Hmmm - clear as mud, heh? I've now read the original space war:

http://community.vsl.co.at/viewtopic.php?t=3640&postdays=0&postorder=asc&start=15

My conclusions: add more predelay to those instruments that are further away. I know the science is complex, but the results of this approach seem to work well.

So, in a concert hall are we talking 10ms for strings, 30ms for woodwind and 50ms for brass and percussion?

Could I get the same effect by setting my single M2000 to 50ms (furthest instruments) and varying the amount of reverb (AUX 1/2) depending on the instrument? I know it would be easier with three units! Could the NFX1 reverb (aux in giga) or my 02R96 internal fx help?


Thanks for your help (esp the demo, Beat!)

Neil

Neil,

I'm sorry, but if you think that longer delays are required, you should also delay the original sound. This is the factor that most people seem to miss in this discussion.

Think of it differently: instruments far in the back of a hall are relatively closer to reflecting surfaces than instruments in the front. So much closer, that the first order reflections will arrive at nearly the same time at the listener as the direct sound.

Now, do this imagery for the instruments at the front of the stage. Here the direct sound arrives at the listener much sooner than the first order early reflections, simply again as a result of relative distances to both listener and reflective surfaces. This is really just a matter of simple geometry.

Take a paper, draw some lines and do some goniometric math. [[;)]]

Beat: spacewars no sound in space there is, says Master Yoda! [:D]

My experiments led to the opposite from your conclusions... No predelays merges an instruments with the ER's, making it appear as a more distant sound source. Longer predelays separate the original sound from the ER's and make it stand out as closer.

By the way, I think you should try to re-do your comparison with the harpsichord and first take out the initial spike of your impulse. IMHO, you are now introducing a first order echo, a bit in the same way as rock guitars a often double tracked. I think this leads to an artificial sense of depth, like with the rock guitar. I get the impression that the first spike is a measurement of the original signal from your graphs and not from the first reflection (but I may be terribly wrong!).

Anyway, let's be happy with what we are learning through fora like this and above all, let's agree to disagree [:D]

Great ambience here!

Peter

Peter, Beat

Before I go any further down this complex road (!) can we agree on definitions:

Predelay?

Early reflections?

Any other important setting?

Thanks

Neil

@Neil C said:

So, in a concert hall are we talking 10ms for strings, 30ms for woodwind and 50ms for brass and percussion?

Neil

I do not maintain, that my theory is the right one, but you will reach acceptable results.
As I always say: Try and believ your ears!
I'm very interested in new experiences!

Try with 50ms? Yes try it.
Try also to produce the 3 different depths with ReverbA (No Reverb only PreDelays > as I recommended in an upper topic). And then thru Aux and M2000 for giving reverb.

I think you're close the solution, I've recommended you first:
Mix all strings together and record them thru your M2000 (10ms) to (in?) a new track called "strings with depth1"... same procedure with woodwinds (30ms) > new track "woodwinds with depth2" and also for percussion (50ms)... Finaly you will have three new tracks with its needed depths. Mix them now (in cubase) to a final-track. Voilà! Quality? Maybe totaly bad...
Thats our (your) way.
(Sorry my English)

Beat Kaufmann

@Peter Roos said:

Neil,
I'm sorry, but if you think that longer delays are required, you should also delay the original sound. This is the factor that most people seem to miss in this discussion.
Great ambience here!
Peter

Hi Peter
Peter: Great ambience here! I agree totaly

Peter:...you should also delay the original sound. I do not agree totaly

This isn't important - from my point of view. We produce musicfiles. And they shall sound like a listener in the concert-room recognize them.
Peter, think you are a listener - 20m away from the orchestra. The "point zero" of recognizing a coming-sound is after the directsound-delay ...yes?. Some ms later - the first reflections will reach you! So important is only the time-difference between direct- and indirect- sound. But: The farther you are the more decrease the directsound-level. This relation of directsound-level and reflection-level gives us the impressen of far away or close. Far = less direct-sound/more indirect-sound and vice versa. Our old fight, isn't it? [:D]

Now - good news! I believe that I understand more and more, what you mean:
Your statement is, that the time between the direct- and the indirect-sound decrease the farther you are away from the origal-sound-source. I agree... when we speek about a "sound-source-listener-distance" around 60m-80m and more. For this situation you are right: There are only some ms between direct- and indirect-sound... and the direct-sound-level is so low, that you can forget it. > We know that sound as mudy cathedral-sound.

On the other hand I think we would like to simulate a studioroom, a chapell, a concert-hall or other concert-rooms - like I've done it with my violin-example. This means a "sound-source-listener-distance" of around 10 - 30 m.
With this situation you have a clear and loud direct-sound-part and also a very clear part of delayed reflections. And for this situation my theory seems to work... as the example2 shows us.
(10ms= depth 1, 25ms depth 2, 40ms=depth 3 > Neil)

But Peter, let me now turn the strategy:
When I'm wrong - then I need a better solution - based on your theories...
Do you agree, that the violin in Example2 sounds farther away then in Example1 [*-)]: (I've changed the harpsi to violin-examples > No more echo!) So or so...

I own the Samplitude Roomsimulator, the Acoustic Mirror and SIR.
Can you give me a setting for one of the upper effect-programs which pay attention to your theory(ies) but with the same effect of my simple predelay-settings [*-)]:
I will love you for such a setting [[[;)]]]

BTW
I've decided to buy the Gigapulse-VST-Effect. ...Christmas you know.
From then on we will have peace in space... [:D]

With love [[[;)]]]
Beat

I've always used predelay as a way to get away with using a lot of reverb while mainting clarity. It certainly does have an effect on the space, but to me it makes it larger more than it moves the instrument farther away.

It's also important to remember that there are frequency cues to tell us how far something is away (brighter = closer). That's one of the reasons a close-miked instrument sounds close-miked even if it's made to seem farther away, i.e. it's a different sound.

So, after much thinking and studying - and typing here! - I've got some great results.

Thanks to everyone - Beat, Peter et al - who helped me to understand much more about this complex issue of reverb. My advice is listen to what everyone thinks and try it! Then modify and LISTEN. I'm very pleased.

Cheers
Neil

All very useful. I've digested it all and definitely improved my mix....

Thanks very much - and the main thing I have been reminded of is to experiment and listen very carefully.....

Neil

@Neil C said:

- and the main thing I have been reminded of is to experiment and listen very carefully.....

Neil

This is the healthiest advice one can get!!

I've enjoyed following along in thread and have also learned a great deal about how other people perceive sound in an effort to recreate it.

Thanks everyone.

@Neil C said:

[...] modify and LISTEN. I'm very pleased.
[...]

That's what mixing is about. Great posting, Neil, even half a year later!

/Dietz