Thank you 
Strange, I had to revert back to RopieeeXL as NAA OS did not start? or was not found by HQP after restart of my system. I had to flash the SD again than it worked. When I turned the system couple of days ago of and today on, again HQP did not find the NAA OS, so I switched back to RopieeeXL (did not try to flash again).
This is fine for me as I also do not have my „crackle problem“ with RopieeeXL, so all fine but still strange.
Why is the output volume much lower with NAA running on Rpi 4 compared to as a daemon on a Windows machine? Using NAA 5 here, with ifi Neo DSD DAC at fixed volume mode, HQP set at -2 dBFS, and identical settings in both cases.
Use alsamixer command , choose soundcard with F6 and compare what the volumes are
Shows 100 (dB gain 0.00, 0.00)
1 Like
Are you using PCM or DSD? At least iDSD Micro has a firmware bug that after power up playing 705.6/768k PCM makes it loud & clipping. Need to play something else first before going to those rates to avoid this anomaly.
DSD. Volume is just ~6-9 dB lower on the RPi NAA compared to windows daemon with identical settings of the DAC
Amazon just raised the price from €103 to €119 
Guess I’ll need to wait it out till I get a cheaper offer…
2 Likes
Strange… Same HQPlayer instance, just output device switched?
Both shows 100? NAA OS and Debian?
Yes sir! Just changing the NAA output on the same HQP instance and of course connecting the DAC to the selected one.
The standard NAA 5.0 OS RPi image; guess it’s Debian.
NAA OS is not Debian or any other regular distribution, it is a custom one.
Both cases do DSD output at the same rate? I mean actual output, not just the settings? And on Windows NAA is set to use ASIO?
Because if you for example have Roon running on Windows, and touching the driver, DSD capability will disappear and HQPlayer reverts to PCM output (“feature” of Thesycon’s ASIO driver). So you really need to check what is the actual output format.
NAA on windows is using ifi’s ASIO driver. Roon is running on a completely different machine.
The actual output is always DSD in both cases (Windows or RPi) as I can see on the DAC display.
What does HQPlayer say about the ongoing output format in both cases?
Since NAA is not capable of doing any DSP and is never touching the data going through it, if there’s such difference, then the ASIO driver is doing something strange.
Since I’m trying the sinc-MG filter, HQP output says either 22 or 24 MHz depending on the input sampling family 44.1 or 48 KHz, in case of input PCM. It’s always 24 MHz for DSD input.
The ASIO driver is indeed the only difference. But I’ve no idea if it tampers the signal in any way as the iDSD is configured to use the Bit Perfect (BP) mode.
You have the latest firmware there? I doubt the ASIO driver would be tampering with the DSD data, but it could be tampering with DAC settings.
On TI/BB chip also the selected DSD filter affects output level, it has four different analog filters (conversion stage configurations), each with different output level.
Have you tested if DSD256 behaves the same way? (because iDSD Micro had some firmware bugs)
You have the latest firmware there?
Yes v3.33.
On TI/BB chip also the selected DSD filter affects output level, it has four different analog filters (conversion stage configurations), each with different output level.
I’m using the BP filter as I don’t the DAC to colour the sound in anyway.
Have you tested if DSD256 behaves the same way? (because iDSD Micro had some firmware bugs)
Not yet but the DAC in this case is iDSD Neo.
This applies only to PCM inputs… And if you input max rate PCM, then it doesn’t have effect since it is bypassed. Unfortunately it doesn’t have control for the DSD analog filter, although the DAC chip has four options.
Yeah, I have that DAC too. But the XMOS firmware it has is likely not so far from the one used in other iFi products…
Ouch! I don’t output PCM at all. So what do you suggest for DSD? Asking especially as well since you’ve the same DAC.