Recording Live Band: Latency and Bleed Questions

Help Help

Member
Hello! I hope everyone reading this is safe and well. I have two fairly simple questions in regards to live band recording.

1) Are there any issues with latency when recording a live band? I've been putting 'highest latency' on when we record. Any overdubs would be done on 'lowest latency'. I've asked this before but have gotten conflicting info in the meantime, and want to be sure. Is there latency between tracks being recorded simultaneously? I'm afraid of dropouts if low latency is on. For any overdubs its no big deal as not much is lost if there's a dropout, and so I do 'lowest latency' etc.

2) If I need to redo something like guitar, is there any issue with bleed I should be aware of in regards to the previous guitar? I used to think bleed was fine, even a good thing, but early on in our sessions I noticed it can become a problem in the overhead mics, diluting the signal and limiting options. I did a bunch of stuff to tame it considerably, but it still comes through a bit on the overheads. People double guitars anyway, and I got the bleed low enough that its no problem in the take itself. But if I need to redo the guitar, is there something I should know or expect? Its pretty inconvenient to get the band to retrack everything with things as they are.

Probably silly questions, but any time taken to respond would be greatly appreciated! Thanks so much!
 
I've done a fair bit of recording live bands. I don't make any latency adjustments, nor have I experienced lags nor dropouts. This is using firewire interface with eight inputs (and in fact daisy-chained to another for 16 inputs).

If you need to redo a guitar, with luck the new guitar should be close enough to the old guitar for bleed not to matter. On the other hand, it may be an issue of the new guitar track is going to be substantially different to the old.

What would make your life easier is that if you know you are going to record live, to arrange the players to maximise the distance between them and minimise bleed. For example place OH mikes in front of the kit and pointed towards the kit. Turn the guitar amp to face away from the kit, likewise the bass. And so on.
 
Thanks so much for the reply!

Yeah this is basically what I figured. If it was a solo or something I could see it becoming a problem. But A/B-ing a new guitar track I did with the old, I don't notice any issues as I play it basically the same. I had maximized the distance in the basement we are recording in, but we found it hard to play together that way, the basement is unfinished with open insulation, captures the sound too well. Was hard to stay in time. So we are playing in the same space with drums and reasonable distance between amps, which has worked out for the most part. I draped a series of blankets in front of amps which has helped a lot for bleed without affecting the sound.

I don't understand why I have dropouts, I've returned a few laptops looking for one without issues that way. I have a Thinkpad P52 so it should be powerful enough. Perhaps its my Tascam US-1800? Regardless, this is what I have to work with as things stand. No problems with dropouts on high latency, only on low latency, so I was hoping it made no difference for live tracking as everything's done at once. Does anyone know if high latency can cause problems in a live band scenario(not overdubbing)?
 
I always set the buffer size as large as I can when recording a live band - just to be on the safe side. If you have a decent control panel app with zero latency monitoring for your interface (or a hardware mixer) then you can also use large buffers when overdubbing and monitor the input through the control panel (or mixer) rather than through your DAW.
 
When recording live, you don't have to worry about latency. Set the buffer size to maximum and jam on.

Latency comes into play when you are listening to a track in the DAW program as you record it. As an example, say you want to record your guitar direct and you want to use a guitar amp sim in your DAW. You plug your guitar directly into your US-1800 without micing an amp or using any pedals. You want to hear the guitar as it sounds with the amp sim, so you monitor the guitar track in the DAW as you record. The amp sim plug-in is cpu intensive and uses a lot of resources, so it has a lot of delay. When you hit a note, it takes half a second for the cpu to process the sound through the amp sim and send it back to the interface and into your headphones. That delay you hear is latency and it is annoying.

For a live band, you do not have the need to listen to a track in the DAW program while recording it. You start the recording, jam out your song, stop the recording, then listen to it.

As a matter of fact, there is no need to use "high latency" or "low latency" for overdubs or anything. If you're going to record a 2nd guitar track to fix mistakes you'll just start in the right spot and play while listening to the other tracks. Again, latency is not an issue here because you're going to hear your both guitar amp directly from the mic and the playback tracks from the DAW, in the headphones.

Hard to explain sometimes, hope this makes sense.
 
Thanks a lot for the responses guys! I feel reassured about latency. I didn't know latency wasn't an issue for overdubs, but it makes sense as you explain it I think. I'm not sure where latency would come in to play for my uses in that case, as I always record while listening to what I am doing live in the room, headphones over one ear. If there is no delay in the playback then maybe it doesn't effect me. For live band stuff we don't even use headphones, just try to play close to the drums while the amps are a bit removed but where we can all hear each other. I'm always a little confused about latency really.

Thanks again everyone!
 
I've recorded live from my X32 and now the M32 for a long time - with the cable plugged into my MacBook, and cubase on the screen, I just push record. Never had a dropout or issue. Never changed a buffer size from the default. USB3 is fine, and firewire before it. 32 channels is bomb proof. It's never even been a problem with a Panasonic tough book running windows 10, although I must admit I've probably never recorded more than maybe 18 tracks.

As for replacing tracks or overdubbing - I find it's often simpler top replace a track entirely than attempt to record in the studio a short piece for a traditional overdub because spill is a big problem. The drums and electric guitar leak into everything, so even close miked, the guitar amp with a 57 hears the drums, and the drum overhead hears everything. I've been able to salvage a few little mistakes using spill from unused tracks - often the mic we use to add into the in ear mix - this picks up everything and can be blended with the studio clean sound to 'cheat'. Sometimes it works, other times, it's the whole track again, I'm afraid.
 
Great suggestion, Rob for dealing with bleed.

Last fall I recorded a live band for the first time. It was my cover band at a gig where we had a video crew tape us. I used the Behringer XR18 into my laptop and Cubase. A 4 hour gig, 14 channels through USB2. Not a single problem, it came out great. I didn't retrack anything or do any overdubs. There were some mistakes in performance, but I could do some subtle editing to hide them. (Except for the ending on one song where the singer completely misses the pitch.... Oh well :rolleyes: )
 
When I recorded our band, I made quite a few mistakes in my bass part playing live - before the band heard the recording, all were fixed and nobody knew! DI'd though - so no spill issues and I don't use effects, so just sat in the studio with my bass and it was simple. The guitar was a devil to fix - not just the spill, but the damn things hums, buzzes and splutters.
 
I had recorded a live jam session a couple year back, and playing bass on a song that I didn't know, totally screwed the pooch on that song. I did the same thing, as you. I turned off the bass track, and plugged in after I learned the song... presto. It actually sounded a little better than the mic'ed bass.

Two years ago, I did Mary Jane's Last Dance, and went completely blank on the words. Since the vocals were fed to one channel, I just redid them, along with harmony vocals. Worked out just fine. You really couldn't hear any vocals in the instrument mics anyway.
 
My suggestions:

DI the bass.

Tent the amps.

Amps put out more sound to the front than to the sides. Open back amps put out almost as much to the back as to the front. Put the amps where they're not projecting toward the drums, also considering any reflections.

Make the amps about the same volume as the drums. That way putting mics closer to their intended sources will optimally minimize bleed. If one instrument is quieter than the others it will be the one most prone to bleed.

Put the mics close to their sources. Use their patterns to optimize isolation. Point the null of the mic toward the most likely source of bleed.

Consider using the overheads as cymbal mics rather than overall kit mics. It's a tradeoff against a more natural sound so you have to decide for yourself if it's worth it.

You might have the singer face the other band members and just mouth the words, or sing just the first word of each line, or just quietly sing.

A really dead room will be a lot better for isolation. A live room will put everything everywhere.
 
I record live bands and I use the Tascam US-1800 and the earlier US-1641. Since you don't have a good zero-latency monitoring option, you have limitations. I keep mine set on the highest latency in the Tascam control panel. I also verify this through my DAW, Reaper.

The last few times I recorded, it was drums, guitar, and vocal. I tried to get 2-3 good takes as a basis. If I had a horrendous take, I redid it. I high passed my overhead mics and just used them as cymbal mics. I faced the null of ALL my drum mics toward the vocal/guitarist. I put a gobo in front of the guitar amp. I turned up the floor wedge and the guitar amp JUST loud enough to be heard by the musicians. I ended up with really good separation. Even in the dynamic vocal mic. Not perfect, but good.

By getting multiple takes, I had some options. For overdubs, I've just been playing back through the studio monitors and haven't been "monitoring" the new part, so I'm still keeping my settings on highest latency.

I have set up a hardware based monitoring solution for my space. I split my mic inputs and have a live sound board that I can use to build a monitor mix. It really hasn't been necessary most of the time because everyone usually feels more comfortable without headphones. I also have a couple of those cheap Behringer boxes that allows "more me". Those are actually pretty handy. I also have a handful of cheap dynamic mics if I need to get a monitor feed that doesn't get recorded....like a drum "heart" mic to blast into the drummers ears.

EDIT: Added video link. You may get a better idea of my setup and how it works while recording.
YouTube
 
Last edited:
Back
Top