TOPIC:

My first post and new to MiniDSP. Some questions maybe basic in this thread. Already tried searching, but I didn't get my answers and hence this thread. I apologize if already answered elsewhere. If so, please point me to the relevant threads.

Goal: 2-Ch ultimate stereo performance combining old and new. Leveraging MQA, Dirac Live room correction, tube pre-amp stage, tube amplifier stage for mids and highs, solid state amplifier stage for lows

I spent a full weekend (our MLK weekend ) and significant hours in calibration and listening. I think I have a reasonable high resolution system (McIntosh C2600 pre-amp acting as DAC and volume control and a McIntosh 275 Amp for mid/high and Audiocontrol solid state for bass). I followed every word in the book as dictated by the MiniDSP manaul. Result, I heard distortion on some tracks. I realized that I have to leave 10dB headroom (I have seen those threads). What is not clear to me is the following. If I do my Dirac Live Calibration not going beyond -10dB, should I still set the master volume at -10dB during playback?
At -10dB the distortion goes away which is good. Can someone, "Dev Team" chime in your best known methods for calibration?
DevTeam, let's say I follow all your rules, I stay within the -10dB during Dirac Calibration, I get all my filters right and loaded. Do I still have to set my Master Volume at -10dB or lower? If so, I don't like this idea. Why, I am not completely convinced that the digital volume control is equal or better than the one offered in my McIntosh C2600.

SHD users, please chime in. I am addressing the DevTeam as they seem to respond.
Regards,
Jai

I missed a couple of components in the chain. I have Roon running on my networked PC and I have the Oppo UDP-203. I have tried the Oppo device settings as renderer only, renderer and decoder and no MQA with identical results. Please factor this additional input into your response. I apologize for missing this in my first post. I did find a way to enjoy MQA tracks though SHD doesn't support MQA (or did that get added recently)?

Can you clarify which miniDSP product you are using and how/where it is connected in your system?

SHD dacs cannot do MQA unfold, and from what I've read on this forum probably never will.

What sort of distortion are you getting?

As a general point, I believe attenuation to avoid digital clipping needs to be done at the inputs, not with the master volume. I also thought Dirac did this automatically though, so that the user didn't need to. Others will know more about this than me.

Not sure what I wrote here was correct - so removed.

If using an analog input, you don't want to clip the A/D convertor. Other than that, it's the net gain (added to the peak input signal level) that will determine whether you get clipping, including the master volume.

In my case the default target curve crosses several valleys of the measured curve. When the difference is > +2dB I get distortions on those frequencies. Lowering the whole target curve helps. But I chose to form the target curve locally to follow the deepest valleys. The result is a less linear target curve but higher overall volume. I like that better.

Thanks folks. Mine is an SHD studio. I use the digital coax input and output. The use case is just Dirac Live and no other uses. Since the SHD comes at the end of the chain where the Oppo digital output goes in to the SHD digital input, MQA is already taken care of by the core decoder and Oppo as the renderer. Am I misunderstanding the MQA flow?

I've been having a bit of a look into the issue of how to avoid digital clipping with Dirac but there seems to be a distinct lack of good information. This old post gave some information:

www.minidsp.com/forum/dirac-series-suppo...nting-clipping#39111

I'm wondering though if a control that was present in Dirac v1 has been removed for Dirac v2? I don't see any way of setting an attenuation to allow for boosting within the Dirac application now?

One other possibly related bit of information that I've come across is point 4.4 in the following link:

www.dirac.com/faq-technical

It is for a different application than the SHD but describes a fixed attenuation applied to both Dirac On and Off states specifically to avoid digital clipping. It would be really helpful if @devteam could respond to comment on if a similar situation applies with the SHD products?

If there isn't any attenuation applied to prevent clipping then it sound to me like what the OP is experiencing is exactly this.

jmcomp124 wrote:

Since the SHD comes at the end of the chain where the Oppo digital output goes in to the SHD digital input, MQA is already taken care of by the core decoder and Oppo as the renderer. Am I misunderstanding the MQA flow?

I've not looked too deeply into MQA but I think what you're doing there makes sense. I tried to be specific above that what the SHD can't do is the MQA specific decoding. Once it has be unfolded and processed to a high-res PCM data stream then the SHD will be able to process this like a normal high-res music file. I guess the only question is if this is definitely what the OPPO sends to it's digital output.

@jmcomp124 one thing that might be interesting is to measure to see what boost levels are actually being applied by Dirac. Are you familiar with using REW? If you are then make a measurements at your MLP with Dirac On and the Off, and divide one by the other to see what the Dirac filter is doing. (Or visually compare the two by dividing them makes it easier to quickly see what is going on IMHO.)

jmcomp124 wrote:

Am I misunderstanding the MQA flow?

I don't think so, but if I understand your setup correctly then the Roon Signal Path will show where it is.

Thank you. I did see that thread with useful information. It is still not clear if the Master Volume needs to be set to -10dB or lower. I think the best bet is to use -10dB or lower (meaning values like -12, -20) during measurements in DLCT and also follow the same rule during playback/listening. It is the latter that I have a problem with. Let me know if you have specific insight into this concern.

jmcomp124 wrote:

Thank you. I did see that thread with useful information. It is still not clear if the Master Volume needs to be set to -10dB or lower. I think the best bet is to use -10dB or lower (meaning values like -12, -20) during measurements in DLCT and also follow the same rule during playback/listening. It is the latter that I have a problem with. Let me know if you have specific insight into this concern.

I see you mentioned DLCT - does this mean you're still using Dirac V1 rather than V2?

Ultrasonic wrote:

jmcomp124 wrote:

Thank you. I did see that thread with useful information. It is still not clear if the Master Volume needs to be set to -10dB or lower. I think the best bet is to use -10dB or lower (meaning values like -12, -20) during measurements in DLCT and also follow the same rule during playback/listening. It is the latter that I have a problem with. Let me know if you have specific insight into this concern.

I see you mentioned DLCT - does this mean you're still using Dirac V1 rather than V2?

I tried Dirac2 and I had a bunch of issues with MacOS Catalina. I then went with Dirac 1 and it works. So I have Dirac 1 and that is what the calibration is from.

My understanding is that the digital output (SPDIF) that comes out of the Oppo is already rendered and there is no more MQA information in it and it is just like any other PCM signal. After a bunch of reading, I thought MQA was not possible due to MiniDSP in the chain. But then it dawned on me that there should be a solution and hence my current config. The MCIntosh C2600 always reports 96KHz when I play MQA tracks. I have set Roon to max 192Khz. Not sure why and I need to look into this.