KTE Spring2 technical measurements

That's my experience, both with the Spring 1 I owned a few years ago and with the Spring 2 I own now (ordered Oct 23, arrived late Dec).

 

This one took 7 weeks to arrive after ordering. Still getting used to it after years with a Yggdrasil, it was not love at first listen. I can post impressions after a few weeks if there's interest.

 

FWIW, I talked to them about ordering last week and the Spring2 has run its course (that's why its out of stock). The Spring 3 is due sometime in April.

 

Did they say what's new in Spring 3? Spring to Spring 2 was quite an improvement, the first NOS DAC I've owned where NOS attributes are not at the expense of clear losses in some situations to competitive oversampling DACs. Of course the following is badly unidimensional and subjective, but here's my multibit DAC hierarchy:

dac1541 < Holo Spring KTE level 3 < Metrum Onyx < Sonnet Morpheus < Yggdrasil A2 < Spring 2 KTE

I currently own the 3 rightmost ones. For them + dac1541, the impressions have been with similar amps (ecp DSHA-3F and EC Af) and headphones (ZMF Verité closed and open). What Spring 2 KTE gets better than everything else above is combining tonal density with crisp detail and convincing dynamics (both micro and macro) that make the Yggdrasil sound like it's trying a bit too hard at times (this is on headphones, never had that impression when I had the Y. on a good speaker system).

 

It will not be back in stock. It’s superseded by spring3 which is released in April. We hoped Spring2 components would last long enough til then but it sold out before our release of its replacement. Hot product!
The spring3
L1 is 1998
L2 is 2298
And
KTE 2798

And each will have an optional preamp module for additional 500usd. Many customers wanted a volume control and this will be a superb solution for that.​

I asked if they anticipated any sonic differences/improvements, but haven't gotten a response yet. I can ask more questions if anyone has them...

 

Would love to hear what you think. Am attempting to decide between the two (well, between Yggdrasil and Spring 3, apparently).

 

Interested. Please do


What I'm most curious about is the difference between the May and Spring 2 or 3 but they never got back to me w/ any info about a May or Spring 2 or 3 beyond an initial email saying (in a round about way) that
1) the only way to hear one (from them) was to buy it and
2) explicitly said no loaners due to prior bad experiences on their end

 

I own both Spring 2 KTE and Yggdrasil, but they are 150 miles apart so I don't really have side-by-side impressions. With similar amps, EC Af with comparable but not identical tubes, I generally prefer the Spring 2 for 88.2 or higher sample rate material: a bit more coherent, never in a hurry but never slow, many layers of detail to the teeniest level. However, for 44.1 or 48 material, Spring 2 is a bit less precise, a bit more smeary than Yggdrasil. I compromise by upsampling 2x all lower rate material for Spring 2 in Roon which cleans up the act but it is not as convincing as the Yggdrasil's proprietary upsampler, which is really special.

 

Here’s an oversimplified summary of my notes from less than a week of listening time:

Spring 2 KTE

• Delicate presentation
• Realistic acoustic instruments and voices
• Wide, deep stage
• Synergy with tubes and dynamics such as DNA Starlett, Utopia and HD800S

Yggdrasil A2

• Powerful presentation
• Satisfying bass texture
• Synergy with planars such as HE6se, LCD-4 and Diana Phi

There’s not a clear winner here IMHO. I think this is a textbook preference / system synergy comparison. If I didn’t care about planars this would be much easier, but I still can’t quit them.

Spring 2 KTE’s character reminds me in many ways of a DAC version of the HE1000. Sometimes a little ethereal or soft.

Oversampling sometimes creates an uncanny valley experience for me when listening with acoustic recordings. (I didn’t even realize this until I tried the Abbas loaner!) NOS at least partially breaks through the valley and makes me feel closer to the music. The impact of NOS was far less noticeable with electronic recordings.

 

Hmm... I think based on my audition of a friend unit in his speaker system the Spring 2 has a strong flavour and very forward sound. Its sound characteristic is not for everyone and needs system matching.

 

I thought Spring 2 Wildism sounded so very layered and natural. That to me was its strongest point. I didn't bother with upsampling etc though, just pure NOS on all. But I never heard an Yggdrasil to compare..

 

Thanks much for those thoughts - helps me to triangulate things. I nearly exclusively listen to redbook, and I’m quite used to the Schiit multibit sound after 10 years of use out of various iterations of Bifrost (now with BF2), so I’ll likely spring for a used Yggdrasil and consider Holo if that proves to be too much of a sidegrade/evolution in my rig from the BF2.

Thanks again !

 

This was what Tim at Kitsune Hifi, the US importer, said 10 Jan 19:

The overall spring1 vs spring2 comparisons are marginal and was not released to be considered an upgrade of the spring1 but rather a continuance or an already great dac with a few improvements. Personally I would not strongly recommend anyone to “upgrade” from spring1 to spring2 since it’s a somewhat small difference imho. Otherwise we would of changed the name if we felt it was a big enough difference or improvement from the spring1.
(From the AudiophileStyle thread about the Holo Spring)

Whenever a salesman tells me not to buy something it always causes me to think. Interesting that some here find significant differences, perhaps this is an illustration of the relatively small margins between DACs as opposed to transducers. Maybe the original Spring isn't much behind the Spring 2 - I certainly hope so as I have one sitting in a box (no system set up at present), I just wish it was smaller and had volume control.

 

I've owned both, top KTE versions. There was a gap of 2 years+ between them, in the meanwhile my amps and headphones changed too, so it's hard to compare. Still, from memory, Spring 2 sounds slightly more contained, slightly less smeary, a tad faster than Spring 1, while maintaining the tonal density I liked in Spring 1.

 

My own opinion - I didn't like the spring 1. I own a spring 2 and honestly went into it expecting myself to not like it.

Spring 1 had some "good stuff" but still had some of the stereotypes of things wrong with some Chinese DACs soundwise (i.e. - not knock off components, questionable BOMs, etc). Was a bit too mushy, soft, "no momentum", undefined, a few odd tonal and timbre issues here and there, etc. If I had to make a wild guess - this stuff is probably great for Chinese ballards, emphasized high female voices, etc. that a lot of people in mainland China like to listen to, especially in the typical settings of most Chinese apartments. Spring 2 cleaned this stuff up a bit while retaining the "good stuff".

 

Did they discontinue the Spring 2?

https://www.kitsunehifi.com/product/holo-audio-spring2-dac-kitsune-tuned-edition/

That's weird.

Edit: Seems like the third iteration is coming soon and it measures perfectly. A glorified Topping? I hope not…

https://www.audiosciencereview.com/...o-audio-spring-3-dac-wpreamp-prototype.22083/

 

Holo Audio KTE Spring2 has been powered up 11770 hours in my lab as of this post. Time to revisit a few select measurements.

Dynamic Range - 44 KHz AES input Bal output

Having settled in the KTE Spring2 is now rivaling the dynamic range measured on the KTE May. This is mind blowing performance.

Distortion - 44 KHz AES input Bal output

Distortion still has quite a few zeros in each of the specific measurements. Look at Hum which is now at 0%.

So one might ask how such high Dynamic Range and Low Distortion measurements are possible to measure on an analyzer, especially one that is not an APx555. There are measurement options on the dScope that apparently aren't easily accessed on the APx555 if they exist at all.

Using an external digital loopback to examine the dScope digital signal generator and sending the received digital signal to the dScope analyzer, a very clean sine to -180 dBFS may be observed. Distortion measures 0% for every metric displayed.


Now evaluating the dScope analog inputs with a 50R input termination, a noise floor of -180 dBFS is also observed, though it does rise a little below 60 Hz to -170 dBFS.


The above two measurements imply the dScope is capable of measuring DACs to a level of 29.6 ENOB across a spectrum. To date no such DAC exists. The yggdrasil, dac2541, wavelight, and a few other of the finest DACs in existence today run 19.6 to 20.6 ENOB. Holo Audio KTE May and KTE Spring2 (now settled) are capable of 23 ENOB.

*edit*
Please keep in mind that ENOB is only a metric, one that often is mis-used and abused on other forums. It also isn't the only metric responsible for the great sound experienced with the Holo Audio KTE Spring2 and KTE May.

Not all that matters can be measured. Not everything that can be measured matters.- Jonathan Novick paraphrasing Albert Einstein.

 

dScope running a one hour residual noise measurement demonstrating ability to evaluate DACs to 30 ENOB:

-180 dBFS at 20 Hz on the unbalanced analog input, closer to -185 dBFS at higher frequencies.

 

Did you ever get a chance to measure the Spring 3? I know you were talking about plans to do so in the loaner thread, but to my knowledge you have not done so

 

I have measured the Spring 3 over a period of time from April to June of 2023. Several comments.

1) When I am able to muster the focus required to review and vette approximately 300 measurement collections I will post.

2) Spring 3 measures substantially similar to Spring 2, meeting or exceeding the Spring 2 performance

3) I bought the Spring 3 for my lab. There are some subtle differences between 2 and 3 that influenced me to keep both.