Reverb Plugins

[cptexas]

OK, I want to learn more about reverb plugins, becuase all these knobs and sliders can get confusing. I understand what most of the knobs mean, but I'm confused about the "size" knob. I assume this controls the virtual room size? If you make the size really big, does that mean your signal needs to be big enough to fill that size?

Like if you take a string quartet and put it in my bedroom--the room's too small and the sound is too loud and, well, bad. If you put that string quartet in a small to moderate sized church, I'm sure it would sound very nice. However, if you put that string quartet in the Atlantic City convention hall the sound would be very thin and echoey. That's why the Atlantic City Convention Hall has the world's largest pipe organ in it--it's the only instrument that can fill such a large room with sound.

So with reverb plugins, should the "size" of your virtual room be appropriate for the size/power of your enseamble? And what exactly does the "size" knob do to the signal?

-Chris

[wst3ae]

geez... I could write a book<G>!

Reverberation describes a very specific behavior of sound energy in an enclosed space. The space has to be sufficiently large to be reverberent, by the way!

The size of a specific space, the shape of that space, and the materials used on, and in all the surfaces, all play a part in the makeup of the reverberent field, if it exists.

Reverb plugins use a variety of methods to approximate what your sound would sound like if it were played in some specific reverberent space.

Some plugins use delay lines, LOTS of delay lines. Some more recent attempts use a mathematical concept called convolution.

The method isn't all that important, though it will limit some of the things you can do.

Most reverb plugins include variables such as the basic space, the size of the space, pre-delay, density, depth, distance from source to pickup, etc. And each plugin uses most of them in slightly different ways.

Without knowing which plugin you are using I can't tell you what the space knob does, but any good primer on reverb will give you a good idea of what the various controls are supposed to do.

There was a very good one on the web, but I can't quite put my finger on it at the moment. I will keep tryin.

The artistic question is much more difficult!

So with reverb plugins, should the "size" of your virtual room be appropriate for the size/power of your enseamble?

Short answer... yes!

You need to pick, or rather build, a reverb that works for the recording. There are two distinct schools of thought on this. The first suggests that we should try to emulate as closely as possible the real sound that would have occured if we had placed real players in a real space. This is pretty popular with folks doing orchestral mockups, and to a degree everyone gives it a try. The second school of thought says that since nothing happened in real time in a real space anyway there are no rules. Put the oboes in a canyon, the horns in a chapel, and use a plate on the strings. You'd probably be surprised at how much fun you can have with this approach!!!

If you can try to pin your questions down a bit I'll try to answer them better.

[cptexas]

Wow!
Thanks for your post, Wst3ae!

Very helpful stuff. In terms of my actual question, I don't really have one in particular...I'd just like to know a bit more about reverb plugins and how to use them most effectively. More of a general question. I use the Cakewalk FX reverb usually if that matters. And sometimes Ambience because the church and cathedral are BEAUTIFUL!!

-Chris

[wst3ae]

Well Cakewalk's FX Reverb plugin and Ambience are pretty different sounding reverbs, at least to my ears! Variety is good!!!

First things first... the most important decision you have to make is the balance between direct or dry and reverb sounds. That's hard!!!

Back in the day<G>, when we used springs and plates mostly, the rule of thumb was to turn the reverb up till it was too much, and then back off ever so slightly. This strategy can work with digital reverbs too, but I've found it is much more difficult to find that point with a plugin!

The key is to know and trust your monitoring environment, and the only way you get to that point is to listen to a LOT of stuff through it. And, comparing the results from your monitoring environment in other environments!

That's this morning's thoughts... I'll have to play around with the two reverb's you mention and see if anything else comes to mind.

(FWIW, I use the UAD Trueverb, the Lexicon Pantheon that came with Sonar 5, and Purespace from Voxengo as my go-to reverbs, but I have a bunch of others that I can turn to. One of my current favorites is a freebie, PSP's Pianoverb... quirky, but cool!)

Bill

[Skysaw]

Just stopping by to add a little more confusion to the mix...

Here are some of my personal "rules of thumb."

1. Never balance the reverb while monitoring in headphones... you will mix in too much.
2. Usually low frequencies don't need nearly as much reverb as mid and high frequencies. Cutting some of the lows and low-mids that feed into a reverb bus can clear up what might otherwise be a muddy mix.
3. Except for traditional realistic mockups, consider using a small amount of digital delay instead of reverb on one or two tracks for busy mixes. The result might be a little clearer, and possibly a little more interesting.
4. When you finally have the mix perfect, trim just a little more off of the reverb. See...? It *was* too much!

[Garritan]

geez... I could write a book<G>!

Maybe you should write a book. You have a lot of knowledge in this area and thanks for sharing it.

We are currently working on a project called "Real Spaces" where we have taken impulses of every player position on the stage of various concert halls (up to 90 positions) and upwards of 15 micophones in the hall. We have been getting some very impressive results.

Gary Garritan

[cptexas]

OK, and this convolution stuff...
How the heck do you sample a ROOM?!?? It's dead...rooms don't make any noise. I downloaded SIR and played around with it. I put in a sample of a BIG church bell and my orchestra sounded very interesting from inside a bell.

So do you sample clicks from different positions on the stage, like clicks of white noise or something? I have no idea...

BTW: about the FX Reverb, there's a little waver of pitch at the begining of the echo. This isn't much of a problem with strings, but when I use a piano or harp or any chromatic percussion, that waver is really annoying. Any ideas?

Thanks for all your info, very helpful!
-Chris

[wst3ae]

Maybe you should write a book.

hmmmmm.......

You have a lot of knowledge in this area and thanks for sharing it.

I still receive a great deal more than I return here, but I'm working on it!

We are currently working on a project called "Real Spaces" where we have taken impulses of every player position on the stage of various concert halls (up to 90 positions) and upwards of 15 micophones in the hall. We have been getting some very impressive results.

Indeed... an intriguing concept.

Having grown up stringing spring and plate reverbs together with analog (and some digital) delay lines, I'm of the opinion that really accurate reverb is cool, but it really is just another color to paint with.

Convolving reverb, even without going to the level of detail that you are going to is a remarkable advance in the art. I really am curious to see how "Real Spaces" translates to the ear.

Cakewalk had a reverb plugin back in the era of CWPA9 I think that allowed you to set the size of the room and the placement of the source and the microphone(s). It was really cool. I still use it. There were a couple of other reverb plugins that worked along the same lines, and they were cool too, but none seem to have survived.

And I don't mean to question the value of a tool like "Real Spaces", or Altiverb or GigaPulse, every tool has it's uses. Like I said, I still use my spring reverb, right along side Pure Space and Pantheon and TrueVerb.

They are all colors with which one can paint! From what I've read so far about what you are planning for "Real Spaces", and being an user of GPO and JABB, I think it will be a very useful tool, and not just as an accurate reproduction of a space<G>!

Bill

[dnortana]

OK, and this convolution stuff...
How the heck do you sample a ROOM?!?? It's dead...rooms don't make any noise. I downloaded SIR and played around with it. I put in a sample of a BIG church bell and my orchestra sounded very interesting from inside a bell.

So do you sample clicks from different positions on the stage, like clicks of white noise or something? I have no idea...
.....
-Chris

There are a couple of convolution reverb impulse response recording techniques that I know of. One is to fire a starter pistol in the room, and record the whole process. The other is to run a sine sweep (low to high frequency and everything in between) using some good quality speaker, and again record the whole process. You feed these recordings into a processor, and voila, you have an IR.

Trond

[wst3ae]

not sure what I did wrong... I wrote this wonderful post, and submitted it, but it never showed up! This will be a much shorted version!!!

OK, and this convolution stuff...
How the heck do you sample a ROOM?!?? It's dead...rooms don't make any noise.

Rooms don't make noise, but things in rooms do!

There are, as already noted, two ways to gather a room signature. The first is an impulse. It could be a starter's pistol, a popped ballon (very effective) or, in one recent case, a yachting cannon! The trick is, the impulse has to be very short so as to include energy in all the areas of the spectrum that are of interest.

The second way is the swept sine wave. This was developed by Richard Heyser as part of Time Delay Spectrometry, an amazing tool that is capable of turning any space into an effectively anechoic space. He was a rocket scientist (really) at the Jet Propulsion Labs, and his hobby was audio. He was looking for a way to make speaker measurements, and from that he came up with TDS.

The really interesting part is that if you have the swept wave response you can calculate the impulse response, and vica-versa. In fact (I hope I'm not misquoting, it was a long time ago) according to Mr. Heyser the energy includes directional information such that a single microphone measurement could differentiate from whence a reflection originated.

If you search for TDS or Richard Heyser you'll find tons of information on the topic.

I downloaded SIR and played around with it. I put in a sample of a BIG church bell and my orchestra sounded very interesting from inside a bell.

A very good start. Keep in mind rule #1 for music production... there are no rules!

So do you sample clicks from different positions on the stage, like clicks of white noise or something? I have no idea...

That is the basic idea. The reverberent field will be different at every location in a space if it is truly reverberent. Acousticians work with average values because, on the whole, that works. Prior to practical convolution (or suitable processing horsepower) that's how digital reverbs worked. Except that they used sequences of delay lines<G>!

BTW: about the FX Reverb, there's a little waver of pitch at the begining of the echo. This isn't much of a problem with strings, but when I use a piano or harp or any chromatic percussion, that waver is really annoying. Any ideas?

That's not a bug, it's an artifact of real reverb as well as digitally emulated reverb, especially if delay lines are used. The principle is called the Doppler effect, and it does mess with the perceived frequency of a sound. I've not noticed it in the FX reverb, but I will definately give it a listen.

Me thinks your ears are pretty good!!

Take care,

Bill