Hey guys, after many searches online I have been able to find a concise explanation of this concept of unity gain so have turned to this forum which has saved my butt countless times. Everyone's talking about how important unity gain is and one guy even named his studio 'unity gain studio!' In specific I'm wondering how I would set it up for an outboard synth (mininova) plugged into a Focusrite 2i2 with the signal then outputting to studio monitors. Using Cubase 5 as my Daw. Can someone please offer us an easy to understand explanation of this clandestine practice?
Unity gain means that you leave the Cubase/Pro Tools volume fader untouched at 0. At this setting you have neither boosted or attenuated the input/output signal level. In Pro Tools and also in Cubase (the newer versions) you can from this level further boost the signal by +12 dB (by +6 dB in Cubase 5).
When at this default setting you import a track of audio, lets say a commercial wav file, then you will notice that at the default unity gain setting, the colored meter at the default metering setting of "sample peak" (in Pro Tools) will show bars that will peak near -0 dBFS because the advanced metering settings for "sample peak" has 0 dB calibrated to 0 dBFS.
Beyond this, the sample peak number indicator will show the true peak value relative to 0 dBFS (not relative to 0 dB). Now, what this means is that you can calibrate the colored meter to show different bar "heights" depending on certain meter behavior, such as e.g. setting the 0 dB to -6 dBFS, which will render the colored bar "heights" at 0 when the audio is at a level of the max output level capacity of your audio interface's output stage, minus -6 dB.
This means that -0 dBFS in your DAW is not necessarily -0 dBFS in my DAW. (in absolute dBu terms, eventhough identical metering calibration configuration is used in both DAWs and eventhough the dBFS values register the same inside of each DAW)
If I would use let's say a FF 800 converter, and you would use let's say an Apogee Symphony, then if we would measure the voltage coming out from my DAW at -0 dBFS and compare that to the voltage coming out from your DAW at -0 dBFS, we would notice a huge difference in that you get much higher voltages out.
This is why it is so incredibly important to have high capacity in the input and output stages of your converter, it's everything (because a lot of non-harmonic noise such as the noise of the noise floor eats up the emotions of your mix).
The more resistance you have there, the more in the dark you are as an engineer.
So, 0 dBFS is a measurement that shows how close you are to the full input/output level capacity of your audio interface. Now, in terms of gain staging ITB, unity gain means little. In Pro Tools the insert tracks are post fader.
This also includes the master fader. So for instance if you have a compressor as an insert effect on the master bus and adjust the master fader up, then you push more signal into the compressor, you do not raise the level of the compressed signal, that's the difference.
In Cubase 5 this might be different, I can't remember, but it is a very important thing to be aware of. Secondly, you should focus on the gain structure of your mix, meaning how the signal is boosted and attenuated at each process stage (e.g. insert fx). If you for instance boost the master fader to +12 dB while the input track faders are very attenuated (lowered in volume), then you are bringing up the noise of the noise floor by +12 dB.
Similarly, when you have a lot of boosts everywhere in the mix because the input level was too low at each process stage, then that will also accumulate into more noise.
The same with compressor make up gain, you always get some added noise in there.
So pay attention to signal boosts at a lot of places, try to narrow down such that you attenuate the input signal and boost the output signal as little as possible at each process stage.
This will ensure you keep as much of the original sound source present in the mix, while providing maximum signal to the fxs, while keeping the noise floor down to max.
When you work in an environment with a lot of "resistance" (which exists in multiple components), then what happens is that you lose track of the gain staging.
What in a low resistance environment would be solved by an attenuation of let's say -3dB could in a high resistance environment result in a -7dB cut.
Similarly, in a pro studio a limiter could be set at threshold of -1.5 dB while doing the same in a home studio environment could result in a threshold of -3 dB.
All of those signal losses accumulate into weak and bad sounding mixes compared to pro mixes.
When you perform gain staging, remember that the gain/volume faders always attenuate/boost all frequencies, indirectly meaning all noise frequencies.
As you can imagine it does not sound sweet to let the mix sink into the noise floor without any control of how much it has done so.
I recommend that you use volume faders to set the max attenuation level of each track at the most minimum possible (for your needs and with a possible post-fader behavior in mind depending on the DAW software in use) and from there use EQs and multiband compressors to further attenuate or boost various frequencies.
This helps to control the noise frequencies of each track. When you don't go this route and especially in high resistance environments, what happens is that the noise eats up your mix.
To get a good understanding about the importance of this, take the most emotional reference mix you know and let it sink well into the noise floor.
Then compensate such that the loudness is the same on both versions. Then bounce, dither, playback and compare. You will be amazed at the difference.
Then do the same however focusing the noise to a particular band. It does not sound as bad although the frequency balance has been damaged it still does not sound as bad, because the noise frequencies in the other bands are less active relative to the other version you compare against.
With great gear and monitoring (and skills) you can hence achieve great signal to noise levels and this will boost the emotion of your final product.
Also remember that you have limited headroom and a summing engine at the end, so for instance a high track count with a lot of parallel processing can yield a lot of noise at the summing stage and not only that but it will give you tons of phase inaccuracies too.
This and especially in a high resistance context will simply not sound good.
Think of the impact of both resistance and noise not only at the stereo level, but also at L, R, M, S individually per band. And remember that it is the hard panned/S component of your mix where the ears are the most sensitive to noise. Noise in general should be focused towards the M component rather than the S component. The S component should be kept as clean from noise as possible, which is the case 1 times out of 10.