TOPIC:

Yes, I have used REW. Actually to set my DIY sub sub-sonic filters. I am not sure if this will answer the clipping question. The applied filter values will not necessarily translate to actual measured values. The REW delta you suggest measuring, will give the delta of before and after in-room response which will be interesting. I will try it out when I get some time.

jmcomp124 wrote:

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.

The v2.0 manual specifies Mojave rather than Catalina so I think it's a general compatibilty issue. This does mean that what you and I see are different though. Do you therefore have a level adjustment you can change in Dirac (not the first one setting speaker volume for the test measurements but another later on)? If so, and you currently have this at -10 dB, you could try reducing this to, say, -15 dB and see if it improves your problem?

jmcomp124 wrote:

The applied filter values will not necessarily translate to actual measured values.

Why not? I think they must do...

There is though a distinction between the maximum boost applied (which tells you what happens at one frequency) and the difference in peak levels of the time domain signal for Dirac on vs off (which is determined by how all of the different frequency components add up).

Ultrasonic wrote:

jmcomp124 wrote:

The applied filter values will not necessarily translate to actual measured values.

Why not? I think they must do...

There is though a distinction between the maximum boost applied (which tells you what happens at one frequency) and the difference in peak levels of the time domain signal for Dirac on vs off (which is determined by how all of the different frequency components add up).

Not necessarily. There are other factors involved in the chain like the speaker driver compression. In theory, a xdB increase or cut should result in the same increase or cut in measured levels if the system faithfully reproduces this. The sure way to verify would be measure the electrical response with a loop back. I have tried this in the analog domain and since my device is purely digital, haven't thought about it. Not sure if I am making sense on how I am explaining this. REW, with measured values will get you close.

I believe there is a level adjustment. Will check this weekend. I appreciate all your inputs. Thanks a lot!

jmcomp124 wrote:

Not necessarily. There are other factors involved in the chain like the speaker driver compression. In theory, a xdB increase or cut should result in the same increase or cut in measured levels if the system faithfully reproduces this. The sure way to verify would be measure the electrical response with a loop back. I have tried this in the analog domain and since my device is purely digital, haven't thought about it. Not sure if I am making sense on how I am explaining this. REW, with measured values will get you close.

Whilst technically true I think these effects should be very small for any half-decent loudspeaker (run at sensible levels). If they weren't very small then the performance of the speaker would be significantly compromised under normal operation, irrespective of any DSP. Thinking about it, I'd guess noise in the measurement may well be the biggest source of discrepancy, which could be assessed by making some repeat measurements.

Overall though I'd have thought a with/without measurement comparison should give at least a pretty good indication of what filter was applied.

Folks/DevTeam,
Let me boil the question down to something simple.

"What is the recommended setting for Master Volume while using a high quality external volume control with Dirac On to get the highest quality digital output from the SHD studio"?

Please provide your reasoning behind the answer.
Thanks,
Jai

Is the MiniDSP SHD Studio limited to 96Khz internally? I tried Quboz and seem to like the 192KHz tracks better than MQA for certain tracks I have listened to so far with the Oppo directly connected to the C2600 preamp. However when I take the chain through the MiniDSP my pre-amp reports 96Khz. Looks like the MiniDSP downsamples and then sends that as the output. Is this indeed the case? I am using SPDIF.
Thanks,
Jai

Yes - there's an asynchronous sample rate convertor on the input that resamples to the DSP's processing rate. Which is 96kHz for the SHD processors.

HTH

john.reekie wrote:

Yes - there's an asynchronous sample rate convertor on the input that resamples to the DSP's processing rate. Which is 96kHz for the SHD processors.

HTH

Thanks John.
Bummer, I had so much hopes. Distortion at no attenuation and being forced to do digital attenuation, not being able to preserve 192Khz due to sample rate converter, No response from Devteam (this is the deal breaker). My exit from this particular product is clear to me.
My thanks to the forum community here for your help and inputs. Will revisit for sure at a later time. My quest continues...
Enjoy the hobby!
Regards,
Jai

jmcomp124 wrote:

Distortion at no attenuation

Um, that's not correct. The reason you are getting distortion is because you are applying gain (via the Dirac Live filter). A digital signal can not exceed full-scale - this is fundamental, a "law of physics" if you like.

Anyway good luck.

jmcomp124 wrote:

I tried Quboz and seem to like the 192KHz tracks better than MQA for certain tracks I have listened ...

This is slightly off the main topic of this thread but this observation is perhaps not that surprising. MQA is a lossy way to try to provide 'high res' data streams using less bandwidth, whereas Qobuz is true (lossless) high-res. If you do some Googling you'll find a lot of technical criticism of MQA.

jmcomp124 wrote:

Distortion at no attenuation and being forced to do digital attenuation

I think you're dismissing the product for flawed reasons. Firstly as @john.reekie has pointed out, management of digital clipping is fundamental to any DSP system that applies boosts. Yes I wish there was more clarity on exactly how this is handled on the SHD but it shouldn't be seen a fundamental flaw. Also I think you're being unnecessarily prejudiced against digital attenuation. I deliberately now use digital attenuation in favour of an analogue pre-amp. In fact my SHD is currently sitting on top of an unused analogue pre-amp! This article may be of some interest on the subject:

benchmarkmedia.com/blogs/application_not...control-technologies

If you're happy to accept the signal processing of Dirac I struggle to see why the more benign change of a simple attenuation is such a big concern?

jmcomp124 wrote:

I believe there is a level adjustment. Will check this weekend. I appreciate all your inputs. Thanks a lot!

Did you try this? It would help understand what is going on, and if it works as I think it might then my thinking was it may enable you to have the master volume higher (possibly 0 dB) during playback.

Finally, have you tried logging your query with support? Bear in mind that this forum is not the official means of support for SHD products.

What I mean by "distortion at no attenuation" is "I hear distortion when I set the Master Volume at 0". If the Dirac filter was applying gain for a certain frequency and Q, and applying cuts and if there are many such data points, that should be factored in and the user should not be allowed to go above that. So basically, there should be a new 0.0 for Master Volume that I don't have to deal with. This is just my opinion and there maybe a lot going on behind the scenes that I don't understand and the very reason I am seeking to.
Thank you for your inputs. Just wanted to clarify what I meant. I do have another MiniDSP product that works in the subwoofer domain where and it is perfectly fine for that domain. Got rid of my old Rane PE-17 analog devices and happy with the small miniDSP box.

Ultrasonic wrote:

jmcomp124 wrote:

I tried Quboz and seem to like the 192KHz tracks better than MQA for certain tracks I have listened ...

This is slightly off the main topic of this thread but this observation is perhaps not that surprising. MQA is a lossy way to try to provide 'high res' data streams using less bandwidth, whereas Qobuz is true (lossless) high-res. If you do some Googling you'll find a lot of technical criticism of MQA.

jmcomp124 wrote:

Distortion at no attenuation and being forced to do digital attenuation

I think you're dismissing the product for flawed reasons. Firstly as @john.reekie has pointed out, management of digital clipping is fundamental to any DSP system that applies boosts. Yes I wish there was more clarity on exactly how this is handled on the SHD but it shouldn't be seen a fundamental flaw. Also I think you're being unnecessarily prejudiced against digital attenuation. I deliberately now use digital attenuation in favour of an analogue pre-amp. In fact my SHD is currently sitting on top of an unused analogue pre-amp! This article may be of some interest on the subject:

benchmarkmedia.com/blogs/application_not...control-technologies

If you're happy to accept the signal processing of Dirac I struggle to see why the more benign change of a simple attenuation is such a big concern?

jmcomp124 wrote:

I believe there is a level adjustment. Will check this weekend. I appreciate all your inputs. Thanks a lot!

Did you try this? It would help understand what is going on, and if it works as I think it might then my thinking was it may enable you to have the master volume higher (possibly 0 dB) during playback.

Finally, have you tried logging your query with support? Bear in mind that this forum is not the official means of support for SHD products.

I am touching on some highly debated topics for example 192Khz vs 96Khz, MQA vs FLAC 192, digital vs analog attenuation and then there are others in the chain (tube vs SS, interconnects, DACs involved...barely scratched the surface). The reasons I gave are some fundamental data points. A lot more reasoning went into it with my ears being the final help for decision making. I have not come to conclusions yet about Dirac Live in a 2-ch stereo setup. So I will wait for the right product or product revision to come up. I do appreciate all your inputs. So thank you for the same.

jmcomp124 wrote:

What I mean by "distortion at no attenuation" is "I hear distortion when I set the Master Volume at 0".

I'd understood this .

I agree with you it would be helpful if there could be an automatic attenuation applied to prevent clipping. I wonder if an attempt to do this has been implemented in Dirac v2 since the control it sounds like you have to prevent this has been removed. That may be wishful thinking though!