargarse previamente el Wav original
Description:
Description: Al final del proceso, los 5 wavs procesados estaran como sigue:
Description:
Description: Wav W en el Clipboard 1
Description: Wav X en el Clipboard 2
Description: Wav Y en el Clipboard 3
Description: Wav WX en el Clipboard 4, Copiar a Disco Duro como WX.wav
Description: Wav YZ en el Clipboard 5, Copiar a Disco Duro como YZ.wav
Mode: 2

Selected: none at 0 scaled 2958028 SR 44100
cmd: Channel Both

Selected: 0 to 2958027 scaled 2958028 SR 44100
cmd: Cut
1: 4

cmd: Paste Special
1: 1
2: 1
3: 3
4: 1
5: 0
6: 1
7: C:\Cool Edit Pro\imps\UHJ_W_4416.wav
8: 2
9: 4

cmd: Cut
1: -1

cmd: Paste Special
1: 1
2: 1
3: 3
4: 1
5: 0
6: 0
7: C:\Cool Edit Pro\imps\UHJ_W_4416.wav
8: 2
9: 4

cmd: Aurora\Convolve with Clipboard
1: 69
2: 69
3: 72
4: 16384
5: 131072
6: 2958028
7: 44100
8: 44100
9: 0
10: 0
11: 0
12:
13:
14: 32767
15: 0
16: 0
17: 0

cmd: Convert
1: 44100
2: 1
3: 32
4: -1
5: 0.5
6: 0.5
7: 1
8: 0
9: 150
10: 1
11: 1
12: 16

cmd: Cut
1: 0

cmd: Convert
1: 44100
2: 2
3: 16
4: -1
5: 1
6: 1
7: 1
8: 0
9: 150
10: 1
11: 1
12: 16

cmd: Paste Special
1: 1
2: 1
3: 3
4: 1
5: 0
6: 1
7: C:\Cool Edit Pro\imps\UHJ_X_4416.wav
8: 2
9: 0

cmd: Cut
1: -1

cmd: Paste Special
1: 1
2: 1
3: 3
4: 1
5: 0
6: 0
7: C:\Cool Edit Pro\imps\UHJ_X_4416.wav
8: 2
9: 4

cmd: Aurora\Convolve with Clipboard
1: 69
2: 69
3: 72
4: 16384
5: 131072
6: 2958028
7: 44100
8: 44100
9: 0
10: 0
11: 0
12:
13:
14: 32767
15: 0
16: 0
17: 0

cmd: Convert
1: 44100
2: 1
3: 32
4: -1
5: 0.5
6: 0.5
7: 1
8: 0
9: 150
10: 1
11: 1
12: 16

cmd: Cut
1: 1

cmd: Convert
1: 44100
2: 2
3: 16
4: -1
5: 1
6: 1
7: 1
8: 0
9: 150
10: 1
11: 1
12: 16

cmd: Paste Special
1: 1
2: 1
3: 3
4: 1
5: 0
6: 1
7: C:\Cool Edit Pro\imps\UHJ_Y_4416.wav
8: 2
9: 1

cmd: Cut
1: -1

cmd: Paste Special
1: 1
2: 1
3: 3
4: 1
5: 0
6: 0
7: C:\Cool Edit Pro\imps\UHJ_Y_4416.wav
8: 2
9: 4

cmd: Aurora\Convolve with Clipboard
1: 69
2: 69
3: 72
4: 16384
5: 131072
6: 2958028
7: 44100
8: 44100
9: 0
10: 0
11: 0
12:
13:
14: 32767
15: 0
16: 0
17: 0

cmd: Convert
1: 44100
2: 1
3: 32
4: -1
5: 0.5
6: 0.5
7: 1
8: 0
9: 150
10: 1
11: 1
12: 16

cmd: Cut
1: 2

*************************************************************

Selected: none at 0 scaled 0 SR 44100
cmd: Channel Both

Selected: none at 0 scaled 0 SR 44100
cmd: Paste
1: 1

cmd: Convert
1: 44100
2: 1
3: 16
4: -1
5: 0.5
6: 0.5
7: 1
8: 0
9: 150
10: 1
11: 1
12: 16

cmd: Cut
1: 3

cmd: Paste
1: 0

cmd: Convert
1: 44100
2: 2
3: 16
4: -1
5: 1
6: 0
7: 1
8: 0
9: 150
10: 1
11: 1
12: 16

cmd: Channel Right

cmd: Paste
1: 3

cmd: Channel Both
cmd: Amplitude\Channel Mixer
1: 0
2: 1
3: 1
4: 0
5: 0
6: 0
7: 0
cmd: Cut
1: 3

cmd: Convert
1: 44100
2: 1
3: 32
4: -1
5: 0.5
6: 0.5
7: 1
8: 0
9: 150
10: 1
11: 1
12: 16

cmd: Paste
1: 2

cmd: Convert
1: 44100
2: 2
3: 16
4: -1
5: 1
6: 0
7: 1
8: 0
9: 150
10: 1
11: 1
12: 16

cmd: Cut
1: 4

End:

[jslombar]

To Matrix,

In step 6: Find for each file the exact start-pulse

I'm not sure if this is correct. After I did the ambisonic method,
My wave files were .3715 seconds longer not .743 seconds The DVD I did was Rush Chronicles. To correct the sync problem (audio following the video), I used AC3 Delay Corrector V1.7 on the ac3 file, and used a -372ms delay. Hope this makes things alot easier.


jslombar

[matrix]

My wave files were .3715 seconds longer not .743 seconds
Yeah, same here. I only cut .3715 from the beginning, and I was OK.

Thanks for your answer

matrix

[hammerstein.dacm]

Thanks, looks like a nice guide!
Only missed the link to the obligatory
ambisonic tools.

Is it thinkable to use also mpeg 1 layer 2
audio as input for the creation of this
artificial dolby surround sound?

It would be nice to burn old vhs material
on dvd with a choice for the mpeg 1 layer 2
stereo-sound and a according to this guide
constructed dts track.

But I understand that the lenght of a whole
concert would give serious desynchronisation
issues?


regards

dh

[Eye of Horus]

Originally posted by hammerstein.dacm
Thanks, looks like a nice guide!
Only missed the link to the obligatory
ambisonic tools.

Yep, they were removed. I don't know why....


Is it thinkable to use also mpeg 1 layer 2
audio as input for the creation of this
artificial dolby surround sound?

Excuse me ? When you call it artificial Dolby surround, then you don't have to read or use it. or you didn't understand it.
Use an AC3 encoder. Bye !!!


It would be nice to burn old vhs material
on dvd with a choice for the mpeg 1 layer 2
stereo-sound and a according to this guide
constructed dts track.


You can use it with mpeg 1 layer 2 , but then you are using already compressed material. Don't expect the end result to be as good as when using CD tracks !


But I understand that the lenght of a whole
concert would give serious desynchronisation
issues?


read the rest !


regards

dh

regards,

EoH

[hammerstein.dacm]

EoH!

Sorry, no offense was intended, had the idea
that only careful mixing of the physical instruments
during a recording session could result in real surround
or dts sound.

But my background knowledge indeed is very modest,
I only seek for a practical way to improve the
soundtracks when burning great old music-videos
on DVD.
So far I only managed to get 6 wav-channels from
the mpa soundtrack I captured with my Hauppage PVR 350
from my old videotapes using Besweet.
This seems to result only in ultra speedy soundtracks.

I can imagine that it is boring to get such an ignorant
newby in this discussion, sorry for that.

dh
PS Would be obliged if you could put the lost link back.

[Eye of Horus]

Originally posted by hammerstein.dacm
EoH!

Sorry, no offense was intended, had the idea
that only careful mixing of the physical instruments
during a recording session could result in real surround
or dts sound.

But my background knowledge indeed is very modest,
I only seek for a practical way to improve the
soundtracks when burning great old music-videos
on DVD.
So far I only managed to get 6 wav-channels from
the mpa soundtrack I captured with my Hauppage PVR 350
from my old videotapes using Besweet.
This seems to result only in ultra speedy soundtracks.

I can imagine that it is boring to get such an ignorant
newby in this discussion, sorry for that.

dh
PS Would be obliged if you could put the lost link back.

I suppose reading the whole thing will bring you where you want :-)
I know it's a lot, but it helps to understand the process and it's dfferences from the "Dolby" way.
The guide is pretty easy to follow. Just give it a try with one of your favorite tracks.

Kind regards,

EoH

[nuked]

I read the whole thing and it didn't describe the specific mix used for stereo compatibility at all. But anyway, if you don't start with something encoded with a real x and y signal then that's that. There is out of phase stuff that can be pulled to the back using the basic dolby surround trick and it sounds really cool and has some actual realistic merit to it, some artificial pro-logic steering and maybe it sounds cooler but you can't really determine where the signal really came from from a regular stereo source at least, unless as said it was encoded with a special mic in the first place. I believe you can add some room characteristics that may sound awesome, but that's a sound effect, not a realistic representation. All there is is sum and diference, nothing else, ***UNLESS***you can fully deconvolve the phase information from the amplitude information, but I think this requires some fancier frequency analysis, not just mixing matrices. That would be impressive. Then maybe I could be convinced some extra information could be pulled out, but I'd still have to be convinced. Is this what's going on? If not basically seems like you're doing a round about dolby surround upmix by using the diference signal (which CAN sound AWESOME! ) , and why bother to encode that in an AC3?

nukeD

[Eye of Horus]

Originally posted by nuked
I read the whole thing and it didn't describe the specific mix used for stereo compatibility at all. But anyway, if you don't start with something encoded with a real x and y signal then that's that. There is out of phase stuff that can be pulled to the back using the basic dolby surround trick and it sounds really cool and has some actual realistic merit to it, some artificial pro-logic steering and maybe it sounds cooler but you can't really determine where the signal really came from from a regular stereo source at least, unless as said it was encoded with a special mic in the first place. I believe you can add some room characteristics that may sound awesome, but that's a sound effect, not a realistic representation. All there is is sum and diference, nothing else, ***UNLESS***you can fully deconvolve the phase information from the amplitude information, but I think this requires some fancier frequency analysis, not just mixing matrices. That would be impressive. Then maybe I could be convinced some extra information could be pulled out, but I'd still have to be convinced. Is this what's going on? If not basically seems like you're doing a round about dolby surround upmix by using the diference signal (which CAN sound AWESOME! ) , and why bother to encode that in an AC3?

nukeD

"Then maybe I could be convinced..........."

The only thing that convinced me when wroting the guide in the first place : my own ears !
I advice you strongly to do the same instead of all this theoretical bla bla !
And yes there are different ways to make sound awesome, if you're a professional or at least know how to use professional software and when you are pretty much into music in the first place.
Ooh, you can get such excellent results with Reason or MX5.1 or.....
Problem is that no one writes an understandable easy to follow guide for that :-) You need to have a technical background to be able to use these "tools" :-) or even read the manuals !

Where is the part of putting that into AC3 ????????????

EoH

[daphy]

@ Eye of Horus

youre right!;)
i dont think the sense of this guide is to convince the whole world ;)

@ nuked
love it or leave it :sly:

CYA Daphy

[Eye of Horus]

Originally posted by daphy
@ Eye of Horus

youre right!;)
i dont think the sense of this guide is to convince the whole world ;)

@ nuked
love it or leave it :sly:

CYA Daphy

THNX !!! :-)

EoH

[nuked]

"Where is the part of putting that into AC3 ?"

I apologize, by ac3 I just meant more than 2 channels, dts, 5 channel wav whatever you like. My point really just was why bother encoding in 5 channels what can be encoded in 2. Don't get me wrong though. If this "effect" sounds cooler than dolby's effects then of course, in the end it's about what makes you smile when you hear it. But don't jump on the defense when people in passing and with no offense call it an effect. Unless you can prove otherwise, I think most of us are just saying it seems like an "effect", which is ok. Personally I believe it's a little more than an effect, as is dolby surround (the pro logic additions are just effects), I just don't believe it's any less of an effect than dolby surround either.

nukeD

[nuked]

Ok, well your link does NOT describe the mixing process in detail, but
I'm starting to find more information that does. Interestingly
enough the calculations are in fact done in frequency space
and true phase information as a function of frequency is used..
ie sine vs cos, 90 degree phase as oposed to just non
linearly-indepent polarity info used in normal time-space mixing(dolby).
I've also found references
saying that ambisonic mics are not used anymore because you can't
really go back from UHJ tyo B-format anyway. This will always be true
of course cause it's impossible to keep some of the original L/R
elements from leaking into the front/back info and that's a problem
with surround sound as well. However, it does seem this may do a be
able to do little better job than surround of doing that, since there is
more information to work with. This is all from a few
google hits of not so official sources of info. Maybe I (or someone) will
find something better and more comprehensive about it. Interesting though.

It's not clear to me personally yet, if real info is gained for
material that wasn't encoded with an ambisonic mic, but maybe it is.
If dolby can get some info from regular stereo, maybe this can do it a
little better.

nukeD

[Eye of Horus]

This is how I came on it.
I read a lot of the websites and most of the times I didn't understand it. My maths didn't go that far :-)

But from what I read, it must be possible to get info from stereo and to add info to stereo.
At that same time I was discussing the possibilities of this with a good friend , who told me about Stereotomy, the Alan Parsons Project album. APP is my favorite group (www.thealanparsonspages.com shameless plug for my own homepage :-)) and that friend told me that Stereotomy was recorded in Ambisonics, which is a system that adds surroundness to a soundstream. I read some old reviews of that album and reviewers stated : this is the best sound ever heard in stereo on vinyl. It gives you a sense of surround sound, without losing the band. ?? = a lot of the quad albums were : voice from left front, guitar from right front, drums from rear left and bass from rear right, but where the heck is the band ??
Well, I thought and so did my friend, if our favorite artist which also is one of the greatest producers of all time (Dark Side of the Moon, Year of the Cat etc.) used Ambisonics, we want to know more about it and try to develop a method so we can use it too on our stereo albums !
My friend developed the method, I wrote a guide for it !
I don't understand everything from the mathemetics, I'm not an audio professor like the guys who developed Ambisoncs. I'm just a user who wants to get the best out of his cd's without having to spend thousands of Euro's on a top system.
And... it works ! On most albums it's definitely an improvement ! Don't ask me why... but it is !

Three months ago I bought a Philips SACD player. This player has an upsampling function for stereo CD's. It upsamples them to 192 KHz and that signal is send to the receiver. They sound a lot better !
So here is another mystery I don't understand : why do stereo CD's made in 44.1 Khz sound better when upsampled to 192 Khz ? (11.025 Hz upsampled to 44.100 Hz does NOT sound better !)

I don't understand it, but I hear the difference !

The same goes for Ambisonics !

And for me that's all that counts !

Another point : I already said it above , but it needs some more information : we all assume that the method of Dolby is surround. But it isn't ! It is separation (at least DD 5.1 is !).
For me surround means : surrounded by sound.
The sweet spot of Ambisonics is so large that you can easily move a meter to the right without losing the effect. Try that with DD 5.1 !

well.... enough for now.....
just some thoughts.... or as they say : my 2 cents

kind regards,

Eye of Horus

[nuked]

Cool, yeah floyd stuff has awesome surround quality even in stereo
playback on good speakers. I'd heard they used some advanced
microphone techniques.

I don't know so don't take my opinions on it. But I'm guessing
frequency upmixing just smooths out the waveforms. I've heard
extreme audiophiles claim that the square edges at 44 khz can produce
audible effects, even if we're not directly sensitive to those
frequencies. It's topic of debate and some people still claim vynil
is better because of it. I don't think my ears are that
good. Or maybe they are maybe I just don't pay quite that much
attention.

[StoneRoses]

Eye of Horus,

Thank you for all your hard work. I really appreciated.

But please correct your guide (on the first page) a bit.

Ambidec.exe rig setting should be "Surround1" instead of "Pentagon". This will save a lot of time for newbies (including myself) to figure it out. Most of us not read this thread to page 12.

[Eye of Horus]

Originally posted by StoneRoses
Eye of Horus,

Thank you for all your hard work. I really appreciated.

But please correct your guide (on the first page) a bit.

Ambidec.exe rig setting should be "Surround1" instead of "Pentagon". This will save a lot of time for newbies (including myself) to figure it out. Most of us not read this thread to page 12.

First : thanks for the compliments.... !!

I don't know if there has been an update for Ambidec, but anyway...
All the other software (the roompulses) etc. have been tested with the Pentagon setting and work fine !
In the time I wrote the guide, "surround" in Ambidec didn't give good results. I can't remember if it crashed or just didn't give what we wanted. Or perhaps it wasn't even there... The pentagon option always gives good results ! And.... 5 speakers is just fine for the ones without a subwoofer !

I don't quite understand what you mean with "this saves a lot of time "..... and with the "should be" remark.....

Can you please explain in what sense it saves time and when you say should be, why should it be ?


TIA....

kind regards,

EoH

[StoneRoses]

I follow your guide and got strange volume problem like Malow got (see Page 10 of this thread) Covak suggest Malow to use "Surround1" rig instead of "Pentagon" and explain a bit more about the rig.

I did some research to confirm Covak's finding and found info about default rig parameters that "ambidec.exe" used for decoding the B-format file. (from the ambidec author's website)

Here is the URL:
http://www.muse.demon.co.uk/ref/speakers.html

The speaker position relative to the listener are present in [X,Y,Z] vector format. (X+ is the direction toward the listener, Y+ is on the left and Z+ is the direction above the listener)

And here is my quick and dirty drawing of "Pentagon" and "Surround1"

http://www.thai.net/stoneroses/ambisonics/Rig.jpg

On that website, there is the nice b-format player called "Ambisonic Player", it have the illustrations of some rigs on its options menu.

[Eye of Horus]

Originally posted by StoneRoses
I follow your guide and got strange volume problem like Malow got (see Page 10 of this thread) Covak suggest Malow to use "Surround1" rig instead of "Pentagon" and explain a bit more about the rig.

I did some research to confirm Covak's finding and found info about default rig parameters that "ambidec.exe" used for decoding the B-format file. (from the ambidec author's website)

Here is the URL:
http://www.muse.demon.co.uk/ref/speakers.html

The speaker position relative to the listener are present in [X,Y,Z] vector format. (X+ is the direction toward the listener, Y+ is on the left and Z+ is the direction above the listener)

And here is my quick and dirty drawing of "Pentagon" and "Surround1"

http://www.thai.net/stoneroses/ambisonics/Rig.jpg

On that website, there is the nice b-format player called "Ambisonic Player", it have the illustrations of some rigs on its options menu.



isn't your drawing incorrect about the pentagon.... ?

shouldn't it be something like this :


Pentagon

0
0 0

0 0

Surround

0 0 0


0 0


And although I'm open to any suggestion : Did you try it out and listen to the differences ?

And.... please keep this in mind : I wrote the guide with the material I got . Over a year ago. I cannot recall exactly why we used pentagon or what was wrong with the surround option, back then.
I have said that before in this long living thread..... I am NOT an expert in Ambisonics. When you point me to a mathematical explanation about why your opinion is right, I can only say : I don't know, because I don't understand those mathematics ! . PERIOD !

When I started this thread it was very simple : a friend let me listen to some music he converted to the format the way described in my guide. I thought it sounded excellent !! (One of the best conversions I did is IMHO still "Jeff Waynes - War of the Worlds").
My goal was to make that whole difficult process as clear as possible to as many people as possible, from beginner to the one that can handle and understand those mathematics. And please don't forget that a year ago, there was nothing else that came close to my guide ! (And I mean the "how to convert stereo to 5 channel surround with Ambisonics") And believe me.... I did search and search and search.... :-) There just wasn't anything that made it clear and that could be used by a beginner. I didn't write Ambidec, I didn't record the pulses, I didn't write Aurora, I didn't write Cooledit. I just wrote the guide that explained the "How to....".

So , to resume...... I don't doubt what you say, I say : convince me with material, not with theory. Let my ears make the decision !

In other words..... post a converted sample to alt.binaries.sound done with the Pentagon option and the same sample done with the Surround option.....

In this case my ears will give better judgement than my brains :-)

with kind regards,

Eye of Horus

[Covak]

If you run "ambidec -lv" it will give you the details on its standard rigs. The coordinate system is as shown in the picture here:
http://www.ramsete.com/aurora/conversion_between_uhj_and_b.htm

StoneRoses drawings are accurate.

I think Visual Virtual Microphone, with its 5.1 preset, is much better than ambidec anyway.

VVMic is here: http://mcgriffy.com/audio/ambisonic/vvmic/

I believe the 5.1 preset is based on G-Format (the "5.1" button used to be labelled "G-Format" in earlier versions), which pretty much makes it THE way to decode B-Format for standard 5.1 setups.

More on G-Format here: http://www.ambisonic.net/gformat.html

[StoneRoses]

EoH,

According to the info on ambidec website. Pentagon rig is like it was in my drawing.

I don't understand all the mathematic behide ambisonic eigther. The reasons why I pointed you to that page are mainly for the standard rig setup information not the equations.

The speaker positions of "Pentagon" rig are:

<X, Y, Z>

1: <0.8090,0.5878,0.0000>
2: <-0.3090,0.9511,0.0000>
3: <-1.0000,0.0000,0.0000>
4: <-0.3090,-0.9511,0.0000>
5: <0.8090,-0.5878,0.0000>

You can compare other rig types with the illustration from "Ambisonic Player"
See this picture for some of them (no Pentagon and Surround1 rig in the player)
http://www.thai.net/stoneroses/ambisonics/SomeRigs.gif

Here is the info on that page
Stereo
1: <0.0000,1.0000,0.0000>
2: <0.0000,-1.0000,0.0000>

Square
1: <0.7071,0.7071,0.0000>
2: <0.7071,-0.7071,0.0000>
3: <-0.7071,-0.7071,0.0000>
4: <-0.7071,0.7071,0.0000>

Cube
1: <0.5774,0.5774,-0.5774>
2: <0.5774,-0.5774,-0.5774>
3: <-0.5774,-0.5774,-0.5774>
4: <-0.5774,0.5774,-0.5774>
5: <0.5774,0.5774,0.5774>
6: <0.5774,-0.5774,0.5774>
7: <-0.5774,-0.5774,0.5774>
8: <-0.5774,0.5774,0.5774>

You can also use the custom rig for decoding (e.g. using real measurements from your room), it's very easy to do.
http://www.muse.demon.co.uk/utils/ambidec.html


BTW, I did not do serious listening test on this subject. (I just read this thread about 2 days ago) :)

I noticed that somethings may went wrong after I have converted a few songs with "Pentagon" rig, many 5 channels songs have significant difference in level between left and right. (The stereo version did not.) If you see the Pentagon's set up you may guess why this happened.

Anyway, I owed you much for the guide. Thank you.

[StoneRoses]

Originally posted by Covak
I think Visual Virtual Microphone, with its 5.1 preset, is much better than ambidec anyway.

VVMic is here: http://mcgriffy.com/audio/ambisonic/vvmic/

I believe the 5.1 preset is based on G-Format (the "5.1" button used to be labelled "G-Format" in earlier versions), which pretty much makes it THE way to decode B-Format for standard 5.1 setups.

More on G-Format here: http://www.ambisonic.net/gformat.html

I've try that and have a question obout the channel order and LFE.

Is this correct?
Ch1 - LF
Ch2 - RF
Ch3 - C
Ch4 - LFE (It's full range, should I have to filtered this channel?)
Ch5 - LS
Ch6 - RS

[Eye of Horus]

Originally posted by StoneRoses
I've try that and have a question obout the channel order and LFE.

Is this correct?
Ch1 - LF
Ch2 - RF
Ch3 - C
Ch4 - LFE (It's full range, should I have to filtered this channel?)
Ch5 - LS
Ch6 - RS

Where do you get that 6th WAV from ????????

I think you better leave out the LFE.
A lot of setups don't have a LFE anyway and a lot have a LFE where the signal fed to that LFE coomes from the two front speakers. (A not amplified LFE !). And beside that every DD and DTS receiver can make it's own LFE from the input.
The 5 speaker options is the best here.

EoH

[StoneRoses]

I used 5.1 profile in VVM for decoding B-format file. There are 6 speakers in that profile. Speaker #4 located at the center [0,0,0] so it may be used as the input for LFE channel.

[Covak]

I'm not sure what's up with that LFE channel. It didn't appear in the beta versions. It does have the advantage of making a wav with all 6 channels in the right order now though. Whether or not that LFE channel should be used, I dunno. I'm inclined to say no, because as EoH said, any system will handle 5.0 material properly, while I don't think all systems handle full-range LFE channels so well.

[StoneRoses]

G-format speaker setting follows ITU-R BS.775-1 recommendation. (Info from Richard G. Elen of ambisonic.net)

here is the speaker setting for use with ambidec.exe

ITU-R BS.775-1
x y z
FL 0.866025404 0.5 0
FR 0.866025404 -0.5 0
C 1 0 0
SL -0.342020143 0.939692621 0
SR -0.342020143 -0.939692621 0


This is the illustration that compare VVM, ambidec's Surround1 and ITU-R BS.775-1 speaker array.
http://www.thai.net/stoneroses/ambisonics/Compare.gif

According to Richard G. Elen, the LEF channel should not be used for G-format decoding.

[Covak]

Ah, now that is some handy info (and a nice diagram), thanks!

A nice bit of B-Format music: http://pcangelo.eng.unipr.it/Public/B-format/Prodi/Bach/B-format/

I was surprised to find that it sounds a good bit better (to my ears with my speaker setup, at least) decoded with ambidec and that G-Format rig you specified versus VVMic's 5.1 preset.

[Eye of Horus]

Originally posted by Covak
Ah, now that is some handy info (and a nice diagram), thanks!

A nice bit of B-Format music: http://pcangelo.eng.unipr.it/Public/B-format/Prodi/Bach/B-format/

I was surprised to find that it sounds a good bit better (to my ears with my speaker setup, at least) decoded with ambidec and that G-Format rig you specified versus VVMic's 5.1 preset.

From the original guide, somewhere at the end........
"Results are a fraction less then the Ambidec/Besplit method, but that's mainly, because there is no gain control. The volume is lower and it sounds a little bit less crispier, brighter."

EoH

[joshbm]

Isn't abfpan a free utility that does this kind of stuff?

This is a quote from the website --->
ABFPAN:
pans a sound Ambisonically in a circle round the listener, between given start and end positions. Multiple rotations are possible.
Output is either a standard (decoded) quad soundfile, or Horizontal B-Format (standard WAVE file). The latter can be converted to the new WAVE-EX format by COPYSFX.
If so-- is there a way to incorperate this method into a freeware utility such as abfpan of MCTools?

Here is a reference:
http://www.bath.ac.uk/~masjpf/CDP/cmcrefmn.htm#ABFPAN

The download site:
http://www.bath.ac.uk/~masjpf/CDP/swfree.htm
(Scroll down to Multi-Channel Toolkit)

This tool set is definately worth the download! I use interlx, channelx, and copysfx constantly.

It would be nice to see a freeware version of this (and totally automated for newbies like me lol).

-joshbm

[Eye of Horus]

Originally posted by joshbm
Isn't abfpan a free utility that does this kind of stuff?

This is a quote from the website --->

If so-- is there a way to incorperate this method into a freeware utility such as abfpan of MCTools?

Here is a reference:
http://www.bath.ac.uk/~masjpf/CDP/cmcrefmn.htm#ABFPAN

The download site:
http://www.bath.ac.uk/~masjpf/CDP/swfree.htm
(Scroll down to Multi-Channel Toolkit)

This tool set is definately worth the download! I use interlx, channelx, and copysfx constantly.

It would be nice to see a freeware version of this (and totally automated for newbies like me lol).

-joshbm


Yeah, not only for newbies !!

Problem is that (as far as I know) there is no freeware utitlity that can handle and process in 32 bits the way CE does !

EoH

[Mug Funky]

@Eye of Horus:

just stumbled upon this (i'm normally rockin' it in the DV or avisynth forums), and i noticed the links for the aurora plugins don't work.

is there anywhere that these can be downloaded from?

another thing (this may have been answered) - i have Dark Side of the Moon in SQ quad format, but only have a regular stereo cartridge. do you think i'd be able to recover the full image using these tools? it would rock to have the album in 5.1ch AC3.

hope people still read this thread...

[SallyDog]

@Mug Funky

It's alive. Try this link.

http://www.ramsete.com/aurora/

SallyDog

[cooper99]

I'm using SoftEncode DD5.1 and I've tested using SurCode DTS, but I'm unable to play the wav files on my pc. All I hear is loud noise, but when I burn the files on cd and play them on stand-alone dvd player, it's fine. Why can't I play the .wav on PC. My board is the A7N8X deluxe and should be able to handle decoding DD5.1 and DTS, if I'm not mistaken.

[wejgomi]

"Here is the software you need.....

- Cooledit
(http://www.syntrillium.com/download/)
- Aurora plugins
(http://www.ramsete.com/Aurora/download/)
- Tools Ambisonics
Included here - see below message"


but the tools aren`t included in the message anymore..
Where can I get that two programs and 3 .wav files included in "Ambisonics Tools" package ?

[matrix]

Try this link

http://home.wanadoo.nl/appyhappy/Tools-Ambisonic.zip

[Eye of Horus]

Originally posted by Mug Funky
@Eye of Horus:

just stumbled upon this (i'm normally rockin' it in the DV or avisynth forums), and i noticed the links for the aurora plugins don't work.

is there anywhere that these can be downloaded from?

another thing (this may have been answered) - i have Dark Side of the Moon in SQ quad format, but only have a regular stereo cartridge. do you think i'd be able to recover the full image using these tools? it would rock to have the album in 5.1ch AC3.

hope people still read this thread...

No, you won't and........

Don't bother ! DSOTM is probably the most posted one in the DTS usenet group. (alt.binaries.sounds.dts)

There are 4 or 5 different versions available !

EoH

[Eye of Horus]

Originally posted by cooper99
I'm using SoftEncode DD5.1 and I've tested using SurCode DTS, but I'm unable to play the wav files on my pc. All I hear is loud noise, but when I burn the files on cd and play them on stand-alone dvd player, it's fine. Why can't I play the .wav on PC. My board is the A7N8X deluxe and should be able to handle decoding DD5.1 and DTS, if I'm not mistaken.

Problem is probably that it only handles 48 Khz surround streams.
Can you play AC3 in 48 Khz on your PC ? If you can, you need special codecs to play 44.1 Khz surround (as on DTS CD !).

EoH

[Eye of Horus]

Hi all,

This method is now more than one year old.
There were quite some new developments and improvements the last year and I decided to make a new guide based on the new methods.
You can find it here

http://forum.doom9.org/showthread.php?s=&threadid=60137

kind regards,

EoH