Optimum levels to capture audio (based on their likely level in the final mix)

[CochraneDrumTutor]

Hi there
Another newbie question if I may...
When recording multiple instruments / vocals, should you always try to capture the strongest audio level for every audio feed in a mix OR can you capture them at levels "similar" to what you expect them to be at in the final mix please? for example...

Snare drum: I'm pretty confident I want the full strong "touching orange" kinda signal for this so thats what I'm gain staging for.

Room mic.: I'm pretty sure I just need a "taste" of this in the final mix, so I capture it solidly in the mid green of say -25Db.

Or should I just always strive to record each audio feed to its fullest, and then just adjust down as required when mixing for playback / final output?

Many thanks in advance.
Ed.

[James Perrett]

Or should I just always strive to record each audio feed to its fullest, and then just adjust down as required when mixing for playback / final output?

I'd say neither.

I aim for peaks on all tracks to regularly get up to around -18dBFS and accept that the odd peak will go higher. For drums that would mean that a typical snare hit reaches around -18dBFS but accents might go up to -12 or -10dBFS.

However, with modern 24 bit recording, I won't worry if levels are slightly lower.

[Hugh Robjohns]

There are two schools of thought here, both dating back to analogue days, and both perfectly valid in the digital domain.

1. Track with good levels for everything.

And by that, I mean peaks to a maximum of -10dBFS and average levels around -20dBFS... but as James says, levels 10dB or 15dB lower are perfectly fine too, and probably a good idea where the sources are unpredictably dynamic.

2 Track loudest items in the mix as above, with sources known to be quieter in the final mix at appropriately lower relative levels.

This is the 'flat faders' approach whereby playing back the raw tracks with all faders at unity gives something closely approximating the wanted mix.

Option 2 is a classic approach that many experienced producers employed in the days of analogue multitracks (and it is still a valid option).

However, it relies upon considerable production experience in knowing how the final mix will work during the tracking phase.

In analogue tape days, deciding to raise something in the mix that was tracked at low level risked a significant noise penalty. Thankfully, that risk is very considerably lower with modern digital recording.

Personally, I favour option 1 for most work but both techniques are acceptable and workable.

[CochraneDrumTutor]

Many thanks all for the responses. They definitely help me grasp the key points to keep in mind.
Ed.