Sound quality comparison

[DavidHB]

(25-08-2014 09:08)best Wrote:  These days when I get a new disc I am in two minds : flac or wav? I then go with wav, my logic being it should sound better, as there is less 'manipulation' of the original that way. Of course the advantage of flac being a smaller size.

My decision was to rip music files to, and store them as, FLAC, and to transcode to WAV24 on playback. This has worked well for me. Here's the logic underlying my decision.

All formats represent some form of encoding, and all lossless formats are capable of being converted back to the 'original' digital encoding of the analogue sound pattern. It is therefore probably not appropriate to argue that any particular form of lossless encoding (whether compressed or not) is 'purer' or 'more original' than another. Where the WAV format seems to win out (based on listening experiences quite widely reported in this forum and elsewhere) is that the player has to do less work than with the compressed format to convert the digital stream back to an analogue signal. But, from the player's perspective, the stream from a .wav file and the corresponding wave transcoding of (say) a compressed FLAC file are the same thing, if the server and transcoder are doing their respective jobs properly. We should not therefore be surprised that there is, reportedly, no difference in SQ between the 'original' .wav file and its transcoded counterpart.

Hard disk space these days is much cheaper than it once was, but replacing and upgrading data storage is a hassle, and therefore to be done as little as possible. My current NAS has a RAID array with a capacity of just under 2TB, of which my music files (some 700+ CDs) take up about 10%. The NAS is used for a range of storage purposes, so I don't want the music collection to grow completely out of hand. Storing the files as compressed FLAC, rather than .wav or certainly WAV24, saves me a useful amount of space.

It is also useful from my perspective that FLAC uses one of the more straightforward (and consistent) tagging schemes. The situation with tagging wave files seems to have improved of late, but remains more complex; complexity makes incompatibility more likely.

All in all, the case for storing files in lossless compressed formats (in particular, FLAC) is respectable rather than overwhelming. At the same time, the case for sticking with .wav as a storage format does not seem particularly strong when (as with MinimServer/MinimStreamer) good transcoding systems are available. But either option can be made to work.

David

[DavidHB]

(25-08-2014 12:09)JimR Wrote:  I fancy trying this, but I use MinimServer to serve music to multiple devices, some of which can render wav24, whist others cannot. Is there any way to tell MinimStreamer to transcode for certain renderers but not for others?

Simon is the authority, but I'm pretty sure that the answer is 'not at present'.

Simon has been telling us for some time now that the capability to run multiple instances of MinimServer on the same device is on his 'to do' list. If this capability were implemented, separate instances of MinimServer serving the same library could have different settings and you could switch from one server to another depending on which player was in use. How practical this would be would depend to some extent on which control point you are using, but it is perfectly feasible with both Bubble DS and Kinsky, for example.

You can already run multiple instances of MinimServer on different devices (NAS, PC, Linux system or Mac) on the same network, and indeed I do this. However, as MinimServer really needs to work with local storage rather than a network drive, you would also need to mirror your music library to ensure that all your music is available on each server. I already do this (using the Microsoft SyncToy software) for backup and system management reasons, and in fact use transcoding with one of my MinimServer libraries and not with the other. This arrangement works very well.

David

[simoncn]

(25-08-2014 14:05)krutsch Wrote:  OK, I was using just aac:wav24 before, but with m4a files, so that should explain what I was doing wrong. However, when I try to add mp4:wav24, I see this:

Incorrect source value 'mp4'; correct highlighted value and retry

When I remove the mp4 entry, leaving only alac and flac, the error clears and all is well. I am also able to add mp3:wav24 and aac:wav24 just fine.

I've verified that I am running the latest version (update-35/0.8.1). I am also running the latest FFmpeg (2.3.3 for Mac OS).

Are you running MinimStreamer 0.4.1? In previous versions, the 'aac' source type applied to all files containing AAC audio (.aac, .mp4 and .m4a). The 'mp4' source type was introduced in MinimStreamer 0.4.1.

Quote:BTW, I can see how you get your interesting result of AAC better / MP3 worse. What I've learned with my Denon player is that there is wide variation with the quality of decoders in the embedded software and it's possible that your Linn DS is doing additional processing on MP3s that is skipped with AAC and WAV equivalents.

This might well be what is happening.

[simoncn]

(25-08-2014 09:08)best Wrote:  Yes my question was whether a file ripped as an original wav file (16 bit) sounds better vs a flac file transcoded to wav24 in minim. The main difference being ofcourse that a flac file takes up a lot less space on the hard drive than a wav file would. Therefore I would like to come to a definite conclusion, but so far my listening tests are not conclusive as to which one is better. But what I do find is that they do sound different!

These days when I get a new disc I am in two minds : flac or wav? I then go with wav, my logic being it should sound better, as there is less 'manipulation' of the original that way. Ofcourse the advantage of flac being a smaller size.

Please do try and make this comparison and lets hear your feedback!

best

I have noticed a difference between 16-bit WAV and 24-bit WAV. To my ears, 24-bit WAV sounds better. I haven't noticed a difference between playing a WAV file and playing a FLAC file transcoded to WAV.

[simoncn]

(25-08-2014 12:09)JimR Wrote:  I fancy trying this, but I use MinimServer to serve music to multiple devices, some of which can render wav24, whist others cannot. Is there any way to tell MinimStreamer to transcode for certain renderers but not for others?

Jim.

The correct way for a UPnP server to support multiple renderers with different capabilities is for the server to offer multiple streams in different formats to all renderers and let each renderer choose which stream format to play. If the renderer is capable of playing more than one of the formats offered by the server, it should choose the first in the list.

Edit: It is actually the control point that decides which stream format to play, based on the capabilities advertised by the renderer.

For MinimStreamer to support this, the current stream.transcode property syntax would need to be extended to allow a single source type to have multiple target types. This is on my to-do list for a future version of MinimStreamer.

[krutsch]

(25-08-2014 17:34)simoncn Wrote:  Are you running MinimStreamer 0.4.1? In previous versions, the 'aac' source type applied to all files containing AAC audio (.aac, .mp4 and .m4a). The 'mp4' source type was introduced in MinimStreamer 0.4.1.

I am running MinimStreamer 0.4.1

[simoncn]

(25-08-2014 21:40)krutsch Wrote:  I am running MinimStreamer 0.4.1

It looks like there is a bug with this in MinimStreamer 0.4.1. I will investigate tomorrow and post back here.

[simoncn]

(25-08-2014 21:57)simoncn Wrote:  It looks like there is a bug with this in MinimStreamer 0.4.1. I will investigate tomorrow and post back here.

This bug should be fixed in the next version of MinimStreamer.

[simoncn]

(25-08-2014 17:55)simoncn Wrote:  The correct way for a UPnP server to support multiple renderers with different capabilities is for the server to offer multiple streams in different formats to all renderers and let each renderer choose which stream format to play. If the renderer is capable of playing more than one of the formats offered by the server, it should choose the first in the list.

This description isn't quite correct. It's the control point that makes the choice of stream based on information provided by the renderer. There are two potential problems with this:

1) If the renderer doesn't advertise its capabilities correctly, the control point might not choose a stream that the renderer would be capable of playing.

2) The protocol for advertising renderer capabilities doesn't always include all relevant information. For example, a renderer can say whether or not it can play WAV streams but it can't say whether or not it's able to play 24-bit WAV streams.

[krutsch]

(28-08-2014 21:59)simoncn Wrote:  This description isn't quite correct. It's the control point that makes the choice of stream based on information provided by the renderer. There are two potential problems with this:

1) If the renderer doesn't advertise its capabilities correctly, the control point might not choose a stream that the renderer would be capable of playing.

2) The protocol for advertising renderer capabilities doesn't always include all relevant information. For example, a renderer can say whether or not it can play WAV streams but it can't say whether or not it's able to play 24-bit WAV streams.

Wow... that all sounds like quicksand for you as the developer/supporter for MinimServer.

If you want to put more time into transcoding features, maybe think about exposing FFmpeg switches for MinimStreamer via the web-based configurator. For example, I would love to be able to invoke SoX with minimum or linear phased filters and forced up-sampling to best match the characteristics of my DAC.

Just a thought...