limiting while recording live...

[keVAn]

Is this frowned upon?

I was always under the impression that engineers dont usually limit when recording in a studio because they want to keep all the dynamics. But in a live application, it is hard to predict the peaks and once it happens there is no "overdub". To prevent the ugly sound of digital distortion, do any of you limit when recording a live gig? (taking signal from the FOH)

Thanks guys,
Kevin

[joelpatterson]

Sure. You better have some kind of compression scheme afoot when you record live. As you say, it's unpredictable when peaks will happen, but you know they will. You can kind of "pre-mix" things if you have a chance to play with compression settings when people are warming up, if you're the daring type.

I look at the whole process like so: you need to transform the dynamic range of "reality" which is, what, 120 decibels from whisper to scream, to about the 20 decibels from digital zero to about 20 dBs down--that's "quiet" on a CD, in my book.

[Gilliland]

Originally Posted by keVAn

I was always under the impression that engineers dont usually limit when recording in a studio because they want to keep all the dynamics. But in a live application, it is hard to predict the peaks and once it happens there is no "overdub". To prevent the ugly sound of digital distortion, do any of you limit when recording a live gig? (taking signal from the FOH)

I don't do it. The actual dynamic range of the performance is much lower than the potential dynamic range of the 24 bit medium. So I just set my levels so that there is plenty of headroom, with peaks reaching perhaps -12db and average levels in the -30 to -20 range. At those levels, the noise inherent in even the quietest live setting is still far, far above the noise floor of the recording medium.

I can then compress or limit or otherwise adjust the levels as I see fit at mix time.

I don't take signals from FOH, though. I split the mics and DIs and use my own mic pres.

[Jules]

One cool live recording trick I saw was the practice of recording TWO lead vocal tracks.

1st = normal level
2nd = 6 db down incase the singer started to scream or shout.

[t_chance]

I suppose you could run a compressor on another channel as an option as well, just to see how bad (or good) the gear/operator hit it.

[Remoteness]

Most of us are recording digital now-a-days...

IMO, leave plenty of headroom and you will not need a limiter.

In the old days of analog I always captured the sound using compressors and limiters.
It's how we did it back then. Some folks still do it that way.
For me I've been recording without any toys, leaving it to be addressed during the mix stage.

Mics -- Pres -- recorders.

With that said, I almost always use every toy I can get my hands on post the multitrack recorder.
I patch plenty of compressors and liniters for the 2mix or monitor mix.
I want the reference (or live feed) to sound good. Right?

[keVAn]

Thanks for clearing that up guys. I asked because I did a recording and had a few clips because I was trying to record way too hot. I guess I learned from my mistake. I am recording digitally but dont have the equipment to take a separate split into my own pres. I will probably track at lower levels since I am recording at 24 bit.

Thanks again,
Kevin

[mpdonahue]

Originally Posted by Jules

One cool live recording trick I saw was the practice of recording TWO lead vocal tracks.
1st = normal level
2nd = 6 db down incase the singer started to scream or shout.

Back in the early days of digital multitrack, when we had at at best 14 bits of dynamic range and a grungy noise floor on out 3324's, we normally printed the main mics of our orchestral recordings with the sensitivity raised 18 dB. For the bulk of the recording they were grossly distorted, but on the quiet stuff we would edit them in and get an additional 3 bits of resolution.
Back in those days we were trying to get every last bit of resolution out of those recorders. We pushed the levels right to the top because you needed to minimize the noise when you went to mix the stuff. The accumulated noise and grunge in classical music was what we were always fighting. We recorded at times with emphasis (what a nightmare in post!) and even used Prism bit splitters to record 18 channels of 20 bit on the 16 bit machines..... Now I complain that I have to run 2 pieces of fiber to get 48 channels of 96k from the stage box into the recorders.....
Long story short, today there is no reason to go into the top 5 or 6 dB. The noise penalty is virtually non existent nowadays.

All the best,
Mark

[Remoteness]

...Ahhh, the "good" old days

[-momo-]

I was in exactly the same boat as you a few weeks ago - I'd recorded a few choirs with compression going in and I really disliked the results. So for my last gig I simply kept the peaks to around -12 or so. Worked like a charm, never had any clips... well, okay - there was one, which came as a complete surprise but it was still my fault - things were just working too well and I wasn't paying attention.

It made working in post a lot easier too. Recording this way was a major "ah ha!" moment for me, and I learned it here!

[mrsteaks]

I second all that here. -6dbfs is my 0 VU. That's on peaks. The noise floor is so low as to be almost non existant, which is proved every time in post when I push the levels up to 0 dbfs for the master. Oh - I record classical - choirs, ensembles, and the such. I did use a limiter back when I was recording to DAT, but never since going to HD.

~~~~Mark
Pretty Good Recordings