WAV vs FLAC vs AIFF

I've listened at length and found I couldn't tell the difference between WAV and AIFF... they were the best. Followed by ALAC and finally FLAC. FLAC just sounded flat in comparison to WAV and AIFF for me. 

At least there is other person out there than thinks FLAC should be called FLAT. My girlfriend screwed up face up when I switched back to FLAC from WAV. Lifeless. 

I'm not evening talking hifi, just musically inept sounding.

AIFF I found a small difference on my Mac with B&W desktop actives, so not the best set up. WAV was move fluid though to out eara even then. 

Anyway, the Zennith folders are easy to access over local network via a computer so all is good. Drag and drop into music folder. 

Can't help but think Innuos are underselling their product with all this FLAC malarky.

Apparantly a firmware update in 10 days will enable direct WAV ripping from CD. Doesn't matter really with remote folder access but would be nice to have proper album artwork, and not .wav at the end of track names. But it's actually fine this way anyway. All in order, under the right album folder name.

Using it with iPeng is a breeze too.

The difference in sound between those lossless PCM file formats will be, for the same source media, entirely down to the renderer. As said elsewhere closed system inter modulation is real world physics and engineering.. and a manufacturer that denies that  I would be wary of, as they clearly have a lack of understanding that could manifest itself with some undesirable outcome in that product like a poor firmware upgrade at some point.

I suspect that they are aware of such things however and would have optmized their development through listening tests using WAV files. Yes a better system should approach zero audible differences in rendering these formats.. but I have yet to hear that anywhere for a closed system

The differences between WAV and AIFF on a quality system really should be minimal however as the only difference between them as far as the encoded media is concerned is the sample word bit ordering.

Simon-in-Suffolk posted:

The difference in sound between those lossless PCM file formats will be, for the same source media, entirely down to the renderer. As said elsewhere closed system inter modulation is real world physics and engineering.. and a manufacturer that denies that  I would be wary of, as they clearly have a lack of understanding that could manifest itself with some undesirable outcome in that product like a poor firmware upgrade at some point.

I suspect that they are aware of such things however and would have optmized their development through listening tests using WAV files. Yes a better system should approach zero audible differences in rendering these formats.. but I have yet to hear that anywhere for a closed system

Hi Simon,

Question. Could the sound of say, a WAV file, vary depending on the program or device it was ripped from CD with?

Assume the rip settings are equal. Sample rate etc.

Ie

WAV itunes rip from CD

Vs

Music server (Like the zennith) rip from same CD

I think it was yourself that mentioned the Micro Rendu? I think I will still go down this route before committing to 2k server. 

Hi, assuming the ripper is not corrupting or altering the sound there will should be no difference.

just a point to stop an unnecessary rabbit hole being followed, different rippers can encode a wav file in one of two ways for Redbook PCM, and this results in different header formats being used, but the resultant media stereo sample data is identically formatted. From memory iTunes does encode WAV with a different header construct than the rippers I am aware of that insert metadata.

iTunes on PC used to also some sometimes have offset errors when ripping and this was dependent on CD reader used, so the first few milliseconds of a rip were absent.. but I'd be suprised if this is not resolved now.

Simon

I haven't found characterized and reproducible differences between WAV and AIFF files.

I have however found notable differences between WAV 192 and DSD64 files via NDS 4.3 update and preferred PCM over DSD in this particular configuration where the DSD64 files are in fact, transferred to and converted to analog as PCM files.

While slightly more airy and articulate in the lower spectrum, the Channels Classics downloaded Podger's L'Estro Armonico DSD64 file is less cohesive and engaging than the Qobuz downloaded WAV 192kHz file (Jared Sacks offers only FLAC versions for download).

Chag -

Simon-in-Suffolk posted:

Hi, assuming the ripper is not corrupting or altering the sound there will should be no difference.

just a point to stop an unnecessary rabbit hole being followed, different rippers can encode a wav file in one of two ways for Redbook PCM, and this results in different header formats being used, but the resultant media stereo sample data is identically formatted. From memory iTunes does encode WAV with a different header construct than the rippers I am aware of that insert metadata.

iTunes on PC used to also some sometimes have offset errors when ripping and this was dependent on CD reader used, so the first few milliseconds of a rip were absent.. but I'd be suprised if this is not resolved now.

Simon

That's a relief! ��

Interestingly, having never noticed much (any?) difference between FLAC & WAV I now find with NDX TT 250DR I favour WAV.

G

All of our music is in FLAC, transcoded to WAV. I can't tell the difference between transcoded FLAC and native WAV. As I understand it, Naim streamers find it easier to convert WAV, which is why WAV sounds better. 

ThatsNotMyNaim posted:

 

Surely others can hear this? Innuos claim extensive testing found no difference im quality / sound repro of formats.

 

Some will some won't. I do. Which is why I store and play native WAV. 

This is a bit moot really. It is simple to interconvert  formats and play back with assorted transcoding. 

If a  difference  can be heard it doesn't matter why or who can hear it. If you can hear it, it's there.

If  you can't, then all the better. No more mucking about if you own files in different formats.

The versatility and flexibility of streaming is a good thing

Ditto HH,  I posted earlier this week on another thread,  I can't tell the difference between WAV & transcoded FLAC,  except all our music is WAV (& DSD) so I don't have need to transcode.      But I can hear a small difference playing straight FLAC compared to WAV;  I concluded at the time its probably something in the NDX design/process as when I tried the same on a friends Linn DS we could not hear a difference.

Mike-B posted:

I concluded at the time its probably something in the NDX design/process as when I tried the same on a friends Linn DS we could not hear a difference.

It's wider than that. If present it applies to DAC, HDX and NDS also. You might be onto something.

Hungryhalibut posted:

All of our music is in FLAC, transcoded to WAV. I can't tell the difference between transcoded FLAC and native WAV. As I understand it, Naim streamers find it easier to convert WAV, which is why WAV sounds better. 

Nor can I on certain UPnP media servers - but on others I can if the transcoding from FLAC to WAV  is performed in real time and affects the inter frame timing of the of the media transferred on the network. I have started to rip in WAV again and convert back to WAV those albums I listen to a lot. Once converted and stored in WAV I can hear no and measure no difference in the resultant rendered audio.

Simon

I also have an Innous Zenith; all files are flac, so this thread caught my eye.

I have just done a, no doubt, unscientific test comparing a short track from a Linn recording of Mozart’s Requiem in flac, wav and studio master, downloaded from Linn.

The Hi Res might have a slight edge but I honestly could not hear a difference between the wav or flac & I suspect in a blind testing I would not even be able to say which one was the studio master.

I would highly recommend the Zenith, excellent sound & backups could not be easier.

Morton posted:

I also have an Innous Zenith; all files are flac, so this thread caught my eye.

I have just done a, no doubt, unscientific test comparing a short track from a Linn recording of Mozart’s Requiem in flac, wav and studio master, downloaded from Linn.

The Hi Res might have a slight edge but I honestly could not hear a difference between the wav or flac & I suspect in a blind testing I would not even be able to say which one was the studio master.

I would highly recommend the Zenith, excellent sound & backups could not be easier.

Funny really. I am in no two minds. And I reckon in a blind test I could tell the WAV from the FLAC. 

WAVs make my hair stand on end. The music floats more. Smoother and more cohesive. Enough to feel it.

I used Enya - Only Time as a test. The FLAC felt flat and uninspiring. The WAV was rousing and capturing.

Night and day for me. More about the feeling though than a measure. 

Agree on the Zenith. It's a nice piece of kit but won't rip WAV from CD and therefore won't auto assign database info. So you have to access the folders from the local server via the iMac and drag and drop to Zenith. Just as good, but no pretty album pictures and all the track names end in .wav - if that sort of things bothers your OCD.

Antony at Innuos says a firmware update is coming on 15th September though for WAV support. 

Will give the Melco Na-1 a go this week as it's cheaper and supposedly also great 'sounding'. No CD drive but I dont need it really. And Melco are offering a free buffalo CD drive at the moment that is supposed to work a trick. And it already rips wav with gracenote.

Sound dor pound though. We'll see. 

Maybe a bit off topic, but I found that what really matters is the quality of the recording, it doesn't matter a lot which format it is.

Moussa posted:

Maybe a bit off topic, but I found that what really matters is the quality of the recording, it doesn't matter a lot which format it is.

Maybe even more off topic as I only listen to physical media, but yes I find recording and mastering quality to be the biggest factors in replay SQ. When I see posts on file format choices I always wonder whether folks aren't listening too hard, justifying their current commitment as to how they rip, or looking at theoretics. In the end I suspect that 'best' format is user, system, and room dependent, and within those factors heavily dependent on a given recording. For a vinyl user it's probably not unlike the choice of which cartridge is best - give and take, but you commit to what works best overall for your listening catalog. Maybe something easier to do, or hear, among LPs versus digital file formats?

Musings from an outsider to ripping

Having originally been sceptical about the possibility of WAV files sounding better than FLAC, I decided some time ago to test a number of albums ripped to FLAC vs directly ripped to WAV (no transcoding) on my then ND5XS/Hugo, and to my surprise found that I did detect a subtle difference in favour of the WAV files. I wasn't entirely sure that I wasn't imagining this, and so I invited a couple of friends to take part in a 'blind' test. They both run top end Linn systems (Klimax DS/1 into Klimax 350 Active speakers for one, and high spec LP12 into Klimax 350 Passive speakers for the other), and neither of them had ever considered the possibility of a difference between FLAC and WAV files.

They both confirmed a consistent preference for the WAV files over the equivalent FLACs in my (Naim sourced) system. Interestingly however, one of my guests (the LP12 user - no streaming as yet) who happens to be a studio sound engineer had insisted that the tests be carried out using directly ripped to WAV albums, because he refused to consider that transcoding on the fly could bring the same benefits.

Also interestingly, they went back and carried out the same tests on the top of the range Linn system. This time, they did not detect any consistent differences between the two, and so my Linn owning friend decided not to transcode to WAV on his own system. However, since I ran (at the time) both Linn and Naim streamers, I made the decision to install MinimServer on my NAS and transcode to WAV on the fly. I have since replaced my ND5XS with a microRendu, but haven't had the time so far to compare the benefits when using this device. For th etime being, I continue to transcode to WAV on the fly.

By the way, have any minimServer users out there tested transposing on the fly using the commands  flac:wav and flac:wav24, and if so have you noticed a difference?      

Yes I have noticed differences with MinimStreamer transcoding on the fly... native WAV sounds marginally best .. on investigation this aligned with more consistent inter frame timing over the network when natively playing wav than transcoding. This will potentially vary from server to server however. I have experienced similar observations with Asset.

i now stream wav from a basic ReadyDLNA media server on my Netgear NAS, this seems to provide the best SQ, and my network traffic timing analysis supports this as it has the most consistent interframe timing over the network. I use a NDX feeding SPDIF into a Hugo.

interestingly when I stream FLAC from this ReadyDLNA media server it sounds better than from Asset or MinimServer with no transcoding .. Again the more consistent interframe timing from the NetgearNAS media server seems to help. I run Asset and MinimServer on RaspberryPi's.. I can't run them on my NAS, and they sound better on the Pi's rather than a Mac or PC.

Simon

Quality of the original recording is a given. This is why it is sometimes pointless but also sometimes easy to compare standard resolution material to higher res. If the material doesn't come from the same source, comparisons are meaningless.  But you'll still likely prefer one to the other. This applies no matter the format. The WAV versus FLAC preference, where it exists is a different matter. If I can hear a difference between the same file played back in each format and I prefer one over the other, why wouldn't I go for the one I liked?

I have albums that are so badly recorded, and remasters that are so badly remastered that they are only good enough to listen to on the MuSo. I can still hear a difference between WAV and FLAC, as I could on our old B&W A5, 

Here in Asgaard we love Aiff 16-24 (DVD or better).

Hungryhalibut posted:

All of our music is in FLAC, transcoded to WAV. I can't tell the difference between transcoded FLAC and native WAV. As I understand it, Naim streamers find it easier to convert WAV, which is why WAV sounds better. 

My music is all ripped as FLAC as this was what my dealer recommended when I started my streaming adventure. Which leads me to follow on from HH.......how do you transcode on the fly to WAV? I take it this is a setting using the uPnP server on the NAS. I  have a Synology NAS using their own music server which seems to work well....apologies if this has been covered elsewhere before

Arnsider posted:

 

Hungryhalibut posted:

All of our music is in FLAC, transcoded to WAV. I can't tell the difference between transcoded FLAC and native WAV. As I understand it, Naim streamers find it easier to convert WAV, which is why WAV sounds better. 

My music is all ripped as FLAC as this was what my dealer recommended when I started my streaming adventure. Which leads me to follow on from HH.......how do you transcode on the fly to WAV? I take it this is a setting using the uPnP server on the NAS. I  have a Synology NAS using their own music server which seems to work well....apologies if this has been covered elsewhere before

I don't think you can do it on the Synology music server, use Minim on your NAS and it can be done.

ChrisSU posted:

Arnsider posted:

 

Hungryhalibut posted:

All of our music is in FLAC, transcoded to WAV. I can't tell the difference between transcoded FLAC and native WAV. As I understand it, Naim streamers find it easier to convert WAV, which is why WAV sounds better. 

My music is all ripped as FLAC as this was what my dealer recommended when I started my streaming adventure. Which leads me to follow on from HH.......how do you transcode on the fly to WAV? I take it this is a setting using the uPnP server on the NAS. I  have a Synology NAS using their own music server which seems to work well....apologies if this has been covered elsewhere before

I don't think you can do it on the Synology music server, use Minim on your NAS and it can be done.

Transcoding can be set very easily on Syn. Media Server, its a simple check (ticK box) in Media Server menu > DMA settings.  The big problem is that it does not play gapless transcoding.   The best solution is to install Minimserver,  its a bit more fiddly but its not difficult.    

As Mike says, Synology media server can easily be set to transcode to WAV, and the sound quality will improve accordingly. But the gapless issue is really significant. On any album with gapless tracks you will get a funny noise between each track, instead of a normal transition, which is really irritating and spoils the experience. MinimServer is really excellent and is a bit of a fiddle to set up. Eventually I got fed up with its quirkiness and when my Synology got a bit elderly I bought a Qnap and loaded Asset, which is a lot easier to set up and use, and sounds just as good as Minim. That said, if you want to keep your Synology, MinimServer is the way to go.