Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > General > Audio encoding

Closed Thread
 
Thread Tools Search this Thread Display Modes
Old 29th March 2004, 14:07   #381  |  Link
kempfand
Registered User
 
kempfand's Avatar
 
Join Date: Oct 2001
Location: Switzerland
Posts: 225
Quote:
On the other hand the Ambiophonics is good for music that needs to have a center.. but misses that nice soundfield that the Ambisonic give you.
I don't won't to be picky, but this statement (as well as the usage of the word Ambiophonic in the last 6 posts of this threat) is just completely wrong, full stop.

Ambiophonic always involves crosstalk-free speaker pairs (Ambiopole or Stereo Dipole) and uses the B-Format approach (Ambisonic) for the surrounds (either by convolution or by panning tools).

To quote from one of Farina's papers:
Quote:
In fact, a modern technique was recently developed by Ralph Glasgal, called Ambiophonics, which aims to the reproduction of a realistic sound field by simultaneous usage of the binaural approach (from which a pair of closely-spaced loudspeakers, with cross-talk canceling filters, can be driven) and of the B-format approach (driving, by convolution with the B-format impulse response, a suitable 3D array of loudspeakers, employing an Ambisonics decoder or other decoding schemes).
Also, Ambiophonics comes in different flavours, but as said, always involves the x-talk cancelled frontal speakers (Stereo Dipole).

Cheers,
Andreas

Last edited by kempfand; 29th March 2004 at 14:10.
kempfand is offline  
Old 30th March 2004, 05:52   #382  |  Link
desertrat
Registered User
 
Join Date: Dec 2003
Location: Germany
Posts: 12
Sorry if I got something wrong here... just trying to figure out things and obviously don't quite have it right. Thanks for pointing this out.

But that still leaves the question... do we have an Ambiophonic mixing method using plogue bidule and if so which one is it? I thought that the ambi 1.5 and the wxyz26 fell under this category (they are both in the ftp under that dir). Guess I was mistaken.

And could someone help with the first 2 parts... if the stereo2dts tests are still available someplace and if there is some information on the other up mixing methods? I've been able to find information on 1. Surroundboy's, 2. Ambisonics (EoH+Kempfand), 4. CS II (commercial product), 6. WXYZ26 (Specise-8472) 9. 3th Order (Specise-8472), and HRTF-XTC-Panorama-WXYZ (Kempfand)(I think this is the one on the ftp) which of course will keep me busy for awhile... now I got more testing to do

thanks,
desertrat

Last edited by desertrat; 30th March 2004 at 08:56.
desertrat is offline  
Old 30th March 2004, 08:53   #383  |  Link
kempfand
Registered User
 
kempfand's Avatar
 
Join Date: Oct 2001
Location: Switzerland
Posts: 225
No problem, desertrat. I was just trying to clarify on what Ambiophonics is. There is also some very good reading on this on Ralph Glasgal's Ambiophonics site.

Quote:
do we have an Ambiophonic mixing method using plogue bidule and if so which one is it?
See http://forum.doom9.org/showthread.ph...435#post451435 , where it mentions specise_8472 x-talk plugin and the set of filters (impulse responses) he created. This will create your frontal Stereo Dipole (L & R speakers). As for the surrounds, you can use the B-Format panner (3_B-Pan VST) and corresponding decoder (3rd_order VST).

I created a different bidule which uses different filters (actually the ones available on Farina's public server). I didn't publish it, but can put it to the server.

In my ears, both specise_8472's and Farina's filters work very well (i.e. there is no "one is better than the other"), yet they use different methods to achieve the x-talk cancel. Up to you to decide which ones you like best.

I am still hoping that someone will write a good guide

---

Quote:
HRTF-XTC-Panorama-WXYZ (Kempfand)(I think this is the one on the ftp)
I think EoH used a bidule which I sent him, which is not on the server, as it uses a commercial product to achieve the x-talk (Spinaudio's 3D Panner Studio). This is yet another method to create the x-talk, but it doesn't give better results than specie's or Farina's filters.

Hope this wasn't too confusing ...
Kind regards,
Andreas
kempfand is offline  
Old 30th March 2004, 13:04   #384  |  Link
Eye of Horus
Banned
 
Eye of Horus's Avatar
 
Join Date: Jun 2002
Location: The Netherlands
Posts: 326
Quote:
Originally posted by kempfand
No problem, desertrat. I was just trying to clarify on what Ambiophonics is. There is also some very good reading on this on Ralph Glasgal's Ambiophonics site.

See http://forum.doom9.org/showthread.ph...435#post451435 , where it mentions specise_8472 x-talk plugin and the set of filters (impulse responses) he created. This will create your frontal Stereo Dipole (L & R speakers). As for the surrounds, you can use the B-Format panner (3_B-Pan VST) and corresponding decoder (3rd_order VST).

I created a different bidule which uses different filters (actually the ones available on Farina's public server). I didn't publish it, but can put it to the server.

In my ears, both specise_8472's and Farina's filters work very well (i.e. there is no "one is better than the other"), yet they use different methods to achieve the x-talk cancel. Up to you to decide which ones you like best.

I am still hoping that someone will write a good guide

---

I think EoH used a bidule which I sent him, which is not on the server, as it uses a commercial product to achieve the x-talk (Spinaudio's 3D Panner Studio). This is yet another method to create the x-talk, but it doesn't give better results than specie's or Farina's filters.

Hope this wasn't too confusing ...
Kind regards,
Andreas
Hi Andreas,

Can you put together a bidule which fits this discussion ? And upload it to the FTP server ?
I was thinking of 2 bidules :
1. Stereo Dipole + Ambiophonics for the rears
2. LCR (Gerzon's ?) and Ambiophonics for the rears

back to work......

our thanks will be great :-))

EoH
Eye of Horus is offline  
Old 30th March 2004, 23:47   #385  |  Link
kempfand
Registered User
 
kempfand's Avatar
 
Join Date: Oct 2001
Location: Switzerland
Posts: 225
Ambiophonics Bidule Prototype

Here is the outline of a true, complete Ambiophonics bidule, which implements:
- A frontal Ambiopole (Stereo Dipole)
- Ambisonics for the surround speakers, 2 in this specific case (LS, RS)

You can download the bidule, filters, and well as 2 pictures for illustration of the settings here: http://www.app.demon.nl/20040330_-_XTC_01.rar (thanks for the hosting w direct link) or from web-based FTP interface.

Please note:

- This is a prototype in order to get you started. I personally use different set-ups, as I use Center Surround speaker(s), but you can easily adjust the posted bidule.

- The left tree of the bidule is for the Stereo Dipole, the right tree is for the B-Format (2nd order Ambisonic) surrounds.

- The Stereo Dipole is created by x-talk cancelling filters (impulse responses) which are from Angelo Farina's public web server. They were employed in the original Stereo Dipole setup by David Wareing (one of the originators of this methodology). These are optimised for +/- 10 deg spacing of the frontal speakers (totalling up to 20 deg), and for "normal" (not binaural) recordings (no HRTF inverse filtering here). Again: If you use them on "regular" speaker setup's (such as +/- 30 deg, they will not work).

- As mentioned before, be aware that a Stereo Dipole is the basis of the Ambiophonics method (look at Ralph Glasgal's great site).

- specise_8472 implemented another methodology for the x-talk cancellation (look at
his posting on this and the following posts). His most recent VST's and filters have been posted on the FTP by Daphy & @ndy, and are now available on the web-based interface. My personal opinion is that species' VST's (x-talk and others) as well as filters work very well, so you might want to also try his setup.

- The Ambiophonic prototye bidule uses all free VST's (i.e. SIR for the convolution). This makes things a bit tricky, as SIR has a 8960 samples latency, which makes it necessary to use 'dummy' instances of SIR in order to keep the Stereo Dipole and the Surrounds in-sync.
Things are much easier if you use commercial convolver VST's, such as the new state-of-the-art Voxengo Pristine Space convolver, which solved all these problems. Voxengo can convolve 8 impulse responses simultaneously, and much more efficiently than SIR.

- Usage of the SIR convolvers: An important point is to disable the autoranging feature of SIR, setting manually the gain, and ensuring that the two instances have exactly the same gain.
-- Always use 0 dB for Wet, and -6dB for Dry.
-- As for the dummy instances of SIR, don't load any filter, and use 'Off' for Wet, and 0 dB for Dry.
-- See enclosed pictures for details.

- Surrounds: I use Gerzonic's Panorama and Emigrator for the surrounds. They both apply 2nd order Ambisonic.
As mentioned above, the prototype bidule only decodes to 2 surrounds (SL, SR), but you can adjust this if you have 1-2 Center Surround speakers. Be aware that for a proper Ambisonic system feeding the surrounds, you will need more speakers (sweet-spot problem with less than 5-6 speakers at the min), but since Ambisonic in this case is 'just' used to 'support' the Stereo Dipole, this might be ok.

- Credits: This 'poor man's Ambiophonic' would not be possible without the pioneering research of Gerzon, Farina, Glasgal, Miller, Kirkeby, Nelson, Hamada, Wareing, Dalenbäck, Leese, just to name a few. Needless to say that I have enormous respect for these people, as well as for the creators of great software tools such as Plogue Bidule, SIR, Gerzonic, CATT.

- Final remarks:
-- I think that you will never find filters that x-talk cancel well on all types music, that is free of coloration on all music, and works for all people. Up to each of us to experiment and see where it works well.
-- You need good (!), matched speakers for the Stereo Dipole to work well (i.e. balanced frequency response curve).

- Suggested Music: I've found that the Stereo Dipole works extremely well on Pink Floyd's The Final Cut, as well as on Roger Waters' Pros And Cons Of Hitchhiking (and possibly also on Amused to Death). Give it a try if you have these handy.

Hope this is helpful,
Andreas

Last edited by kempfand; 2nd April 2004 at 15:19.
kempfand is offline  
Old 1st April 2004, 06:59   #386  |  Link
daphy
Miles Freak
 
daphy's Avatar
 
Join Date: Apr 2003
Location: Europe
Posts: 652
Hi folks,

I have uploaded the file to www.needfulthings.webhop.org , you will find it in folder ambisonic>latest!

kempfand & me are thinking about a change in structure of this website - mixing the methodes leads to sorting problems

-> So this target folder could be moved very soon on an other place

Thx to EOH for quick help!

CYA Daphy
__________________
CYA Daphy
daphy is offline  
Old 1st April 2004, 17:00   #387  |  Link
PeterC
Registered User
 
Join Date: Nov 2003
Posts: 12
Hi all

I seem to have completely lost the ball with all these
new bidules flying around, haven't read this thread for a while
. I tried the "complete Ambiophonics bidule" with Rocket Man by Elton John, I don't know if I did something wrong, I had trouble locating
the vocal when listening, the rears sounded very loud and it was like Elton's voice was floating around me. Perhaps somebody slipped something in my tea ;-)

Anyway, could anybody please tell me what is the current so-called "best" or preferred method that utilizes the center channel?

I also tested Desertrat's V2 and 3, but is the Besweet
commandline output the usual one for this? this confused me.
Surely I did something wrong since both output a file with
..well one surround was like a phantom effect - very very quiet
, the other surround was loud as the front. Rest were normal
sounding. Might this have something to do with also my program
not having a recognizable 5.01 to WXYZ_O plugin . The bidule still didn't recognize it after installing and I chose to replace it from the list (I have it installed), and I tested each of the three settings and always got the same oddly quiet rear.

What am I doing wrong :S ?


EDIT: Addition, on desertrat's it's the right-surround
that always comes out as a total fuzzy quiet piece :S


Last edited by PeterC; 1st April 2004 at 17:03.
PeterC is offline  
Old 4th April 2004, 03:49   #388  |  Link
Shayne
Registered User
 
Shayne's Avatar
 
Join Date: Jan 2002
Location: northern canada
Posts: 215
Quote "the vocal when listening, the rears sounded very loud" = new amp

all is here and you should add help. Please read and look for linkz and all methods are there ........... the first page and 25 after is all you have to read here.

All sounds so good at the farm

Peace
Shayne is offline  
Old 5th April 2004, 12:58   #389  |  Link
Lex55
Registered User
 
Join Date: Mar 2004
Posts: 31
kempfand new method

Nice results

Tested with Eric Clapton Layla unpluged

Trying to Add LFE AS Well Thinking on crossover off 60hz.
Lex55 is offline  
Old 5th April 2004, 13:43   #390  |  Link
kempfand
Registered User
 
kempfand's Avatar
 
Join Date: Oct 2001
Location: Switzerland
Posts: 225
Quote:
thinking on crossover off 60 Hz
Actually, if you use the bidule for the Stereo Dipole I posted in the first place, you should put things to 0 Hz (i.e. no cross-over at all).

Reason: Someone kindly made me aware that I made a mistake in the x-talk bidule I posted.

Actually, x-talk cancellation is needed at low frequencies but not so much at the high frequencies. At high frequencies the head itself inhibits the x-talk since it is larger than the wavelength. If at all, the cross-over filters should be used at high frequencies, not at low frequencies as outlined in my original bidule. Guess I had too much Wisky when I did it ...

Anyhow: I took the cross-over filters out for now (post corrected, as well as bidule in the zip-file).

Kind regards,
Andreas
kempfand is offline  
Old 6th April 2004, 12:23   #391  |  Link
Lex55
Registered User
 
Join Date: Mar 2004
Posts: 31
Nice Results

Got some nice results using your new enlightments

My next test will be combining First method with last one

Finale objective is crating good 5.1 DD from my Produced films.
Lex55 is offline  
Old 7th April 2004, 11:54   #392  |  Link
specise_8472
Newbie
 
specise_8472's Avatar
 
Join Date: Oct 2002
Location: New Zealand
Posts: 176
Have just uploaded into FTP server under Incoming\Ambiophonics

Allinone.dll

My latest offering.
Basically it is as stated, All in one package.
Stereo in 5 channel out.(Usual ordering L,R,C,SL,SR)

It uses my own custom method to get true? stereo surrounds. Not the same sounding as is the case now with the Ambi methods.
It still uses Ambisonics to compute the surrounds and Center Front. But with my own twist.

BTW for the Neo6 fans, the SC is mixed into the surrounds as I normally do.
specise_8472 is offline  
Old 7th April 2004, 12:53   #393  |  Link
daphy
Miles Freak
 
daphy's Avatar
 
Join Date: Apr 2003
Location: Europe
Posts: 652
Hi folks,

thx specise_8472 for the new methode -> the old FTP is unbelieveable slow - stay tuned - I will upload the file to new webspace within a couple of hours

CYA Daphy
__________________
CYA Daphy
daphy is offline  
Old 8th April 2004, 02:39   #394  |  Link
Tantulus
Registered User
 
Join Date: Sep 2003
Location: Massachusetts
Posts: 34
I'm completely at a loss!

I have been following this thread for several months now and I'm afraid that I'm becoming hopelessly confused.

Part of the problem was that I was unable to download the test files from the newsgroup (I never used one before) nor could I obtain them from donkey2000.

I would appreciate a little hand-holding in my attempts to master converting stereo into surround sound.

Essentially, I am interested in encoding my stereo orchestral CD's into dts. I wish to emulate a symphony or chamber hall with a bit of ambience.

What I have been doing is making an image file of my CD, run it through Bidule and use wav2wav6 to convert the resultant file into 6 mono files and then use softencode to encode into dts. I use Wave2Wave6 because the CD images are usually to large for besweet to handle. I understand that the ultimate result is a matter of taste but I would appreciate some suggestions as to which method will give me a somewhat broad space with a hint of reverb but not to much gain in the rear surround.

I've noticed that most tests were done on vocal or band tracks and I find that with symphonic music I'm swamped with sound from all directions. I've developed quite a coaster collection and I would like to narrow the possibilites a little.

If anyone has experience with orchestral music and has some recommondations as to which methods would be most useful I will be greatly appreciative.

Thanks again and keep up the good work.
Tantulus is offline  
Old 8th April 2004, 08:52   #395  |  Link
kempfand
Registered User
 
kempfand's Avatar
 
Join Date: Oct 2001
Location: Switzerland
Posts: 225
First of all, if you don't know (yet) which is a good set-up for you, I'd recommend that you use small extracts of the music (1 minute max; 30 secs is usually enough). You can collate some snippets (i.e. 10 of them coming from various CD's, having 30 secs each) and use these for testing purposes until you have a setup which is good for you. This will save you lot's of time, plus you will learn to get a feeling what different bidules do.

As for ochestral music, I don't want to discourage you, but you might be out of luck here ... Reason is that the music you use is not anechoic (dry) and already has the characteristics (reverb, ambience) of the listening space / concert hall where it was recorded. This is especially the case for orchestral music.

Another thing to keep in mind is, that some people will not like the room-sound that Ambisonics creates (even if it is correct!) ... This is because many people are too much used to what what the "Tonmeisters" (German expression for soundmasters) usually produce on current multichannel dolby mixes.

Also keep in mind that most people are so used to stereo, such that hearing too much of the room won't sell (even if it is right). Just my opinion ...

You might want to try the VST created by specise_8472 (either the original one called 'WXYZ_to_5.0', or the latest one just announced a few posts above), as it keeps the stereo image for L & R speakers, yet adds some nice "room" through C, SL, SR.

Good luck,
Andreas
kempfand is offline  
Old 9th April 2004, 09:04   #396  |  Link
Lex55
Registered User
 
Join Date: Mar 2004
Posts: 31
Advice from newbie

Few weeks ago i was in the same feeling you are now.

As my recommendation is just start at the first method as guided in the first page by Eye of Horus

Play with it so you will learn the pros and cons of that guide.

Only after you familiar with that you might think of starting upgrading your results following expert's remarks.

Good Luck
Lex55 is offline  
Old 9th April 2004, 13:34   #397  |  Link
Shayne
Registered User
 
Shayne's Avatar
 
Join Date: Jan 2002
Location: northern canada
Posts: 215
Re: I'm completely at a loss!

Quote:
Originally posted by Tantulus
I've developed quite a coaster collection and I would like to narrow the possibilites a little.
Well i would buy a 60 buck Hercules sound card for my computer and optical out to my amp. ie use my computer as the dvd player. This would produce absolutely no coasters and give you 100 gig dvd's which makes changing a lot less frequent.
Shayne is offline  
Old 15th April 2004, 16:33   #398  |  Link
rc
Registered User
 
Join Date: Feb 2004
Posts: 3
RE: specise_8472 new method

Hi Specise,

I dl your new method All In One.Rar and it came out only a
file "allinone.dll". Pls let me know how to use it. Is it just
a plug in for Winamp????

Thank you


rc
rc is offline  
Old 15th April 2004, 21:26   #399  |  Link
Tantulus
Registered User
 
Join Date: Sep 2003
Location: Massachusetts
Posts: 34
Ambiophonics Bidule Prototype

First of all, thanks for the earlier advice, Andreas, I will start experimenting with the other VST's. I guess what I'm trying to do is take away the stereo sweet spot and translate the recorded ambience to give the recording a fuller feel. The first thing I did was to adjust the initial stero2ambi bidule with LFE and 6 channel gain to decrease the surround. Still not satisfied.

BTW, I read Ralph Glasgal's article and decided I am asking too much out of 5.1. However, I did try your Ambiophonics method even though my room couldn't accomodate a 20 degree separation of my speakers. I used your first bidule (the server was down so I couldn't use your second bidule). Anyway, the results still allowed me to localize a solo instrument (although the instrument seemed somewhat broader than life) and the surround effects were wonderfully warm, giving a decent illusion of ambience.

If possible could you please PM me your bidule with the corrected X-talk while the server is down. Also, any more info on this method would be very interesting. I am especially interested in the use of impulse responses in your method. I want to apologize in advance but I'd like a little more hand-holding until I can get better familiarized with ambisonics/ambiphonics.

Sincerly,
Scott

Last edited by Tantulus; 15th April 2004 at 21:33.
Tantulus is offline  
Old 19th April 2004, 15:01   #400  |  Link
kempfand
Registered User
 
kempfand's Avatar
 
Join Date: Oct 2001
Location: Switzerland
Posts: 225
Quote:
If possible could you please PM me your bidule with the corrected X-talk while the server is down
The hosting with EoH now contains the corrected bidule and pictures: http://www.app.demon.nl/20040330_-_XTC_01.rar. But it will only work well for +/- 10 deg speaker setup.

On howto create the B-Format:
Quote:
I am especially interested in the use of impulse responses in your method.
I often use the ones outlined on Angelo Farina's public server. The link is in his guide: Conversion between UHJ and B-format, and it uses the SIR convolver.

State of the art these days is the Voxengo's Pristine Space 1.1, which can convolve simultaneous 8 indepdendent filters, and even has a "zero latency mode". There is a nice outline for AudioMulch, also on Farina's public server: Voxengo-Bformat.

Quote:
I want to apologize in advance but I'd like a little more hand-holding until I can get better familiarized with ambisonics/ambiphonics.
No problem, but I suggest you start with EoH's excellent guide on the 1st page of this threat. Having done so will help you to understand the basics (i.e. what is done why), and you will later be able to compare (by listening) where more advanced methods differ.

If you really have time (now or later), I highly recommend tha CATT-F (FIReVerb) package. For pure Ambisonics, it currently delivers the best results (at the expense of slower speed, and no real-time processing as with Bidule). You can get a fully-working test-license, which is valid for 1 months (I believe). The package also contains an excellent Help-file and even a small Ambisonics tutorial (of which I learnt alot). Link: CATT-F / FIReVerb

Good luck & in a hurry,
Andreas
kempfand is offline  
Closed Thread

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 06:57.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, vBulletin Solutions Inc.