I did some further debugging. I disconnected the external ADC/DAC and ran a single AES cable from a 642 output to a 642 input. The signal was clean and channels are properly going where they should at 192 kHz.

This exonerates SMUX and the MADI as far as I can see. But it DOES point to the possibility that the RME ADI-642 PLL is having trouble locking to the AES preamble from my external converter at 192 kHz. All lock lights are good however, but still this points to a problem. I've had exactly this problem with two different models of ADC/DAC, by the way.

Is there any other logical explanation?

Thanks, Mattias, for your quick reply. Glad to see you are alive and well!

I've tried this at 192 kHz in the past with a Cranesong HEDD quantum. And recently a Grimm Audio UC-1. In both cases the ADC is the master clock. Currently the WC from the ADC is going into two 642's on high impedance with a T connector and then to the WC input of the MADI card where the WC is terminated. All the sync lights are performing correctly on the ADI-642 and everything is perfectly well at 96 kHz. Only at 192 kHz is there this problem.

When I go to the "About" menu it does not tell me which version PCB it is. All I get is "RME Hammerfall DSP. Windows XP/Vista/7/8/10 driver. Driver date 6/7/2021. Version 4.38." I'm not sure I can find it in Device manager. What kind of device does it show up as in Device manager?

looking for the board version I can find the serial number in the HDSPe Flash update tool: HDSPe MADI (22694227). Firmware or "revision" is 33. Maybe that's the board number.

Thanks for all the information. You have the latest driver, and also the latest firmware in the HDSPe card (v33). And I don't think this is an issue of the card being defective.

The card should be visible in the Device Manager, under Audio, Video and Sound controllers. It's there anyway, otherwise it wouldn't work.

To further track this down I like to come back to your first description - the external DAC-ADC loop. The 642 has phones monitor, which can easily be routed to any input channel. If you have distortion on one channel use that tool to check that the 642 does not have (or has) this distortion at its input. If so then check the ADC's output on another receiver. If it's ok there plug it back to the 642. Distorted there then this would strongly point to an issue with the 642 AES receiver.

Another possible issue could be the Matrix having some overlooked routing active that causes issues. Not very likely, as the Matrix is not a mixer, and you should get immediate full level feedback when doing so. Still it might be an idea to use the MIDI Remote app to get a full view on all active routings.

In the same way you should check that TotalMix doesn't have any overlooked routings for the MADI card. Here, due to level control, feedback can indeed be metallic sounds, depending on the level.

HDSPe MADI cards should be in stock, as updated version with a new FPGA and therefore different firmware number (210). No changes or fixes that would explain your issues, though.

Hi Bob (or should I say Mr. Katz),
glad to see it has improved, although it sounds like it has changed now “out of the blue” or did you change something in the setup/routing?
I don't mean to intrude on Matthias' excellent support and knowledge here …. I can't give an exact technical reason for this, pure gut feeling: I would try out if changing the MADI frame size from 48 to 96k causes a change. Even if you find that it doesn't change anything, that would also be a statement that you know it's not related.

The distortion is present in the phones monitor of the ADI-642 as well. I checked all stages of the route and it is distorted. Some kind of clocking or locking problem?