Minimserver users...transcode FLAC to WAV?

[tboooe]

Just curious if Minimserver users are transcoding FLAC to WAV or just leaving it alone. Is there any sonic benefits to WAV before sending the file to the renderer (in my case Jplay). I know there are many many debates on the possible sonic differences separate FLAC and WAV files of the same song but I am asking about the benefits of transcoding a FLAC file to WAV. If you are transcoding FLAC to WAV are you changing the bit depth and upsampling as well? I am just getting started with Minimserver and messing around with the transcoding settings.

[Skywatcher]

I'd say it depends on the system.

 

Simon (minimserver/minimstreamer programmer) tested it on his system and he sees an improvement on transcoding FLAC to Wav first and then sending the Wav to the DAC. He stated that in a thread at Minimserver.

 

Obviously I take his word for it, but in my system I have the opposite result. The difference is immediately noticeable on the bass that dies a bit.

 

I guess you'll have to check with minimstreamer transcoding on and off and check how your system responds to it.

 

Cheers,

[Audio_ELF]

I'd say it depends on the system.

Could you tell us what your system is?

[Skywatcher]

Netgear NAS (w/ MinimServer)-> CA UPnP/OH renderer (BeagleBone Black) -> Emotiva Stealth DC-1 DAC -> Sennheiser HD 650.

 

Controlled via BubbleUPnP.

[lmitche]

Transcoding Flac to Wav 24 improves SQ.

[audiventory]

Just curious if Minimserver users are transcoding FLAC to WAV or just leaving it alone. Is there any sonic benefits to WAV before sending the file to the renderer (in my case Jplay). I know there are many many debates on the possible sonic differences separate FLAC and WAV files of the same song but I am asking about the benefits of transcoding a FLAC file to WAV. If you are transcoding FLAC to WAV are you changing the bit depth and upsampling as well? I am just getting started with Minimserver and messing around with the transcoding settings.

 

In binary form no difference between FLAC and WAV. However some people listen difference.

 

"There are difference or not" know exactly developers of playback software and hardware only.

 

Transcoding with increasing of resolution don't improve quality of file. However offline transcoding to "native" resolution of DAC (without inline resampling and re-bit depth) can sometimes improve playback quality.

[tranz]

In binary form no difference between FLAC and WAV. However some people listen difference.

 

"There are difference or not" know exactly developers of playback software and hardware only.

 

The difference is due to having another process, the FLAC converter, running in the background. Better to have the files native for optimal audio.

[nyatousga]

Copy:running in the background. Better to have the files native for optimal audio.

[audiventory]

Hi Tranz,

 

From point of view audio player's developers I don't see any reason apply different processing for any lossless file type.

 

Especially for bit perfect playback.

 

Possible bugs in some versions of FLAC library. I suppose, all developers use standard FLAC library. Different versions, of course.

 

I checked last (or almost last) version in my audio converter software: it give absolutelly idendical binary after decoding like source (coded) WAV.

 

In this case DAC's get identical audio stream from FLAC and WAV.

 

Possible developers of audio player software apply some additional processing. However, I can't understand: why need process FLAC and WAV differently? There are exists the processing or not in real software?

 

Unambiguously (without knowledges what inside of player) get answer possibly by bit prefect capturing input of DAC.

 

Or connect audio player to virtual audio device. And capture stream from the device.

If streams by WAV and FLAC will identical - no real sound difference.

 

In this case DAC must work also fully identically, isn't it?

 

Best regards,

Yuri

[tranz]

Hi Yuri,

 

Having the Flac converter processes run in the background causes more electrical pulses and more read and write activity. I prefer uncompressed. With disk space prices these days, there is no reason to compress.

 

On my setup even cutting down on processes using the CAD script helped. But every setup is different, so it is best to test it out for yourself.

Granted this is not heard on all systems hence the sometimes heated arguments.

 

Converting Flac to Aif and back makes no difference to the file itself, you are correct. Many years ago I transferred back and forth dozens of times with no resulting change just to test it out.

 

Please keep in mind that a Flac file will always need to be un-FLACed before it is sent to a DAC. Just like you always need to unzip a file to read it. Now you can even have uncompressed Flac, but it will still require the extra process to un-Flac it causing the same potential audio impact.

 

I do use Flac in the CD ripping stage as it captures the meta data. Then I translate it to native PCM when adding it to the library. This way no Flac decoder process is running during playback.

 

There is another recent thread discussing this too.

 

http://www.computeraudiophile.com/f10-music-servers/do-music-servers-improve-sq-over-spinning-discs-24067/

 

Cheers

[Music Matters]

 

I do use Flac in the CD ripping stage as it captures the meta data. Then I translate it to native PCM when adding it to the library.

 

Cheers

 

There is no Flac in ripping stage, because you can extract audio only in WAV format. So you have WAV -> FLAC -> WAV.

[tranz]

There is no Flac in ripping stage, because you can extract audio only in WAV format. So you have WAV -> FLAC -> WAV.

 

Bit pedantic. To capture and add metadata easily during the ripping process, it saves me time using EAC to capture this and have it saved in the Flac files. Of course it is just a wrapper/zip file packaging of the ripped WAV. I thought that was obvious.

[Pepsican]

The difference is due to having another process, the FLAC converter, running in the background. Better to have the files native for optimal audio.

 

WAV needs decoding too as the metadata needs to be stripped before you end up with pure LPCM. And there are manufacturers that actually advice to use FLAC over WAV. Linn Audio does that. I would say it is more a matter of the file format used when equipement was developed if there is a difference at all.

[Pepsican]

Hi Yuri,

 

Having the Flac converter processes run in the background causes more electrical pulses and more read and write activity. I prefer uncompressed. With disk space prices these days, there is no reason to compress.

 

You can have uncompressed FLAC. Just like you can have compressed WAV (WAV is a container, not a format, just like FLAC).

 

When you compare uncompressed FLAC to WAV, it is both LPCM. The only difference is the blocking in FLAC.

[Music Matters]

Bit pedantic. To capture and add metadata easily during the ripping process, it saves me time using EAC to capture this and have it saved in the Flac files. Of course it is just a wrapper/zip file packaging of the ripped WAV. I thought that was obvious.

 

I don't remember how it's with EAC, but I use dBPoweramp, and you can skip the flac and save time. As you said that you are using PCM with dBPoweramp you can rip to WAV and have the tags written into the WAVs.

[wanta911]

This is interesting. I notice a subtle but very real difference when playing music through my Auralic Aries between the following two options:

 

1. FLAC files from a directly connected 128GB USB stick

2. FLAC files transcoded to WAV24 via Minimserver from my NAS (WiFi to Aries)

 

I'm not sure if the difference is the transcoding or just the different delivery methods? I guess I'll test it out by disabling the transcoding - one day when I'm motivated.

 

The USB route is crisp and sharp and via Minimserver transcoding it is a touch softer and smooth in comparison, it is particularly noticeable on drums. I am not sure if I have a preference and it probably depends on the particular music. The difference I hear is similar to how people describe the difference between PCM & DSD.

[tboooe]

This is interesting. I notice a subtle but very real difference when playing music through my Auralic Aries between the following two options:

 

1. FLAC files from a directly connected 128GB USB stick

2. FLAC files transcoded to WAV24 via Minimserver from my NAS (WiFi to Aries)

 

I'm not sure if the difference is the transcoding or just the different delivery methods? I guess I'll test it out by disabling the transcoding - one day when I'm motivated.

 

The USB route is crisp and sharp and via Minimserver transcoding it is a touch softer and smooth in comparison, it is particularly noticeable on drums. I am not sure if I have a preference and it probably depends on the particular music. The difference I hear is similar to how people describe the difference between PCM & DSD.

Interesting observations! I am in the process of testing out the different delivery methods. Right now, my music all reside on a NAS. I have a separate PC connected to my wireless router (the NAS to also connected to the same router) running Minimserver. I am going to see if it sounds any different with my music stored on a USB drive connected directly to this PC. According to the developer of Minimserver:

 

"The SMB/CIFS protocol for network file access adds a large overhead to MinimServer scanning performance. This might be related to the way Java implements directory scanning and file I/O. There should be no effect on sound quality when playing a file."

 

 

[Cebolla]

Interesting observations! I am in the process of testing out the different delivery methods. Right now, my music all reside on a NAS. I have a separate PC connected to my wireless router (the NAS to also connected to the same router) running Minimserver. I am going to see if it sounds any different with my music stored on a USB drive connected directly to this PC. According to the developer of Minimserver:

 

"The SMB/CIFS protocol for network file access adds a large overhead to MinimServer scanning performance. This might be related to the way Java implements directory scanning and file I/O. There should be no effect on sound quality when playing a file."

Presumably MinimServer's developer said that with reference to your far less usual situation of running MinimServer on a computer to use the music files stored on a separate network device. This would not apply to most users of MinimServer who for example run it on the NAS itself, ie, MinimServer runs on the same device where the music files it uses are stored.

[sydlow]

...

2. FLAC files transcoded to WAV24 via Minimserver from my NAS (WiFi to Aries)

...

 

Quick question,

Does your Lightning DS show 16 or 24bit when it's transcoding.

Mine still shows 16bit as per screen below:

[Distinctive]

The redbook to wav24 in minimserver is just padding zeroes to the LSB's.

[Cebolla]

Nevertheless, the wav file data being received by the Aries should still be 24-bit, assuming that MinimServer's transcoder (MinimStreamer) has been properly set to wav24. It's possible the Lightning DS app might be caching older links to MinimServer's 16-bit transcoder URLs, so will require a rescan of the Lightning DS app to apply MinimServer's latest settings of links to the 24-bit transcoder URLs.

[lmitche]

My minimstreamer settings run with flac to wav24 transcoding and resampling turned with convOut=-af aresample=resampler=soxr:precision=28 in the streamer options string.

 

These settings yield the best SQ to my ears.

[Distinctive]

My minimstreamer settings run with flac to wav24 transcoding and resampling turned with convOut=-af aresample=resampler=soxr:precision=28 in the streamer options string.

 

These settings yield the best SQ to my ears.

Interesting.

Especially intriguing that minim understands emoticons!

[lmitche]

Oops, lets try again: convOut=-af aresample=resampler=soxr{semI-colon}precision=28. You will need the ffmpeg-full library for this to work.

 

Damn emoticons!

[Distinctive]

Probably not a coincidence in this case that the emoticon was 'Stick Out Tongue'....