Thank you for the report and for your ongoing patience. We’ve escalated this case to QA and Development and will return with a more thorough response as soon as possible.
Automatic diagnostics for your RoonServer instance have been enabled - we shouldn’t require any additional technical or descriptive information to take action from here. Thank you again and we’ll have next steps shortly.
Does it help if I try and reproduce the problem while diagnostics are running? Otherwise I don’t use Roon everyday so I want to make sure you can capture the problem while I am actively using
No worries, Roon logs only capture Roon data when it’s running, and there are several files of data to analyze at a time, so usually the error is captured before the log file overwrites itself. I wonder if a RAATServer refresh could help with this HDMI issue.
You can generate a new RAATServer instance on your device by following these instructions, but please be aware that this will reset your Roon Settings → Audio Tab to factory settings and I would advise making a backup of any custom DSP settings you have:
I appreciate your testing of the issue I am having. As you can see from my write up’s I have tested in the same way you have though I have not tested as many formats as you but that is a none issue here.
I’m trying to digest what you said. Are you saying that when you play via your Denon (ie Target) you are indeed getting the downsampling below 192/24.
I’m just trying to determine whether I am the only one with this problem or there are others.
Clarifying: when I play any track that exceeds 192/24 from the Nucleus One via HDMI to the Denon AVR-A1H, Roon is downsampling the track to 192/24 or lower. In the case of a 352.8/24 track, it is downsampled to 176.4 rather than 192/24. The same happens with some DSD64. The only track that downsampled to 192/24 was the crazy 768/24 sample track I have.
For the 352.8 → 176.4 it is dividing the rate by 2. For the 768 → 192 it is dividing the rate by 4 to meet the input requirements of the Denon via HDMI.
I have tried this test with renaming the RAAT Server and I am getting the same downsampling result. The pics I originally posted are still exactly what I am seeing now.
I’ve done one more new test. I connected HDMI A of the Nucleus Titan to the Denon as my target, I then connected HDMI B to a Bluesound with 192/24 sampling capability.
If I play HDMI A (Denon) using a particular song that is 192/24 it gets downsampled. If I change over to HDMI B and play the exact same song it does not downsampled.
I fully understand these are different hardware components but both have the same spec of 192/24 but anything played on the Denon as the target always gets downsampled.
If this was a music format issue on a certain song then I should see the same results on both hardware platforms but I don’t. So this takes music or the library out of the equation.
So far I can only reproduce when the Target is my Denon AVR. And I can even choose any Denon model in the device setup and it still downsamples everytime,every time.
If this helps anyone looking into this for me, the downsampled music always downsamples to the same format of 48khz/16bit. I hope maybe this is meaningful somewhere***
So I don’t have much music over 192 so I am getting downsampled on any track 192/24 or higher. And my downsampling is major is brings it down to like 48khz at 16 bit. Nothing even close to what the DAC can handle.
I had another forum post a while back under different ID. There was a big debate on how good the DAC was in the Denon AVR-110. Many people chimed in. All I wanted to know was that the DAC was reasonably quality which is what the results seemed to indicate. Actually many people suggested it was an above average DAC. As long as it was not a cheap DAC that couldn’t handle common formats I was happy and I am still happy.
I actually upgraded to the. Nucleus Titan from the Nucleus+ I believe. I don’t remember seeing this in my first gen Nucleus+ but I cannot confirm this for certain so I did not include in the details to the support engineers.
I really just want this working properly because this is in essence why we all by these Roon products. Without this working it is pointless to have high quality/lossless music in your library.
One more thought here, why is the setup window changing as well when the downsampling is occurring? I lose the MQA and DSD capabilities parameters in the device setup on the device with an output target of the Denon AVR-110. So I know if this is going to be a downsampling issue even before I play any music. You can see this in the pics I attached in the original posting.
I do really thank everyone on this forum for their responses and even testing on my behalf. I know the answer is out there somewhere and someone will post a fix for me soon. This is my hope anyway.
That 16 bit 48 kHz limitation looks like a lot like a TV EDID. Many operate exclusively at 48 kHz because that is the de facto standard sample rate for video sources. Is there a TV involved?
No TV involved. This is a Nucleus Titan directly connected to a Denon Receiver using an HDMI v2.1 cable.
But, this is partially my point here. My setup is very straightforward and pretty standard. The HDMI cable connects the two devices together straighthrough and there is absolutely nothing else in play here.
What perplexes me is what triggers the downsampling process?
I’ve eliminated the HDMI cable and receiver being the culprit as I’ve tested my setup with the Nucleus HDMI port connected to no HDMI cable and thus no receiver and I still get a downsampling. I know many of you tried the same test and get the same result.
I’ve already ruled out the actual music being the culprit (see prior post) testing this.
So it appears somewhere in the Room engine the downsampling is being triggered. And for full disclosure I do not have much expertise of the Roon software so I rely on the experts in this forum and Roon support for help with this.
Also important to note this problem occurs whether I play Roon via the App installed on an Apple IOS device (I’ve tried several different devices) or running the Roon app on a Win11 desktop.
And as of today (Jan 16) I am running Roon 2.0 (build 1496).
Well I misunderstood your question. Of course there is a TV connected to the TV monitor port of the Denon but what would that have to do with testing an audio input port?
Also, I can’t stress enough that I can reproduce this problem without the Nucleus connected to the Denon merely by disconnecting the HDMI cable from the Nucleus.
Through my posts I have methodically eliminated every component in the system and am left with Roon running on Nucleus by itself and I still can produce this problem. I would think this becomes an easier problem to solve with just Roon running and no 3rd party components to deal with.
For support:
Do you think there is value in reinstalling the Roon software then doing a restore from one of my backups as an ultimate test of the software. I know we did the RAAT test but would reinstalling the entire system test anything more that is worth doing?
As a last resort I tried renaming the RAATServer after stopping the software again and this time it worked! But unfortunately as I tested Roon at some point it went back to downsampling. So it looks to be something within RAATServet that is causing my problem.
Also tried reinstalling the software today which did not help. I’m not sure if the monitoring you added to my Nucleus needs to be readded or not.
Please let me know if there is anything else I can do to resolve this issue once and for all.
Thank you for your patience. Please navigate to the Nucleus Web Admin interface and reinstall the OS from the button listed.
We’re seeing the [HDMI A] port erroneously report 48KHz as the available max to ALSA in diagnostics. Please let us know if the reinstall allows you to reproduce the downsampling - if the issue recurs, engineering will circle back with a next step.
I reinstalled the OS as instructed but am still having the problem.
Are you saying the 48k is just reporting wrong and it is actually playing at the correct sample rate? If so that is better than the sample rate actually being 48k.