EQ of reverb

Hey dog,
 
I usually cut both high and low frequencies, to some extent. 
It's mostly important cutting the lows to keep them tight and punchy.  For vocals I find myself cutting much more of the lows to keep part of the vocal frequencies dry, while allowing the upper mids/highs to have some space.

Hi Reece,

Got any rain lately??  We're having some here on the other side of the Metroplex.  From reports, it looks like you may have gotten significantly more.

On high freqs and reverb: Although you asked Tarsier, I also have some knowledge of this subject.  Yes, this is a fairly well understood problem in acoustics wherein higher frequencies are absorbed more quickly by many materials and media so that the reverb time (e.g. "T60") varies with frequency.  In my reverb codes, I followed some advice that was probably from one or more of the classic texts (e.g. Morse and Ingard, Pierce, Kinsler, etc.) and used a linear approximation for the decay exponent determined by two T60's, one specified at 125 Hz (longer) and the other at 4000 Hz (shorter) --- at least this appears to be what I did, written several years ago and not particularly well documented at that.  Actually I thought it was piecewise linear, but doesn't look that way now!

Edited: I should point out that this is really not the same thing as simple EQ'ing of the reverb, however.  Proper modelling incorporates dynamic EQ-ing of the reverb or the IR.  Early reflections will have more high frequency content than later reflections and the so-called "reverb tail."

Best regards,
Dave
 

hey, guys, thanks for chiming in!

jamz - what kind of range are you talking about - and also - since you specifically mentioned vocals - does that mean you use different EQ for instruments?  or even a different verb?  I tend to use perfect space on a reverb bus and run sends to that bus from everything to try to keep the verb cohesive.  is that a bad approach?

dave (papa) - that's pretty much what I was thinking.  thanks!

dave clark - actually I thought of you when I posted this as well!  I just haven't seen you a lot on the forums lately, so I didn't specifically mention you.  yeah, we got some good rain yesterday and it was about time!  it's been what over a week that the humidity and clouds have been hovering over us?

anyway - that's great info and I know your modelling does much more than just static EQ (or at least I think I remember that from our earlier conversations!), but I just use the tools available to me like I described above.  but your post was very helpful, tho, thanks!

Albert - which verbs do you use and how do you use them?  do you put them on each track and use the presets for dry/wet mixes?  or do you use a bus like I described above?

jimmyman,

What's up with the formatting of your posts? Ten words per line?

Think about this: the concert hall is very spacious. What is the likelihood of high frequencies having the energy to bounce back at all? They will bounce and lose their power quickly. Lower freqs will seem to sustain a lot longer and that energy is what you hear in a church or hall.

Stand outside a dance club: hear any highs? No. Feel a ton of lows? Oh yeah.

Bit covers it all pretty well there.

I'd like to go into a little more detail. I think it's easiest to understand how reverb works by thinking about impulses. Imagine you have an impulse of pink noise in a room. That sound has a flat frequency response (perceivable, approximately). Lets imagine it's from a three dimensional source, so it is not directional. The sound starts moving spherically outwards. As bit said, some of the highs are attenuated. This is because the air acts as a dampener. It absorbs the really high frequency content a lot, and the high just a bit, and so on. It keeps traveling till some of it hits a wall. Some of the sound is again absorbed. Probably more highs. Then it travels back again, with some of that energy going back to the source, losing highs as it keeps traveling back. It hits the source. The first early reflection. Other refecltions quickly follow. This tells the brain most of the detail about the room. By now, there are so many reflections going on that they diffuse into a smush of sound, the reverb component, all the while losing those highs in the air and probably lots from the surfaces they are refeclting off too.

So if we record all that sound that gets back to the source, that's the reverb impulse. Lets say it's 1 second long. What if we break it up into octave bands? We get many impulses. The 100-200 impulse might still be 1 second long. The 200-400 might be 950ms, the 400-800 is 900ms, the 800-1600 is 800ms all the way to the 10,000-20,000 (which I know is not in the same octave band, just for illustrative purposes) which is only 100ms long.

So what does an eq do to your sound? Is it natural? If you cut the 10-20khz band by 6db then that band will now only be 50ms long IF it has a linear response. It could be natural, but it limits you. Lets say at the start of that octave band, the level was -10dB and the end is -60dB with a linear response. Cut that band with eq by 6dB and it now starts at -10dB and ends at -66dB for the same length. But reverb time is to -60 (we are assuming the full impulse is normalised) so we shorten the reverb time back to the -60dB point. This effectively dampens  the high end, but also dampens those early reflections. In a reverb emulator, it could keep that band still at -10dB at the start, but shorten the decay, thus making a steaper slop. One reverb might be linear. One might act more like a fast or slow curve. All can happen in reality. Eq of a reverb can't change these things. It is very restricted. But you can edit that impulse. You could make two copies, one the same, but the other one with the highs all cut. Then cross-fade the two together. Keeps the high end high at the start of the impulse, but dampens those highs later on in the response.

Every reverb emulator sounds different because of so many different things, one being how it deals with the decay of the different specral components. They all use shortcuts of some sort. Real life is so much more complex. Does that make eqing of a reverb pointless? Certainly not. If the emulator provides some sort of dampening control, as bit has already mentioned, using that and before and after eq gives you plenty of flexibility in changing the way the reverb breathes, among many other things. If it doesn't have a dampener control then you are much more limited, but you can still get a great sound.

Reverb is so complex, but understanding how it works and what all those controls are trying to emulate in real life is very handy.

As for specific frequencies to cut it at, it all depends on the source and the song and artistic considerations. Certainly some low frequency and high frequency eq can be very helpful in getting the reverb right in your mix. Just play around till it sounds right.

"which verbs do you use and how do you use them?  do you put them on each track and use the presets for dry/wet mixes?  or do you use a bus like I described above?"

I have pared down my vst's and other toys over the last few years. I found that having too many choices was worse than having not enough, since you can audition one after another after another until you get nothing accomplished. So, now all I am using are the Kjaerhus Classics, a few other toys, and the effects that come with sonar and sound forge. So, I will use the studio verb, the impulses in sound forge and the classics, depending on the desired effect.

I usually work track by track, seldom use sends, and will use bus reverb if I have several things going to the same sounding reverb. I like "simple, easy to duplicate, easy to remember" things.

In that sort of mindset, and I don't want to minimize the amount of information, the thought process, or the value of the knowledge that people like mattplaysguitar have at their disposal, but for me, it is too much information. I want something that works, sounds good, can be adjusted, and has few buttons and switches. I know how an internal combustion engine works, I understand hybrid, hydrogen, solar and electrical powered vehicles, lithium batteries and that nasty dirty gas we use for the fuel, but I really just want to get in the car and go.

One other thing on the reverb, I am not married to the idea that all of the reverb for the different instruments and vocals in a particular song match up, I don't need to have the song sound like it was recorded in the same room at the same time. The mixing process is as much of the creative process as the recording, playing and writing in my mind.

Unless, of course, I want the song to sound like it was recorded in the same room all at once, or the client tells me to do it "his way", and then all bets are off.

Later
Albert

I PMed you.

LOL younger?
 
I think most of us know that Bit probably didn't invent the idea...  More like he reminded us to try it.

anyone care to comment on this? tarsier - I'd be especially interested in hearing more details on your thoughts about this. at approx what frequency should the cut off be?

Cut it off where it sounds good.

Ok, flippant response aside, others have covered it pretty well and I don't have a whole lot to add.  Some people like to eq the signal before it hits the reverb, some after. I tend to eq before.

I did find a nice Atmospheric sound absorption calculator where you can play around with choosing frequencies and finding out how much the atmosphere will absorb them over distance. Combine that with charts of sound reflection characteristics of various materials at various frequencies and you can start to get a better understanding of how reverb decays.

And everyone should read Spaces Speak. Are You Listening? by Dr. Barry Blesser of Lexicon and EMT fame. And read his whole website. And listen to spaces. Neat ones show up in the oddest places.