How much reverb? (samples)

[jnorman]

these days, i am hearing a variety of approaches to reverb levels on live recordings, from nearly dry to very cathedral-like. i am including 6 samples of a short clip of flute and piano for your consideration. bricasti IRs prepared by Acousticas (thanks guys!)

clip1 is the dry track
clip 2 is a small amount of a bricasti ambience IR in SIR
clip 3 is a bricasti echo plate IR at -8dB
clip 4 is the echo plate IR at -12.5dB
clip 5 is echo plate IR at -17dB
clip 6 is echo plate IR at -20dB

what is your preference? any other comments are welcome. thanks.

[BrianHanke]

Number 4 gets my vote!

[jnorman]

thanks brian - i sort of liked #5, but the tail is rather short.

here are 3 more: (sorry, i dont know who prepared the 960 and tc6000 IRs, but they are great)

first is lex 960 medium hall at -18
second is poor plate by vacuumsound at -7
third is TC6000 medium hall at -18

[Douglas Whates]

For me, it's either clip 2 from your first batch, or clip 2 from your second batch. Anything else has that annoying "I-can-hear-the-original-less-than-ideal-room-and-slapped-on-artificial-reverb" thing, which tends to remove me from any sense of being in a real space.

[7rojo7]

Sorry, I'm responding without listening.
However, I feel that it's important to realize something.
When a musician is performing, it plays first; responding to what it needs to output and secondly; to what it hears in return to that output. To try to seperate these two factors is a mistake. When synthesizing a reverb to put on top of a reverb (naturally recorded) you must take into consideration the tempo of the music, the size of the space you would like to reproduce and several other factors. It's not as simple as choosing the Carnegie Hall patch and then call it a night. There's a good chance that there's a fair amount of ER with diffuse reverb in the recorded signal, then sending a signal that is ER heavy into a reverb that has a hefty dose of ER already is going to give you some heavy comb filtering in the first 60 ms (attack time=muddy attacks with the addition of a ER heavy reverb), specially if you can't match them. Pre delay? Wait for the original ER to decay and start your reverb from there. Otherwise things can get complicated and overblown rather quickly.
More than the question: How much reverb? is: WTF did I record and what kind of reverb with what level of ER and how much pre delay will reinforce, change or fix what I've already recorded and how much time do I want to spend to try to find it again for this recording, time forgiving. Every case is different
There have been month long editing and mixing projects where I hadn't nailed the verb til the end of the project. It just sounds right when you get it.

[jnorman]

7red - thanks for your comments. please do listen to a few of the clips, and let me know what you think.

[7rojo7]

#3B with a little less, the halls are sounding right, I'd try a smaller hall (warm baroque) or a theater (operetta size). Do these mixes have the verb sent from an aux or are they just mixed with the wet dry mix on the verbs themselves?
I noticed differences in all the pieces, specially the level of dry sound hence the overall level. I might try to give it a bunch of reverb but not push it to the back.
The Bricasti sound great (the best in this context) but the patch is not for me, a little too bright. I hear more mystery for this. nice piece. nice recording.
I just spent 10 mins to get this, it's close enough for 30 secs
I shall not be held responsable for double MP3 encoding

[mixedupsteve]

I like 1B. Intimate and kind of dark. Let me know when the record is available.

[Casey]

Originally Posted by 7rojo7

There's a good chance that there's a fair amount of ER with diffuse reverb in the recorded signal, then sending a signal that is ER heavy into a reverb that has a hefty dose of ER already is going to give you some heavy comb filtering in the first 60 ms.

I agree with all of your points, including this one.

I felt it worth mentioning though, that this exact problem was one of the first issues that I had to overcome in the development of the Bricasti algorithm.

Try any recording, in any room, turn off the main reverb, just leaving the ERs. Adjust the earlies any way you need, you will never hear any combing.

Overall, the algorithm works very hard to avoid fighting the room.



-Casey

[jnorman]

casey - thanks for commenting. always appreciate your input.

7 red - excellent remarks, thanks. very thoughtful repsonses.

steve - this is one of my wife's students, and she is very good for a 17 year old. i will let her know someone liked her work. she jsut won both state and regional MTNA contests, and is now heading for MTNA finals in albuquerque, NM. the piece is "medieval suite" by katherine hoover.

[Casey]

@jnorman, I noticed in another thread that you mention that these samples were done with reverb IRs?

I would very much appreciate the appropriate attribution in your initial post if this is true in the case of the "bricasti" files. I probably missed some other context where this was made clear!

Thanks!



-Casey

[jnorman]

casey - you are correct. i usually try to indicate that i am using a convolution reverb with sampled IRs - sorry i overlooked that. these are the samples of the M7 prepared by Acousticas. they are very nice, and certainly make me wish i could obtain a real hardware bricasti unit - i believe the bricasti is probably the best reverb unit made at this time.