Two Convolution Filter

[Simon_Arnold3 (Simon Arnold)]

Just set Roon to convert DSD to PCM or disable native DSD playback it will then convert to PCM and stay PCM.

[brian (Brian Luczkiewicz)]

@Karl-Heinz_Fink

I have another question about DSD: Yes, we need to make it PCM in order to add the filtering, but why is it not possible to stick with the PCM instead of going back to DSD. Can we fool the system by telling it the D/A cannot do DSD and so we only convert it to PCM?

No, it is not necessary to make it PCM in order to do the filtering. Multi-bit yes, PCM, no. There is no sample rate conversion involved.

If you want DSD to be converted to PCM, processed, then outputted like that, disable the “Native DSD processing” setting and that is what will happen.

Parallize Sigma-Delta Modulator: then the sigma delta modulator is still working? but uses more cpu core?

Yes. Normally we only use one CPU core per zone. That setting lets it use more cores, which can make more expensive DSP configurations possible when DSD is involved.

Enable Native DSD Processing: Then the signal will be converted to PCM for DSP purpose and then back to DSD for output? and if we turn it off we get only PCM as output?

No. The signal will be converted to a multi-bit representation at the original sample rate, then modulated back to 1-bit (DSD) for output.

If you turn it off, the signal will be converted to PCM, processed, and output as PCM.

[Karl-Heinz_Fink]

Hi Brian,

thanks…makes it a lot clearer to me. We need to read more about the DSP and DSD thing.However, as our server(player is a bit weak, we convert to PCM right now. I should switch back on the unit I’m using to an endpoint only and let my NUC do the hard work. We had the system running today and it works very well. In theory, we could make the filters for almost every speaker, but so far, we will offer only data for our BORG. We tried the bottom end filtering today on a new model that is on the way and it did very well with corrections. More work to do
Best regards

KH

[alec_eiffel (Rémi)]

For what is worth : I am using two convolution filters, in series. The first is a stereo --> stereo “HAF” filter for speaker/phase/room/reverberated field correction. The second is a stereo --> 2 x 4 = 8 channels for active filtering of 4 way speakers. Works like a charm. To my knowledge Roon is the only audio player allowing to “pipe” convolution filters, HQPlayer, Jriver, ConvolverVST can’t do it. Saves the hassle of manually convolving all filters, which is always feasible but painful with Convolver command line utility and Audacity (probably easier with Matlab, but I don’t have a license for it).

[Shauk_Khan]

Hallo Brian

So the settings for the DSD processing are independent from the fact that Sample Rate Conversion is turned on or off?? In our case even when the sample rate conversion is turned off, changing the DSD settings work, is it the correct behavior?

[brian (Brian Luczkiewicz)]

Yes. Those settings apply whenever there is a reason to process DSD.

[Harald_Pott]

Reading a recent article in the online magazine LowBeats you were successful with your work.
Do you plan to publish the results?
I would be interested in the files for the Q Acoustic C500 speakers.