Jump to content

Using delays in mixes


Recommended Posts

  • Members

I've read that a lot of well known mix engineers use barely audible short delays (16th-note and shorter) in their mixes as an ambience effect. One example that I've heard of is that Bob Clearmountain sets up sends for different reverbs and then a send with a 16th-note delay and another one for 32nd note delay.

 

So far I've only used reverb in my mixes and I'm very interested in how this works and on which instruments/sources it might work well. If the "big guys" are doing it there must be something in it. Can anyone offer some advice on how this works?

Link to comment
Share on other sites

  • Moderators

Delays are the way and the light. A lot of vocals that sound bone dry really have several very short delays. 20ms, 40ms, etc. Then on top of that you might run the 32, 16 or 8th note stuff. Being timed to the track like that, they tend to disappear, in a good way. Now... send that timed delay to a plate reverb, keep the plate fairly short if you're trying to go covert. .8 seconds or less. Mix it so you don't here any of this... just below the threshold of awareness (hey good album title).

 

Now turn that voice up. It sounds dry and in your face and yet it has a polish to it. S'cool.

Link to comment
Share on other sites

Delays are essential for me, and I tend to use them as much as, if not moreso than I use reverb. Reverb is fine, but it can muddy things up if applied too liberally and / or to too many things in the mix, although some HPF on the reverb return can definitely help with keeping the bottom end more clear...

 

As far as delay times, I like to time them to the track's tempo, and the formula for that is pretty simple: 60,000 / tempo (in BPM) = 1/4 note delay. If the song is at 120 BPM, that means a 1/4 note is 500ms long. Divide that down for shorter times - in other words, if you want a 1/8 note delay, divide that number in half, etc.

Link to comment
Share on other sites

  • Members

 

Delays are essential for me, and I tend to use them as much as, if not moreso than I use reverb. Reverb is fine, but it can muddy things up if applied too liberally and / or to too many things in the mix, although some HPF on the reverb return can definitely help with keeping the bottom end more clear...


As far as delay times, I like to time them to the track's tempo, and the formula for that is pretty simple: 60,000 / tempo (in BPM) = 1/4 note delay. If the song is at 120 BPM, that means a 1/4 note is 500ms long. Divide that down for shorter times - in other words, if you want a 1/8 note delay, divide that number in half, etc.

 

 

yea. i use at least 4 delays on any typical mix. most are 1/8, 1/4 , 1/2 notes. you gotta watch the regen though, much like decay can muddy things up on verb. one repeat for me usually on vocals. more on solo instruments like sax and guitar.

 

i usually set them up by soloing the track and delay return and cranking it up, setting the delay time by feel...then back down to just where you can hear it..then a lil' extra more. they're like spices in a good spag sauce, you don't add them until you can taste them, you trust they're in there:)

Link to comment
Share on other sites

  • Members

I like using stereo delays to carve out more of a presence within a stereo field, especially on center panned mono sources like bass or vocals: original panned center, 4ms delay left, 8ms right.

 

Want something to sound bigger, reach for a delay!

Link to comment
Share on other sites

  • Moderators

Regarding regeneration...

 

I like to set my main delay, let's say the timed 8th note delay, I like to set that up flat, no feedback on the delay plugin itself. THEN... use the aux channel's send "to send itself to itself" (is that the name of a George Harrison song?). So you send it out to the very same aux track, back to itself. Now, put an eq before the delay. See, it's not really going to be flat, you're just not going to use the LPF on the delay itself. Now with the eq, everytime the delay 'regenerates' back into itself via the aux send, it goes through that eq as well. So get nasty. Roll of the lows as high as 350Hz. Roll off the highs as low as 3k. Boost a narrow band at 2k...

 

Each time that signal gets regenerated it degrades into a messy funk. NOW... automate the regenerating fader. Bring it up to feedback then back down to nil. When the music peaks, get it feeding back in time, riding with the drums, distorting and adding an almost pad like sound. But you can still keep it undetected. If you mess up, just edit you automation data. Mute the aux track itself at stops so there is absolutely no delay at all... when you want that. Automate the send mute to only effect certain words if that's your cup of tea.

 

Lot's of silly stuff you can do. And BTW, don't limit yourself to just eq before or after the delay on that aux track. Try flange, chorus, a tape emulation, an early refection portion of a reverb algo in mono.

 

Go crazy baby!

Link to comment
Share on other sites

  • Members

not a bad Idea to eq and even de-ess those delay returns on vocals ! keep it from getting spitty.

 

If more people would compress ( to make it "stick" to the source better) and eq and , when necessary , duck the returns on reverbs , then things would'nt always be so muddy.

Link to comment
Share on other sites

Also, don't be afraid to track with delays. If you have a cool delay pedal, don't be afraid to use it while tracking - especially if the delay is an integral part of the "sound", like with a rockabilly guitar part.

 

If you have an analog tape recorder with three heads and a varispeed (even a cassette deck with these features will do), the delay / modulated delay fun really begins. ;)

Link to comment
Share on other sites

  • Members

Awesome! Thanks for all the advice guys. Much appreciated. I've got a couple of gigs on at the moment, but I can't wait to get back in my studio next week to try them out.

 

I agree with Phil: Keep those tips coming.

Link to comment
Share on other sites

  • Moderators

Here's a fun delay tip.

 

Copy your lead vocal track to another track. Then delay (slide back I mean) that track by a musical note value. 1/8, 1/4, whatever. Now... go about mangling certain phrases in unimaginable ways destructively. Meaning just go ahead and write the data. Commit. You mute the delay track when you want. You soak in 6 second verb on a phrase, you distort and slide the delay back in time to double for a phrase, you cut and paste a word but in a cool syncopated way that no delay plug could come up with.

 

In other words, make that copied vocal track a playground to spice up the meaning of the song and it's words or just to provide ear candy.

 

Fun stuff.

Link to comment
Share on other sites

  • Moderators

For the regeneration routing scheme I mentioned earlier, a fun thing to do is ride the send up gradually during a build in the song. The feedbacked delay is getting lo-fi'ed to death. Distorting, bouncing in time, squawking, and it's getting louder, almost too loud, right at the climax and... mute the aux.

 

STOP. Silence.

 

This works well where the band's climax results in a nice muted sting with choked cymbals etc.

Link to comment
Share on other sites

  • Moderators

Or...

 

I remember back in the late 70's, 79 maybe, an engineer using a delay (Delta Lab Prime Time?) on snare. I really hadn't thought about it for all these years. But the other day it came back to me... he was putting a timed delay, timed by ear mind you, on the snare. ???

 

But I thought about it and remembered how cool it was. I remember as he set it up it was totally hokey. SMACK smack smack... We were a little worried.

 

Then he eq'ed the return and turned it down (that's key, turning down) and sent it to a plate. The song just grooved like a mother. So I tried it. I mean I've done all this with more techno type stuff since, but not on a more standard type rock or pop track. But I did and it just swings.

 

Just mute the return in the breaks so you never actually hear it.

Link to comment
Share on other sites

  • Members

Here's one of my quick delay tips:

 

I like to run my send to the delay pre-compression. If you have an aux that is pre-insert that's one way to do it, or assign to a buss and compress the buss...

 

 

What that does for me, is that when something get's big it sends a little more to the delay, be cause the dry sound is getting compressed...it works like an automatic accent...

Link to comment
Share on other sites

  • Moderators

 

Here's one of my quick delay tips:


I like to run my send to the delay pre-compression.
If you have an aux that is pre-insert that's one way to do it, or assign to a buss and compress the buss...



What that does for me, is that when something get's big it sends a little more to the delay, be cause the dry sound is getting compressed...it works like an automatic accent...

 

 

Nice... never tried that but I will shortly.

Link to comment
Share on other sites

  • Members

Yeah, set your delay so it's just barely under the threshold...then the accents will pop anytime the singer leans into a line...

 

Granted this doesn't necessarily work if your singer is all over the place. In that case I'll compress a touch before the delay send to get things under control, and then send to the delay, and then do a little more squeezin' on the buss...

 

I can never have enough compressors in the rack...luckily nowadays I don't need to...it's all DSP...:thu:

Link to comment
Share on other sites

  • Members

TC units have a feature called "dynamic delay".

Basically the level of the dealy returned to the mix depends on the level of the dry signal its mixed back in with.

 

So, if A vocalist is holding a long note the delay is ducked down to a certain level. If the vocals have a sharp loud note followed immediately by silence then the delay that follows come in at a higher level.

 

This automagially raises the level of the delay to fill in open spots in the mix. You can hear this on many commercial releases.

 

You could also do this manually by automating the delay return. However, the automatic application is most helpful if you have many such fader moves to make.

 

I suppose you could also configure a sidechain compressor duck of the delay to accomplish this even without the TC implementation.

Link to comment
Share on other sites

  • Moderators

 

TC units have a feature called "dynamic delay".

Basically the level of the dealy returned to the mix depends on the level of the dry signal its mixed back in with.


So, if A vocalist is holding a long note the delay is ducked down to a certain level. If the vocals have a sharp loud note followed immediately by silence then the delay that follows come in at a higher level.


This automagially raises the level of the delay to fill in open spots in the mix. You can hear this on many commercial releases.


You could also do this manually by automating the delay return. However, the automatic application is most helpful if you have many such fader moves to make.


I suppose you could also configure a sidechain compressor duck of the delay to accomplish this even without the TC implementation.

 

 

This is interesting to me... I actively try to do just the opposite. I like to turn the delay up on sustained notes and turn it down and staccato stuff or anything that brings attention to it. So the sustained stuff sounds godly without giving the trick away...

Link to comment
Share on other sites

  • Members

 

This is interesting to me... I actively try to do just the opposite. I like to turn the delay up on sustained notes and turn it down and staccato stuff or anything that brings attention to it. So the sustained stuff sounds godly without giving the trick away...

 

 

To claify: the higher level of the delay is applied when there is open space after a sound.

This is independant of whether the note preceeding the space is a long note or a stacatto note.

 

Cheers,

Link to comment
Share on other sites

  • Members

 

Regarding regeneration...


I like to set my main delay, let's say the timed 8th note delay, I like to set that up flat, no feedback on the delay plugin itself. THEN... use the aux channel's send "to send itself to itself" (is that the name of a George Harrison song?). So you send it out to the very same aux track, back to itself. Now, put an eq before the delay. See, it's not really going to be flat, you're just not going to use the LPF on the delay itself. Now with the eq, everytime the delay 'regenerates' back into itself via the aux send, it goes through that eq as well. So get nasty. Roll of the lows as high as 350Hz. Roll off the highs as low as 3k. Boost a narrow band at 2k...


Each time that signal gets regenerated it degrades into a messy funk. NOW... automate the regenerating fader. Bring it up to feedback then back down to nil. When the music peaks, get it feeding back in time, riding with the drums, distorting and adding an almost pad like sound. But you can still keep it undetected. If you mess up, just edit you automation data. Mute the aux track itself at stops so there is absolutely no delay at all... when you want that. Automate the send mute to only effect certain words if that's your cup of tea.


Lot's of silly stuff you can do. And BTW, don't limit yourself to just eq before or after the delay on that aux track. Try flange, chorus, a tape emulation, an early refection portion of a reverb algo in mono.


Go crazy baby!

 

 

Would you mind explaining how you route this up in your mixer with some greater detail Please?. when it gets sent via aux back unto itself--how does the routing happen and go from there? I am also a little bit confused when you mention "flat"--it may be misleading me in a sense, because i tend to think that you may be referring to flat eq here, but on the other hand, i don't ever think of the delay unit as a "flat" processor---actually I think of a delay unit as a "sparkle" or "shimmer" and never ever as "flat"......

 

would you also please explain in a little more detail how you coordinate or automate the "nasty" to be in time with say the drums? (or a single drum). Very interesting post, and I appreciate your sharing these neat tricks!! Thank you Lee!

Link to comment
Share on other sites

  • Members

 

Delays are essential for me, and I tend to use them as much as, if not moreso than I use reverb. Reverb is fine, but it can muddy things up if applied too liberally and / or to too many things in the mix, although some HPF on the reverb return can definitely help with keeping the bottom end more clear...


.

 

 

 

Share some numbers with me Please ! Where do you like to start rolling? what frequency(s)?

 

And What thought process are you going through and what factors are affecting your decisions here? Just trying to learn how you pro's think ---thats all.....Are there emotional considerations with where you start rolling for example?

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...