Taiko Audio SGM Extreme : the Crème de la Crème

[Christiaan Punter]

...do vowels have a negative impact on software naming SQ? You guys must have experience working in government! And we still have beta and "final" releases to come. OMG LMAO. Love ya...

Oh boy, I'm afraid that we did not test that!
But we did test for alliteration!

 

[nenon]

"XDMS NSM Alpha Version 0.0.1+53" sounds a lot more sophisticated than "Taiko Player".

 

[joet]

Hi Joet, can you measure the sound level with an iPhone app? I use "dB Volume". I can imagine a perceived difference but there really should not be an actual 2dB difference. At least I can't think of a reason why that could be the case. But if there was then such a difference should be measurable.

Hi Christiaan, Thanks for the follow-up. Per Ed's advice I'm going to limit posting my listening impressions for a couple of days to allow things to settle in. I appreciate your patience in having to evaluate early posts. What I will say is I've been playing a system disc non-stop for about 18 hours now and things already are already sounding very different (better) this morning than they did last night. That includes the perceived volume differencs which seem to have normalized, at least to my ears. However, I took your advice and downloaded a measurement app, Decibel X. I made two sets of recordings of the same song for the same duration. For the first recordings the avg DBs was exactly the same. For the second recordings, which ran 30 seconds longer and included a louder passage, the LAN card was 1.5 DBs louder. I don't know why there was a difference the second time. Maybe I held the phone in a different location? I'll try it again later today and report back.

 

[microstrip]

That's using a second network card, going to a Ravenna based DAC, compared to using the LAN port to go to the DAC. However, it could be the sonic artifacts of the pre-break-in period. I'll listen to it for another few days, keeping it constantly running, and get back. As an FYI when I installed the first network card there wasn't an comparative change in volume.

No break-in of network card can make a 1-2 dB change in sound level. However even very small differences in sound quality can trick our mind, suggesting level changes. If you have doubts on it play a test a 1kHz track and measure the level with an AC voltmeter.

 

[joet]

No break-in of network card can make a 1-2 dB change in sound level. However even very small differences in sound quality can trick our mind, suggesting level changes. If you have doubts on it play a test a 1kHz track and measure the level with an AC voltmeter.

Yes, I agree, that is what I was trying to say. Albeit poorly.

 

[ACHiPo]

Yesterday in Oldenzaal was a big day for XDMS Sound Quality. After many months, and hundreds of coding hours, Wilson sent us the test mule of the new XDMS with a changed internal architecture which we have given the codename XDMS NSM

The difference between NSM and the current XDMS is huge. Emile carried out a complete rework of the CPU core allocations to take advantage of the new architecture and then voila, we heard detail, tonal body, layering of texture and detail that we did not know existed on the test tracks. The sound stage was spectacular, wide and realistic with seamlessness and integration a new level. PRAT was superb with flow and swing

In short another big leap up in total realism. There is still a lot of work done integrating the new architecture into the current eco system, but all in good time it will happen.

Ed,
That's pretty exciting. No one commented on my observation last week after returning from several weeks of travel that my sound quality had increased, and that was WITHOUT NSM (not sure if there have been other improvements in sound quality with releases of XDMS?) ! Looking forward to getting to listen to NSM!

Evan

 

[EuroDriver]

Ed,
That's pretty exciting. No one commented on my observation last week after returning from several weeks of travel that my sound quality had increased, and that was WITHOUT NSM (not sure if there have been other improvements in sound quality with releases of XDMS?) ! Looking forward to getting to listen to NSM!

Evan

There was a jump in SQ in the 27 April backend build, which everyone liked

 

[Steve Williams]

There was a jump in SQ in the 27 April backend build, which everyone liked

Well I'm told that the word around Taiko headquarters over the past few days is that this new XDMS NSM is the biggest jump in SQ yet. That is something we will all be patiently awaiting. Kudos to Wilson and the Taiko team

 

[DW101]

There was a jump in SQ in the 27 April backend build, which everyone liked

Ed, do you mean March rather than April? Major FOMO kicking in here...

 

[Levitator]

Ed, do you mean March rather than April? Major FOMO kicking in here...

It was definitely 27 April

 

[Christiaan Punter]

Hi Christiaan, Thanks for the follow-up. Per Ed's advice I'm going to limit posting my listening impressions for a couple of days to allow things to settle in. I appreciate your patience in having to evaluate early posts. What I will say is I've been playing a system disc non-stop for about 18 hours now and things already are already sounding very different (better) this morning than they did last night. That includes the perceived volume differencs which seem to have normalized, at least to my ears. However, I took your advice and downloaded a measurement app, Decibel X. I made two sets of recordings of the same song for the same duration. For the first recordings the avg DBs was exactly the same. For the second recordings, which ran 30 seconds longer and included a louder passage, the LAN card was 1.5 DBs louder. I don't know why there was a difference the second time. Maybe I held the phone in a different location? I'll try it again later today and report back.

Hi Joet, nice to hear the components are improving as they settle in and you have taken measurements. Since there is no conversion or DSP in the Extreme Switch or Extreme Network Cards, it should be impossible for the volume level to change. I thought that measuring the level would be an easy method to prove/disprove any differences but it seems the tests so far are inconclusive. It's probably best done using a test tone rather than a dynamic piece of music. In any case, I can imagine that you may have more than average sensitivity in your ears/hearing and are picking up on clues related to running in. It's entirely possible that hardness translates to a feeling of loudness.

 

[engadin]

Something went wrong with roon when I substituted the switch’s wall wart with my new Sean Jacobs mini arc6 dc4.
So, once again, Ed came to my rescue.
He also installed XDMS.
So there is now the impact of the power supply and XDMS both.

While everything is still running in … I am already floored.

Many thanks to Ed (who hopefully gets a bonus for dealing with digital dorks like me) and to Nenon.

I don’t want to sum up audiophile stuff all improved (it is), but simply say that music has become much more interesting, fascinating, meaningful and a few more adjectives I now lack. Oh, and still improving.

Many thanks for these truly great products,
even more thanks for a fast and effective support,
for ever and ever improving these products and for a helpful community!

 

[John T]

Hi Joet, nice to hear the components are improving as they settle in and you have taken measurements. Since there is no conversion or DSP in the Extreme Switch or Extreme Network Cards, it should be impossible for the volume level to change. I thought that measuring the level would be an easy method to prove/disprove any differences but it seems the tests so far are inconclusive. It's probably best done using a test tone rather than a dynamic piece of music. In any case, I can imagine that you may have more than average sensitivity in your ears/hearing and are picking up on clues related to running in. It's entirely possible that hardness translates to a feeling of loudness.

Absolutely perceived db increase can translate as a result of a number of factors...

 

[Steve Williams]

Something went wrong with roon when I substituted the switch’s wall wart with my new Sean Jacobs mini arc6 dc4.
So, once again, Ed came to my rescue.
He also installed XDMS.
So there is now the impact of the power supply and XDMS both.

While everything is still running in … I am already floored.

Many thanks to Ed (who hopefully gets a bonus for dealing with digital dorks like me) and to Nenon.

I don’t want to sum up audiophile stuff all improved (it is), but simply say that music has become much more interesting, fascinating, meaningful and a few more adjectives I now lack. Oh, and still improving.

Many thanks for these truly great products,
even more thanks for a fast and effective support,
for ever and ever improving these products and for a helpful community!

This is the very reason why every Extreme user who has Roon/TAS need to put in their request for XDMS installation. The change is so profound that once heard you cannot unheard it. Plus every now and then changes are made which have a big uptick in SQ so that just when you think it is so good, it only gets better.
'For my ears it is the emotional involvement I feel with XDMS that I just don't with Roon . XDMS has a way of drawing the listener in and maintaining the hold on you. You'll hear things in music that you've played for years and never knew they were there and for me the beauty is how uncolored and neutral the music is

 

[DW101]

It was definitely 27 April

Thanks - I didn't notice the new 27th April thread on discord, I've just returned from almost 5 months away from my system so only looked at the original backend thread on discord and got the end March update .

Which is actually good news as there is an even more improved build! My quick first impressions of 27th April combined update vs my listening yesterday to March's update are very worthwhile improvements in noise floor and attendant musical insight & engagement, bass better control.

I should get my Switch & NIC in a week or so, and Router when shipped. I'm viewing these as a long term investment in Taikos' processing eco-system that will provide a foundation for further software SQ enhancements, as per Ed's news about XDMS NSM. What a journey!

 

[engadin]

Steve Williams wrote: 'For my ears it is the emotional involvement I feel with XDMS that I just don't with Roon .

This.
This sums up what I meant.
I think it is a new wealth of microdynamic gradations I previously thought 16bit could never achieve due to too few dynamic steps.
And a wealth of tonal shadings, colours, too.

 

[ReformedNiceGuy]

Apologies if this question has previously been posed in this thread. I did search with keywords, but couldn't find much.

I currently use an upstream Uptone EtherREGEN, which in turn is fed optically from a Startech ET91000SFP2 FMC. I use Finisar SFPs in single mode. In addition, the EtherRegen is clocked with an AfterDark King 10MHz clock. It's a good setup, though as ever I'm looking for possible improvements.

Feeding the EtherRegen optically rather than from ethernet definitely improved my sound quality. However, I've read that though an optical bridge can eliminate electrical noise, it can still be prone to phase noise and jitter. I'm unsure how the Startech clocks its SFP output, but it's not intended for audiophile use.

Is it likely that swapping the Startech for a Sonore opticalModule Deluxe would bring benefit to my system?

I ask as I get the impression that most users of the opticalModule use it for converting optical to ethernet, rather than vice versa. I'm hoping the better clock in the Sonore might pay dividends, or is it so far upstream that differences will be negligible. Remember, the downstream EtherRegen is being re-clocked by the AfterDark King.

 

[DW101]

Apologies if this question has previously been posed in this thread. I did search with keywords, but couldn't find much.

I currently use an upstream Uptone EtherREGEN, which in turn is fed optically from a Startech ET91000SFP2 FMC. I use Finisar SFPs in single mode. In addition, the EtherRegen is clocked with an AfterDark King 10MHz clock. It's a good setup, though as ever I'm looking for possible improvements.

Feeding the EtherRegen optically rather than from ethernet definitely improved my sound quality. However, I've read that though an optical bridge can eliminate electrical noise, it can still be prone to phase noise and jitter. I'm unsure how the Startech clocks its SFP output, but it's not intended for audiophile use.

Is it likely that swapping the Startech for a Sonore opticalModule Deluxe would bring benefit to my system?

I ask as I get the impression that most users of the opticalModule use it for converting optical to ethernet, rather than vice versa. I'm hoping the better clock in the Sonore might pay dividends, or is it so far upstream that differences will be negligible. Remember, the downstream EtherRegen is being re-clocked by the AfterDark King.

Hi

I used to use Startech optical SFPs in single mode with optical Rendu and EtherRegen (no external clocking) before I got my Extreme, and similarly thought the optical and EtherRegen were each very benefical. My optical was downstream from the EtherRegen.

At the time I received my Extreme the recommendation from Taiko changed from optical SFP to copper Ethernet and my Extreme shipped without the SFP card. I tried direct with copper, and with an opticalModule Deluxe converting optical to a short copper Ethernet cable into the Extreme. I found it was easy to hear the additional noise the optical SFPs introduced (per Taiko the optical / electrical conversions are unavoidably more electrically noisy than copper).

So I am currently using EtherRegen from the 'fast side' switching only direct into Extreme (as I found fast to fast switching on EtherRegen sounded better than fast to the slower 'less noisy' side).

I'm due to receive a Taiko Switch & NIC soon and I understand the received wisdom from those on this forum is that the Taiko Switch + NIC outperforms the EtherRegen so am looking forward to playing with that.

Not sure that is exactly what you are asking, but hope its helps anyway

 

[oldmustang]

Apologies if this question has previously been posed in this thread. I did search with keywords, but couldn't find much.

I currently use an upstream Uptone EtherREGEN, which in turn is fed optically from a Startech ET91000SFP2 FMC. I use Finisar SFPs in single mode. In addition, the EtherRegen is clocked with an AfterDark King 10MHz clock. It's a good setup, though as ever I'm looking for possible improvements.

Feeding the EtherRegen optically rather than from ethernet definitely improved my sound quality. However, I've read that though an optical bridge can eliminate electrical noise, it can still be prone to phase noise and jitter. I'm unsure how the Startech clocks its SFP output, but it's not intended for audiophile use.

Is it likely that swapping the Startech for a Sonore opticalModule Deluxe would bring benefit to my system?

I ask as I get the impression that most users of the opticalModule use it for converting optical to ethernet, rather than vice versa. I'm hoping the better clock in the Sonore might pay dividends, or is it so far upstream that differences will be negligible. Remember, the downstream EtherRegen is being re-clocked by the AfterDark King.

Somewhat different context but many months ago in this thread I compared the original opticalModule to a Startech FMC to an unclocked EtherRegen and found the opticalModule was an audible improvement. No reason to believe the opticalModule Deluxe would be a step backwards.

However, even before the introduction of the Taiko switch and new network interface card to my system I went back to using UTP copper ethernet without any fiber conversions from my home LAN router direct to the Extreme and found it to be a significant step up in sound quality.

Steve Z

 

[joet]

Hi Joet, nice to hear the components are improving as they settle in and you have taken measurements. Since there is no conversion or DSP in the Extreme Switch or Extreme Network Cards, it should be impossible for the volume level to change. I thought that measuring the level would be an easy method to prove/disprove any differences but it seems the tests so far are inconclusive. It's probably best done using a test tone rather than a dynamic piece of music. In any case, I can imagine that you may have more than average sensitivity in your ears/hearing and are picking up on clues related to running in. It's entirely possible that hardness translates to a feeling of loudness.

Hi, as promised I've waited a few days and burned in the new card before reporting back. I've run a system disc for 4 days straight and things seem to have normalized.
First of all, I want to apologize for creating churn while reporting on the way things sounded while it was clearly breaking in. I thought people might benefit from the data points. Once again, apologies.
Here's what I've found:
1)There is no longer a perceived difference in volume between the cards and the issues with sibilance have subsided.
2) To me, the differences are subtle, perhaps very subtle, between the network card and the LAN port.
3) The network card seems more controlled and balanced. For example, in Norah Jones' "Come Away With Me" there is restraint when she hits a high note. You don't get the distortion in that upper register that I've heard in every system I've had before this, including with the LAN port. To be clear, it's not as if the note is clipped, it's more as if the system has control of it and is gently guiding it. Overall, you hear the resolution, but it's not calling attention to itself as a unit that is biased toward the top-end might.
4) Some might consider the LAN port to have slightly more sparkle. What it gives up in control it makes up for in liveliness. It calls attention to the high-frequency detail, which often brings a sense of excitement. However, what it gains in excitement might be offset by fatigue.

All this said the difference is small, perhaps very small. For those connecting via Ethernet who have a choice of using the network card to connect to their DAC, it's a matter of taste. If I'd done this experiment before purchasing the card I might not have thought it was worth the effort and $2K. However, now that I have it installed and paid for, my tastes run along the lines of the network card. I will likely keep it because I find it more controlled and less fatiguing. However, I can see others preferring the LAN port and there's a large part of me that believes I'm being too extravagant by making a large purchase for such a small gain.