Log in

View Full Version : Mono to Stereo to 5.1


drpaulng
6th December 2004, 02:30
I found some plug-in works quite well with MONO-to-STEREO. However, the tracks derived from these plug-in adds some reverb to the tracks that would be undesirable for further processing for STEREO-to-5.1 conversion. Is there anybody having any idea how to make stereo from mono without adding too much reverb? I myself found that Adobe Audition can generate a pseudo-stereo effect without much additional reverb but the stereo effect is not very dramatic.

ursamtl
6th December 2004, 14:38
One of the biggest problems with applying reverb is that we all have a tendency to apply too much. We keep pushing up the level until it sounds really apparent. If we come back the next day and listen to it, chances are it's overdone. If we try to run it through one of the upmix to surround methods, the reverb really overwhelms the whole thing.

Some may suggest working with stereo impulse responses. These can "graft" the sound characteristics of a space onto a sound. The only problem with this is that it will end up sounding like a signal sound source playing in a space. For instance, if you had a "room" type impulse response, the result would sound like a recording of a mono speaker sitting in a room!

Other than using reverb, the other usual mono-to-stereo conversion tricks are delaying one channel and/or applying comb filtering. One free plugin that combines these two quite nicely is the mda Stereo VST plugin available at http://mda.smartelectronix.com/effects.htm.

You might check out specise_8472's Mono-to-Stereo work at http://forum.doom9.org/showthread.php?s=&postid=498287#post498287. I belive you can find the Plogue layouts and groups in the download of all layouts at Daphy and @ndy's server at http://www.needfulthings.webhop.org/

One recent popular technique among those trading recordings of concerts is called the "TISDU" method. This stands for "Teased Into Stereo Down Under", since it was presumably popularized by a concert recording trader from Australia. It consists of delaying one channel by 5ms then cutting the sound level of the non-delayed channel by -3dB to help defeat the Haas effect (when two sound are delayed by less than about 35ms, the first sound masks the second).

It's been my experience that a combination of these effects usually gives the best overall results. For example, you could try combining TISDU with a room impulse, plus another larger impulse for the surround feeds. You might take the original mono for the center at -6db, delay the left channel by 5ms, the right channel by 10ms, apply a room impulse to the front left and right but keep the dry/wet balance fairly dry so as not to overdo it. Then take the different signal from the left and right after the impulse processing, and derive the surrounds by taking the difference signal as the left surround, the same signal inverted by 180° for the right, then running them as a stereo pair through a larger stereo impulse response for the surrounds.

This might sound complicated, but it's really quite simple in Plogue. Give it a try, play with your levels and see what you get. Then come back and tell us!

Steve.

drpaulng
8th December 2004, 03:56
Very enlightening!

Boulder
22nd March 2006, 10:03
It's been my experience that a combination of these effects usually gives the best overall results. For example, you could try combining TISDU with a room impulse, plus another larger impulse for the surround feeds. You might take the original mono for the center at -6db, delay the left channel by 5ms, the right channel by 10ms, apply a room impulse to the front left and right but keep the dry/wet balance fairly dry so as not to overdo it. Then take the different signal from the left and right after the impulse processing, and derive the surrounds by taking the difference signal as the left surround, the same signal inverted by 180° for the right, then running them as a stereo pair through a larger stereo impulse response for the surrounds.

Hi Steve,

could you please explain this to me..

1) are you supposed to use the delayed wave file for FL and FR or the one that has had room impulse applied?

2) exactly how do you derive the surround sounds? Is there some specific program/technique that must be used to get them?

It would be nice if you could write a small guide, possibly with a generic Bidule layout regarding this kind of mono to stereo conversion :)

I've got an old project to finish in which some parts of the audio track (99% dialogue) were recorded in mono, some in stereo. I planned that I'll do two different audio tracks for the DVD: one that is in 5.1 and one that is as close to the original as possible (with either dual mono or stereo depending on the original audio). I need to do a mono to stereo conversion (and then to 5.1) for those odd mono parts for the 5.1 track and I was hoping that I could get to creating the 5.1 track at the end of the month when I have the time to drag the PC next to my 5.1 system and start testing different settings.

Thanks in advance!

ursamtl
22nd March 2006, 14:54
Hi Boulder,

It's been nearly a year and a half since I wrote my response here and I've experimented a bit with other methods. I'll start by answering your questions.
1) are you supposed to use the delayed wave file for FL and FR or the one that has had room impulse applied?Both processes should be applied before using them as FL and FR. Experiment to find the order that works best for your material. It might work better if you applied the impulse first and then the delay, or perhaps delay followed by impulse. See what works best for you.2) exactly how do you derive the surround sounds? Is there some specific program/technique that must be used to get them?In my Dec. 2004 message that you quoted, I suggested doing the same to the surrounds as to the fronts, but using a larger impulse. This would have to be at a lower level to be effective as a surround, perhaps -3dB. The one problem with this is that any dialog or vocal sounds will also appear in the surrounds. Some people don't like this as much. However, depending on the material, it might be inevitable. You could try reducing the freqencies of the dialog from the surrounds with EQ. Again, it would depend on your source material.

One experiment I did do recently that gave me some interesting results was to run a mono source in Bidule to both input channels of a tremendous (and free) reverb plugin called Glaceverb and then in turn run this through my V.I layout. After playing around with the settings, I ended up with a surprisingly convincing 5.1 mix. Then I tried a plugin called Convo Boy Lite. One problem I had with SIR as a free convolution plugin was that it seemed to treat the two input channels as dual mono channels without any crosstalk in the reverb effect. Convo Boy seemed to work properly with stereo impulse files. With this fed through the V.I layout, I got really tremendous 5.1 results from a mono source.

As always, it depends on your source material and then the settings you use. You have the right idea to drag your PC next to your 5.1 system. You really do need a monitoring system to try different settings and achieve the best results. I haven't had time to write a guide yet for this. I've actually got about five different surround-related projects in mind right now, but I haven't had time to develop them and keep up with my day job!

For now, I recommend that you start with my V.I Guide (http://forum.doom9.org/showthread.php?p=570432#post570432) and the bidule layouts I created for it. Then add one of the following plugins between the file player and the two V.I inputs: Glaceverb (http://www.dasample.com/index.php?show=glaceverb)
Ambience (http://magnus.smartelectronix.com/) (another tremendous donationware VST reverb plugin. Both this and Glaceverb sound as good as a lot of expensive commercial reverb plugins).
Convo Boy Lite (http://www.paulrharvey.co.uk/elevayta/product7.htm)Now, if you want a bit more separation, put the Elevayta Wide Boy (http://www.paulrharvey.co.uk/elevayta/product9.htm), a mono-to-stereo plugin on the FL & FR V.I outputs. You could also add another Wide Boy to the sL & sR V.I outputs. This will mess with the natural reverb fields a bit, but with the ight adjustments, it will give you a huge, open surround field, the kind you often hear in big 5.1 movie mixes.

I also have some more complicated approaches to this in mind, but I need to experiment some more first and then I'll write some guides.

For now, tak this and experiment, adjust and see what happens. Then of course, come back and tell us about it!

Finally, if you find these plugins useful and can afford it, please donate something to these talented programmers or consider purchasing their full version plugins. Some of them are surprisingly inexpensive and their features rival those of expensive plugins. They deserve our support. I've personally started making it a point of purchasing at least one useful plugin a month. It doesn't ruin my budget, and I'm ending up with some incredible tools!

Regards,
Steve.

Boulder
23rd March 2006, 07:45
Hi Boulder,

It's been nearly a year and a half since I wrote my response here and I've experimented a bit with other methods.

Hi Steve,

I was kind of hoping that something new had evolved:)

One experiment I did do recently that gave me some interesting results was to run a mono source in Bidule to both input channels of a tremendous (and free) reverb plugin called Glaceverb and then in turn run this through my V.I layout. After playing around with the settings, I ended up with a surprisingly convincing 5.1 mix. Then I tried a plugin called Convo Boy Lite. One problem I had with SIR as a free convolution plugin was that it seemed to treat the two input channels as dual mono channels without any crosstalk in the reverb effect. Convo Boy seemed to work properly with stereo impulse files. With this fed through the V.I layout, I got really tremendous 5.1 results from a mono source.

This sounds very interesting, I'll definitely have to check it out. The problem I've had trying to create some sort of stereo effect is that it either sounds too artificial or doesn't have enough separation.

For now, I recommend that you start with my V.I Guide (http://forum.doom9.org/showthread.php?p=570432#post570432) and the bidule layouts I created for it. Then add one of the following plugins between the file player and the two V.I inputs: Glaceverb (http://www.dasample.com/index.php?show=glaceverb)
Ambience (http://magnus.smartelectronix.com/) (another tremendous donationware VST reverb plugin. Both this and Glaceverb sound as good as a lot of expensive commercial reverb plugins).
Convo Boy Lite (http://www.paulrharvey.co.uk/elevayta/product7.htm)Now, if you want a bit more separation, put the Elevayta Wide Boy (http://www.paulrharvey.co.uk/elevayta/product9.htm), a mono-to-stereo plugin on the FL & FR V.I outputs. You could also add another Wide Boy to the sL & sR V.I outputs. This will mess with the natural reverb fields a bit, but with the ight adjustments, it will give you a huge, open surround field, the kind you often hear in big 5.1 movie mixes.

I need to add a bit of music and a few sound effects to some parts of the audio track. I was considering creating an empty audio track which would be exactly as long as the one I'm converting to stereo, then placing the music and the F/X onto the empty track and using a second audio file player (or an appropriate plugin) to feed it to the same V.I input pins as the fake stereo file. Does this sound OK?

I also have some more complicated approaches to this in mind, but I need to experiment some more first and then I'll write some guides.

I'm eager to test them for sure. I'll archive the original video capture files anyway so I can always remix the audio tracks if I feel like it and have the spare time for it.

For now, tak this and experiment, adjust and see what happens. Then of course, come back and tell us about it!

For sure - if everything goes according to my plans, I'll have the whole track done by the end of the month. Then I'll upload some before-after samples for feedback and the layouts as well. That's the only possible way to develop one's own skills.

Finally, if you find these plugins useful and can afford it, please donate something to these talented programmers or consider purchasing their full version plugins. Some of them are surprisingly inexpensive and their features rival those of expensive plugins. They deserve our support. I've personally started making it a point of purchasing at least one useful plugin a month. It doesn't ruin my budget, and I'm ending up with some incredible tools!

Looks like it might be a good point for me getting a credit card at last..not to mention that buying DVDs abroad would be a lot easier :) I've recently been involved in IT systems development and have noticed how difficult and frustrating it can be at times. I'm sure a donation would help the developers over the worst times. I'll have to do something about that as soon as my options for the summer are clear.

A big thank you for all your help :)