DACs, DIRAC, Preamplification - What am I not understanding?

[mdsimon2]

I had a few that I use, power physics, Class D Audio, and D-Sonic. The SHD manual states not to go above -10db of attenuation, so i dont.

I asked for the specific amplifier you are using as I am interested in gain / power output (aka input sensitivity) of the amplifier as it would give an idea how much output you are really losing with 10 dB attenuation. If output voltage is quite a bit higher than input sensitivity in the base case you will be able to tolerate more attenuation.

That being said as mentioned by others using the Okto upstream of the SHD makes no sense, especially if you add -10 dB attenuation in the SHD. If for some reason you continue with this setup it would be better to use 0 dB attenuation on the SHD and use the volume control of the Okto or SHD. Remember that whether you actually have digital clipping in the SHD depends on what the input level is, what processing you have applied in the SHD and the volume control position of the SHD.

In general I would never add permanent attenuation in the SHD unless you are running the SHD at a fixed volume level and are doing volume control in a downstream multichannel DAC / preamp. If you decide to go this route it would be a good idea to measure the actual response of the Dirac filters being implemented in the SHD using the bi-directional USB audio. This will give you a better idea of how much boost you actually have and you can apply more exact attenuation rather than a blanket 10 dB.

Michael

 

[dualazmak]

Hello OP @newrival,

I use OKTO DAC8PRO in my multichannel multi-SP-driver multi-amplifier stereo audio project.
My latest setup here and here (with TT in digital setup) would be of your reference and interest.

As you may find in my system diagrams, I use DSP software EKIO as system-wide DSP (XO/EQ/delay) center, and use JRiver MC for music player and master volume controller, both on a Windows 11 Pro PC.

 

[dualazmak]

Even in an audio setup of single stereo DAC and passive SP system having LCR-network, we can implement XO/EQ/delay features if we would use software DSP in upstream PC or Mac computer as shown in this diagram;

Here, I write "Quasi-Relative-Time-Alignment" because we still need LC(R)-passive network between the amp and the SP drivers in this case.

I once have tested this approach in my reference singel-DAC+single-amplifier passive audio setup having LCR-network, and objectively confirmed that "Quasi-Relative-Time-Alignment" could be achieved.

Actual real setup example? Yes, you may find it here on my project thread; with AHB2 or Accuphase E-460 amplifier.

I assume, therefore, this approach would be practically useful/worthwhile if you would like to establish time alignment between sub-woofer(s) and group of other SP drivers in ms precision/accuracy in singel-DAC+single-amplifier passive audio setup.

After that experiment, I quickly moved/departed forward, into my long journey of multichannel multi-amplifier fully active audio system with complete elimination of LC(R)-network and SP attenuators.

 

[phoenixdogfan]

Welcome to ASR.

Placing the Okto DAC upstream of the miniDSP SHD would be a waste. The SHD will reconvert the analog output signal of the Okto DAC with its onboard ADC back into digital for processing, then output the DSP'd signal using its onboard DAC.

Why not put the Okto DAC downstream of the SHD, and feed the Okto DAC with the digital output from the SHD?

Streamer >>> SHD digital out >>> DAC8 >>> Preamp (if necessary) >>> Amplifier >>> Speakers
                                                     └─ analog out >>> subs

If the DAC8 output is insufficient and you need to amplify it, I think the Topping Pre90 is a much better choice than the Audio-GD.

Problem is he has the stereo version of the DAC 8, not the Pro so he has only two outputs which means he can not feed it four digital channels and output channels to the subs, so he has to send the Octo output to the SHD, so it can do the 2 x 4 crossover.

 

[dualazmak]

Problem is he has the stereo version of the DAC 8, not the Pro so he has only two outputs which means he can not feed it four digital channels and output channels to the subs, so he has to send the Octo output to the SHD, so it can do the 2 x 4 crossover.

Yes, I also noticed your point, and it is why I suggested an alternative setup (as an example) in my post #23 above...

 

[phoenixdogfan]

In personally don't see where you are deriving any benefit from having the DAC 8 Stereo in the chain. Either you will do an A to D conversion with the DAC 8 and feed it into the Analog Inputs of the Shd where it will promptly do a D to A conversion, perform all of its DSP operations (PEQ, Xover, Dirac Correction) do a D to A conversion on the resultant four channels and output them to the amp. Alternatively you can output a digital signal from the DAC 8 via AES/EBU where the SHD will do all the digital operation and convert the four channels to Analog before outputtin them to the amps. Either way the bottleneck will be the very good D to A converters on the SHD which are not quite as good (at least by measurement) as the two channel DAC on the OCTO.

I think ultimately it makes more sense to put 2-4 channels of analog gain ( preamplification) at the end of the signal path (after the SHD but before the amps). That way you can continue to use the SHD for your analog inputs as well as your digital and have a quality electronic crossover, DAC, Dirac, and enough gain as well. I have an Octo DAC 8 Pro, but if I had your use case (I don't), I would not hesitate to jetison the Octo for the Shd. The SHD is a quality unit, and despite it's slightly inferior measurement, I'm sure your won't notice an audible difference.

 

[phoenixdogfan]

Yes, I also noticed your point, and it is why I suggested an alternative setup (as an example) in my post #23 above...

Obviously, if he has a sub such as an original SB 2000 which has both a low and high pass filter, he could sent the output of either DAC (probably the miniDSP b/c that's the one where Dirac resides) have the sub do the xover, and send the hi-passed signal to the stereo amps. But even doing that raises the question of what advantage derives from having both the Octo and the miniDSP in the signal path. Ultimately, I think, one of them should come out. I think the best and simplest solution is to just use the miniDSP b/c it will do A to D conversion for any analog inputs as well as doing all the High/low pass, Dirac, PEQ, routing, ect. with that superb interface. I think it's just fine as a four channel DAC for a 2+2 system.

 

[dualazmak]

In personally don't see where you are deriving any benefit from having the DAC 8 Stereo in the chain. Either you will do an A to D conversion with the DAC 8 and feed it into the Analog Inputs of the Shd where it will promptly do a D to A conversion, perform all of its DSP operations (PEQ, Xover, Dirac Correction) do a D to A conversion on the resultant four channels and output them to the amp. Alternatively you can output a digital signal from the DAC 8 via AES/EBU where the SHD will do all the digital operation and convert the four channels to Analog before outputtin them to the amps. Either way the bottleneck will be the very good D to A converters on the SHD which are not quite as good (at least by measurement) as the two channel DAC on the OCTO.

I think ultimately it makes more sense to put 2-4 channels of analog gain ( preamplification) at the end of the signal path (after the SHD but before the amps). That way you can continue to use the SHD for your analog inputs as well as your digital and have a quality electronic crossover, DAC, Dirac, and enough gain as well. I have an Octo DAC 8 Pro, but if I had your use case (I don't), I would not hesitate to jetison the Octo for the Shd. The SHD is a quality unit, and despite it's slightly inferior measurement, I'm sure your won't notice an audible difference.

I fully agree with you.
As you may well aware, I fully utilize DAC8PRO (please refer here and here) as 8-Ch excellent multichannel DAC.

Since it looks OP @newrival has only one DAC8 Stereo, however, I suggested the PC-based DSP (XO/EQ/delay) setup for single-stereo-DAC plus single-stereo-amplifier for passive-SP-system. This suggestion is just one of the many possible directions the OP may move forward at least for a while.

 

[DWPress]

@phoenixdogfan has the best solution I think with adding preamps after the SHD and I agree that the Okto doesn't need to be in the chain. Finding a pair of volume controls with gain that measures transparent shouldn't be to expensive and, after levels have been set in Dirac or REW on the outputs, they don't even have to be accessible or visible. At least this way you're assured you won't be digitally clipping but the amps and speakers still have limitations before they introduce noise of course.

When I use Dirac on my computer the filters chew up 10.5dB of headroom which is probably similar to your observations with the SHD. Both the SHD and Okto are the same voltage on the balanced outputs so even if you ditched the SHD and used a computer for Dirac you'd still end up with about the same output levels in the end.

Dirac and other room correction come at the price of headroom so if you want it louder and distortion free the only real way for that to happen is bigger speakers and more capable amplification.