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. |
24th May 2004, 13:21 | #461 | Link | |
Registered User
Join Date: May 2004
Location: Montreal
Posts: 729
|
Re: name
Quote:
|
|
24th May 2004, 13:25 | #462 | Link | |
Registered User
Join Date: May 2004
Location: Montreal
Posts: 729
|
Re: Re: Re: Re: MatrixMixer bidule
Quote:
If this is the case, then it would be better to convert back to 16 bites with dithering before loading the 6 files into Surcode. I'll see what I can find out. |
|
24th May 2004, 14:10 | #463 | Link | |
Registered User
Join Date: May 2004
Location: Montreal
Posts: 729
|
Re: Re: name
Quote:
http://www.geocities.com/ursamtl/audio.html |
|
24th May 2004, 14:16 | #464 | Link |
Registered User
Join Date: May 2004
Location: Montreal
Posts: 729
|
By the way, for those of you looking for impulses to experiment with for your bidules, www.noisevault.com has just put up a new complete set of 44.1k, 32-bit impulses recorded from a Lexicon 960L! It's 38MB download, but it seems worth it.
These should fit in well with the discussion of 32-bit files we're having! |
24th May 2004, 21:54 | #465 | Link | ||
Registered User
Join Date: Oct 2001
Location: Switzerland
Posts: 225
|
@ursamtl: Just had a quick look at your bidule and MatrixMixer group. 2 questions regarding Kelly's Bass Manager which you use inside the MatrixMixer Group:
1) The manual says: Quote:
2) The manual says (and I can confirm this by a quick test I made): Quote:
Not sure, but I think something is not correctly linked. Kind regards, Andreas Last edited by kempfand; 24th May 2004 at 21:56. |
||
24th May 2004, 22:17 | #466 | Link | |
Registered User
Join Date: Dec 2003
Location: Germany
Posts: 23
|
Re: Re: Re: name
Quote:
The new website is still under construction (panta rei ; for the greeks between us) but you have full access to every file you`ll find at my ftp- server. Use my FTP- Server to upload your files Don`t forget: needfulthings is still alive
__________________
If you`re not a part of the solution, you`re a part of the problem. a mirror of needfulthings.webhop.org and |
|
24th May 2004, 23:31 | #467 | Link | ||
Registered User
Join Date: May 2004
Location: Montreal
Posts: 729
|
Quote:
Quote:
Thanks, Ursa Last edited by ursamtl; 24th May 2004 at 23:34. |
||
24th May 2004, 23:32 | #468 | Link |
Registered User
Join Date: Sep 2003
Location: Massachusetts
Posts: 34
|
Impulses
I did download the Lexicon impules and did a quick test by adding SIRs to the SAD5.1 bidule of the one called ambience. Sounded great so far.
BTW what are the SFK files for? Still planning to use the matrixemulator after I finish a two disc set I've already started. I plan to experiment with the Lexicon impulses in your bidule. Regards, Scott |
24th May 2004, 23:52 | #469 | Link | |
Registered User
Join Date: Oct 2001
Location: Switzerland
Posts: 225
|
Quote:
Will do some extensive tests coming long weekend here, with various types of music. Kind regards, Andreas |
|
25th May 2004, 00:47 | #470 | Link |
Registered User
Join Date: May 2004
Location: Montreal
Posts: 729
|
I've uploaded version 2 of MatrixMixerEmulator to the FTP server's incoming folder and also posted it at http://www.geocities.com/ursamtl.
This updates the entire signal path to L, R, C, LFE, Ls, Rs as per ITU standards. It's designed to work with Kelly Industries Bass Manager v0.5 and SIR 1.005. I've also replaced the dummy SIR with a native Plogue delay line set to 8960 samples as per SIR 1.005 specs. I haven't tested it, but I imagine this will be less CPU and/or memory intensive than creating an extra instance of SIR just to compensate for the delay. Thanks for the feedback. Ursa |
25th May 2004, 23:40 | #471 | Link | |
Banned
Join Date: Jun 2002
Location: The Netherlands
Posts: 326
|
No good results
Quote:
The rears sounded terrible ! There was a definite delay, so bad that it sounded like echo !! I will recheck all the settings, but I'm sure I got them right this morning. In one of your previous mails here, (the long one) you made quite some comments about Ambisonics and that you want to achieve more separation. That's what we're working on too. SAD51inBidule gives at this moment better results than Matrixemulator ! Of course all IMHO, because I only tested with two tracks. I will do some more tests ASAP to see if I can get rid of the echo in the rears. Perhaps it's just an overseen setting in your bidule or in the new Plogue version, but for now I was not impressed...... I take for granted that you tested it thoroughly, so maybe you can point me to what can be wrong ? kind regards, EoH |
|
26th May 2004, 01:07 | #472 | Link | |
Registered User
Join Date: May 2004
Location: Montreal
Posts: 729
|
Re: No good results
Quote:
Also, make sure you have the center level high enough because what appears in the front left and right channels is basically what is unique to each channel. The center needs to be high enough to re-construct the original soundstage. This also reveals the one downside with my approach in that one really needs to be able to monitor the mix in surround and not simply monitor everything to the two Microsoft Mapper channels as in your original guide. It's not necessary to have a proper 5.1 computer setup. I`m just using a 4.0 setup and mixing the center and LFE channels into the front. So far, the tests I've done when burning DTS CDs and playing them on my true 5.1 system have shown that I`m getting close with the monitoring. As for ambisonics, I was playing around with a 3rd-order setup last night and while it still was terrible in terms of separation, it does some interesting things to the music, bringing out instruments that were buried in the stereo mix. I`m thinking of trying a combination of the MatrixMixer Emulator with the ambisonics. By the way, have you tried the real MatrixMixer DirectX filter from SourceForge? I'm very impressed with its sound. If we could just fit it into a bidule! It provides exactly the kind of big, expansive sound I'm looking for in an upmix. Last edited by ursamtl; 26th May 2004 at 02:50. |
|
26th May 2004, 04:32 | #473 | Link | |
Guest
Posts: n/a
|
Re: Re: No good results
Quote:
Don't forget that there are two versions of Ambiophonics. The version I use that takes WXYZ and computes speaker feeds. Or the Stereo Dipole method. Which is why I created the X-talk VST. Which I am overhauling to produce binaural and transaural mixes and conversions. Try my Allinone VST for rear channel seperation. It is 3rd order and computes the rears independant of each other. What I have done to really enhance the seperation all round is to use my x-talk impulses. Just use three instances of SIR. Left and Right Fronts using one of the front impulses. Left and Right Surround using one of the Rear Impulses. Center using the Zero impulse. Really spreads out the sound image. And by feeding all signals through SIR's there is no delay. At the moment, among other things, I am working on a better V3 ambiophonic mix using some enhanced methods to extract better channel seperation. |
|
26th May 2004, 13:23 | #474 | Link | |||
Registered User
Join Date: May 2004
Location: Montreal
Posts: 729
|
Re: Re: Re: No good results
Quote:
Quote:
By the way, instead of loading a SIR with no impulse, why not try what I did in v2 of my MM Emulator? Add a Plogue Delay Line and connect a Plogue Constant to its right pin entering the SIR delay value (8960 for v1.005, or 16384 for SIR 0.86). This does the same thing without having to load in an entire instance of SIR. Quote:
|
|||
26th May 2004, 14:08 | #475 | Link | |
Banned
Join Date: Jun 2002
Location: The Netherlands
Posts: 326
|
Re: Re: Re: Re: No good results
Quote:
I have some old quad conversions here, where the level of separation is unbeatable. A guitar from FL, a singer from FR, drums from RL and another guitar from RR. What I miss is the band Realistic space can only be achieved by Ambisonics and certainly not by separation. There is nothing realistic about separation at all, actually it will only be "realistic" when we have 5.1 ears ! IMHO separation is artificial ! Just simply because our ears are not build that way. We will hear from all directions at once, thus a left channel also influences what we hear with our right ear ! In other words, even if you can completely separate 5 sounds, there is still the fact that we will not hear it completely separated. So why this search for complete separation ? Are we that much indoctrinated by Dolby ?? You mentioned the excellent results on DVD's. But...... are you comparing methods like Ambisonics with a music DVD or with a DVDA ? Or were you talking about movie DVD's ? I am just wondering because you will never get that separation from stereo to 5.1, whatever method ! You can come close, but not with the bidule you presented us ! I have tested it to it's limit this morning and it's not what I hoped it would be ! But I have to confess that I used a very difficult piece of music. When I use DPL II, this happens : a guitar starts in the front right channel. The guitarplayer starts to sing. His voice is in the Center channel and the guitar remains in the right channel. The separation between C and R is perfect here with DPL. Playing it in regular stereo you get the guitar from the right channel and the singer from both. And a very small bit of the guitar from the left, but you really have to play it vey loud to hear that. With all methods I used on this piece of music, not one gave the same results. They just don't do a good job on this separation ! The only thing that came close was to use Gerzon LCR, and process the L=R from there and leave the Center as is. All others made a mess of it ! Including yours. For Ambisonics this piece is also difficult, but although no perfect separation, you still can sense where the singer comes from (Center) and where the guitar comes from (C+R). It's sounds more natural in Ambi because it's more fitted to our 2 ears. It makes a complete soundfield, which actually resembles what we hear with our 2 ears. Also I would like to point out the zillion of music DVD's where the rears are nothing more than some weak reverb from the fronts. I even have this on a few DMP demo SACD's ! We can do better :-) So there's still work to do ! kind regards, EoH |
|
26th May 2004, 19:06 | #477 | Link | ||||
Registered User
Join Date: May 2004
Location: Montreal
Posts: 729
|
Re: Re: Re: Re: Re: No good results
Quote:
All I'm saying is that I'm extremely disappointed with the results from Ambisonics so far. I came upon your guide quite by accident and read with enthusiasm its promise. I kept redoing the connections thinking surely I must be doing something wrong because the resulting 5.1 recording sounded small and boxy. I took Floyd's Comfortably Numb, Sarah Brightman's Captain Nemo, Marillion's Lavender, Zeppelin's Gallows Pole, Deep Purple's Lazy, Elton's Tiny Dancer, plus a couple of classical recordings, and tried over and over again to get some sense of the acoustic space you say Ambisonics gives. They all sounded like I remixed the tracks to mono and then routed the same mix to the rears. There was no sense of a realistic soundstage as if the real instruments were being played in front of me. The closest I achieved with Ambisonics was with the Surround configuration in Emigrator and only after tweaking the bpan controls (and obviously without the 36° rotation). This was on the classical guitar intro to Rush's A Farewell to Kings. It almost sounded as if the guitar were there right in front of me. Once I took this same track and played it through the MatrixMixer filter, I got a much better sense of realism. I put together my bidule with a similar approach and with similar positive results. Whereas Comfortably Numb sounded really small and dull when rendered in Ambisonics, it became massive through my bidule. I dug out the remastered version of Deep Purple's Lazy and reveled with the sound and realism through my bidule. Blackmore's guitar sounded as if it were emanating from the amp in that Swiss hotel corridor so long ago, with gigantic Hammond groans coming from the right and sweeping through the room. In Ambisonics, this same song sounded like a 60's mono single mix (albeit with improved fidelity). Quote:
Quote:
Did it ever occur to you that perhaps you're expecting too much from Ambisonics? The theory seems good and probably sounds quite stunning if one is listening to material recorded with the proper miking, and correctly encoded. However, most music that was recorded for 2-channel stereo reproduction was not properly recorded for Ambisonic reproduction. Moreover, many modern recording studios record in rooms specially treated to add no coloration to the sound whatsoever. Therefore, any soundstage produced by its decoding method is going to be artificial. Perhaps Ambisonics does work wonders on a properly recorded sound that was recorded in an acoustic space capable of being reproduced, but even with proper recording, if an sound is recorded in a space that has little or no reflective capability--as is the case with many recording studios--there is no natural spatial information present in the recorded sound. Quote:
Regards, Ursa |
||||
26th May 2004, 19:09 | #478 | Link |
Registered User
Join Date: Oct 2001
Location: Switzerland
Posts: 225
|
@ puzio: I agree with a BUT: It's not easy to come up with a method that works well on many pieces of music. I still appreciate that ursamtl shared his method, which allows us others to learn.
Ambisonics / AmbioPhonics is truely exceptional and hard to beat Just to repeat that Ambisonic's goal is not to bring the orchestra into your living room, but bring you into the concert hall, so no need to look or search for single instruments coming from defined speakers. For non-Ambisonics, I recently posted LCR Upmix Bidules. Amongsts these is the Gerzon LCR, but the others are also worth checking as they produce some reasonable LCR from LR. Regards, Andreas Last edited by kempfand; 26th May 2004 at 19:13. |
26th May 2004, 19:11 | #479 | Link | |
Registered User
Join Date: May 2004
Location: Montreal
Posts: 729
|
Quote:
Ursa |
|
26th May 2004, 19:16 | #480 | Link | |
Registered User
Join Date: Oct 2001
Location: Switzerland
Posts: 225
|
Quote:
Regards, Andreas |
|
Thread Tools | Search this Thread |
Display Modes | |
|
|