DAW Mixing: Lower Levels=Better Sounding Mixes?

[Nescafe]

Just want to share my opinions, I also feel that my ITB mix is better when I ended using lower level (in every track, master bus fader still in 0 position), everything become better sounding, reverb, compression, etc, less annoying artifact, my glorious number around -8 peak in master bus meter (Cubase 4 here), never null the file, but my ear is the final decision here.

Cheers

[trapper]

I think there's a lot of confusion in this thread so here's my tuppence: keeping levels not super hot when tracking is one thing, but once in your DAW it's a different kettle of fish. Peeder's on the money. A 32bit float DAW has NO problems with gain until it hits the master bus, where upon the converter will react accordingly to the signal level. Keeping your channels at -6db or whatever is a waste of time as this is 32bit audio we're dealing with! The whole point (pardon the pun) of using a floating point engine is they are ideal for gain stages and summing as there is no "ceiling" (in useful terms), as such. Try cranking a channel down by 30db then applying 30db of gain... the result will sound the same as the unprocessed signal.

How plugins handle >0db is really down to the plugin, so unless there's some sort of saturation schnizzle going on there it shouldn't be a problem. A fixed-point audio engine like PT is a different kettle of fish and something I have no experience so can't really comment.

[kurt]

Originally Posted by trapper

I think there's a lot of confusion in this thread so here's my tuppence: keeping levels not super hot when tracking is one thing, but once in your DAW it's a different kettle of fish. Peeder's on the money. A 32bit float DAW has NO problems with gain until it hits the master bus, where upon the converter will react accordingly to the signal level. Keeping your channels at -6db or whatever is a waste of time as this is 32bit audio we're dealing with! The whole point (pardon the pun) of using a floating point engine is they are ideal for gain stages and summing as there is no "ceiling" (in useful terms), as such. Try cranking a channel down by 30db then applying 30db of gain... the result will sound the same as the unprocessed signal.

How plugins handle >0db is really down to the plugin, so unless there's some sort of saturation schnizzle going on there it shouldn't be a problem. A fixed-point audio engine like PT is a different kettle of fish and something I have no experience so can't really comment.

This is not at all thread about Cubase versus ProTools.
You miss the whole point. Try before judge.

[trapper]

Originally Posted by kurt

This is not at all thread about Cubase versus ProTools.
You miss the whole point. Try before judge.

what on earth are you talking about??

[emkay]

Originally Posted by not_so_new

Hi Charles....

We are on different pages here sorry to disagree but I disagree 100%...



The noise in front of the mic will always be there. This is not "noise in the digital domain" this is noise in the analog domain. When you are going into the AD conversion process the concern is the headroom of the converter NOT the headroom of the source signal.

Lets say that you have 60 cycle hum in a guitar rig. That 60 cycle will be on the recording medium no matter what. The 60 cycle hum will be on the analog tape or the digital converter either way. If I have that 60 cycle hum on an analog tape and I bring up the fader on the console I am bringing up the 60 cycle hum from the original source. This is no different than the digital domain.

The point is that analog tape ADDED to the noise. 60 cycle hum + tape hiss. In the digital domain the 60 cycle hum is added with the noise floor of the converter and the noise floor of the converter is MUCH MUCH less than the noise floor of the tape.

All the noise that adds up in the analog domain is cumulative, amp noise + line noise from the cable + power supply noise of the preamp + maybe more line noise from the cable again + the converter noise = the whole noise floor. Once the signal is in the converter there is no more added noise, distortion yes but noise no.

Yes, I agree you are on two different pages.
One is referring to the whole chain in either medium. The other is referring to the recording medium itself.
I think what we are talking about here is not the "front end" (instruments,mics,pres,dynamics etc. the "whole chain" and it's inherent cumulative noise),but the actual noise floor of the recording medium itself. There is no comparison between the noise floor of analogue vs. the noise floor of digital. In audible comparison, there is no noise floor in digital. Noise floor in analogue is very audible.
I am sure if you could "isolate' and listen to the two different mediums in a perfectly grounded anechoic chamber this would be very evident both to the human ear and the measuring devices..

[mdoelger]

Originally Posted by trapper

what on earth are you talking about??

I don't know either. I guess he's missing the point of your post.

Keep levels moderate while tracking. Make sure you don't clip the output while mixing.
Know your plugs.

As simple as it gets.

[GeorgeHayduke]

In case anyone would like to check out SSL's free "intersample peak" meter, it is here:
Solid State Logic | Music

-hayduke

PS: get the trick or treat plugin, too!

[mdoelger]

Originally Posted by GeorgeHayduke

In case anyone would like to check out SSL's free "intersample peak" meter, it is here:
Solid State Logic | Music

-hayduke

PS: get the trick or treat plugin, too!

Peakmeter is nice. What rocks is the listen mic compressor. Go get that!

[smoke]

Originally Posted by dodittydada

Just read the entire thread. very enlightening!!!

ONE question however...

Tracking and mixing at the discussed lower levels means the operating range of plugs, say for example the threshold of a compressor is greatly reduced. I find I will have to crank the threshold near full just to get a few db gain reduction.

I find this with all my plugs. Seems not a high enough input to drive plug ins through their usuable range, or much range at all really!

WHAT am I missing on this one??

Yes this definitely happens when you have plug-ins where you want to load the input to cause overdrive ie: saturation plugs. It depends on the plug, some plugs have workarounds.

[prismtheory]

I think just like with analog gear, digital domain has it's sweet spot. I almost always track with the levels just into the yellow in pro tools (except for drums sometimes)...I don't even really look at it, it just ends up being that way. I developed that technique before I got real converters.
I always thought that driving shitty converters hard didn't do anything but add artifacts from the analog part of the converter. Afterall -18dbfs is unity in analog world. Or close to it. I think analog guys tend to record hotter (to retain the bit depth or some crap like that)...but with 24 bits, and I'm speaking on behalf of pro tools HD...you would need to lower the fader -96dbfs in order to JUST reach 24 bits. I think...if there's someone nerdier than I, feel free to correct me if I'm wrong. My mixes end up peaking somewhere around -12 to -8dbfs and I've had a top mastering guy master my work and he said that that was perfectly acceptable.

As a matter of fact, any mix i've done OTB has been that quiet too...I guess I'm just really sensitive to distortion. Like, some guys like API driven hard...it makes me cringe when that signal starts getting into the yellow. My 2 cents