Threaded Mode | Linear Mode

simoncn · (This post was last modified: 10-01-2016 09:56 by simoncn.)

(10-01-2016 01:47)tboooe Wrote: Right but according to my tests, this speed by running Minimserver on the nas comes at the expense of sound quality.

I have been thinking about this. In your previous posts about this, I thought the comparison was between (1) playing music from the NAS via MinimServer and (2) playing music locally from the PC. From what you have said in this thread, it seems that in case (2) you are not playing music locally from the PC but you are running MinimServer on the PC and playing music remotely from the NAS. Given this, I am not sure why there is a difference in sound quality because both cases involve streaming music from the NAS to play it on the PC.

Mihaylov Wrote:OK. Then use at least gigabit interfaces in the local network.

I don't think this would make much difference. I have the impression that the overhead with remote scanning is somewhere in the SMB protocol (or the way it is used by Java to scan directories and read file attributes) rather than in network data transfer speed. If anyone has has done a comparison of remote scanning using different network speeds, I would be interested to know the results.

Mihaylov · 10-01-2016, 14:48

(10-01-2016 09:45)simoncn Wrote:
(10-01-2016 01:47)tboooe Wrote: Right but according to my tests, this speed by running Minimserver on the nas comes at the expense of sound quality.

I have been thinking about this. In your previous posts about this, I thought the comparison was between (1) playing music from the NAS via MinimServer and (2) playing music locally from the PC. From what you have said in this thread, it seems that in case (2) you are not playing music locally from the PC but you are running MinimServer on the PC and playing music remotely from the NAS. Given this, I am not sure why there is a difference in sound quality because both cases involve streaming music from the NAS to play it on the PC.

I think tboooe means that ethernet interface between file storage (hdd) in NAS and MinimSerever in streamer have advantage over ethernet interface between MinimServer in NAS and streamer in SQ. In my opinion there is a sense in this.

simoncn · 10-01-2016, 16:32

(10-01-2016 14:48)Mihaylov Wrote: I think tboooe means that ethernet interface between file storage (hdd) in NAS and MinimSerever in streamer have advantage over ethernet interface between MinimServer in NAS and streamer in SQ. In my opinion there is a sense in this.

Do you have any ideas why SQ might be better when streaming from the NAS using SMB/CIFS than when streaming from the NAS using HTTP?

tboooe · 10-01-2016, 17:11

Just for clarity here is my hardware and network setup

NAS>>Wireless router-----Wireless bridge>>PC#1 (Audiophile Optimizer, Jplay, Fidelizer) >>PC#2 (Audiophile Optimizer, Jplay, Fidelizer)

>> = wired cat 6 cable
--- = wireless AC

All of my music is on the NAS. I've been testing running Minimsever from either the NAS or PC#1.

Minimserver on NAS = music has more detail and more dynamics but sound is thin and a bit bright.

Minimserver on PC#1 = music is full bodied and warm

Perhaps the difference in sound has to do with the transcoding I am doing to Wav24? I am using different versions of ffmpeg on the NAS and PC#1. On the NAS I cannot use libsoxr since that version is not supported by ARM processors.

simoncn · (This post was last modified: 10-01-2016 17:42 by simoncn.)

(10-01-2016 17:11)tboooe Wrote: Just for clarity here is my hardware and network setup

NAS>>Wireless router-----Wireless bridge>>PC#1 (Audiophile Optimizer, Jplay, Fidelizer) >>PC#2 (Audiophile Optimizer, Jplay, Fidelizer)

>> = wired cat 6 cable
--- = wireless AC

All of my music is on the NAS. I've been testing running Minimsever from either the NAS or PC#1.

Thanks for clarifying this.

Quote:Minimserver on NAS = music has more detail and more dynamics but sound is thin and a bit bright.

Minimserver on PC#1 = music is full bodied and warm

Perhaps the difference in sound has to do with the transcoding I am doing to Wav24? I am using different versions of ffmpeg on the NAS and PC#1. On the NAS I cannot use libsoxr since that version is not supported by ARM processors.

This is quite likely. It you are doing transcoding resampling without libsoxr, sound quality will be adversely affected.

The resampling transcoding setting in MinimStreamer is intended only to enable you to play files that would otherwsie not be playable. For example, you might have a renderer that is limited to 24/96 and you might have some 24/192 files. Any other use of resampling is very unlikely to improve sound quality and is more likely to reduce it.

What settings do you have for the stream.transcode property on the NAS and PC#1? What format (e.g., FLAC, ALAC, DSF) are the music files that you used for the comparison and what is the original resolution of these music files (e.g. 24/96 or 16/44.1)?

tboooe · (This post was last modified: 10-01-2016 17:51 by tboooe.)

Simon, all files I am using for comparison is flac 16/44.1. I am transcoding everything to wav24 with no upsampling. I have read (and you even recommend) transcoding to wav24. In addition to that, on my PC#1, I am also using some convout settings as described in the Resampling thread.

So based on what you are saying I should perhaps disable transcoding entirely on the NAS since I can't utilize a binary of ffmpeg with libsoxr?

simoncn · 10-01-2016, 18:18

(10-01-2016 17:51)tboooe Wrote: Simon, all files I am using for comparison is flac 16/44.1. I am transcoding everything to wav24 with no upsampling. I have read (and you even recommend) transcoding to wav24. In addition to that, on my PC#1, I am also using some convout settings as described in the Resampling thread.

So based on what you are saying I should perhaps disable transcoding entirely on the NAS since I can't utilize a binary of ffmpeg with libsoxr?

Please post the exact settings for stream.transcode (including all punctuation) that you are using on both the NAS and PC. Thanks!

tboooe · 10-01-2016, 18:45

(10-01-2016 18:18)simoncn Wrote: Please post the exact settings for stream.transcode (including all punctuation) that you are using on both the NAS and PC. Thanks!

This is exactly what I have on both NAS and PC.

Flac:wav24, mp3:wav24

Since I use jplaystreamer I have to transcode to at least wav.

Mihaylov · (This post was last modified: 10-01-2016 18:53 by Mihaylov.)

(10-01-2016 17:11)tboooe Wrote: Just for clarity here is my hardware and network setup

NAS>>Wireless router-----Wireless bridge>>PC#1 (Audiophile Optimizer, Jplay, Fidelizer) >>PC#2 (Audiophile Optimizer, Jplay, Fidelizer)

>> = wired cat 6 cable
--- = wireless AC

All of my music is on the NAS. I've been testing running Minimsever from either the NAS or PC#1.

Minimserver on NAS = music has more detail and more dynamics but sound is thin and a bit bright.

Minimserver on PC#1 = music is full bodied and warm

Perhaps the difference in sound has to do with the transcoding I am doing to Wav24? I am using different versions of ffmpeg on the NAS and PC#1. On the NAS I cannot use libsoxr since that version is not supported by ARM processors.

PC#1 have the soft network bridge?

tboooe · 10-01-2016, 18:59

(10-01-2016 18:53)Mihaylov Wrote: PC#1 have the soft network bridge?

Hi I am not sure what you mean by soft network bridge. PC#1 it's hard wired to a Dlink wireless bridge.