The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
how much delay should you add to dry tracks.. with respect to wet tracks?? Virtual Instrument Plugins
Old 30th March 2006
  #1
Gear Maniac
 

how much delay should you add to dry tracks.. with respect to wet tracks??

Hi all ..

this is an important question that I am trying to find an answer to ..

lets say you simply have one track with a dry vocal recording ...
and a second one with the same vocal recording but with added reverb using an external effect box..

there will be a delay for the signal to go to and from the effect box and then to your computer .. while the dry track goes straight to your computer...

I can see the wave in both tracks .. and I can find the exact delay (lets say in samples)


the question is:
for the correct timing between the two tracks .. should I compensate the 2 way trip of the signal to the effect box .. by delaying the dry signal by the same amount ??? is that what I should be doing??
(I use 2 busses .. one for dry signal .. which I will insert a delay and a second bus which will route the wet signal)

any comments?..

thanks
John
Old 30th March 2006
  #2
Lives for gear
 

Not really a high end question - expect this to get moved.

When using an external reverb, sure there will be some latency. Normally, you would use the reverb 100% wet, and this latency becomes part of your pre-delay. Since you normally need some pre-delay on reverb, it's not a real problem - just adjust any further pre-delay according to taste.

For other effects, like chorus maybe, you probably need to nudge it back otherwise it will be a rather different effect than intended.

The original dry track will be synced to the backing track (hopefully), so you wouldn't be moving that. You would move the wet track backwards. (Or is that forwards - depends on your choice of terms). Moving it back it time - negative delay.

Some DAWs, like Cubase SX, have a hardware latency compensation option you can use. The actual latency depends on your soundcard and drivers. A lot of people set latency extremely small, and put up with poor performance. Unless you are tracking virtual instruments, you don't need such small latencies. You can always set your buffers larger (greater latency) for more performance at mix time.

So in answer to your question "how much delay" - latency is variable - your choice.

Since you will need 'negative delay' - if you can't shift your whole wet track back by the right number of samples, you can always use a delay plugin. Shift the whole track back by a beat or a bar, if you have to, so you can use 'positive delay' to place the track where it should be. But most hosts let you shift tracks by samples, so you shouldn't use a delay if possible.
Old 30th March 2006
  #3
Gear Maniac
 

Quote:
Originally Posted by Kiwiburger
Not really a high end question - expect this to get moved.

When using an external reverb, sure there will be some latency. Normally, you would use the reverb 100% wet, and this latency becomes part of your pre-delay. Since you normally need some pre-delay on reverb, it's not a real problem - just adjust any further pre-delay according to taste.

For other effects, like chorus maybe, you probably need to nudge it back otherwise it will be a rather different effect than intended.

The original dry track will be synced to the backing track (hopefully), so you wouldn't be moving that. You would move the wet track backwards. (Or is that forwards - depends on your choice of terms). Moving it back it time - negative delay.

Some DAWs, like Cubase SX, have a hardware latency compensation option you can use. The actual latency depends on your soundcard and drivers. A lot of people set latency extremely small, and put up with poor performance. Unless you are tracking virtual instruments, you don't need such small latencies. You can always set your buffers larger (greater latency) for more performance at mix time.

So in answer to your question "how much delay" - latency is variable - your choice.

Since you will need 'negative delay' - if you can't shift your whole wet track back by the right number of samples, you can always use a delay plugin. Shift the whole track back by a beat or a bar, if you have to, so you can use 'positive delay' to place the track where it should be. But most hosts let you shift tracks by samples, so you shouldn't use a delay if possible.
you are right .. I was not clear ...
let me try again..

you have a dry track that you recorded earlier


consider this with NO delay compensation :

time:- - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
dry track starting point: D- - - - - - - - -
wet track starting point: - - - W - - - - - -


is The following the OPTIMUM starting point of BOTH tracks ..

time:- - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
dry track starting point: D- - - - - - - - -
wet track starting point: W - - - - - -

in such case .. BOTH tracks start (the signal magintude will start to increase)
at the EXACT point in time .. i.e you have shifted the wet track back in time ..

do you think this is correct?????

otherwise .. where in time should the wet track start???
Old 30th March 2006
  #4
Lives for gear
 

I have no idea why you can't understand my original reply.

The dry original track should not move at all. Especially if it part of a multitrack recording. If there are no other tracks to consider phase relationships with, then I guess it doesn't matter. But why would you want to move it at all?

The Wet track is your psychoacousticly modified track. There is no particular right or wrong amount of delay - it's whatever sounds good. Reverb should usually have some pre-delay. If you want to fake a 'natural' acoustic space, then your wet sound should occur after the dry sound. But reverse reverbs are cool too - there are no rules when it comes to effects processing.

But if your effect includes any dry sound at all - or if it's an insert type effect - or if you are doing parallel processing, the phase relationship with the original dry track needs to be preserved. That means you have to nudge it back in time, to compensate for the latency of your i/o and the processor itself.
Old 31st March 2006
  #5
Gear Maniac
 

Quote:
Originally Posted by Kiwiburger
...

But if your effect includes any dry sound at all - or if it's an insert type effect - or if you are doing parallel processing, the phase relationship with the original dry track needs to be preserved. That means you have to nudge it back in time, to compensate for the latency of your i/o and the processor itself.
hi ... you hit the right point ...
however in my example I did not move the dry signal .. I only shifted the wet signal back in time .. such that BOTH the dry and wet tracks will start at the same instant in time ... I was not sure if that is what supposed to happen at best conditions .. i.e assuming that there is NO latency ..

so basically I nudged the wet track back in time .. which is what I need to do ..

but then by how much??.. that was the question .. and what is the optimum amount that you need to nudge back the wet track .. to make both tracks sound as best as it can
Old 31st March 2006
  #6
Lives for gear
 

That's going to depend on the combined latencies of your soundcard and the effects processor, and any inbetween converters or DSP.

Here is how I would set it up:

Create a "dry effect" in your external reverb - purely 100% dry sound straight through - no eq, no boost or cut.

Then, process a sample of audio and re-record it. This processed 'dry' sound will be delayed from the original dry track, and some flanging should be heard.

Flip the phase of the processed track, and now start shifting it back in time. Get to the point where the two signals null. (Make sure they are peaking at identical levels, or they might not null). When you've found the spot where they null as well as they possibly can, you've found your perfect sample offset.

Bear in mind that anything that changes your latency will require you to do this test again.

You could also try lining the two tracks up visually, but I like this phase nulling trick best. It will be more accurate than trusting published latency figures.
Old 31st March 2006
  #7
Gear Maniac
 
aevan's Avatar
 

Umm...depending on the program you use it may have a way to check the latency - like cubase sx has a "check user delay" option. If not, just set the reverb time to zero second, record a click out through the reverb and back in then zero in and see how much you have to nudge it to get it in time. Keep in mind, most reverb is phase related anyway, so it might not really matter. Latency is only an issue if you have the dry signal coming back through the reverb as well as on the track, but as mentioned if you're using it on a send you should have the reverb return set to wet only anyway.
Old 31st March 2006
  #8
Gear Maniac
 

Quote:
Originally Posted by Kiwiburger
That's going to depend on the combined latencies of your soundcard and the effects processor, and any inbetween converters or DSP.

Here is how I would set it up:

Create a "dry effect" in your external reverb - purely 100% dry sound straight through - no eq, no boost or cut.

Then, process a sample of audio and re-record it. This processed 'dry' sound will be delayed from the original dry track, and some flanging should be heard.

Flip the phase of the processed track, and now start shifting it back in time. Get to the point where the two signals null. (Make sure they are peaking at identical levels, or they might not null). When you've found the spot where they null as well as they possibly can, you've found your perfect sample offset.

Bear in mind that anything that changes your latency will require you to do this test again.

You could also try lining the two tracks up visually, but I like this phase nulling trick best. It will be more accurate than trusting published latency figures.
nice trick .. I guess it would help more if you use some test signals like square waves.

do you record the wet track of each track separately then sum all the wet and dry tracks for the final wave file?.. ..
or its just as good to record 2 wet tracks from the output of the effect box .. and it will contain the wet signal of all the send tracks.. and sum these 2 tracks with all the dry tracks?..
Old 31st March 2006
  #9
Lives for gear
 

The audio signal that you use for that null test doesn't really matter - just whatever you have on hand.

With a multitrack DAW lack of tracks is hardly a problem, so why not just keep your effect return track right up to your final mix down.

Or if you are using analog summing - don't record it all (assuming you are happy with the pre-delay without correction). Or make a delay track just to feed the reverb, so the latency can be compensated for on playback. Cubase SX3 has a latency compensation feature to accomodate outboard effects.

The big reason to record your reverb 100% wet is that you can maxmise your use of your available 24 bits to maintain signal integrity.

The level of reverb usually required can be fairly low, so if you record it along with your dry signal at the level actually required, the reverb can be basically very 'under-recorded' and will lose some detail. Compression or boosting at a later date might expose the graininess due to the low effective number of bits used. If you have a good strong reverb track, you can always attenuate it.
Top Mentioned Products
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
cc1 / So Much Gear, So Little Time
52
Peakly / So Much Gear, So Little Time
1
Mark / So Much Gear, So Little Time
2
ttauri / So Much Gear, So Little Time
21
makoto / So Much Gear, So Little Time
128

Forum Jump
Forum Jump