Genelecs fed with AES digital - unwanted clicks in audio

[sleepy.sock]

Hi folks, just to let you know that this issue has now been resolved. The occasional clicks were indeed caused by buffer underruns inside the RME ADI-2. I increased the buffer size in Roon to 100msec; now there are no more clicks. Hurrah! I'm delighted to have put this issue to bed. It also turned out it wasn't anything to do with the Genelec DSP.

I have to praise Genelec and their customer support (thanks also to @RobL for nudging me to talk directly to them). They were absolutely outstanding: within minutes, they worked through various possible options and told me how to solve the problem (despite it being not being with their own product at all). Really, really outstanding customer support, engineering knowledge, and engagement. Very reassuring to have access to this as an owner.

There is the remaining question of why I didn't hear clicks when using the headphone output of my RME unit. I am unsure of the answer to this. It could be that the RME's DAC (which is not engaged when driving the AES output) does some reconstruction work which either removes or attenuates the effects of buffer underruns. Or it could be that the Genelecs are more revealing than my headphones.

I hope that this thread may be helpful to anyone who encounters similar issues. If you use an RME ADI-2 over USB with Roon and experiencing rare clicks, consider increasing the buffer size.

It would be useful if the RME unit could report or warn about buffer underruns. This would have helped diagnose the issue quickly and give certainty that it won't reoccur. It might be a nice feature to add to the unit. (I believe that you can visualise this on the computer in USB interface mode, but not on the device when it is used as a class compliant USB device).

 

[Norcal]

Hi folks, just to let you know that this issue has now been resolved. The occasional clicks were indeed caused by buffer underruns inside the RME ADI-2. I increased the buffer size in Roon to 100msec; now there are no more clicks. Hurrah! I'm delighted to have put this issue to bed. It also turned out it wasn't anything to do with the Genelec DSP.

I have to praise Genelec and their customer support (thanks also to @RobL for nudging me to talk directly to them). They were absolutely outstanding: within minutes, they worked through various possible options and told me how to solve the problem (despite it being not being with their own product at all). Really, really outstanding customer support, engineering knowledge, and engagement. Very reassuring to have access to this as an owner.

There is the remaining question of why I didn't hear clicks when using the headphone output of my RME unit. I am unsure of the answer to this. It could be that the RME's DAC (which is not engaged when driving the AES output) does some reconstruction work which either removes or attenuates the effects of buffer underruns. Or it could be that the Genelecs are more revealing than my headphones.

I hope that this thread may be helpful to anyone who encounters similar issues. If you use an RME ADI-2 over USB with Roon and experiencing rare clicks, consider increasing the buffer size.

It would be useful if the RME unit could report or warn about buffer underruns. This would have helped diagnose the issue quickly and give certainty that it won't reoccur. It might be a nice feature to add to the unit. (I believe that you can visualise this on the computer in USB interface mode, but not on the device when it is used as a class compliant USB device).

I think RME has a presence here. @MC_RME - any thoughts?

 

[MC_RME]

Not many - apart from pointing out a complete misunderstanding on how USB interfaces work. Setting buffer sizes in Roon changes the way data is sent to the ADI-2 within the computer, not within the ADI-2. The ADI hardware does not care nor act in any way differently. It is also impossible to have clicks on AES but not on phones if these are generated within the computer (not as stated above 'inside the RME ADI-2'). They most probably randomly did not occur while using phones, or were less noticable then.

I also have to say 100 ms on the application is a very big buffer size that should not be necessary at all. Which buffer size was selected in the ADI-2 Settings dialog? That one is much more relevant for clicks and should have been raised as first step. Then followed by an analysis with Latency Monitor etc why the computer is so slow in processing.

Application buffer underruns can NOT be detected by the audio interface, only the application , and should be reported by the application (here Roon). USB errors ARE shown in our Settings dialog.

My comments describe Windows - the OP did not reveal the OS that he uses.

 

[sleepy.sock]

The OS is Linux (RoPieee), device is a Raspberry Pi. I am using the USB interface in stereo class compliant mode over ALSA. This is why I could not use LatencyMon or the ADI-2 Settings dialog box on the computer.

I understand from your response that there is no way for the ADI-2 to know if there has been a buffer underrun on data it receives, that is fine. I did not know whether it was possible. I now know that it is not.

BTW, I had not intended to suggest that the ADI-2 was in any way responsible for the clicks. I understood that the issue was with the computer feeding it.

 

[MC_RME]

That info should have been in the first post - it explains the high buffer time set for Roon. When I read Genelecs and AES I immediately thought of studio work, not home hifi listening.

 

[sleepy.sock]

I posted this info earlier in the thread, shortly after my initial post.

I did not post it in the first post, as I only noticed these clicks once I had introduced the Genelecs into the system. I wrongly assumed they had somehow caused them. As you say, the clicks were there all along; the monitors are just exceptionally revealing of them. Technical support at Genelec was excellent and identified the problem was buffer underrun on the computer.

 

[MC_RME]

No problem, glad it is fixed!

 

[sleepy.sock]

I can confirm that my issue with the occasional clicks have been completely solved. However, because I'm now sensitised to hearing them, anytime I do hear a click or tick I immediately scrub the audio track to confirm that it was on the recording and not produced by my system. In all cases now, I have found the click was on recording -- great news for me!

What I learned from this is that is it remarkably common for digital ticks and clicks to sneak onto a record -- even recordings that are otherwise well produced and mastered. The Genelec 8331As are simply incredible at unearthing these. Anyway, despite spending a good chunk of my life listening to music, this information was news to me.