Sign in to follow this  
Followers 0
dylab

Gain Structure and recording levels

6 posts in this topic

I was reading this rant from Todd about gain structure and old VU decks and tape and headroom etc

http://www.toddburke.com/tb_pages/Todd_Burke/Entries/2010/9/7_Gain_Structure.html

Which has got me thinking about my setup at home and levels I record sounds from my gear into computer.

I record things pretty hot into my MOTU between - 4 and - 8 sometimes a little louder and then monitor with my headphones turned down a fair bit , usually around -10 or so.

So maybe I need to have my headphone at 0 db and record at lower volumes , around the -16 db mark.

What do you guys that have gear do?

Is quieter recording the secret to clearer better sounding final products?

Share this post


Link to post
Share on other sites

Yay, open up a can of worms dylab! There's several opinions on the matter. Some say make the most of your bits and do everything at high levels. Others say leave some headroom and record your tracks at lower levels.

It seems the prevailing recommendation is to record multitrack files at lower levels around -15 to -18dB, so long as you do it all in 24 bit resolution. Roughly, -15dB still uses around 21 of the 24 bits, so the resolution is still about 32 times better than CD/16 bit (without using the calc - someone correct my figures if you want ;) )

When you mix those tracks together, you are not pushing the master stereo output too hard into - and potentially over - zero. Eg. two tracks summed at -6dB each results in 0dB at the output. Only one more track pushes the level into clip, requiring you to lower the master fader or insert a brick wall limiter. 64 tracks and it sounds like shite.

So without getting too complicated, yes lower levels will give you a cleaner and more dynamic sound as you prevent clipping and limiting, and the resolution is still good enough for the noise floor to be inaudible.

Also inserting gain-lifting plug-ins like compression and boosted eq are likely to add more dB's to the channel's overall level anyway, so you're buying some headroom here too.

Just keep an eye on the master output at all times, and ride the master fader to bring the final output close to zero. The your final hard limiter can do its job of loudening without stressing it too hard ;)

Share this post


Link to post
Share on other sites

cheers RB

I moving to record at lower levels - I always did 24 bit but now will be a bit quieter

Share this post


Link to post
Share on other sites

Yeah, I'm kinda with RB's thinking here.

My view might be way out of date, and I'm led to believe that floating point maths somehow negates much of the internal clipping concerns but my personal experiences lead me to stop hammering the inputs at just under 0dBFS, preferring to go in at around -16dBFS.

It could be that the analogue components of the hardware audio interface's inputs simply don't like HOT signals, or that the DAW's sends/insert paths didn't like the HOT signals, or the plug-ins I was using didn't like the HOT signals, or there's some merit in not pulling back the master fader to restore an unclipped state from the sum of a couple of dozen HOT channels, but the result of the lower inputs appeared to me as breathing clearer, less harsh, more detailed, increased sense of depth.

It also made sense to me, as far as how the integration of a full-scale digital scale should appear against a regular analogue dBu scale.

And it made calibrating my mixer against my DAW simple: by lining up the 0dBu on the hardware mixer's meters to hit -16dBFS on the DAW, I can now simply set input gain levels for any source on the mixer to peak around 0dBu, knowing it'll be perfectly accommodated in the DAW and with plenty of room to breath.

Things get a little clunky when 'mastered' stuff comes out of the DAW (following compression and maximiser-style limiting) and now cranking at fraction under 0dBFS, but I manage this by having my DAW's return to the hardware mixer for monitoring purposes turned down a touch so it doesn't peak off the scale.

Share this post


Link to post
Share on other sites

cheers spectrum - i must say it makes it a lot less of a bun fight trying to pick out drum machines and synths with everything a fair bit lower -

Share this post


Link to post
Share on other sites

Yeah, absolutely D.

And I find this way also makes it far easier to take hold of something special in the mix and bring it up to the forefront to jump out and startle the listener if everything else isn't already cranked up to "11".

Start out conservative with levels, work with the idea of just getting a sweet, balanced mix, then move on to worrying about how to get it LOUD (if that's the agenda).

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0