Altabay

Ha, import/export hiding right under my nose. Thank you for the tip!

I have a feature suggestion. It would be very convenient to have configuration sets (or "profiles") to easily switch between favorite settings. This is similar in spirit but independent of Kaplesh's idea of whittling down the list of filters. My actual need is switching between NAAs for the loudspeaker system and the headphone system whose DACs have very different requirements. Currently I would have to change various items in HQPlayer settings including target NAA, filter, modulator, PCM vs SDM, max rate, adaptive rate, and so on. It would be nice to save the settings as descriptively named profiles. All configurable items should be included in a profile definition, even things like convolution and pipeline. Profiles can also be used to easily switch between favorite filter/modulator settings depending things like genre and recording quality. This might be a bit crazy, but one thing that would be really cool is being able to instantaneously switch between two profiles for easy A/B comparison. Imagine HQPlayer splits the input signal to two and applies filters/modulators from two chosen profiles. There would have to be some restrictions like using common output device and some other settings. This will double the load on the HQPlayer host, so it would be necessary to dial down things like max rate. There might not be enough general value in this capability, so maybe not worth the development effort, but I just thought it could be really cool to instantly hear the impact of switching to a different filter or modulator.

Please help me pick the right GPU for my new HQPlayer machine. I have a NUC13 Extreme i9-13900K running Windows 11 as the dedicated HQPlayer Desktop machine. Based on chassis constraints I think the safe GPU choices are RTX 4070 and RTX 4070Ti. I prefer 4070 because it has lower TDP and is an easier fit in the chassis, but I'm concerned it might be underpowered for the job and I'm also wondering if I'm leaving too much performance on the table by forgoing the 4070Ti. My DAC is Holo Spring (v1) which goes up to DSD512 but doesn't do 48k DSD, so rate family conversion is necessary for 48/96/192k sources. Would the 4070 be powerful enough to run all filters to go from 192k PCM to 44.1k DSD512? Or is that too much even for the 4070Ti? I will be upgrading the DAC at some point to something that can do DSD1024 and 48k DSD. Would the 4070 be adequate at that rate? I'd guess it should be able to handle upconversion to DSD1024 without rate family change if it can do DSD512 with rate family change. The i9-13900K can run all EC modulators at DSD512 without any BIOS tweak. But can it do DSD1024? Appreciate all thoughts. Thanks!

Yes, it seems to be a GUI behavior only. I tested with HQPlayer Desktop 4.3.1 for Windows. Adding 5 dB steps would be really nice, and I hope other users agree. Roon UI is also limiting the adjustments to 1dB steps. Is it a limitation in Roon or is it somehow discovering HQPlayer is only allowing 1dB increments?

My apology if this was discussed before, but is it possible to change HQPlayer volume in 0.5dB steps? Currently it seems to allow 1dB steps only, and depending on the tracks I sometimes wish a half step change was possible. I don't think a smaller step would be necessary. And would Roon allow half step changes if HQPlayer supported it?

Good point. I will try with the same setting as v3. I was using DSD512, ext2, DSDv2 with HQP3, compared to DSD128, ext2, DSD5EC with HQP4. Do you have any thoughts on allowing longer buffering to absorb any temoirary interruptions from Roon?

I've been having dropout issues too after upgrading to HQP4. There were very occasional dropouts with HQP3, but definitely more common with v4. I didn't find anything obvious in the HQP log. I'm running Roon and HQPlayer on the same Windows 10 machine, with files on NAS. Playing through NAA (NUC running Miska's NAA image) over wifi connection. Is there a way to view the logs on the NAA? Would it be possible to add configurable buffering to HQPlayer? I'd be willing to live with 10, 20 seconds delay if it results in more stable playback. I'm already dealing with 7-8 seconds of wait to start the music to use the EC modulators.

I am also in the Windows NAA camp. My NAA is a NUC running WinServer 2012R2 Core with AO, going into SU-1 and Holo Spring DAC. In my system the Windows NAA is clearly superior to microRendu. I know it's not a defective mR because I have two of them. I can't explain why, but the difference is obvious. This experience has stopped me from getting interested in an upgrade to ultraRendu. Maybe I'll tweak things again at some point, but I am very happy with the Windows NAA for now. Disclaimer: I have not tried any other Linux based NAA.

I am also using microRendu 1.3 board, 2.5 software. I have two of them and both exhibit the problem. My chain is normally NAA > IR/LPS-1 > SU-1/LPS-1 > Holo Spring L3 DAC. IR and SU-1 use separate LPS-1s. I've taken IR and SU-1 out of the chain to eliminate them as the source of the problem. I also have a NAA on NUC5i5RYH running WinServer 2012R2 with AO core mode, and this sounds substantially better than microRendu, in my system. HQP runs on another NUC (NUC6i7KYK), running Windows 10 with no optimization software. RoonServer runs on this machine as well. The CPU stays 50% or less when upsampling to DSD12, so no problem there. The fact that NUC NAA sounds great also helps eliminate HQP machine as the problem in my setup. I'm wondering if the microRendu hardware is underpowered to handle DSD512. Or it could be the HQP NAA software, or a combination of both. Too bad there's no way to monitor CPU usage of the microRendu.

I'm very interested to hear what you find out after debugging with Sonore. I'm having this problem as well and I've put my plan for microRendu 1.4 or ultraRendu on hold. Are you using microRendu 1.3 or 1.4?

Thanks for correcting me, Scan. I retract my comment that NUC7i7BNH can handle DSD512 upsampling. I'll stick to describing what my NUC6i7KYK can do. Upsampling redbook to DSD512 with poly-sinc-xtr-2s shows CPU fluctuating between 25% and 50% in Task Manager. It sits around 36% if I change Task Manager to slower refresh interval. (View > Update speed > Low). Memory usage is only 2.8GB.

I don't remember the % CPU, let me get back to you on that. I'm running Windows 10 without any optimization software. I went through settings and turned off whatever I could find, but honestly I don't think these matter much because there is enough headroom and I'm using a NAA. My NAA is a NUC5i5 running Windows Server 2012 R2 and AO core mode. You might be able to hear the fan noise during upsampling, but you can easily solve this by running it in a different room, assuming you are using a NAA. I tend to listen at high volume and I can't really hear the fan even when I have the NUC next to the DAC.

Other than the 16MB typo, I think you'll be fine. At a glance the NUC7i7BHN sounds more powerful than the NUC6i7KYK I'm using. I am also running RoonServer and HQP on the same NUC. I can upsample everything to DSD512 with poly-sinc-xtr-2s filter. Or non-2s when upsampling DSD64 to DSD512. PCM to DSD512 only works with -2s. I have 32GB memory but it's an overkill. 16GB should be plenty.

+1 !

Can you show your HQP settings? Does DSD work with any other player? (foobar, roon, jriver)