join

Latency As Vocal Producing Obstacle Part 1

Brandon Drury —  July 22, 2010

For years I was stuck using mediocre headphones when tracking. I’m such a fan of the Audio Technica ATH-M50 studio monitor headphones that I decided to order 4 more pair so I could let clients use them. I’ve had singers complain that they can’t hear themselves for a while now. I just figured the old Behringer headphones I was making them use was the culprit. (Singers, as a rule, are crazy! Don’t forget that one.) It often required me bumping their vocal a good 5-6dB higher than I thought it should be (I’m wearing the same model headphones with the same mix) and half the time I’m only 4 feet from the singer.

I’ve noticed quite a few singers lately have needed to resort to the one-ear headphone trick. (The one-ear button on my trusty, crusty ol’ Behringer headphone amp certainly comes in handy.)

Then yesterday I was working on a new propaganda video for Killer Home Recording. It required my EXTREME voiceover talents. (Caps indicate sarcasm). Anyhow, at 192 samples, I couldn’t tell what the hell I was saying. Okay, that may be a stretch, but the latency induced delay made it an AWFUL experience. In time, I gradually decided to ignore it. I can’t say I ever got used to it. Switching down to 128 samples was noticeably better, but still “weird”. At times it seemed like I was hearing distortion. …the kind of thing my Distressor does in “British mode”. However, on playback, there was none. It was clear that this is FAR from an ideal vocal monitoring situation!

Conclusion: Singers aren’t bitching entirely because they are crazy. They hate the latency when it mixes with the sound of their real voice. They can’t hear what they are doing! With the voiceover work I was doing, I really didn’t NEED to hear myself. I didn’t care about my pitch or anything like that. In a real deal singing situation, this sound is entirely unacceptable.

Latency Causing Plugins

Another thing that shocked me was how random plugins added dramatically to the latency issue. For example, the stock channel EQ in Cubase 5 added significant latency. This got me wondering if using a brickwall limiter on the 2bus (which I use to save time as a way to keep song to song levels consistent in the headphones) is such a good idea. In fact, I wonder how many plugins I’ve stacked on vocal channels over the years. It’s amazing that any singer managed to come even close to a performance. I worry that I’ve made a lot of really good singers sound bad.

Quick Points Of This Blog

1) Basically, when I used the same latency settings I’ve been given my clients for longer than I care to admit, I HATED it!

2) I wonder if I can improve things with a zero latency mix. Presently, I’m only aware of two interfaces in home recording land that can do a zero latency headphone mix with compression and reverb. One is the Steinberg MR816csx. The other is the MOTU 828mk3. I’m fairly confident I can scrape up a method with the DSP routing matrix in my Presonus Firestudio.

3) You MUST put yourself in the position of the people you are recording on a regular basis, particularly if they are complaining. There may be a problem and it may be your fault!

So that’s all I got for this blog. The plan is to do a few A/B tests and report back to you guys. I’ve got a chick singer coming in who is notoriously crazy about here headphone mix. Maybe going with zero latency will do the trick.

Brandon

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

26 responses to Latency As Vocal Producing Obstacle Part 1

  1. The EMU interfaces like the 1212m do zero latency with compression and reverb. However, EMU’s support sucks and they do not work with Windows 7.

  2. Matthew Tryba July 22, 2010 at 8:47 am

    As a vocalist and an engineer this is one of those situation in which a board is useful. I use my Presonus D8 to split the signal (ADAT to Pro Tools to get the dry signal) and then use the analog out to my Presonus Studio Channel for monitoring compression and eq. I then run into my board and use a hardware reverb post-fader. I monitor Pro Tools through the board as well. Thus, I get ZERO latency in my vocal mix and I can run as large a buffer in Pro Tools as I need. This works really well and I do all of my monitoring in this fasion.

  3. Benson Russell July 22, 2010 at 9:28 am

    Zero Latency mixing is da BOMB for this kind of stuff!! Just make sure that you turn off all monitoring in CuBase except for your 2 bus output. Hence you’ll be doing your monitor mix in the interfaces mixing panel software and no longer through CuBase. There’s a lot of great interfaces out there with great control software. BTW another one that has reverb and compression is the TC Electronics Konnekt 48 (the other unit I was considering along with my Liquid Saffire 56).

    Also you can pickup a cheap external FX box if you don’t want to get a new interface. I picked up a Lexicon MX400 (which is way overkill BTW) for a few hundred. So in the Saffire mix control, I route the mic pre to an additional output feeding the MX400, and then I just blend it back into the monitor mix. Works great for guitars as well if you need to track with a delay but don’t want to keep it. :)

  4. Dave Kauffman July 22, 2010 at 9:31 am

    I use the low-latency mode in logic that helps, but what about this – give the vocalist realtime zero-latency analog feedback, the adjust the latency at mix time by moving the track the few ms it is out by?

  5. Poonna Yospanya July 22, 2010 at 11:18 am

    Another one would be the Focusrite Saffire Pro 24 DSP, which has built-in compressor/eq/reverb for zero-latency monitoring.

  6. Agree, I have the Konnekt 24 (actually two..). the nice thing with the Konnekt24 is that you can choose to record the reverb/compr or not. and then you can record the reverb/compr and the clean signal! and you can balance the outlevel of your DAW and the output of your zero latency soundcard. that’s all I need. warning! it seems that with the Impact Twin (the successor) you can only monitor the reverb, without recording it. that’s why I choose to buy another Konnekt24 (daisy chaining) for when I have to record a live show for example,which I’ve done by now. they don’t make em anymore but you can still find them.

    Any questions, just ask, but keep in mind I’m no allknowing record dude like Brandon. I just try to work comfortably with my own stuff (and sometimes with the Control24 wich is very nice but out of my budget)

  7. I keep hearing of latency issues from a lot of people. A lot of the time it can really come down to what kind of power you computer has to handle its work load. It can sometimes be you setup of course, but i have to say also certain daws are just a pain to deal with. Cubase for instance has never been a favorite of mine, and i too had a lot of latency issues with it. Tho i know a lot of people who use it. I must say that even tho i have protools and know protools is “industry standard” bla bla, but we wont get into that load. acid pro7 is really a good way to go also.. I And i never have any latency issues with it either, even when recording with several plug gins going on vocals in realtime. I always say first hings first ..make sure your recording computer is beefed up..I personally wouldnt not go with anything less than 8 meg of ramm, a quad and fire wire to connect any interfaces etc. its very important when processing effects in real-time and then getting them back to the monitors with zero latency, that your computer can make it happen with ease.When doing a lot of recording its well worth the money. I have seen a lot of really nice home studios with the wrong setup and yes people dropping sample rates to try and compensate..but obviously the quality will suffer. It is almost never the headphones or wiring ..it is almost always your setup, daw, and the power and speed of the workhorse processing the audio. Just my two cents..take it easy.

  8. If you have not tried the Focusrite Isa One mic pre, then this is the gear for all vocal chains. It is worth the money just for the headphone amp in it. It has enough grunt to make your ears bleed and it gives you zero latency real time monitoring. As well as giving you two channels of Isa pres, for the price this is a must for vocal monitoring and recording in home studios for vocal tracking.

  9. yeah a little latency can make a big difference. i’m always impressed by what the human ear can notice in terms of ms or fractions of a ms.

    i use hardware monitoring with my DAW (Samplitude 11 pro/RME Fireface 800). no latency or so little you can’t hear it that’s for sure. i’m not entirely sure but i believe i can do a hybrid monitor…hardware and then track fx. so the processing is maybe 13 ms (depending on buffers) later then the hardware signal. kinda like a predelay on a reverb.

    but i always track my own vocals dry so i can hear everything well. this doesn’t work for everyone i know.

    and i’ve also had most singers ask for so much volume in their headsets it hurts my ears (i even carve out a nice spot on the master eq for them). i have a headphone amp and low resistance cheap-o headsets from radioshack. this set up gets very loud.

  10. Hi Brandon – good point – which is why I dont monitor the people tracking through the box at all (do some tests to ensure sound coming in great and ensure levels arent clipping).
    My audio interfaces (focusrite PRO 40) and SM EP480 (linked via adat) – all can be set to have their 8 channels out as hardware monitors (and hardware controlled), while still feeding the DAW through the PRO40′s firewire.
    So I take the channels I am tracking out into an Alto mixing desk (only because it was cheap, had 16 ins and a few aux and control room etc out, and has on board effects) – and use that to set up monitor/headphone mixes.

    This means zero latency and if need to can feed up to 4 headphones through the aux channels and set up specific mixes for each set and some reverb on the vox etc- still getting nice clean/raw tracks into my system. (also I can have one set of headphones on the desk checking what the singer etc are hearing and another through the interface ensuring the sound in is maintained)

    Anyway just thought I would throw those thoughts into the mix :-)

    Cheers

  11. I always say first hings first ..make sure your recording computer is beefed up..I personally wouldnt not go with anything less than 8 meg of ramm, a quad and fire wire to connect any interfaces etc. its very important when processing effects in real-time and then getting them back to the monitors with zero latency

    While I do agree that a hot rod computer is always a good thing, Macs and PCs can’t process audio with zero latency. If there is a process to be done, it will take time and there is no way around it.

    In terms of the actual plugins, I’ve found the Oxford plugins to have extremely low latency even though they use more CPU power than other plugins with higher latency. It seems this latency business could be considered part of the “quality” of the plugin in terms of real world usability.

    Brandon

  12. If you have not tried the Focusrite Isa One mic pre, then this is the gear for all vocal chains. It is worth the money just for the headphone amp in it. It has enough grunt to make your ears bleed and it gives you zero latency real time monitoring. As well as giving you two channels of Isa pres, for the price this is a must for vocal monitoring and recording in home studios for vocal tracking.

    I saw this in a magazine a while back. I don’t recall it having a method for dealing with reverb in the headphones, but I’ll need to verify that.

  13. Ok, don’t get mad, but did you activate the “constrain delay compensation” button before recording? This bypasses all high latency effects and is essential if you monitor through Cubase.

    You can look up the latency of every plugin in the plugin info window and there are a lot of plugins with zero latency for a headphone mix if too much gets lost.

  14. fHumble fHingaz July 22, 2010 at 8:38 pm

    From the moment I started using a computer to record (as opposed to a purpose-built stand-alone recorder – yeah, I know, they are computers too – but they have no noticeable latency), I couldn’t stand any kind of latency – As soon as I could I got into a zero-latency enabled setup.(I’m using the Steinberg MR816CSX now). When I was dealing with a latency-prone setup (M-Audio Projectmix), I found that the best way to enable recording at the lowest latency was to mix down to a “backing track” so you’ve only got one stereo track running (with no plugins)running alongside whatever else you are recording.

  15. Well, as many have stated, as well as Ruud, you could have used an interface with its own DSP like TC Konnekt Live, Konnekt 48, 24D etc.
    Latency during tracking is not an issue anymore.

  16. I’ve always used a mixing console to completely eliminate latency of any kind. I read posts by people that constantly complain about it and just shake my head and think to myself “I’m so glad I don’t ever have to deal with that in any form EVER!”

    Some might say that a console may alter the sound. If it does, I don’t hear it at all. I use the mixer to send into my soundcards, this way I can kill “direct input monitoring” through my cards. I don’t use any eq on the mixer and most times only have to feed a very slight amount of the pre-amp to get the signal up to my recording level…and I’m talking ever so slightly.

    I’ve done comparisson’s using the board and going directly into the soundcards. I hear 0 difference in sound…the only difference is, I need to alter my latency settings to 3ms using one method, I leave them alone at 100ms for the other. The sound is identical to my ears though, so for me, the console wins as I never have to touch a latency setting or ever worry about it.

    Another great thing about having the console is, I can load up loads of instrumentation in the board and press one button and it goes to disc. I have my entire studio routed through my consoles here. Cd players, DAT tape, cassette deck, multiple guitar pre-amps, bass pre-amps, V Drums, assorted samplers, keyboards, the list goes on and on. If I want to send one of those devices to my DAW, I press an assignment button and bang, it’s sent with 0 latency and no issues at all. Why people deal with this latency thing, is beyond me…but hey, whatever works. :)

  17. MOTU has a cuemix console where you can set up monitor mixes. I use Sonar 8.5 and had trouble with vocal latency when the track was dense. Now I disable the monitoring through Sonar and go straight out of the MOTU. If I need reverb or compression I do it through hardware in my mixer.

  18. Any digital latency and too much space between lips or instrument and mic are fouling up control of both intonation and dynamic by vocalists and players of wind instruments, violin, viola, and mouthdrums. Therefore I split the mic signals via ART S8 and PreSonus DigiMax D8, and using a PreSonus HP60 I feed a pre-digital cue signal to the artist’s headphone (in case of singing myself to mine). For vocalists it is best to position the mic used for cueing close to the forehead. Using an omnidirectional mic will make the overall frequency range more linear, and using a cardioid one may sound a little phasy but provides more fundamentals due to the near effect. I decide according to the artist’s choice; my personal one is using the omni. I am recording this signal and one or two more mic signals on different tracks regularly.

  19. I just upgraded my setup to a Digi002 rack and Pro Tools – as a singer, for the first time ever, I can actually *hear* my voice, and with no audible latency! It’s a lifechanger.

    As for adding a little monitoring sweetness with plugins, I did some research and found a trick to avoid latency (in Pro Tools low latency mode): Add an aux track, with the same input as the vocal track, and then bus that to an additional aux track with the fx inserts you want on that. I then mute my original vocal track as I’m recording, so I don’t hear both of my vocal tracks (slightly misaligned, due to the latency).

    The latency doesn’t occur on the aux tracks, so I can monitor my voice through them pretty perfectly.

  20. Johns recording equipment August 8, 2010 at 9:36 pm

    I bought a lexicon and It has been great for me.

  21. baby clothes wholesale September 6, 2010 at 12:25 am

    Agree, I have the Konnekt 24 (actually two..). the nice thing with the Konnekt24 is that you can choose to record the reverb/compr or not. and then you can record the reverb/compr and the clean signal! and you can balance the outlevel of your DAW and the output of your zero latency soundcard. that’s all I need. warning! it seems that with the Impact Twin (the successor) you can only monitor the reverb, without recording it. that’s why I choose to buy another Konnekt24 (daisy chaining) for when I have to record a live show for example,which I’ve done by now. they don’t make em anymore but you can still find them.

    Any questions, just ask, but keep in mind I’m no allknowing record dude like Brandon. I just try to work comfortably with my own stuff (and sometimes with the Control24 wich is very nice but out of my budget)

    http://www.danikids.com

  22. All the more reason to use a firewire attached multi channel mixer like the Phonic or it’s new clone the Mackie.

    0 latency… the phonic let’s you put effects on the vocals for monitoring only and put clean into the computer.

  23. Tshepo (means Hope) October 3, 2010 at 12:29 am

    Hi!Guys,

    Firstly, I respect all of you for your selfless suggestion on this website.

    I am new to home recording…I mean very new!

    My set up is a basically, Cubase 5, an ESI 1808 audio ineterface, a pair 5″ inch monitors, and a dynamic mic…

    NOTE: My monitor speakers are plugged on to the 7 (left) and 8 (right) jack holes on the interface (also, displayed at the back of the audio interface)

    Question…

    In my baby-stage thinking, On the VST connection tab I have put “OUT 1″ for left and “OUT 2″ for right.

    And on the “Studio Tab”, on the “Monitor 1″ bus I have put “OUT 7″ for left monitor and “OUT 8″ for right monitor…

    QUESTION…

    1. Why is it that when I change the “Monitor 1″ bus OUTputs to “OUT 3″ and “OUT 6″ for left and right respectively, which have nothing to do with where plugged my jacks for speakers (speakers are plugged on OUT 7 and 8 on the audio interface), the VOLUME of the monitors is increased drastically?

    Thanks you in advance for your input.

    Regards,
    Tshepo, South Africa (means Hope)

  24. This routing stuff gets a bit tricky in Cubase 5. Their added control room feature can be handy, but it can be a giant ass pain.

    In your case, if you are getting a significant volume boost from using different outputs, this leans in the direction of the control panel for your audio interface needing some tweaking. That’s where I’d look first.

    Brandon