TOPIC:

Firstly my apologies if this has been covered elsewhere but I’m a bit of a rookie and just want to make certain I understand how I need to set up my new SHD for gain and volume when implementing DIRAC. It’s a fairly simple setup initially - active stereo speakers (with digital inputs) and a sub, using the SHD as a crossover and preamp to control the volume (everything else is fixed volume).

I’m taking a digital spdif output from a streamer (Sonos Port) at fixed level into spdif input 1,2 on the SHD. Also using the network Ethernet input for Volumio.

The digital spdif outputs 1,2 of the SHD feed active L and R speakers (with their own DACs). Analogue RCA output 3 of the SHD goes to an analogue input of an active subwoofer.

My question is what volume setting (and where is it applied) do I need to set the SHD at? Do I adjust the the output channels 1,2,3 on the SHD by eg 10dB? Or can I run these at 0dB but make sure I never use the SHD at over -10dB on the master volume (assuming it’s in the digital domain and processes before DIRAC so there’s enough headroom)? Remember that outputs 1,2 of the SHD are digital and the output 3 (the sub) is analogue, but I’m assuming that shouldn’t make a difference.

There’s so much stuff on the forums on gain and headroom that I’m getting a little swamped, so I’m just after a simple reply or explanation. Thanks again.

There is a way (discussed in this thread www.minidsp.com/forum/dirac-series-suppo...eadroom?limitstart=0) that you can work out exactly what headroom you need to leave, however the general rule of thumb seems to be never go above -10db with Dirac on. How you make sure of that is up to you both ways you suggest can work.

Personally I've applied -10db to the output channels. I'm using a ddrc24 without display so applying to the outputs means I'm free to use the MV without worrying about ever exceeding -10db (though my ears keep me a fair distance away also).

TLDR: start with SHD gain at -20dB and adjust until you get a suitable sound pressure level.

Either method you suggest is fine. The former is arguably safer but the latter is easier to change. There's actually no reason why the SHD gain shouldn't go over -10dB. It just may not be able to if Dirac is using 10dB gain and your audio signal hits 0dB.

If your audio signal only hits -4dB your SHD may be able to go to -6dB.
If your audio signal only hits -4dB and Dirac uses no boost your SHD may be able to go to 0dB

The point is the SHD control is a gain, not an output level, so the actual output level is the input level multiplied by Dirac gain multiplied by SHD gain. The only constraint is that the output level should not go above 0dB. You will likely never know for sure where your input level peaks or what gain Dirac applies so -10dB SHD level is just the worst case that guarantees no clipping at the SHD output.

I believe the SHD gain is at the end of the processing.

Note that if your power amplifier is very sensitive and you find SHD gain needs to be set significantly below -10dB it is worth transferring some of that attenuation to the individual channel output gains in the SHD plugin. This ensures safety even if SHD gain is increased to 0dB suddenly by accident.

entripy wrote:

You will likely never know for sure where your input level peaks or what gain Dirac applies

Hi, I'm not sure that's true anymore. @Tw99 figured out a way to measure what Dirac does and a few of us have done it (including @Ultrasonic and @Tony_J).
It's in the post I linked. REW > MiniDSP > REW . Run a sweep with and without Dirac on. Divide one result by the other and you know what Dirac has applied. Take a look at that thread.

asx77 wrote:

entripy wrote:

You will likely never know for sure where your input level peaks or what gain Dirac applies

Hi, I'm not sure that's true anymore. @Tw99 figured out a way to measure what Dirac does and a few of us have done it (including @Ultrasonic and @Tony_J).
It's in the post I linked. REW > MiniDSP > REW . Run a sweep with and without Dirac on. Divide one result by the other and you know what Dirac has applied. Take a look at that thread.

Well OK, it is possible

But I still think it's not likely that many folk will bother to do that, I'm certainly too old and lazy to bother.

For one of my SHD systems I can reason that Dirac applied all 10dB gain at certain frequencies because some notches I had got improved by 10dB but no more.

In terms of signal level its a bit harder. Old loudness wars mastering will hit 0dB almost continuously but any more enlightened mastering should arguably be to EBU R128 by now. All the mixes I do certainly are and not much of the music I listen to gets anywhere near 0dB because of that. Well that's maybe only 4dB to 6dB below 0dB but that's significant in this context.

Thanks everyone. So just to be super clear. If I leave everything at 0dB in the chain I described in post 1, AND I never turn the master volume knob of the SHD over -10dB, I should broadly be ok for headroom and avoid digital clipping when Dirac is doing its stuff?

Yes