Goodbye WAV, hello FLAC

God knows, to my ears its identical.

But wheres the fun of the audiophile without a hierarchy? My ears simply are not golden enough.

Hi Scott.

Two different "theories" at work here.  With your mp3 example, clearly info has been lost when ripping to mp3 and that info cannot be re-created out of thin air by 'converting' to flac.  But with wav and flac, no info has been lost; it's all still there.  BUT the theory goes that the cpu in the player has to work harder to unpack flac, and that that hard work can lead to 'stuff happening' that degrades the sound.

I wasn't intending to open up a WAV/FLAC debate, merely recounting my experience. I can clearly hear the difference, and I have no hearing above 4k, or so the audiologist told me, so I certainly don't have golden ears.

My experience is the same as HH -- I cannot hear any difference between wav rips and flac transcoded to wav.  So I'm happy with all flacs in my MQ folder!

Backup your WAV library first, then change the transcode settings (as this may take a while).  You then have the luxury of a WAV library backup from which you can easily compare, in real time, FLAC transcoded to WAV vs native WAV from the UnitiServe.

Different streamers provide different results (I've had the luxury of a few manufacturers' models 'round these parts lately); if the transcoded FLAC sounds perfectly acceptable, it may end up being your best bet from an archival perspective.  As always, your mileage may vary.

I'm planning on using my transcoded FLACs with Asset UPnP and / or MiniServer on a spare NAS or two; that said, I have no idea yet as to the FLAC compression level used by the UnitiServe (which still rips to WAV as an intermediate step, then transcodes to FLAC)–I'm assuming FLAC Level 7 or 5.  (I may choose to transcode this to uncompressed FLAC as well, just to see....)

Well, I'm WAVless. The master copy in the Serve, the backup and the Raid copy are all FLAC now, apart from the 10 that got away. 

Hi I just use a PC and a nas drive to flac saves me a lot of money sounds fine regards.

I transcode with LMS on the Vortexbox from Flac (level 5) to PCM on my UQ. PCM sounds better, but it's subtle. But considering the gobs of $ people spend for subtle upgrades, checking a box on the server cost absolutely nothing.

Yep, I transcode on the fly from FLAC to WAV here, and sounds identical to streaming WAV. In fact I did a Wireshark trace sometime back, and could see the WAV data being transferred by the UPnP server whether transcoded or played natively indeed was identical... Well at least for the first sequence of packets.

indeed it is the processing of the FLAC data by the network player to extract the PCM that does indeed cause the sonic side effect. I can also hear a less obvious difference between different UPnP servers that use different TCP window sizes, and therefore it would appear a  different frequency of ACK processing. 

So far I have found the Unitiserve and the Raspberry Pi (subjectively) sound best.

Simon

Scott, in my experience (and seemingly of the experiences of several others on this forum) then yes.

It does cost nothing to transcode, assuming your UPnP server supports it, so if it makes a positive difference then great... If not you can switch transcoding back off.

Simon

Simon - you are losing me...... What are PCM, TCP and ACK? My mum used to put TCP on my cuts and grazes when I was small, but somehow I suspect it's not that.....

HH -

TCP stands for Transmission Control Protocol.  It is one of the core protocols used for network communications, and is responsible for things like flow control and error checking.  It's actually part of a seven layer cake known as the Open Systems Interconnection model, where it assumes responsibility for layer three processing.  Layer one is the physical interconnect, layer seven is the human interface, and in-between is all the code that allows your US to talk to your SuperUniti.

It's called an ACK protocol because part of it's bi-directional processing is the positive acknowledgement of data frames that arrive without errors.  Other protocols are silent in this regard, and only negatively acknowledge problems (and are called NACK protocols). TCP has to deal with the chaos of layer two processing, where frames arrive at varying rates and even occasionally out of order.  Basically, it makes our network connections...reliable.

Simon has long posited that by correctly setting the frame size for TCP, network processing can be more efficient.  When network cards do less work, they generate less noise, and they have the potential for producing better sound. 

Or something like that.  

Am sure 'da man will be along shortly to explain further!

ATB.

Hook

PS - And oh yeah, under the logical wrappers of WAV or FLAC lie musical data.  That data is in a format called PCM, or Pulse-Code Modulation.  It's a method used to digitally represent sampled audio signals.

Hook- thanks - perfect reply.

Simon

The student blushes at the praise of the teacher...

Thanks Simon!

In the White Paper on the NDX from 2010, Naim writes:
"Naim’s UPnP™ servers deliver the uncompressed audio data ripped from CD using the Naim ripping engine to ensure the best quality reproduction. Uncompressed audio data will always give better results than compressed. Even lossless compression may not reproduce audio with equivalent quality to the uncompressed original as the processing required to uncompress the data increases the computational load. This raises the power supply noise floor, which detracts from the sound quality."

I didn't know when Simon has become a professor in audio. 

The reason WAV sounds better than FLAC, even if it's FLAC converted to WAV, is the same reason as music sounds better if your amp sits on a fraim, as opposed to being on a shelf from pottery barn.

Greg,

Choice between Fraim vs Pottery Barn can be a matter of cost. Doesn't really apply with WAV vs FLAC does it?

I've footered about with 'My Book Live' but cannot fathom this 'transcoding' lark at all.

hmmmm...

G

Now there is a thought, perhaps time for a career change.....