If running DSD1024 with the DAC200, are there any filter/modulator combinations that should be avoided or require special consideration?
I picked up a new tube preamplifier and SET amplifier and I’m hearing more distortion on some tracks than on my old 45 SET amplifier and I don’t think it’s because I’m overdriving the output tubes. Either it’s a bandwidth issue hitting the transformers or the system is now so revealing that I am using a combination that inherently is causing microstutters on my computer or there’s another issue somewhere in the chain. If I set the volume of the DAC to -25 instead of line I can get it to be not so noticeable, but I can still hear it happening.
I’m running DSD1024, ASDMEC7v3 and closed-form-16m, DAC Correction on 60kHz, DAC200 set to NOS1. Computer is 14900KS, 48GB,RTX5090 CUDA Offloading and I have Process Lasso only using P Cores with HQPlayer. It’s running on Windows 11. Holo Red as NAA with HQPlayer image loaded.
I have a setup similar to yours apart from the valves. 14900 processor, 64gb RAM, 5080 card for CUDA offload. Same DAC correction. The biggest software difference is that I use Linux for the server and an Intel NUC NAA. I have not experienced any problems with DSD1024 using the closed form filter with ASDM7ECv3 . But I have recently moved to DSD512 and poly-sinc-gauss-long using Qobuz or my own local NAS and not experienced any issues with that combination either. I’ve purposely stayed away from Windows 11. I am using HQPlayer version 5.
Changing anything on the DAC200 made no difference to me.
Do you get the same issue streaming DSD128 or 256?
I had the DAC200 connected to a T+A S10 power amplifier, which has 4 tubes at the input and otherwise delivers 35W per channel in Class A (switchable to Class A/B), and there was never even the slightest bit of noise. Absolutely no distortion.
The DAC200 ran at a 120 kHz bandwidth because the power amplifier supports it. Jussi recommends 60 kHz, but I preferred it this way. I have a DIY fanless audio PC with an Intel i9-14900KS CPU. No GPU. The connection was made either via NAA (DietPi Nano3) or via USB using an Intona USB isolator.
To narrow down the problem, I’d play back PCM and also try removing Holo Red from the setup. The fewer components in the chain, the easier it is to troubleshoot.
@hal9001 That’s tonight’s project. I have to drop things down and see.
I don’t know if the Holo Red is causing the bottleneck. I have read conflicting things. I think someone had said that at DSD1024 it converts to DSD over PCM with the DAC200, but I’m not sure.
I have an etherregen in line, too. It’s passing the signal but maybe it’s also a point of bottlenecking everything. I can go optical Ethernet to eliminate the possibility of the line carrying noise, but last time I tried it I didn’t like the way it sounded. Things got very thin. But for diagnostic purposes I’ll put it in and use a lower sample rate.
@StreamFidelity I was afraid to run things at the wide setting because I didn’t want to oversaturate the input transformer of the amp. I’ll have to ask Aric if it can handle the bandwidth. I would need to get a really long usb cable to connect my computer to the DAC200, it’s in a different room, but it may be worth a try just for diagnostics.
This will cause performance issues with HQPlayer. HQPlayer does it’s own core allocations, and trying to affect those from outside will make things misbehave in HQPlayer.
No, it doesn’t.
Are you using RedOS or NAA OS on it?
I would suggest the newer generation modulators, such as ASDM7EC-fast. You will also get lower CPU load thanks to better optimizations.