AudioEnthusiast
Member
- Joined
- Feb 8, 2020
- Messages
- 55
- Likes
- 44
Have there been any objective measurements of the Roon resampling algorithm, similar to what Archimago did for the macOS Core Audio resampler?
Thanks.
Thanks.
I asked the same question in the ROON forum. No answer. Silence.... I transcode my FLACs to OPES for my phone. OPUS has it´s own resampler inbuilt and I wanted to know if ROON uses the inbuilt OPUS-resampler or its own resampler, or any other third-party resampler.Have there been any objective measurements of the Roon resampling algorithm, similar to what Archimago did for the macOS Core Audio resampler?
Thanks.
I ended up not using the Roon resampler and just letting macOS do the resampling. I figured a tested resampler is more of a known quantity than an untested one. I doubt I could hear a difference anyway, but this keeps the load on Roon low.I asked the same question in the ROON forum. No answer. Silence.... I transcode my FLACs to OPES for my phone. OPUS has it´s own resampler inbuilt and I wanted to know if ROON uses the inbuilt OPUS-resampler or its own resampler, or any other third-party resampler.
Yes, I run a DSP pipeline with Home Audio Fidelity’s room shaper (time domain correction for long duration room reverberations, which can’t be done via convolution). The loopback audio device used to connect Roon’s output to the pipeline cannot be run in exclusive (hog) mode, so Roon cannot set its sample rate. It has to be fixed. I fix it at 96 kHz, since the majority of music I stream from Qobuz happens to be 96 kHz. Anything else has to be resampled. I need a resampler somewhere. I just let macOS Core Audio resample internally, between Roon and the loopback device.Any reason to resample?
The media player I use plays everything at its native sample rate.
I wanted to know if ROON uses the inbuilt OPUS-resampler or its own resampler, or any other third-party resampler.
I‘m generally happy with all things Roon — they’ve clearly put a good architecture and design together — but I was wondering if there was an independent test of their resampler anywhere. As a software developer myself, I’m always wondering if there are (possibly inaudible) bugs. Since Apple’s algorithm has been tested by Archimago, I decided to use that.According to this thread they made their own resampling code, and it looks like they've put a lot of attention into it https://community.roonlabs.com/t/what-library-does-roon-use-for-src-mac/7943.
At a certain point further improvements are inaudible, and I expect them to be at that level (they're using SOX as a benchmark).
At least earlier on Apple Core Audio was not as good as for example SoX. I don´t know if Apple have done any remake of their resampler since then. It is of course transparent anyway, but I don´t think CoreAudio is better than ROONs resampler. I transcode my FLAC files to 48 khz OPUS for smartphone use, but I use dbPoweramp which seems to be among the best resamplers.I ended up not using the Roon resampler and just letting macOS do the resampling. I figured a tested resampler is more of a known quantity than an untested one. I doubt I could hear a difference anyway, but this keeps the load on Roon low.
Thanks for showing those comparisons. It’s exactly that kind of comparison between Core Audio and Roon that I’m looking for. It’s entirely possible that Roon is as good or better as SoX, and therefore ”better” (at an inaudible level) than Core Audio, but in the absence of data showing that, I can’t rule out the possibility that there’s a coding error in Roon making their implementation worse. I generally don’t accept manufacturer claims about performance without independent verification, and unfortunately I lack the equipment to verify performance myself.At least earlier on Apple Core Audio was not as good as for example SoX. I don´t know if Apple have done any remake of their resampler since then. It is of course transparent anyway, but I don´t think CoreAudio is better than ROONs resampler. I transcode my FLAC files to 48 khz OPUS for smartphone use, but I use dbPoweramp which seems to be among the best resamplers.
View attachment 359437
Thanks for showing those comparisons. It’s exactly that kind of comparison between Core Audio and Roon that I’m looking for. It’s entirely possible that Roon is as good or better as SoX, and therefore ”better” (at an inaudible level) than Core Audio, but in the absence of data showing that, I can’t rule out the possibility that there’s a coding error in Roon making their implementation worse. I generally don’t accept manufacturer claims about performance without independent verification, and unfortunately I lack the equipment to verify performance myself.
Thanks again, I appreciate the information.
Thanks. I'm not familiar with Deltawave, but I will check it out. In any case, I am happy with the solution I currently have, so it probably won't be a high priority for me.Not a Roon user, but isn't this simple to do? Even if Roon won't resample into a saved file just send the digital out passing to an audio interface with digital in. Record it and dump into Deltawave. You can compare two resamplings one from Roon and one from whatever else as well as the original bitstream prior resampling.
Seems some of those Windows virtual source softwares might even let you do it without leaving the PC.
Not a Roon user, but isn't this simple to do? Even if Roon won't resample into a saved file just send the digital out passing to an audio interface with digital in. Record it and dump into Deltawave. You can compare two resamplings one from Roon and one from whatever else as well as the original bitstream prior resampling.
Seems some of those Windows virtual source softwares might even let you do it without leaving the PC.