Taiko Audio SGM Extreme : the Crème de la Crème

jbrrp1 said:

Hi Steve,

Thank you for sharing your evolving experience using the alpha TAS. The buffer adjustments sound intriguing, a bit like tube rolling, to make a crude analogy. It is valuable to have elements for system tuning, as the sum of all parts are different for each of us.

Regarding the overall improvements in sound quality, could you please elaborate more on what makes the music "more real", "more natural"? Is it more informative? Better microdynamics? Better delineation of the harmonic envelope timing? Etc. I'm curious to understand what it is that is fleshing out better than the Roon experience. Thanks!

Click to expand...

All of the above IMHO lol

sadly I have not been in Hengelo for about 7 weeks, so I am really behind on the SQ story

I do have a 4 core i5 desktop computer feeding some lowly electronics an a pair of KEF LS50’s playing as nearfield monitors.

From the time that the first command line version of XD player, all of the characteristics you list indeed seem to have been quite audibly improved.

I have been immersed in computer audio for 5 years now, the last 10 days have been the most amazing in terms of SQ advances.

I am sending applause emojis to Emile and Wilson 7 days a week and more than once a day

jbrrp1 said:

My Extreme ships this week, it seems - - very exciting! I would like to add one feature request for TAS that means a lot to me: can you have an easy way to select Pausing play after a track finishes? I got so used to this feature when using Foobar2000 a while back, and Roon stubbornly refused to implement this rather simple feature, which I have missed dearly.

I love listening down a playlist with a friend, and we like to pause between tracks at times to talk about the music. This feature makes it a snap, and I as the host can simply relax into the music, enjoying the full track to completion. With Roon, I have to be alert to hit Pause, which is just flat out disruptive to the enjoyment.

Nice if it can be selected with an Alt key sequence or something, establishing it for the track currently playing. Thanks!

Click to expand...

Noted,

we will find a way to implement
This is what’s so great about having one’s own software

EuroDriver said:

All of the above IMHO lol

sadly I have not been in Hengelo for about 7 weeks, so I am really behind on the SQ story

I do have a 4 core i5 desktop computer feeding some lowly electronics an a pair of KEF LS50’s playing as nearfield monitors.

From the time that the first command line version of XD player, all of the characteristics you list indeed seem to have been quite audibly improved.

I have been immersed in computer audio for 5 years now, the last 10 days have been the most amazing in terms of SQ advances.

I am sending applause emojis to Emile and Wilson 7 days a week and more than once a day

Click to expand...

I tend to agree. I have continued to play with the buffers. I liked 1024 wat better than 256 and I Boulder play all PCM resolutions and DSD to 512. For DSD my favorite is 8192

1024 I can play all PCM resolutions

Adaptive-1 however I just discovered will play only 44.1 BUT it was the best sounding 16/44.1 I have ever heard in my system

Adaptive-2 buffer plays nothing in my system

Steve Williams said:

Adaptive-2 buffer plays nothing in my system

Click to expand...

Hi Steve,

There is clearly no one size its all. As we gather more data points, we will map out possible strategies to give good sounding buffer size matches for the different formats for particular DAC’s

At the end of the day, there are not that many high end DAC’s, and we will have enough data points to map out what works with which DAC


Taiko’s custom USB driver Is undergoing a rewrite to broaden the buffer size choices and tuning possibilities. Hopefully this will be in time for the Beta

EuroDriver said:

Noted,

we will find a way to implement
This is what’s so great about having one’s own software

Click to expand...

Lovely! Thanks!

DAY 3 BUFFER PREFERENCE

I’ m sticking with 8192 for DSD but now I’m preferring Adaptive-1 for PCM

tonight i've been exploring, first in TAS for a few hours, both in Quboz and in my files, now later in Roon. i'm not yet proficient in switching back and forth from TAS to Roon and back again quickly. so i have not tried to quickly A/B cuts like that.

as an overall comment; everything sounds better. and in these first few hours i'm struck with a couple of consistent steps up across the board, both with TAS and also with Roon. the sound is less processed, more disconnected from the speakers, and more dimensional. and then also it's more continuous and absent of grain. i consider grain to be an artifact, and so the lowering of that = more true detail. these are relative gains to my previous listening perceptions. so higher degrees in positive directions. these are not big differences objectively, but significant musically.

and understand that Emile did my previous set-up, and now has done this set-up.......so what he changed all around is un-known to me. i like sausage, but don't need to watch it being made or know all about it.

in TAS with Quboz i listened to most of a Collection of Aria's by Mezzo-Soprano Marianne Crebassa, "Oh Boy" in 24/96.......except at this point it was in 16/96 (a problem.....i'm getting no 24 bit or DXD files in TAS......yet) in TAS. could be my buffer settings which i've not yet played with? in any case, it was really fine. i was captured. later i sampled this in Roon/Quboz in 24/96 and from my recollection i preferred it in TAS (seemed more dimensional and expressive......and 'flowing').



then i ventured into Roon. in TAS i was struggling to find my touch-stone references. these are my go-to cuts for getting my bearings. so i spent the last hour or so sampling them from my files in Roon, and i am really stoked by what i'm hearing. this is a new reference for these cuts to my ears. period. i did my Anna Netrebko and my Nils Lofgren and a few others and we are in a new realm. i've just done a 20 cut playlist that is running now as i'm typing.

i do look forward to finding all these touch-stone reference cuts in TAS and sampling them. hopefully soon. i expect them to be even better.

i know the gang has work to do but i'm eating this up.

wil said:

Thanks for the further explanation of what XD is doing with the buffers.

Is the XD auto a static buffer size or is it variable depending on the file format and sample rate that's playing?

And for users who choose specific buffers for PCM or DSD at various sample rates, can those choices be programmed in to automatically switch back and forth depending on what's playing?

Have you found that buffer choices vary by DAC used?

Click to expand...

I'm all for furthering SQ, but as a new Extreme user, I admit I have zero interest in changing any settings, buffers or otherwise, depending on the source format. It would be more fun to go to the dentist. I'm sure I'll venture forth with TAS someday, but not until it is plug and play with the best optimization automatically for whatever is playing. Reminds me of the early Lampizator days when you had to flip a switch to go from playing PCM to DSD. I didn't buy one then for the same reason.

wil said:

Thanks for the further explanation of what XD is doing with the buffers.

Is the XD auto a static buffer size or is it variable depending on the file format and sample rate that's playing?

And for users who choose specific buffers for PCM or DSD at various sample rates, can those choices be programmed in to automatically switch back and forth depending on what's playing?

Have you found that buffer choices vary by DAC used?

Click to expand...

marty said:

I'm all for furthering SQ, but as a new Extreme user, I admit I have zero interest in changing any settings, buffers or otherwise, depending on the source format. It would be more fun to go to the dentist. I'm sure I'll venture forth with TAS someday, but not until it is plug and play with the best optimization automatically for whatever is playing. Reminds me of the early Lampizator days when you had to flip a switch to go from playing PCM to DSD. I didn't buy one then for the same reason.

Click to expand...

What we have in the alpha is unlikely to be in the final release.

We currently have the following options:
1) Auto, which follows vendor supplied DAC driver defaults, this is a "safe" mode which will work reliably on all types of installations, all types of computers running all kinds of Windows OS versions, and all kinds of playback software.
2) Adaptive, which is our own algorithm adjusted to what we feel sounds best for each sample rate on the Extreme, you can consider this optimised for the specific OS and hardware we run, so far this seems to be the preferred option amongst the alpha testers.
3) Manual, where you can specify your own, this is really just there to verify operation of the Adaptive mode and test what a wider range of DAC (drivers) can support.

This will likely go in one of the following directions:
1) A choice between Auto (DAC vendor default), or Adaptive (optimised for our hardware/software).
2) An optimised profile for the DAC used, for example, MSB, DCS, Lampizator, Totaldac etc, where you select the DAC you use and it will load the optimal settings for that.
OR 3) As we are developing our own generic driver which can support a very wide range of DACs we may just end up with an optimum for that driver with no need to change anything.

Basically we know what the Extreme can handle, which is any size, we're now gathering data on what the vendor supplied DAC drivers can support, and if the optimum depends on the driver(software) or the DAC hardware/firmware.

The Alpha and Beta rounds will give us the feedback we need to nail this down.

Mike Lavigne said:

in 24/96.......except at this point it was in 16/96

Click to expand...

Ok, it looks like the MSB driver/firmware has an alternate settings specifically for 24 bit files we need to force. Pretty sure Wilson can implement that rather quickly.

EuroDriver said:

Hi Kingsrule,

the feedback from the Alpha testers here on this forum is actually a key part of the development process for us.

I will give you an example, Romaz’s and Ray Dude’s very comprehensive reports. We re-read those reports many times and also received feedback from other WBF members who had read the reports. These data points paint for us a well populated landscape of what Audiophiles are hearing, and what are important/ key drivers for their musical enjoyment.

Hardware and Software development in a bubble / echo chamber is not productive in our experience.

This forum is key resource for us and we thank all posters and question askers for the time they take to do this

Click to expand...

I must have missed Romaz's and Ray's reports on TAS. Were they posted here?

Mike Lavigne said:

then i ventured into Roon. in TAS i was struggling to find my touch-stone references. these are my go-to cuts for getting my bearings. so i spent the last hour or so sampling them from my files in Roon, and i am really stoked by what i'm hearing. this is a new reference for these cuts to my ears. period. i did my Anna Netrebko and my Nils Lofgren and a few others and we are in a new realm. i've just done a 20 cut playlist that is running now as i'm typing.

i do look forward to finding all these touch-stone reference cuts in TAS and sampling them. hopefully soon. i expect them to be even better.

i know the gang has work to do but i'm eating this up.

Click to expand...

Hi Mike,

What do you mean by "sampling them from my files in Roon"?

Are you suggesting that Roon can be used with TAS?

Geoff

heebrog said:

I must have missed Romaz's and Ray's reports on TAS. Were they posted here?

Click to expand...

I was referring to there original Extreme reports

They are now diving into TAS and no doubt will have sonic observations to share

heebrog said:

Hi Mike,

What do you mean by "sampling them from my files in Roon"?

Are you suggesting that Roon can be used with TAS?

Geoff

Click to expand...

The TAS Alpha operating environment allows easy switching between Roon and TAS. Enabling this functionality meant running roon in a different way which actually sounds better than Roon installed alone by itself

a nice bonus

heebrog said:

Hi Mike,

What do you mean by "sampling them from my files in Roon"?

Are you suggesting that Roon can be used with TAS?

Geoff

Click to expand...

Roon and TAS software are totally independent from one another and never run concurrently

EuroDriver said:

The way TAS works is definitely not gapless. There is just so much important preparation going on before a track starts to play which makes a very audible difference to the sound quality.

A vinyl playback analogy would be washing or carefully cleaning the disk before playing. A brush mounted on the cartridge shell is very definitely audible, a similar story for digital with processing environment preparation activity during playback.

However, we hear you and understand how some live concerts which have been split into tracks need to be played gapeless. We have some possible solutions in mind where tracks in a playlist can be gapless.

what we are not inclined to do is fade out and fade in, that does not increase realism. ;-)

Click to expand...

hopefully this can be worked out in the future as this is an absolute show stopper for me.
in addition to live concerts (certainly more than 1000 here), i can't imagine listening to side 2 of Abbey Road without gapless playback (actually, i can imagine it as i've had a few
streaming solutions over the years that didn't play gapless-------the result was not pretty)

fading in/out is, of course, not a sane solution in any reasonable universe.

i trust you'll have this problem solved and integrated into your buffering somehow by the time, sometime next year, i finish saving up enough free cash to pull the trigger on one of these

cat6man said:

i trust you'll have this problem solved and integrated into your buffering somehow by the time, sometime next year, i finish saving up enough free cash to pull the trigger on one of these

Click to expand...

did i really say that out loud?

cat6man said:

hopefully this can be worked out in the future as this is an absolute show stopper for me.
in addition to live concerts (certainly more than 1000 here), i can't imagine listening to side 2 of Abbey Road without gapless playback (actually, i can imagine it as i've had a few
streaming solutions over the years that didn't play gapless-------the result was not pretty)

fading in/out is, of course, not a sane solution in any reasonable universe.

i trust you'll have this problem solved and integrated into your buffering somehow by the time, sometime next year, i finish saving up enough free cash to pull the trigger on one of these

Click to expand...

when mine was installed 4 days ago the first and most glaring concern was indeed the long gap waiting between songs. In fact the first time I played some music and encountered the long pause I thought something was broken and then suddenly music played. I gather this gap has something to do with an algorithm written that removes network induced noise when changing songs. I gather this is something that will be worked out going forward.

Not sure what Emile's plan is with these buffers or whether more will be added but the amazement I had from day1 to day 2 when I started experimenting with them proved to be a game changer

I remain sold on Adaptive-1 for PCM and 8192 for DSD BUT I can only play 44.1on Adaptive-1 but it was the best 44.1sound I have ever heard in my system

For me Emile delivered on his promise of giving us the best sound quality and clearly TAS is a step beyond what I hear when listening to Roon

How long of a gap between tracks are we talking about? Is it variable depending on track length?

I'm guessing they'll come up with a solution, but, personally, I'd thankfully trade a few seconds of silence between tracks for better sq any day!