Threaded Mode | Linear Mode

dtune · 11-03-2022, 22:07

Simon, I guess persistence pays off. I had not tried 8Player Pro before since I read there is no gapless support. But I tried it anyway hoping my DSF pops would disappear and they did. If you still have those tracks I sent to you try it for yourself!

So for now if I need gapless I'll use mconnect but for DSF files 8Player seems to do the trick.

I'm hopeful the new control app from JPlay (over) due since Christmas will be a complete solution. Why it's been delayed so long nobody knows.

Thanks for your prompt support, Dale

simoncn · 11-03-2022, 23:00

I tried this with 8player Lite and I am still getting the pop. I think this might vary depending on which renderer is being used.

simoncn · 12-03-2022, 00:27

I converted the first file to WAV offline using FFmpeg and played it without the second file. There was a pop at the end of the first file. This seems to point to FFmpeg as the cause of the problem, as you suggested. I will do more investigation tomorrow.

simoncn · 12-03-2022, 14:27

I tried some other DSF files and they don't have this issue. I also noticed that the pop at the end of this track is on the right channel only. Looking at the audio levels in Audacity, the right channel jumps to full volume about 0.005 seconds before the end of the track and stays on full volume until the end.

To check whether this is a problem caused by the FFmpeg converter, I have tried converting the file to WAV using a different converter (DSD Master for Mac). This produces the same issue with the the right channel, so the issue seems to be with the DSF file rather than the converter. The WAV encoding produced by DSD Master is not identical to that produced by FFmpeg, which means that DSD Master is not just using FFmpeg "under the covers."

dtune · 12-03-2022, 20:01

Simon, I have many other dsf files that exhibit the same thing so don't Focus too much on that strange volume change on that particular track. Suffice to say there is some weird interaction between controllers and renderers that is not fully understood nor fully dictated by the upnp standard. The difference between my machine in yours is significant I'm sure since I'm using a very stripped-down version of Windows 10 ltsc and running Jplay Femto at a high resolution setting with USB reclocking. Even if you ran JPlay as a renderer our systems would still not be the same exactly.

dtune · 12-03-2022, 20:03

I definitely suspect ffmpeg to be doing things we don't particularly understand either. That is why I was hoping to find an alternative transcode to compare.

simoncn · (This post was last modified: 12-03-2022 21:43 by simoncn.)

I have found an alternative converter (DSD Master) and it behaves similarly to FFmpeg, at least on this test file.

Using a converter is a valid comparison, as the renderer receives exactly the same audio data.

I agree that there could be many factors involved in this, some of which I can reproduce here and some that I can't.