join

Latency As Vocal Producing Obstacle Part 2

Brandon Drury —  November 22, 2010

This is a followup to Latency As Vocal Producing Obstacle Part 1

Back in July, I came up with this hypothesis that latency was TOTALLY screwing with my vocalists after doing a few little voiceover things myself. I felt that the ability of the singer to hear themselves was all screwed up. The delay from their own voice in their head and the headphones causes a distinct comb filtering that makes everything sound totally NOT distinct.

Well, I’m happy to report that after testing this over and over that zero-latency headphone mixes are the ONLY way to go. I will never go back to recording vocals with any noticeable latency. Singer comfort seems to increase dramatically, complaints about headphones have dropped dramatically, the results seem to be better, and good singers who struggled with pitch seem to make miraculous improvement. (Crappy singers are still crappy, unfortunately.)

There are still some singers who aren’t in love with using headphones as this can is not necessarily a person’s natural way of hearing themselves. I’ve found the singers that HATED singing with 96 samples of latency in their headphones now seem about 10x happier

Getting FX In Zero-Latency Land

DAW Monitoring
Many DAWS will allow you to record vocals via direct monitoring and then route that signal to the reverb. In this case,the latency will delay that reverb signal. This is the exact same thing as pre-delay, which is a very common reverb control anyway. In fact, for big vocal reverbs (Celine Dion-type stuff, it’s required anyway) so a little more won’t hurt.

Hardware
I bit the bullet and paid $220 for a Kurzweil Rumour hardware reverb. This requires the ability to route signals out of your interface or it requires a hardware mixer. In either case, you need the ability to send signals and bring them back in on a new channel that can be heard. Users with 2-channel audio interfaces are out of luck.

Conclusion

Find a way to record vocals with zero latency and never look back. It works.

Brandon Drury

Posts Twitter Facebook Google+

Brandon Drury quit counting at 1,200 recorded songs in his busy home recording studio. He is the creator of RecordingReview.com and is the author of the Killer Home Recording series.
join

6 responses to Latency As Vocal Producing Obstacle Part 2

  1. Brandon, you are correct. I just shut off any power hungry plugins not necessary to the vocal recording process. (Cubase 5.5.2)
    I set the latency to a buffer rate of 64, enable a built in reverb from cubase 9if my vocalist insists on a wet monitoring) and have at at.

  2. Brandon,
    as a singer, it’s a no-brainer. You absolutely need zero latency otherwise pitch will be flat and timing will be variably delayed. The outboard reverb is really the right way to go. But another approach that I use is to send the vocal signal into another channel and only get wet signal through the headphones (with the dry out of the mix, there’s no worries about its latency). hope that helps

  3. I know that some interfaces (like the metric halo stuff) with built in DSP and signal routing allow for ZLM + effects. I’ve been looking into picking up a MH 2882 even though they are not cheap :)

  4. So, the solution here is all about signal routing? This has been what has turned me off with computer recording…..the latency. So with this said, I know I want an interface with the ability to record vocals, acoustic instruments, or harmony tracks…..without the latency.

    Brandon, you love the RME products, but these are way over my little home studio budget. Any suggestions for the guy with the under $3-400 dollar budget? Or do I just keep saving?

    Thanks Gents!

  5. So, the solution here is all about signal routing? This has been what has turned me off with computer recording…..the latency. So with this said, I know I want an interface with the ability to record vocals, acoustic instruments, or harmony tracks…..without the latency.

    The latency is only an issue with vocals. That is because the sound exactly resonates out of the singers head. The delay from the throat to the ear is nada. So any signal that is just a hair behind is going to induce comb filtering…..which basically makes the whole thing weird and chorusy in a bad way.

    On the other hand, I can play guitar all day long at 128 samples of latency not flinch. That equates to 3ms, which is the same thing as sitting 3 feet from an amp. The same goes with pretty much every instrument I can think of.

    Brandon, you love the RME products, but these are way over my little home studio budget. Any suggestions for the guy with the under $3-400 dollar budget? Or do I just keep saving?

    Dave, for what it’s worth my RME card was under $400, but it’s a stripped down model. The interface is a very specific thing for each person. Take a look at the Soundcard Wizard http://www.recordingreview.com/soundcard/soundcard_wizard.php and then after you’ve narrowed down, make a post in Home Recording Equipment. We’ll ehlp you out.

  6. I use an M-Audio profire 2626 for all my recording. Vocals get tracked through a hardware channel strip with gate, comp, eq, etc. I use the 2626′s routing matrix to send this vocal signal (zero latency, zero additional processing) to the same output as the rest of the recorded tracks and the reverb from my DAW. Since I record mostly metal and I have to compress the crap out of vocals anyway, this chain works fine for me. I never have any problems with latency. I get an added bonus of having drastically less processing to do with the vocals ITB since they’re basically ready to mix once recording is done.