I’ve confused myself into a possibly needless tailspin! While mixing, I usually monitor with the control room level pot on my analogue console set between about four and five (out of 10). This usually equates to around 76dB SPL (C‑weighted, slow response) from my main studio monitors. It’s a comfortable yet loud enough level to work with, and my mixes have generally seemed consistent in terms of overall level.
I should probably have left it there! The trouble began when I decided to try to measure the output from the main monitors when playing pink noise from the DAW, and I think my confusion may be due to my convoluted signal flow and hybrid studio setup.
I’m sending one track of mono uncorrelated pink noise to one input channel of a Soundcraft Series 600 analogue console. The channel fader is set at unity, with no gain, and it’s bussed to the main mix with the master fader at unity. The signal goes from the main mix to an RME ADI‑2FS interface, which converts the stereo analogue signal to a S/PDIF signal and routes that back to the interface and a DAW stereo track running a Sonarworks plug‑in. The signal then goes out through the D‑A converter to a two‑track input on the console, and that’s what I am monitoring. It’s a complex setup, but it allows me to hear the output of the console with Sonarworks ‘correcting’ for the deficiencies in my room.
Because of that signal flow, and the different maximum input and output levels of the DAW/interface, the analogue console and the RME converter (as well as the 7.3dB level loss from the Sonarworks correction software), I’m getting confused about what it’s important to measure and where. For example, ‑20dBFS pink noise from the DAW doesn’t come back at the end of the line and hit the stereo track running Sonarworks that I am monitoring at ‑20dBFS. It’s lower — but I can increase the output of the pink noise from the DAW above ‑20dBFS so that it does come back in at ‑20dBFS, and then measure the SPL from the monitors at that point.
But is that what I should be measuring? If I do this, then, coincidentally, I’m back at around 4.5 on the console’s control room pot again, and measuring around 76dB SPL.
On top of all that, I’ve noticed that when I am getting ‑20dBFS out of the Sonarworks track and monitoring that on the console, the console’s VU meters hover around ‑12 or ‑13 VU. Does that seem correct? In other words, it’s as if 0VU on my console corresponds more to around ‑10dBFS than ‑20dBFS in the DAW. For what it’s worth, I just sent a sine wave down the same path and when the VU meters read zero on the console I’m measuring that output at around 1.2V RMS, so that seems correct.
SOS Forum Post
Hugh Robjohns, SOS Technical Editor replies: I think ‘tailspin’ is the correct description — you have several different problems in the gain structure of your signal path, and that’s clearly confusing you!
76dBC is a perfectly acceptable reference listening level in a small room (equating to ‑20dBFS in your DAW). So that’s great. Unfortunately, having the monitoring volume control on the desk around 4/10 or 5/10 is not so great, as in that region you’ll have stereo mistracking and poor level resolution issues. It would be far better if you could reduce the sensitivity of your monitors to allow that control to reside around 7/10 when delivering your comfortable (76dBC) reference level from the speakers. If you can’t turn the speakers’ input sensitivity down, you could buy or make inline attenuators to solve the problem, or perhaps increase the attenuation using Sonarworks.
As for the rest of the alignment process, all that matters is that ‑20dBFS on your final mix bus equates to your acoustic reference level. It doesn’t matter what level you’re pushing through the analogue console at this stage, nor what level the pink noise generator has to run at to achieve the desired goal. That said, you’d make life much easier if you could bypass the console part of the input signal path for this alignment process.
You can’t measure the true RMS level of pink noise on a DAW’s sample‑peak meters, so it’s better to use a file that has been properly designed.
You can’t measure the true RMS level of pink noise on a DAW’s sample‑peak meters, so it’s better to use a file that has been properly designed. I use the Blue Sky calibration files, which I know to be correct: https://abluesky.com/support/blue-sky-calibration-test-files. All you need to do is load the file into your DAW and play it directly through your output monitoring chain (and Sonarworks) into the console’s monitor section and your speaker, and adjust the speaker’s sensitivity accordingly with the monitoring volume control at its reference position (ideally around 7/10, as mentioned). You may well find that using a band‑limited (500Hz‑2.5kHz) pink noise file gives more reliable results with your SPL meter as it neatly avoids problems with LF room resonances and HF splashes from hard surfaces like consoles.
Once you’ve established a suitably calibrated relationship between the DAW’s output and your speakers, you can do away with the pink noise, turn the monitors down, and align the rest of the signal path using sine‑wave tones. (I prefer working with 400Hz to 1kHz, since it’s less painful on the ears, though the BlueSky files only offer 1kHz.)
Then work through your interface I/O level settings and your console input gains to achieve a unity gain signal path all the way through the chain. This may or may not be possible, depending on the options in your interface, and what your console’s VU meters are aligned for. In normal professional circles 0VU should equate to +4dBu and ‑20dBFS (meaning that 0dBFS = +24dBu), but there are various other ‘standard’ calibrations that are all perfectly workable. You mentioned an RME converter, but only the very latest ones can cope with +24dBu. Most of the older ones have an option for 0dBFS = +19dBu, and are designed to conform to the European broadcasting standard alignment where 0dBu = ‑18dBFS (or +4dBu (0VU) = ‑14dBFS).
Anyway, check the alignment of your console meters, decide on a suitable alignment between the console meters and the DAW, and adjust gains and sensitivities accordingly so that the headroom margins through the whole chain are sensible and everything works as you expect it to.