Flac vs Wav audio quality

Just a note - all DLNA renderers must be able to play back 16-bit linear PCM as a requirement, so I'd be careful not to paint all UPnP renderers with the same stripe.

If there is a difference in sound quality between FLAC (converted to PCM data) and WAV playback, that's a hardware-design trait. Some players do not exhibit this behavior, though that does not make them broken or compromised at all.

I know no players which sound worse with uncompressed PCM data, except in situations where the network bandwidth is marginally insufficient to properly fill the buffer (at which point uncompressed hi-res is necessarily out of the question).

In cases such as these, FLAC encoding may actually function better where the FLAC decoding takes place on the player itself, provided that there is just enough network bandwidth to keep the buffer filled. That said - that is a case of functionality, and the power supply noise floor is an unmovable constraint.

Mike , very good

sjbabbey, this is interesting, I wonder if this is from NASs very close to the Naim equipment.

Of course when media is streamed, the file isn't. Its the actual media payload [a subset]  that is transferred across the network. When this media is transcoded from a system and network level the data is identical - I have measured and analysed this. (using Asset as UPnP server and transcoder and Wireshark)

However if you look at disk I/o and I/o load on a NAS and on UPnP server, the load is higher when reading WAV and streaming WAV PCM  than reading FLAC and streaming FLAC encoded media. Again I have measured this. (using Asset on a Linux box using SMB mounts)

So if the NAS/UPnP was closely coupled to the Naim audio equipment - perhaps through powersupply or radiated RFI then from system theory I can see that cross talk from the NAS may provide a minute impact on the Naim audio equipment. This is possible, although I feel unlikely - and I would suggest if this is happening reading FLAC and sending PCM should have a less sonic footprint than reading WAV and sending PCM. So transcoded FLAC should sound 'better'

For my sins I tend to regrettably hear things that most others don't (if I trust that people really cant hear what they say they can't hear that I read on this forum and elsewhere) - its really annoying. For what its worth I cant hear any difference between WAV->PCM vs  transcoded FLAC-> PCM - but my UPnP server and NAS are quite separate and in a different room from my audio equipment.

Simon

Simon, my testing was done quite a long time ago and before I upgraded from my ND5 to NDS. However, I can say that the NAS drive I used then (a WD MBL) was remote (upstairs)  from the streamer.

I mentioned the 'perceived' difference in loudness since I know this can have an influence on how people rate different hifi equipment during comparitive demos.

Hi - yes if there is a perceptional difference in loudness - then the transcoding process you used was  somehow modifying the audio data itself  - then clearly all bets are off - as the data would be very different

Simon

I think it cost about 100 times more than the marker pen cost without the new label!

Err... So players can happily unpack FLAC (or even mp3) and play the resultant PCM stream correctly but they cannot correctly play an uncompressed WAV stream properly?  That makes sense NOT!

Or is that not what you meant?

Please note that my use of the word "pathetic" was directed toward the content of a post you made, not you personally.  However, it did rub me up the wrong way for some reason; I am not usually quite as reactionary in my replies.  I therefore apologize.  

Hi Alan

I am actually just using a PC as a streamer in to a DAC-V1.  This is in a main system though, not a desktop-type arrangement.  The results are very good in my view, but primarily versatile.  It is therefore quite a different animal to a Naim ND/HDX---- systems.  

When I'm talking about decoding on the fly, I simply mean where FLAC files are played directly via media software, and real-time decoded into a PCM audio stream that should match the original file, i.e WAV, CD audio etc.  

Today I took the Naim test files for the DAC-V1, and did the obvious experiment of converting them to FLAC from the original WAV format, and playing them via my usual bit-perfect media player, and let the DAC run it's test routine on them.  Guess what?  The results are the same with FLAC compression level 5 decoded on the fly, as the original WAV file, i,e both are announced to have generated no errors and are bit-perfect.

I think the reason I came back to this thread to announce my recent findings, and the reason I then bothered to conduct the experiment above, is because the idea of there being any audible difference between the two does not really make sense.  Yet, my ears disagree.  At least for the moment, and I reserve the right to change my mind at any time.....as I often do.

Having a morning full of nothing-in-particular to do, and inspired by Songstream (whilst nodding apologetically towards any Ena Sharples channelling scientists) I conducted my own eyes wide open experiment (I failed to persuade either of the cats to swap media while I closed my eyes).

Same piece of music played in Wave, FLAC 0 and FLAC 5 formats, with a slight preference for the Wave file over the two identical-sounding FLACs. Then transcoded both the FLACs and compared with the Wave with no discernible difference.

On one level I'm surprised, as I thought they'd be identical, but on two other levels, not: 1) Nothing should surprise me, as in the past I thought sound improvements through mains leads etc. was utter bullsh1t 2) I'd pretty much been told there'd be a difference and it was a pretty crap test

I am with you on the cable malarkey Gary.  But with analog there is a valid reason for paying attention to cables.  That is, of course, that an analog is what it is, all we do is amplify (maybe) and then get to drive our transducers (speakers).  So it is easy to get people to spend vast sums of money on interconnects, speaker and even mains cables.  Silver mains cable at $2,000 anyone?

Digital is so very different and we should not get hung up on the things we might have worried about in an analog World.  It's sample and hold man!

I think that you and I having disbelieving attitude towards a lot of tis stuff has got to help - be a doubting Thomas all the time, it makes sense.  Which is probably more than this post does.

ps I mentioned Magic Marker pens sold at vastly inflated prices earlier but I forgot to mention my favourite.  It was the Japanese Magic Wood - oh yes step up folks and buy a few blocks of this wood at £80 a pop, spread them round your listening room and totally transform your HiFi.

Is there any other industry that has such bull%*^%t in it?

I'm a green marker pen-wielding (until I got a CD2), blindingly expensive cable-buying loon myself. 

Be be interesting to see how the wind blows re: "it can't take that long to burn in" and "a wire costing that much can't be worth it" comments that us Sarum buyers saw, when the Lumina cables become more common currency.

I'd try them myself if I had any money left!

and my favourite :

I even painted the inside of the draw green in my Marantz KISig CD Player many many years ago. :-D  That's actually true!  And I also buy silly expensive cables, mains, digital the lot.  I'm also a little doubtful of the sonic benefits to some, but if you buy an 'audio grade' USB cable for example, at the very least they tend to be very well made, fit properly in the sockets etc, and this is not always so with cheaper alternatives.  Some of my friends laugh themselves to tears over barmy ideas I might implement and purchases I make, believing of course no difference could possibly result, and they are probably right at least occasionally, but my reply is always the same.  It's better to be safe.

Don't the top ad and the third one down have their slogans transposed?

G

There may be a load of bull in this industry and because of this the people who get caught up in it tend to have a great sense of humour.  As you guys just proved!

@Songstream

Thanks for your note.  Much appreciated.  Sometimes we don't quite say quite what we mean to say or hear what we were meant to hear. 

Back to the discussion

JSH