Taiko Audio SGM Extreme : the Crème de la Crème

[steve williams]

I go to sleep every night with visions of NSM in my mind. According to Emile that is the 2nd best uptick in SQ behind only the pending release of the new Extreme with BPS.

Giddee up Taiko team. We are all. saddled and ready to ride

 

[JackD201]

waiting......

 

[Taiko Audio]

Is it possible in the future Taiko [internal alpha code] programs will begin to exhibit anywhere near that same level of focus to allow determining what the best version between local and streaming files is?

Meaning aware if a streamed version is different from the downloaded or ripped copies in ways that would allow making some high level determinations on quality before listening. Well established fact streamed files move around a lot and there is zero chance of notice being given if rights owner chooses to deliver a different mastering multiple times within a short period.

Your most recent post noted being detail oriented enough to capture the best (purchase download within window it is available) sounding copy to be archived locally. Weekly. Routine polling at time of loading file for playback with any changes detected given discrete notice in UI feasibly only happens on the order of months and years.

That would be very difficult, there's an online database, called accuraterip, which gathers information from people ripping CDs and provide a checksum which would give you an indication if your rip was close to accurate as in near the same content as a certain number of others (no rip is actually exactly the same surprisingly).

For this we would need something similar with people rating versions of tracks I guess. But it's just a "problem" in general. For example what I do nowadays is purchase files from Qobuz, but I always select the 16/44.1 uncompressed .wav version in the downloads section. Most of the time these sound better then the 24/96 etc versions. The downside of this approach is that you lack good metadata, so in Roon they are very difficult to find again due its library being metadata based, in XDMS you can use the browse folders navigation option (which Roon lacks) to find them relatively easy. But this makes it a rather unfriendly option to use.

 

[Di-fi]

Is it possible in the future Taiko [internal alpha code] programs will begin to exhibit anywhere near that same level of focus to allow determining what the best version between local and streaming files is?

Meaning aware if a streamed version is different from the downloaded or ripped copies in ways that would allow making some high level determinations on quality before listening. Well established fact streamed files move around a lot and there is zero chance of notice being given if rights owner chooses to deliver a different mastering multiple times within a short period.

Your most recent post noted being detail oriented enough to capture the best (purchase download within window it is available) sounding copy to be archived locally. Weekly. Routine polling at time of loading file for playback with any changes detected given discrete notice in UI feasibly only happens on the order of months and years.

Nice. In the future, this implies that the Extreme server (or any server for that matter) will possess the capability to recognize distinctions between streamed, downloaded, or ripped versions, enabling it to autonomously assess quality prior to playback.

 

[Taiko Audio]

Nice. In the future, this implies that the Extreme server (or any server for that matter) will possess the capability to recognize distinctions between streamed, downloaded, or ripped versions, enabling it to autonomously assess quality prior to playback (or even 'listening').

But that does not actually tell you what the best sounding version of that file is..

 

[cat6man]

but I always select the 16/44.1 uncompressed .wav version in the downloads section. Most of the time these sound better then the 24/96 etc versions.

my observation also, and a shame on the industry

 

[Di-fi]

But that does not actually tell you what the best sounding version of that file is..

Exactly, best version equals not necessarily best sounding version. So it seems we are still stuck with a personal rating system.
Still, the idea of auto-recognizing some file parameters and incorporating them in the user interface, maybe even in a new (visual) way, could be worth exploring.

 

[cat6man]

Emile,

Do you have an approximate target date for shipping the router that you can share with us?

 

[Taiko Audio]

Emile,

Do you have an approximate target date for shipping the router that you can share with us?

Yes that is a bit of a bummer as we are still waiting for the chassis to return from the finishing company, we just received the batch of switches we shipped them before the routers / DC distributors. Basically we paid a rush fee for the new router PCBs for nothing as they would have arrived by now on a regular supply schedule. We are just getting an as fast as possible reply now, they do actually have a valid reason for their delays as part of their facility burned down in May, they even had an employee die, which is all very sad, we're doing the polishing ourselves now which they used to do, but they are still behind schedule. We are also waiting on samples to return from a new company we found who promise a 2 week turnaround but that would only help for future batches.

 

[Taiko Audio]

my observation also, and a shame on the industry

Indeed, it would be nice if the highest res version of a file would automatically be the best sounding one, but that is absolutely not the case.

 

[steve williams]

Indeed, it would be nice if the highest res version of a file would automatically be the best sounding one, but that is absolutely not the case.

will things change with the release of NSM

 

[rando]

That would be very difficult, there's an online database, called accuraterip, which gathers information from people ripping CDs and provide a checksum which would give you an indication if your rip was close to accurate as in near the same content as a certain number of others (no rip is actually exactly the same surprisingly).

I understand. Thank you for considering the implications and how they would mesh with this discussion.

Nice. In the future, this implies that the Extreme server (or any server for that matter) will possess the capability to recognize distinctions between streamed, downloaded, or ripped versions, enabling it to autonomously assess quality prior to playback.

I don't think we are there yet. Inside his server attaining SQ is the only job. My only purpose was noting how continuation of established behaviors might be portrayed in the Taiko server environment. Where a separation occurs.

 

[Taiko Audio]

will things change with the release of NSM

Well, indirectly I'd actually say yes as improving certain aspects of playback fidelity can make you care less about "worse" recordings / file versions.

An interesting observation I'm now more comfortable in sharing regarding the interface part of the upcoming TACDA/TACDD is the effect of data rate continuity on sound quality. I currently run 16 bit recordings about 250 times "faster" (smaller data packets / intervals between packets) then USB can (due to its 125uS packet interval) and am working on about 1250 times faster. 24 bits halves "speed" as you basically use 32 bits for that. The interesting part of this that I currently just prefer realtime downsampling 24 bit files to 16 bits in Roon as that doubles the interface "speed", it just sounds more expansive, with better flow, less edgy and most certainly less of everything people tend to associate with "digital sound". Using this interface is much more forgiving on "bad recordings" while it's actually significantly more revealing at the same time, so this is not a "smoothing things over" effect...

 

[SwissTom]

An interesting observation I'm now more comfortable in sharing regarding the interface part of the upcoming TACDA/TACDD is the effect of data rate continuity on sound quality. I currently run 16 bit recordings about 250 times "faster" (smaller data packets / intervals between packets) then USB can (due to its 125uS packet interval) and am working on about 1250 times faster. 24 bits halves "speed" as you basically use 32 bits for that. The interesting part of this that I currently just prefer realtime downsampling 24 bit files to 16 bits in Roon as that doubles the interface "speed", it just sounds more expansive, with better flow, less edgy and most certainly less of everything people tend to associate with "digital sound"

It's a truly remarkable discovery.

Do you think that by further increasing the speed we'll reach a point where 24-bit recordings will take over???

This new interface will consign all the others to oblivion!

Have you been in contact with other DAC manufacturers other than LampizatOr and MSB regarding the potential adoption of this new interface?

Cheers,

Thomas

 

[wil]

It's a truly remarkable discovery.

Do you think that by further increasing the speed we'll reach a point where 24-bit recording will take over???

This new interface will consign all the others to oblivion!

Have you been in contact with other DAC manufacturers other than LampizatOr and MSB regarding the potential adoption of this new interface?

Cheers,

Thomas

I believe Areis Cerat as well.

 

[steve williams]

I believe Areis Cerat as well.

I thought TotalDac

 

[Taiko Audio]

I believe Areis Cerat as well.

Correct.

I thought TotalDac

Yes but after a brief discussion it appears there’s little benefit to be gained over just using AES/EBU as the Totaldac uses that “internally” already.

 

[steve williams]

Emile.for my understanding , will your new interface connect to the Horizon via i2s

 

[Taiko Audio]

Emile.for my understanding , will your new interface connect to the Horizon via i2s

That would be the “plug&play” solution, we’ll just have to try and see if a full implementation is worth the effort, that would require a PCB installation inside the DAC.

 

[steve williams]

That would be the “plug&play” solution, we’ll just have to try and see if a full implementation is worth the effort, that would require a PCB installation inside the DAC.

is there a sonic difference between plugnplay vs a PCB installation in the DAC