Proper Gain staging in a world of synths, outboard, DAW's

[Bath House]

I've been reading up on gain staging and the difference between 0 DBVU and DBFS and all of that. http://rhythminmind.net/1313/?p=3224

I tend to try to track so that my signals are averaging -18 and peaking at -6 at the most, but I do wonder if I couldn't just record hotter since I'm mixing in software anyway and can pull down those channels.

How do you guys handle proper gain staging?

[Graham Hinton]

I've been reading up on gain staging and the difference between 0 DBVU and DBFS and all of that. http://rhythminmind.net/1313/?p=3224

I tend to try to track so that my signals are averaging -18 and peaking at -6 at the most, but I do wonder if I couldn't just record hotter since I'm mixing in software anyway and can pull down those channels.

How do you guys handle proper gain staging?

The ideal match between analogue and digital audio is when they both clip at the same level which would be about +22dBm. That makes studio line level of +4dBm = 0VU = -18dBFS.

As both synthesizers and mixers are essentially the same technology running on similar power supplies don't worry about it, but you don't really need it hotter. If anything you need to turn synths down by 6 to 10dB because the common +/-5V signal level is equivalent to about +14dBm (dependent on waveform). You have enough headroom on pro audio equiment, but a lot of semi-pro computer soundcards won't handle that level.

[Babaluma]

i tend to track and mix looking for an average RMS value of -18dBfs, like you. this usually equates with 0dBVU, which is the analogue standard, so if i want to run things between digital, analogue outboard, and back again, everything is at the perfect level for the best s/n ratio and the least distortion.

there are lots of previous posts here regarding "gain staging", if you want to use the advanced search function.

no need to record hotter if you are recording with 24 bit, you could be 30dB RMS down and still have a better S/N ratio than 16 bit audio. if you tax all the analogue stages less, you'll end up with loads more headroom and a cleaner sound overall, win/win!

worry about loudness in the mastering stage (or not...) i have 75dB clean analogue class a gain on tap, so i'm in no short supply.

[Bath House]

i tend to track and mix looking for an average RMS value of -18dBfs, like you. this usually equates with 0dBVU, which is the analogue standard, so if i want to run things between digital, analogue outboard, and back again, everything is at the perfect level for the best s/n ratio and the least distortion.

there are lots of previous posts here regarding "gain staging", if you want to use the advanced search function.

no need to record hotter if you are recording with 24 bit, you could be 30dB RMS down and still have a better S/N ratio than 16 bit audio. if you tax all the analogue stages less, you'll end up with loads more headroom and a cleaner sound overall, win/win!

worry about loudness in the mastering stage (or not...) i have dB clean analogue class a gain on tap, so i'm in no short supply.

This is exactly what I'm doing, but I'm getting pushback from other people when I'm working collaboratively or for clients because its' "too low, bro." Synths plus pres plus outboard into +4 audio interface = -18 RMS in my book, because the noise floor is pushed sooooooooooo far down, so you can record way quieter.

[Babaluma]

just stroke your chin like a hipster and be safe in the knowledge that it's you who is doing it right. your recordings and listeners will thank you for it in the end.

[Graham Hinton]

no need to record hotter if you are recording with 24 bit, you could be 30dB RMS down and still have a better S/N ratio than 16 bit audio. if you tax all the analogue stages less, you'll end up with loads more headroom and a cleaner sound overall, win/win!

There are two fixed points in recording: the analogue clipping level at about +22dBm and the analogue noise floor at about -95dBm. The noise floor is often higher, -95dBm is what a single quality op amp would produce. This gives a snr ratio of 117dB. Recording at 16 bit has a dynamic range of 96dB so that is clearly not enough to capture everything accurately, recording at 24dB with a theoretical dynamic range of 144dB does not give a better snr, it just gives better waveform resolution (= lower distortion) at the lowest levels. In analogue a -96dBm sinewave is still a sinewave buried in noise, in digital it is a severely quantised signal. You have to decide where you map the analogue range into the digital range, 24 bit is better than 16 bit, but not 30dB better. The noise floor is set by the worst analogue equipment in the chain and can only be lowered by denoising software processing.

worry about loudness in the mastering stage (or not...) i have 75dB clean analogue class a gain on tap, so i'm in no short supply.

If you have a digital recording with an analogue noise floor of -80dBm and then put 75dB of analogue gain on it you will have a noise floor at -5dBm.

Don't get confused between dBm which is an absolute voltage and dB which is a power ratio and commonly used to play numbers games with specifications.