Taiko Audio SGM Extreme : the Crème de la Crème

[Christiaan Punter]

So the conclusion might be as long as streaming services use FLAC (or MQA) they will never be able to deliver the same (or better) quality as the best local files.

Matt

Indeed, that sounds to me like a logical and reasonable conclusion. That is unless we find a way to make the unpacking of FLAC truly transparent. After all, the format is supposed to be lossless and it's assumed that it is the unpacking that creates the sonic difference.

 

[Christiaan Punter]

Thanks to each for your responses to my post.

Steve - your detailed reply gets to many of the individual sub plots that crossed my mind resulting in my original question.

Since I'm just now contemplating a server, DAC, and ADC my head is reeling from the myriad of choices. Where to start? My experiences are limited but because of exposure to a Wolf server and Bricasti M1 SE DAC I can appreciate the possibilities beyond my standalone Vitus CD spinner.

But again, where to start?

As far as servers go, will XDMS conquer ROON, TIDAL and the like? Seems there is a lot of momentum building here for XDMS.

With respect to the Taiko Extreme, I can tell you that we are very firmly focused on surpassing the sound quality of Roon with XDMS. Tidal and Qobuz are services that are implemented into a platflorm such as Roon, TAS or indeed XDMS. As such, XDMS will not replace streaming services, but it will certainly aim to make the most of them, whilst also making the most of your local library.

 

[Cyrus]

It's interesting the reasons why people make different selections.

Indeed.

Scale isn't a factor in your setup. Nor video in line of sight.

And yes, the Extreme is front and center and prominent, in my setup.

The power/on light is an issue (especially at night) when I watch movies, etc.

Given the provenance of my rack, floor positioning is superior for sound quality.

 

[steve williams]

Beautiful set up Cyrus. I love those tubes. So you are saying the little white light from the Extreme bothers you but not the glow of all those tubes

 

[Cyrus]

@Steve Williams Thanks. The tube glow is very soft and diffuse. The white light, not so much.

 

[wil]

Indeed, I also hear a slight advantage for local files, and a big difference between WAV and FLAC, either streaming or playing locally.

@Christiaan Punter Are you saying that you prefer a streamed file in WAV to a local file in Flac? And comparing a WAV streamed to WAV local, only a slight preference to the local file?

For some reason, I just assumed all Qobuz streaming was Flac. Is it a mix of Flac and WAV?

 

[Cyrus]

Am I the only one having trouble understanding comparison of local files versus those (even when equivalent) fed from afar, over multiple lines, and any number of components in the chain...before arriving at the Taiko Extreme?

I would not expect the results to be the same. The fact that they are close is worth celebrating.

 

[cat6man]

Thanks for sharing,
even on a lesser device than the Extreme I prefer WAV

Matt

a friend did the test yesterday, jriver to rossini, and he came to the same conclusion.

 

[cat6man]

I totally second these test results. I was about to write a post to be careful to compare the EXACT same versions locally and streaming but you have already done the test in the same way that I did (purchasing albums in FLAC and WAV from Qobuz and comparing the same tracks local and streaming). Indeed, I also hear a slight advantage for local files, and a big difference between WAV and FLAC, either streaming or playing locally. A friend of mine actually really likes FLAC, either streaming or played locally, precisely because it sounds more fluid to him. To me, however, it sounds more rounded than it should. In my opinion, streaming via Qobuz is good enough not to worry about but when you really love an album, it's best to get the CD and rip it or buy the WAVs via Qobuz.

just to clarify, my FLAC version was created here with the dbpoweramp converter.
neither file was download but began as a local CD rip.

 

[cat6man]

Am I the only one having trouble understanding comparison of local files versus those (even when equivalent) fed from afar, over multiple lines, and any number of components in the chain...before arriving at the Taiko Extreme?

I would not expect the results to be the same. The fact that they are close is worth celebrating.

Hi Cyrus

My observation is that FLAC v. WAV sounds different when both are local. In fact, here local FLAC and streamed FLAC are close (well worth celebrating). Why FLAC sounds worse that WAV is problematic.

Please try the track I linked to and let us know if you think there is an SQ difference.

 

[jbrrp1]

Hi Cyrus

My observation is that FLAC v. WAV sounds different when both are local. In fact, here local FLAC and streamed FLAC are close (well worth celebrating). Why FLAC sounds worse that WAV is problematic.

Please try the track I linked to and let us know if you think there is an SQ difference.

I'm a bit challenged on file organization, or something, but I cannot seem to locate these files after saving them into my Extreme SSD folders. Any hints on this? Thanks!

 

[Malcng]

Hi Cyrus

My observation is that FLAC v. WAV sounds different when both are local. In fact, here local FLAC and streamed FLAC are close (well worth celebrating). Why FLAC sounds worse that WAV is problematic.

Please try the track I linked to and let us know if you think there is an SQ difference.

When getting into the detail of comparing FLAC vs WAV, should we not be describing the degree of compression applied to the FLAC rip being compared? To me, zero compression is favourite but I ripped all my CD’s way back when disk space was expensive so I used compromise compression 5. In fairness, my system back then would probably not have highlighted a difference anyway.

 

[oldmustang]

Future resting place for the Extreme when it arrives next week. This technique is the same as making concrete counter tops. I build the form, pour the concrete, and then seal. The concrete takes on a cool patina. Pretty labor intensive. My component shelving is half the thickness, that sit on custom brackets that are lagged into a 6" thick log wall. For the platforms I have adhered 3/4" cork sheets to the bottoms. The Mephisto sits on a larger platform approx. 250lbs. Taiko platform weighs approx. 200lbs. In addition I use Isoacoustics which is most likely overkill? You can literally jump up and down on the shelves without any movement. Granted this would not work in all decor's, for this type home it works. I showed it to Emile and his reaction was Whoa!

Nothing succeeds like excess! (One of my DIY mantras. . .)

I like it, John.

Steve Z

 

[rando]

When getting into the detail of comparing FLAC vs WAV, should we not be describing the degree of compression applied to the FLAC rip being compared? To me, zero compression is favourite but I ripped all my CD’s way back when disk space was expensive so I used compromise compression 5. In fairness, my system back then would probably not have highlighted a difference anyway.

Emile responded to my questioning (generalizing from here on) certain aspects of files that he expended a great amount of time and energy minimizing impact. The amount of compression that may have bogged down older hardware unpacking is trivial on an Extreme. File handling (embedded pictures etc) should also be no source for concern.



Opinions vary widely on if older or lower quality optical drives provided the best possible rips back when or currently. You can always select a few albums and use a collection of hardware and software you feel offers the highest quality and rip them to WAV.

I have suspicions older files and especially those transferred around over the years lose magic. Nothing lasts forever. Including strategies to provide storage for an entire library of ripped files, in some homes.

 

[John T]

@oldmustang: Thanks Steve, I like to go over and above too. Very good mantra! It was actually a fun project. You use silicone to round the edges inside the form, Also a releasing agent so the form slides off. Its a quickset so you have to have everything ready.

 

[Malcng]

Emile responded to my questioning (generalizing from here on) certain aspects of files that he expended a great amount of time and energy minimizing impact. The amount of compression that may have bogged down older hardware unpacking is trivial on an Extreme. File handling (embedded pictures etc) should also be no source for concern.



Opinions vary widely on if older or lower quality optical drives provided the best possible rips back when or currently. You can always select a few albums and use a collection of hardware and software you feel offers the highest quality and rip them to WAV.

I have suspicions older files and especially those transferred around over the years lose magic. Nothing lasts forever. Including strategies to provide storage for an entire library of ripped files, in some homes.

Useful info, thanks.

 

[matthias]

Indeed, that sounds to me like a logical and reasonable conclusion. That is unless we find a way to make the unpacking of FLAC truly transparent. After all, the format is supposed to be lossless and it's assumed that it is the unpacking that creates the sonic difference.

Christiaan,
do you know which FLAC compression level Qobuz applies for streaming?
AFAIK, the default FLAC compression level is 5.
Thanks

Matt

 

[Cyrus]

My observation is that FLAC v. WAV sounds different when both are local.

I would expect them to sound different, especially when local. As others have reported (including you), I also have a preference for WAV. I also find FLAC more variable.

WAV is generally (not always) non-compressed. It was developed by IBM and Microsoft and is the 'preferred' format for Windows. Since the Taiko Extreme is Windows based, this may be a contributing factor (among others) to WAV file superiority and preference (by many) as inferred by final sound quality vis a vis the Extreme [as reported in this thread].

FLAC is cross platform and open source with open implementation (READ: variability) not to mention the complexity: frames, compression / decompression // encoding / decoding, algorithms used, etc., ...despite "perfect reconstruction."

I expect that ANY well implemented audio system with high levels of transparency can and does differentiate between file formats. What is preferred by the individual audiophile is another matter.

-----------------------------------------------------------------------------------------------------------------------------------------------------------
Note re. FLAC [from Wikipedia]: "The amount of compression is determined by various parameters, including the order of the linear prediction model and the block size. Regardless of the amount of compression, the original data can always be reconstructed perfectly.

For user's convenience, the reference implementation defines 9 compression levels, which are presets of the more technical parameters to the encoding algorithm. The levels are labeled from 0 to 8, with higher numbers resulting in a higher compression ratio, at the cost of compression speed. The meaning of each compression level varies by implementation.

FLAC is optimized for decoding speed at the expense of encoding speed. A benchmark has shown that, while there is little variation in decoding speed as compression level increases, beyond the default compression level 5, the encoding process takes up considerably more time with little space saved compared to level 5."

 

[microstrip]

(...) I expect that ANY well implemented audio system with high levels of transparency can and does differentiate between file formats. (...)

I have a different perspective - I expect that a well implemented digital audio playing system sounds the same in any bit exact file formats.

I hope that XDMS will make them sound very similar. Otherwise I will ask Emile for an operator that changes all files to the preferred format at a single click!

 

[cat6man]

When getting into the detail of comparing FLAC vs WAV, should we not be describing the degree of compression applied to the FLAC rip being compared? To me, zero compression is favourite but I ripped all my CD’s way back when disk space was expensive so I used compromise compression 5. In fairness, my system back then would probably not have highlighted a difference anyway.

I didn't mention it explicitly in my post (sorry) but the filename *-flac5 indicates I used FLAC compression level 5. I'm just guessing that qobuz uses something like level 4 as their bit rate was showing up as a little bit higher than my local file.