Best Recording Levels

[Top Jimmy]

Originally Posted by booter

I would bet that if you did a true A/B X test of signals recorded at -6, -18, and -24 with levels matched you wouldn't be able to tell the difference.

You would if you were following the industry practice of placing the signal RMS at those specified values.

Don't apply a peak interpretation to an RMS standard.

[janmuths]

Love this discussion!

Here my tip:

1) figure out the noise floor of your recording gear. Eg, set up a vocal mic, set gain as you'd usually do for vocals and record 'silence'. Use Audiosuite 'Gain' and use 'find peak'. Write down peak and rms values. Using the same settings record some vocals, find peaks again and write down the numbers.
2) By subtracting your results you'd find your recording-chain's signal to noise ratio. Owners of high end gear should come up with lower noise-floor, if you use entry level gear the values will probably not be so impressive.
3) Lower the mic pre gain by 10dB. Repeat step 1) and 2). Compare the results. Chances are that the S/N ratio hasn't changed (ignore fluctuations of less than 1dB for the sake of this experiment)
4) Keep on lowering the gain in 10dB steps and repeat the procedure until you realise that the S/N ratio changes to the worse. You've probably hit the noise floor of your interface. The noise floor of your interface is now higher than the noise-floor of your mic/preamp.
5) You've now hit the 'bottom' of your dynamic range. Do always record at higher levels than that. No big surprise here. Chances are that this level would be ridiculously low anyway and nobody would consider to record that low. Common sense here.
6) Using normalise, bring all vocal recordings up to the same RMS level (e.g. RMS=-20dBfs depending on how dynamic the performance was)
7) Ask your assistant to do a blind test with you. Let the assistant solo one voc recording at a time. Listen. Focus on S/N ratio, and also the 'detail' of the recording. All kinds of technical (noise, distortion) and aesthetic aspects (tone, balance, feel)
8) Decide for a recording level, based on the sound that you like best.

I am pretty confident that different ears listening to the sound of different devices will result in different 'optimal' recording levels.

I've done the test. Let me know what you've figured out for yourself.

What I've learned is that I have to re-adjust my workflow with the gear I use. Example: I used some quality condenser mics in a very quiet recording space. Using my external preamps, I recorded the signal thru and TLA valve console to PT and monitored back thru the console. Most of the noise was introduced by the console. Therefore, I kept the preamp gains a bit higher than usual, so that the signals traveled higher above the noise floor thru the console.

Usually I work without that console, my preamps directly into my interfaces.

If the microphone was noisy, the same workflow wouldn't work as the preamp gains up signal and noise equally. Long story short, I've figured out for myself that the biggest noise issues are usually caused in the live room or in the analogue domain.

My preferred workflow is:
1) a quiet room
2) a low noise microphone
3) quality shielded mic leads
4) quality preamps
Once I've got that, it really doesn't make a difference to my ears whether the peak in PT is @ -12dBfs or -24dBfs.
Anyhow, I do hear a difference in sound when cranking up my preamps to get close to 0dBfs. They clearly operate in their headroom and the increasing THD is clearly audible. My preamps sound best around ... (goes what!!!) +4dBu. TATA!TATA!TATA!
Once I figured that out I started calibrating my 192i/os. Tried +4dBu from my preamps and cranked up the calibration screws at the back. And than down. I even tried that at 192Khz.
My conclusion: once the different recordings are normalised to the same level, they all sound identical to my ears.

After testing my own gear inside-out I've come to a very simple conclusion:
1) my preamps sound best around +4dBu
2) my interfaces sound equal at any calibration level (and therefore varying bit resolution). Consequently I calibrated all I/Os to -24dBfs @ 0VU (=+4dBu) to gain maximum headroom.
3) The 'bottleneck' regarding noise-floor is still the preamp, not the interface, unless I record levels below -35dBu (roundabout), which I won't.


Please do not take my words for granted. This is my setup, this is my ears. Test your own gear. Draw your own conclusions.
I'd be very curious to find out where you agree and where you have different findings. I am sure there is more than one truth here.

Jan

[janmuths]

One last thing:

"Directly or indirectly, all questions connected with this subject must come for decision to the ear, as the organ of hearing, and from it there can be no appeal." Lord Rayleigh, "The Theory of Sound" (Strutt, 1896)

OK, now I shut up.

Jan