I figured since a lot of questions have been asked, it was time to just post up with some tips for the peoples. Everyone has a different take on how to create space and environment. This is just my approach.

What is the major advantage to miking an instrument vs. going line in? Guaranteed, it is the sense of real space. Space is as much a part of the music as the music itself. It's the illusion on record that not just presents the sound, but takes you to another place, allows you to "zone out." Literally, this means to leave the zone, or space, you are in, and go somewhere else.

Spatial processing is a rough science. There is a mathematical component, and an artistic component. The advantage to artificial reverb is that you can negotiate the math to bring out the art, by create surreal environments.

But too often, we underplay the value of good reverb or we get lazy in our approach. Or, we overplay the value of reverb, and turn our reflection up way too high. Lastly, we forget about the musical context of reverb, and focus too much on reality - disregarding that reverb is still a sound element in our mix with rhythm and tone.

Let's start with the overplay of reverb as this is an easy topic. Less is more. In a live setting we tend to hear a lot of room sound - but realism is probably not our primary goal. It can be - but often times we want the music to be crisp and defined. Lots of reverb is the enemy of crisp and defined. But a little reverb can add depth and energy.

Artificial reverb uses a series of mathematical algorithms to generate new sound based on the source material. This includes any reverb that is already in the source material. Artificial reverb sounds better when you have a bit of real reverb feeding it. Why? Because it ends up bringing out that real space sound. This doesn't mean record in a church, and then add subtle artificial reverb (you can of course). Rather, it means, don't expect a totally dead room mixed with an artificial reverb to sound good. This is a common problem I call "studio dark;" that endless dark abyss that shows up on too many studio recordings. We all know that an amateur mistake is to over pad a vocal booth with acoustic foam - you get a tight, dead room, and a muffled, dead sound. We're much better off with a decently sized room with appropriately placed absorbers and diffusers. We can use artificial reverb to exentuate and control the tone and rhythm of our natural reverb! - Oh, side note - if you have a synth going line in, you may want to consider miking your monitor to get the actual capture you're going to use, just so you can have that real space in there.

Now, you need to have a good idea of how each instrument will function in the song. The prominence of the instrument will determine how live the reverb should sound. Why? The reverb is going to reinforce whatever is feeding it. If you have a lead vocal generating a flat reverb, and the drums generating a bouncing reverb, your drums are going to get the attention - which can be good or bad, depending on the scenario. That being said, let's say you want the lead vocals to be the focus. Instead of throwing a compressor on your lead vocals, bus your lead vox to an aux channel, and then compress. Go to the uncompressed vocal channel, and throw an aux send to your reverb channel. This will give you controlled vocals, but live space!

We will be sending other instruments to our reverb channel, but let's focus on getting the reverb right on just our lead. If we can get it right here, everything else falls into place. To do this, we need the reverb loud, the diffusion and decay/density and room size all the way down, the time, and the predelay all the way down. You may have to turn the time up a little so you can actually hear the reverb, but you want to hear it as close to a pulse as possible.

Start with the predelay. This is the first rhythmic element of your reverb. Let's say your bpm is 120. That means you have one quarter note every 500ms. You have one eighth note every 250 ms. You have one sixteenth note every 125ms, you have one 32nd note every 63ms. 64th notes at 32ms. In order to have the predelay trigger the reverb in a rhythmic fashion, it needs to be at one of these measures. I'd go with 32 or 63ms, because we want the reverb to still feel attached to it's source sound. Also, you notice how I rounded up? That's to put the reverb "behind the beat." This helps create a rhythmic pocket. I might even suggest moving the predelay higher a couple ms, just to make that pocket a bit more open, and so that the hit of the reverb isn't directly overtop the next part of the music.

Now move on to the Duration. Using the same time rubric, we can determine how long we want our time to be. Texturally, we want our reverb to be clean. This means a long time is going to create a wash sound, and defeat the work that we did on the pre-delay. The reverb time in a small room is very quick. A quarter of a second, give or take. But rhythmically, we want our reverb to pull us into the meter of the song. So we want our length to line up along the same rubric we found for the predelay. Let's say 250ms if we're going for a room slap effect. This way, we're moving right along to the eighth note pulses. BUT! Consider that our predelay is set to, say, 34ms. This means we have the tail of the reverb pulling us to the beginning of the next reverb, not the beginning of the next pulse. Subtract your predelay from your time, 250-34ms. Set your reverb time to 216ms. Remember, even if you don't hear the difference between 216 and 230, if the track is going to get heavily compressed, that difference WILL stand out. Now, for a lot of applications you might want a longer tail reverb - less of a room slap and more of a bigger resonance. A good meter to use for this 1 bar measurement - which at 4 beats per measure, 500ms per beat = 2 second reverb tail.

EDIT: From three years in the future here. Choosing a duration for rhythmic purposes is about deciding where you want the reverb to pull the ear into. If you are doing a dance song, with a snare on 2 and 4, a short slappy reverb sound on the snare that rhythmically connects the 2 beat to the 4 beat might be good. From the example above that a 1 second duration. But let's say you have a vocal - you may want the phrases to connect which would require a longer reverb - this has a lot to do with feel because phrasing can vary a great deal - but a 1 measure reverb may in fact be better.

-Side note- predelay and time are subject to taste. Use the math to get where you want. I usually find the math gets me right there. But make adjustments and go with what sounds best.

Size. Size will effect the tightness of the sound. This is more creative, but here's a good starting place: Divide your time by ten, and round to the closest integer. That's a good number of ft for your room size. Adjust up or down according to tone and texture. Too large will get a spacey characterless sound. Too tight will sound almost more like a echo/delay.

Decay/Density. This is somewhat ambiguous, and different processors will give you different results. Basically, I equate this to presence. Remember you have your reverb up loud now, so adjust the presence to match with the instrument that's feeding it. You probably want your reverb a nudge less present than that. Consider it like a "tone" control. If you hear a metallic oil tank sound you may want to turn the density up, but if you hear the reverb washing out or masking other sounds in the mix, you may want to turn it down.

Diffusion. Diffusion is the scattering of sound waves. A highly diffuse room tends to make things sound very distant, open, or haunting, whereas a non-diffuse room makes the reflections sound more like one unified echo that washes back at you. I find diffusion can help add a sense of being "further away" without actually changing time constants.


Damping. Picture the material you want your room to have. Is it metal walls? Is it oak wood? Picture your ideal room and how it sounds, then adjust the damping to meet that. Damping is how long frequencies persist during the period the reverb is presence. Basically fiddle until you get it right. Your going to eq in a moment anyway.

BUT FIRST

Compression? Sometimes, there is a lot of attack sound in reverb. Especially when hot transients from your drums are feeding into it. It might be worth taking a compressor and setting as fast an attack as possible and a really quick release, just to ease off the transients. But be careful. Compression will eat the energy out of your reverb (if there's too much energy afterall, that's the point). Err on the side of caution, and if you're not sure, leave it out. I find I don't usually need to compress out the transients in the reverb. But every now and then, I do.

Ok, now EQ. The idea of the reverb is to lie underneath your sounds. You will find that as other instruments start feeding the reverb, you get fatty build up somewhere in the high mids, somewhere in the mids, and sometimes in the bass/sub bass. Your goal is to find the fattest point in these three sections and use a wide bandwidth to ease this build up away. A HP filter may not be bad to get rid of the bass build up. You can generally cut out anything below 70Hz. Be honest though, that rule doesn't apply 100% of the time - maybe 90% of the time.

FINALLY, start lowering the sends on each instrument individually. Solo one instrument at a time. Lower the send volume until the reverb exists, but the total clarity of the instrument is still present. After this is done for all instruments, start adjusting the send levels by little bits based on what sounds best: Here's a tip: Normally we like to put heavy reverb on things like strings and pads. While that seems logical, it's actually not the way to go. Think about this, if you have a constant tone playing at one level, even if it's way down in the mix, won't it stand out? Take instruments that naturally sustain at one level, or have lots of compression on them, and reduce the amount of reverb. The more dynamic instruments like vocals, should have the most reverb. That means drop the reverb in your heavily compressed drums, and your sustaining pads - not boost the reverb on your vocals. Less is more.

NOW, the final test.

Mute your audio. Mute your reverb aux channel. Wait a minute, shake your head out. Stand up, get a drink of water. Come back and unmute your audio.This is the dry sound, no reverb. Listen for a moment, then unmute the reverb. Does it sound like the song just came to life!?

If it does - your done. If it doesn't, it's back to the drawing board.

sounds like...
Wall of Sound - Wikipedia, the free encyclopedia

is there a chart that gives your the break down of the bpm vs correct pre delay.

Interesting approch.

Allthough i always do the math for delay i usually go by ear for reverbs.
I'm gonna have to try it

good thread

-Alxi-

Great post, i'm going to print this out. I just realized i have no clue what i'm doing.

I'm ending up using more delay than reverb but there are some pretty great pointers in here that I will add to my arsenal...thanks
cam

Delay Chart

Wall of Sound - maybe in some respects. The overall goal is different though. I'm going for a subtle but strong reverb, one which will not result in blurring between instruments - where as Spektor was going for an orchestral sound.

Chart - nice chart. The math ain't too tough if you don't have the chart off hand.

Math - be careful with mathematics on a delay. If it's too metronomic it can have a negative effect on the music. With live drums for example, if your player is a bit behind the beat on his/her downbeats and the delay is set to tempo, you might end up with your delay actually making the track sound less cohesive. That being said, if you have a good number of instruments playing around time, but one specifically that's dead on, an perfect delay might help show the contrast around the tempo. I wouldn't say the math should EVER be the final step anywhere. The math is helpful to get you to a starting point, the ear is ALWAYS the final judge.

No Clue - Yeah, reverb is tough cause there's a whole lot of knobs that have really subtle effects. My post above is just one of countless approaches to reverb. Ultimately you find your own system. The one above is mine, and took a great deal of time to come about.

Delay vs. Reverb : Interesting what makes us choose one over the other. I find that people go to delays because reverb seems inferior: It muddies the mix/ washes the transients/ adds nothing rhythmic. Hopefully my post above will allow reverb to be a viable and exciting option in place of delay, as a well used reverb shouldn't do any of those terrible things I just mentioned. However, this is not to say that a well placed delay doesn't kick ass. I find myself hitting the reverb more than the delay, even in Hip Hop tracks.

Nice! Thanks for taking the time to do this.

I always though diffusion controlled density vs. sparseness of the wet signal (but not "mix"), but the way you explain it makes a lot more sense. Thinking of it in terms of distance gives me some new ideas.

As an aside I have had some fun lately with putting a reverb (with generous pre-delay and approx 70% damping) after a slapback delay. Panned the buss about 80% the way left, for instruments that were slightly panned right. It sounded like things were bouncing off a wall in a long hallway!

Used Virsyn Reflect and Virsyn delay, which work PERFECT together for this.

I dont know WHAT I would do without both these tools

BPM Calc

And

MusicMath

They have got EVERYTHING you need. And they are FREEEEEEE

Any tips on panning the reverb?

Thanks for this. I've been meaning to really get to know all of reverb's ins and outs for a while now so maybe this'll get me on track. thumbsup

Also any tips for setting a sound far back without making it appear quieter?

And maybe a tip for creating "faux bleed" between instruments?

don't get hook on that. Sometimes using the incorrect predelay tempo works well also.

Absolutely. First understand that a panned reverb is NOT a realistic effect, it is a creative one. Then again, using reverb as a rhythmic devise falls into this category too . I pan reverbs rarely, but there are three occasions where it comes up in my mixes:

1) I want to give a mono instrument more stereo content. Panning a mono sound helps to create a stereo image, but sometimes that stereo image is not concrete enough. For a WIDE feeling, pan the reverb slightly to the OPPOSITE side, and keep your predelay tight. For a SOLID STEREO IMAGE, pan the reverb in the... OPPOSITE side as well. A little counterintuitive, but remember that the sounds are all relative. If you have a longer predelay, and a longer decay time, this will exaggerate the position of the source sound in the stereo field. Panning to the same side will create a fractured stereo image, almost like two locations clashing together... which can be disasterous, or really cool.

2) Subtle panning to flavor up the space. Panning multiple reverbs with similar settings a little away from the center won't be heard as panned reverb. It will be heard as one reverb, with a bit more space in the center. This will really make your center instruments pop through the stereo field. Major panning here will have the same fracturing effect as above.

3. Creating a surreal landscape. Remember that panning reverb by it's nature is Dali. Don't ever be afraid to push boundaries. Make a mono mix, and pan everything to the left, then throw all your reverb to the right. It'll sound weird. Or automate the movement of the reverb. For that matter, automate and increase and decrease in the predelay. The effects will be wonky, but if your mix isn't too complex it could be exciting.

Decreasing the volume and increasing the reverb, and increasing the diffusion will help create a far back sound without too much volume loss, but there's much more than reverb at play here.

Think about sound over distance. What sounds lose the most energy as they travel... Low frequency, or high frequency? High frequency requires more energy to travel the same distance as low frequency. The morale is, if you shelve down everything above 2k, and you will create the illusion of a being farther back. You need to create a curve shape in the eq, so you have more loss as frequency increases. A wide Q on a bell with the center set to 16k might give you this.

In addition, the appearance of transients disappears over distance, and sustaining sounds tend to stick around longer. Makes sense, since a transient literally means "moving on" and a sustain means "sticking around." Less transient + more sustain = compression. If you compress away some of the transient your sound will appear further back. Bare in mind that pushing further back does not mean increasing your ratio, or lowering your threshold! To create the illusion of going back further, increase your release time.

Pan. If you have something panned hard left or right, it will never sound far away. Why? Think about it. If you are a foot away from someone who is three feet to your left, you clearly hear that they are coming from your left. If you are forty feet away from someone who is three feet to your left, it basically sounds like they are standing directly perpendicular to you. As things get farther away, they approach the center point. That's why having a lead vocal both up front and dead center is so prominent - it's really like someone is standing directly in front of you.

Now the reverb properties of something far away: The DRY SOUND realistically does decrease. It's just how it is. A doubling of distance results in a 6db decrease in volume. So volume should come down. The high frequencies damp off quicker, but this is a bit more complex than simply setting a faster hi damp - I'll explain in an example. The diffusion is higher up. The Early Reflections become more prominent: This will help make the sound louder even though we decreased volume (though we still might need to turn the dry sound up at the end of the day). Increase the predelay to move farther back. The more the reverb is heard as an "echo" the farther back it appears. At super distances the lows start to roll off to, because they tend to either pass through things and scatter, or couter phase with each other.

Example: You have three backup singers who sing AS LOUDLY as the main singer, and stand halfway left of the main singer, but also 10ft back. You envision your audience as being 20ft. away.

Formula: Repan the backup vocals to be 25% left of the singer. Drop the volume 3db. Increase the predelay to 30ms. Why? Speed o'sound = about 1ft. per ms. 20ft. from audience to lead singer, 10ft from lead singer to backups. Increase the early reflections by 3db. EQ out a db from everything above 2/3k. Maybe a drop more at the highest frequencies (this might give you a little more than a 10ft apparent change, but hey, you can always eq out less). Now because there is a bit less hi content, and the sound of the lead is in the song as well, you don't actually need to change the hi-damp. Keep it the same. Here's why: If everything moved back, you would have more hi-damping. Because the only one element is moving back, and our perception of change in volume in high frequencies is exponential we will percieve more hi-damping if there is less content there to begin with. It's a little complicated, but take my word for it. It also helps keep the environment consistent. Increase the diffusion %. Compress the transients out of the backups a bit on the way in to the reverb. Transient heavy reverb will not make the backups sound farther away.

great post storyville really interesting.

Keep them coming!!!

^^ Storyville, I'm going to read that and absorb it, but 1st thanks for taking the time to answer. That is appreciated.

Faking bleed: You might be better doing this with a delay. If you send the instrument to a delay, coming in only a few milliseconds later, eq out a lot of the high content and comb out some of the mids, then send that delay to the reverb you would get a bleed effect.

The delay should be very low volume, and the amount that is sent to the reverb should be equivalent in volume.

Bleed shouldn't be audible - it should be textural, and in generally characterized by phase issues, so handle artificial bleed with care.


Hey, and your welcome.

Yes indeed. That was helpful.

Basically, instead of turing knobs until it sounds good (which to a dergee i find myself doing), the focus should be to recreate the space and the signal from origin to the ear. Well explained. It all makes a lot of sense, especially about the absorption of transients over time and distance, and the little bit about compression release (but not ratio and thresh).

Lot to learn in this thread!

Can you elaborate a bit on how you calculate the Difussion settings.

And if you can elaborate a bit on Early reflexions... theire timing and theire volume... if that's not asking too much

I frequently use Epik Reverb by Bootsy cause i think it sound very good and because it's free.

The setting for the ER are making me a bit inconfortable

“epicVerb” digital reverberation simulator « Variety Of Sound

Thanks

-Alxi-

I'm not familiar with epicVerb but it sounds interesting, especially the "detailed control of the first reflections" part.

Imagine standing in a large room with a friend on the opposite side of the room. Your friend says something. The first thing you will hear is the direct sound or the dry sound. The next thing you will hear is a combination of the sound bouncing off the left and right wall at the proper angles to hit your ears, and the sound bouncing off the back wall. Together these form the early reflections. They are the most distinct reflections and tend to reveal the room's spacial characteristics. The late reflections are all the sound waves bouncing off multiple walls until they eventually reach your ears. These reflections sound more like a stream of reverb and reveal the room's tone, shape, and absorbing properties.

Diffusion represents a tonal effect of the room material and shape, and an effect of increased distance between the listener and the sound source. This is done a variety of way- higher diffusion will have more space between the transients in the early reflections, and a more echoey set of late reflections.

Diffusion is an odd setting, I generally use the texture to find the setting I like, rather than take too much stock in it show how far away something is. Too low the reflections sound buzzy. Too far and they sound kind of metallic. I usually go for somewhere in the middle. Diffusion works in conjunction with density. Density is the space between the early reflections and the late reflections, and again, supposedly indicate the distance from the listener to the source.

The reason diffusion and density don't clearly indicate distance though, is because the room shape effects these things as well. A round room will have a very low diffusion and density, regardless of where the source sound comes from. And we rely on a lot more cues, such as volume and hi-freq content to hear sonic distance.

Thanks for taking the time storyville .. Very generous of you.

Yes the detailed controls of the ER are the separate time and volume of the two first Early Reflexions.

Volume seems to react a bit like density... making sound appear closer but the time is really trowing me of. I just turn the knobs til it sounds good but i'm not even sure i'm listening for the right thing. I will listend in solo to see what i can learn.

Thanks for the clarifications though, very helpfull

-Alxi-

Why would you pan a send? I noticed a setting in Nuendo that says link send panner to channel pan pot.

No problem. I have a nasty cold right now, so this thread feels like a good way to occupy myself. And the questions that everyone is asking are forcing me to think and reinforce my own knowledge, so it feels pretty productive.

I'll have to test drive the plugin to be on the same page. Theoretically, and in my experience - the louder the early reflections, the more cavernous and distant the sound source should feel. If it feels like it's getting closer you may be thinking of bigger - the two are easily confused by our ear. Time control between sets of early reflections should essentially act like Density/Diffusion. But I'll go back to my original post and say that it's probably better to think of reverb in terms of tone and rhythm, rather than just the illusion of distance. Turning the knob until it sounds right is the bottom line. There are three periods for the engineer: Beginner, where you play with stuff until you make it sound the best you can, but without an experienced approach: Intermediate: Where you learn all the theory and apply it regardless of how it sounds because even if it doesn't sound right to you, you justify it by saying my ear probably just doesn't recognize good sound: and Advanced: Where you know exactly which knobs to turn, but you turn them to where it sounds good regardless of technicality.

Turning the knobs until it sounds good is how you discover your own approach, then I get to learn something from you.



As for why you would pan a send:

Let's say you have multiple instruments going to the same reverb, and you want to create a panned reverb effect. Let's say you have a trumpet panned medium left, and you want the reverb to pop up medium right, but you have a cello medium right and you want the reverb of the cello to show up hard left. You can't pan the reverb channel, because then everything goes in the same direction. Therefore, you have no choice but to pan the individual sends.

Then rest well. You gave us enough to justify a day off or two

On an other note. I feel a bit stupid. EpikVerb is supplied with a manual

Here's a quote from it about the ER settings:

Control about the very first early reflections is given with this parameter
section: Changes timing and level of the early reflections behavior - use
this to obtain a more focused or diffuse sound. It can change the overall
sound as well ranging from a more 'colorful' to a rather 'transparent'
sound. Changing the first affects the second and changing both affects
other reverb details under the hood as well. This way altering the timing
can affect the overall reverberation even if both 'LEVEL' controls are set to
minimum (left most position). Timing is displayed in ms. Level control is
unity in upper middle position and increases clock-wise (decreases
counter clock-wise)


Wich to me makes it a bit more clear

-Alxi-

Hey!! Great break downs! Appreciate it.

Enjoyed your reply on "distance" too.. that is relevant to part of a mix I will be starting soon that requires from abnormal reverb/eq processing to create realistic sense of space.

Great thread storyville. AA++

It´s all about to be in the "zone"...

Storyville,

I really dig/appreciate you sharing your approach to reverb... I've read it over and over... and have found it very useful... I have a couple questions... I understand this isn't an exact science, and my Ears will ultimately tell me if something sounds right or wrong... I just need a little more clarification.

Start with the predelay. This is the first rhythmic element of your reverb. Let's say your bpm is 120. That means you have one half note every 500ms. You have one quarter note every 250 ms. You have one eighth note every 125ms, you have one sixteenth note every 63ms. 32nd notes at 32ms. In order to have the predelay trigger the reverb in a rhythmic fashion, it needs to be at one of these measures. I'd go with 32 or 63ms, because we want the reverb to still feel attached to it's source sound. Also, you notice how I rounded up?

- Are you programming beats in double time? Or am I just confused...

Rhythmically, we want our reverb to pull us into the meter of the song. So we want our length to line up along the same rubric we found for the predelay. Let's say 250ms. This way, we're moving right along to the quarter note pulses. BUT! Consider that our predelay is set to, say, 34ms. This means we have the tail of the reverb pulling us to the beginning of the next reverb, not the beginning of the next pulse. Subtract your predelay from your time, 250-34ms. Set your reverb time to 216ms. Remember, even if you don't hear the difference between 216 and 230, if the track is going to get heavily compressed, that difference WILL stand out.

- Did you mean to say "the difference between 216 and 250"? And... does this apply just to compressing the reverb or Group, Master Bus Compression or @ the Mastering stage?


Size. Size will effect the tightness of the sound. This is more creative, but here's a good starting place: Divide your time by ten, and round to the closest integer. That's a good number of ft for your room size. Adjust up or down according to tone and texture. Too large will get a spacey characterless sound. Too tight will sound almost more like a echo/delay.

- Can you elaborate on Ft vs Room Size and how this translates to ms (and when you earlier said a Small Rooms is like a quarter of a second)... or maybe a link that I can read up on this...

Decay/Density. This is somewhat ambigous, and different processors will give you different results. Basically, I equate this to presence. Remember you have your reverb up loud now, so adjust the presence to match with the instrument that's feeding it. You probably want your reverb a nudge less present than that.

- What are the benefits of using Decay/Density vs... EQ and Filtering... for clarity and presence

Diffusion. This basically refers to how far the source sound is to your ear. Unless you plan on sending each element to its own reverb (using the same processor with the same settings and using different diffusions is not a bad idea if you have the time) set the diffusion as the average distance between the furthest element in the mix and the closest. If you picture your sound stage as having the drums at the other end of the room as the audience, and the vocalist being halfway toward the audience, you want your diffusion to be 75%. There will be other ques from the mixing of the instruments themselves to put them in their own space.

- Kinda get the gist of what your saying... I guess envisioning the Drums at the other end of the room, the vocalist center stage, and 75%... is throwing me off... Can you give an example with the drums behind the lead vocalist... Maybe I'm confused once again....

Compression? Sometimes, there is a lot of attack sound in reverb. Especially when hot transients from your drums are feeding into it. It might be worth taking a compressor and setting as fast an attack as possible and a really quick release, just to ease off the transients. But be careful. Compression will eat the energy out of your reverb (if there's too much energy afterall, that's the point). Err on the side of caution, and if you're not sure, leave it out. I find I don't usually need to compress out the transients in the reverb. But every now and then, I do.

- How does all this play out, when using Parrallel Compression... and Tricks/Tips or things to look out for....

- Do you reply more on the use of reverb or panning to frame your mix....

_ Any thoughts on applying delay to verb and vice versa?

Any audio your particularly proud of, that I/we can check out?

Thanks again for your insight!

Thanks again for this post.

But.....

Could you please review your numbers?

All the logic makes sense, but in your examples, some of the math seems questionable (as suggested in finsta's post)...

Thanks for any changes/clarifications you can offer