HOLO Audio MAY DAC

[ted_b]

Powerman, if you'd done your homework as you say, you'd understand that us May owners determined the dac's now-famous value is as an NOS R2R/native DSD combo dac that excels by a large margin when it is sent files in NOS mode.  Combine that with the fact that upsampled and properly filtered and/or modulated music sounds MUCH better than raw recordings...you get the fact that the Holo May is easily at its best when playing, in NOS mode, very high sample rate PCM or very high bit rate DSD.  Hence, the proliferation of HQPlayer settings/options/feedback on the thread.  I mean, other than boring specs, what else do you want to know about the May?  The second best set of answers is simply the list of companies that offer the May on demo.  :)

[sdolezalek]

1 hour ago, ted_b said:

he list of companies that offer the May on demo.  :)

Ha!  There's a thought...  I could make money by lending out my May for demos...Hmmm...Nope, I'm keeping this secret for myself :-)  

 

Actually, at the moment I'm really more interested in how we test whether those that LOVE the May at DSD1024 are doing so due to euphonic distortions or whether there is actually some higher truth to be found beyond DSD512 (even if it costs me more than the May in a new computer capable of doing 1024 with the AMSDM7EC - 512+ modulator and the latest Gauss filters...)

 

My money is on what Jussi has suggested; that the extra "air" in particular that listeners love at 1024, is some form of distortion rather than that next layer of clarity...

 

Obviously, we already know that distortions are much preferred over the more complicated scientific truth in so much of our daily lives these days, so why not in audio? ;-)

[fas42]

45 minutes ago, sdolezalek said:

Actually, at the moment I'm really more interested in how we test whether those that LOVE the May at DSD1024 are doing so due to euphonic distortions or whether there is actually some higher truth to be found beyond DSD512 (even if it costs me more than the May in a new computer capable of doing 1024 with the AMSDM7EC - 512+ modulator and the latest Gauss filters...)

 

My money is on what Jussi has suggested; that the extra "air" in particular that listeners love at 1024, is some form of distortion rather than that next layer of clarity...

 

Obviously, we already know that distortions are much preferred over the more complicated scientific truth in so much of our daily lives these days, so why not in audio? ;-)

 

Ummm, in 30 years I have never heard, "euphonic" distortion ... apart from, say, a Marshall guitar amp. Also, late 50's pop songs sometimes have a syrupy, golden glow to them, from lotsa valves and tape - this works. But extra "air" is just what's on the recording, period. It's the lack of air, er, that's the distortion - that is, the sound of boring, conventional, stereo playback ... .

[sdolezalek]

44 minutes ago, fas42 said:

It's the lack of air, er, that's the distortion - that is, the sound of boring,

For some, silence is golden, for others it is clearly boring.  As per your tag line, you have weighed in on this on the basis that in your "universe, ego is a mighty power..." Admittedly, I was looking for a bit more science and a bit less ego...

[Quadman]

1 hour ago, sdolezalek said:

Actually, at the moment I'm really more interested in how we test whether those that LOVE the May at DSD1024 are doing so due to euphonic distortions or whether there is actually some higher truth to be found beyond DSD512

As I have quoted many times in these forums "you don't know what you don't know until you do know"  You need to hear DSD1024 and make the decision yourself.  I own Audiomatica's Clio system and have REW but neither of those systems is capable of measuring what you seek.  The top of the line AP stuff may be able to measure it, I don't know.  What I do know is nothing brings me anywhere near as much musical enjoyment as DSD1024 and I play with 512 and all the EC filters and DSD256 warms my system up and when I go back up to 1024 I wonder how anyone can really enjoy this (256).  But they probably don't know what they don't know yet.

 

My listening time per month has doubled since I got up to 1024 That's one of the most important measures to me.  I never want to turn the music off even at 4AM.  The Holo May is a truly amazing dac and at its best being sent high rate up sampled DSD content in NOS mode via HQPlayer or possibly some other very good upsampling software, I only know HQP.

 

You'll probably have to wait years and years to hear it with multichannel that your system has.  Stereo, it is available now with the right server setup. I live in Knoxville you get here you can see.

[fas42]

1 hour ago, sdolezalek said:

For some, silence is golden, for others it is clearly boring.  As per your tag line, you have weighed in on this on the basis that in your "universe, ego is a mighty power..." Admittedly, I was looking for a bit more science and a bit less ego...

 

You 'ask' for responses - but you don't like it if you get them . Whether a rig uses standard PCM D/A converters, or something way up there, like DSD1024, what counts is whether the chain produces less distortion than usual. For any number of 'silly' implementation reasons. People want to believe that the magic is in big numbers ... well, no. A big number may allow a particular rig to to reproduce better, but it's purely a fluke of the implementation ...

 

Science will only tell you useful answers if you know how to measure things. But the history of audio is that almost no-one does, properly - that good old standby, ears, does a heck of a good job in the meantime. So, now we go down the rabbit hole of DBTs, etc - and everyone knows how great they are in determining, Truth ... .

 

 

[sdolezalek]

4 hours ago, fas42 said:

You 'ask' for responses - but you don't like it if you get them

Actually, I liked this one quite a bit. 😀 Thanks   I completely agree that bigger isn't necessarily better -- we had the same thing in the mexapixel race in digital cameras; often adding lots of pixels also just adds lots of noise (which then has to be cleaned up in software).  Until recently, it was easier with consumer sized sensors just to go with fewer pixels and cleaner images.  But technology has moved on and my 50 Megapixel R5 is as clean or cleaner than my 1Ds from 5 years ago.  I also agree that we often don't know what to measure or whether it matters, I learned that in trying to build speakers with very low harmonic distortion (until we realized) that none of our customers could tell the difference -- we wanted .0001% in amplifiers but 10-20% was just fine in speakers, I still don't know why.

 

I also know that with my May KTE DAC there are HQ Player settings that sound great for an hour or two but wear on me after a few days, the ones that seem really good even after a few weeks of listening are often not ones that stand out when doing instant comparisons...

[fas42]

24 minutes ago, sdolezalek said:

I also agree that we often don't know what to measure or whether it matters, I learned that in trying to build speakers with very low harmonic distortion (until we realized) that none of our customers could tell the difference -- we wanted .0001% in amplifiers but 10-20% was just fine in speakers, I still don't know why.

 

I also know that with my May KTE DAC there are HQ Player settings that sound great for an hour or two but wear on me after a few days, the ones that seem really good even after a few weeks of listening are often not ones that stand out when doing instant comparisons...

 

It's the type of distortion that matters ... human hearing can adjust for certain 'anomalies', but not others, IME. As an example,  frequency response is way, way down on the importance meter; I once used a standard, old style amp with all the controls for treble, bass, etc - but could nurse it into giving decent SQ for very short periods of time. During those times, I could spin those tone controls from one end to the other, and not 'hear' the sound changing - my brain was, unconsciously, compensating. When that amp slumped back to its normal standard, it then became obvious what turning the knob was doing; too much distortion now, and the ear/brain just rejected it.

 

Good SQ is, for want of a better word, 'natural' - it doesn't stand out as being "special!", if you're not used to what happens  ... but then you raise and raise the volume, and realise that you're not flinching; it becomes immersive, flows over and surrounds you effortlessly - yep, just like the "real thing", .

[James Stephens]

On 2/3/2023 at 8:12 PM, Extreme_Boky said:

Has anyone kept the original USB module? 

I still have the Gen 2.0 module.

 

 

[Miska]

10 hours ago, James Stephens said:

I still have the Gen 2.0 module.

 

I also still don't have any device with the new one:

NAA output network endpoint: Holo Audio UAC2.0 Gen2 Standard: USB Audio

 

No complaints, objectively or subjectively, so I haven't rushed changing the module.

 

Of course likely the next one will be with the new one!

 

Although right now I'm playing with a kind of optionally alternative - Holo Red.

 

[Allan F]

17 hours ago, sdolezalek said:

I also know that with my May KTE DAC there are HQ Player settings that sound great for an hour or two but wear on me after a few days, the ones that seem really good even after a few weeks of listening are often not ones that stand out when doing instant comparisons...

 

In my experience, whether from a component or a recording, sound that really stands out when listening in the short term often tends to become fatiguing in the longer term.

[davide256]

On 2/6/2023 at 10:38 AM, Miska said:

 

0 - 22.05 kHz sweep at 44.1k NOS:

 

0 - 22.05 kHz sweep, with upsampling to 705.6k:

 

0 - 22.05 kHz sweep, with upsampling to DSD256:

 

Does this change if the input is coax vs USB?

[Miska]

1 minute ago, davide256 said:

Does this change if the input is coax vs USB?

 

No, type of input doesn't have any impact on this. (it is the same data at same sample rate after all)

 

[Kalpesh]

On 2/8/2023 at 5:37 PM, Miska said:

 

You cannot have proper noise shaping without upsampling. Higher the output rate is, better the noise-shaper works.

 

If you are familiar with statistics. Higher your sample set is, better accuracy your statistics also has. Comparing NOS to statistics, is like randomly picking one item from the set and then assuming it to correctly represent the set it is from.

 

 

That is wrong, usually that is weak point of R2R, although as I have shown, you can help it with high rate, less bits (using only linear region of the ladder) and suitable noise-shaper.

 

thank you,

I learnt many interesting things and you clarified false assumptions. Funny thing is that to start with, and triggering this search crisis,  I had a mess of live vs neutral wiring/plugging consecutive from migrating from several dedicated electrical lines/breakers (not a bad idea per se I think but a I had a mess of extension cords) to a neater organization around the Supra mains block + dedicated (well, just normal sockets on different breakers but no more extension cords !) lines/breakers for computers and amplifiers. Wonder how people design reliable testing conditions for power cords and mains blocks but I sure could hear negative effects of not having the neutral of devices plugged in the neutral line. Sorted things out, redid convo filters just to be sure and now I'm happy for several days with Sinc Mx and ASDM5EC (V2 is too much load) with sometimes the need to switch to PCM/LNS9 with 192 files while keeping Sinc Mx as the filter.

Questions :

how "low" in SDM modulators would you go before switching to PCM due to cpu overload ?

 

A quick parsing seems to reveal that even with a latest Mac Studio there would remain frustrations with regard to HQP ; correct ? or is there yet a worthy alternative to original MM1 ! Gb (I'll keep a Mac for side usages) ?

[James Stephens]

13 hours ago, Miska said:

 

I also still don't have any device with the new one:

NAA output network endpoint: Holo Audio UAC2.0 Gen2 Standard: USB Audio

 

No complaints, objectively or subjectively, so I haven't rushed changing the module.

 

Of course likely the next one will be with the new one!

 

Although right now I'm playing with a kind of optionally alternative - Holo Red.

 

 

Very interested to hear how you make out with the Red.

 

I haven't touched my system in 2 years but I have one on order to experiment with as an NAA to potentially replace my Aries G2.1. 

 

Not going to purchase HQPlayer Desktop just yet though ... 😉

 

James

 

"The rabbit hole went straight on like a tunnel for some way, and then dipped suddenly down, so suddenly that Alice had not a moment to think about stopping ..."

 

[Miska]

3 hours ago, Kalpesh said:

Questions :

how "low" in SDM modulators would you go before switching to PCM due to cpu overload ?

 

A quick parsing seems to reveal that even with a latest Mac Studio there would remain frustrations with regard to HQP ; correct ? or is there yet a worthy alternative to original MM1 ! Gb (I'll keep a Mac for side usages) ?

 

Regular ASDM5 or ASDM7 is also fine. If you need to go for PCM mode, 16x or 32x output rate are best choice (minimum 8x), helped with noise-shaper at 20-bit.

 

Mac Studio with Ultra can do a lot, but not necessarily DSD512 with EC modulators, although there's one user report that it would work too. But Mac Mini with M2Pro should be pretty good for DSD256 output.

 

[KenMoreira]

@Miska

I can run my pc at ASDM7 with DSD 512 
or ASDM7ECv2 at DSD 128

which one is the better choice ? a better filter or a higher DSD Sample rate

[Miska]

1 hour ago, KenMoreira said:

I can run my pc at ASDM7 with DSD 512 
or ASDM7ECv2 at DSD 128

which one is the better choice ? a better filter or a higher DSD Sample rate

 

I would personally go with at least DSD256. DSD128 is on the low side.

 

Can you run for example ASDM5EC at DSD256?

 

[KenMoreira]

1 minute ago, Miska said:

 

I would personally go with at least DSD256. DSD128 is on the low side.

 

Can you run for example ASDM5EC at DSD256?

 

seems i can with Poly Sinc Gauss Hires Ip (or LP cant tell what letter that is)
with nx Sinc L

not the ECv2 though.  But yeah otherwise that does seem to be working.
stick with that then?

[Miska]

Just now, KenMoreira said:

seems i can with Poly Sinc Gauss Hires Ip (or LP cant tell what letter that is)
with nx Sinc L

not the ECv2 though.  But yeah otherwise that does seem to be working.
stick with that then?

 

Yes, any ASDM5EC or ASDM7EC is big improvement over regular non-EC. v2 is heavier than "v1", but the quality difference is rather small.

 

[semente]

On 2/9/2023 at 10:32 PM, fas42 said:

 

Ummm, in 30 years I have never heard, "euphonic" distortion ... apart from, say, a Marshall guitar amp. Also, late 50's pop songs sometimes have a syrupy, golden glow to them, from lotsa valves and tape - this works. But extra "air" is just what's on the recording, period. It's the lack of air, er, that's the distortion - that is, the sound of boring, conventional, stereo playback ... .

Try shelving up the treble above 10kHz by some 3dB, maybe 6dB if your extreme top listening capabilities have been crippled by age and/or abuse, and you'll hear (added) 'air' that's not on he recording.

[fas42]

12 hours ago, semente said:

Try shelving up the treble above 10kHz by some 3dB, maybe 6dB if your extreme top listening capabilities have been crippled by age and/or abuse, and you'll hear (added) 'air' that's not on he recording.

 

Probably would work, because it emphasises the spatial cues that are in the recording - for many systems, this is two edged sword territory; it will also highlight all the weaknesses in the replay chain, and make listening a real chore, for many albums.

 

Only solid long term solution is to sort out the rig. Then the frequency curve becomes irrelevant, and volume dictates the level of involvement with the musical goings on. Just like experiencing live, acoustic instruments, if you are right up close to the action, you hear so much going on, the tiniest inflections of sound, of all types, are obvious; move right away, a good distance from the musicians, then the fine stuff is no longer apparent, but the sense of the musical occasion is still fully in place - you don't have to physically move with competent audio playback, the gain control performs the same function.

[Tihon]

On 2/11/2023 at 12:45 AM, Miska said:

Although right now I'm playing with a kind of optionally alternative - Holo Red.

What do you think about Holo Red vs opticalRendu (assuming it would be used with extremly good PSU)? Red is newer, but oRendu's network isolation is still superior.

[Tihon]

On 1/17/2022 at 8:45 PM, Bertel said:

Bottom line:

Should you encounter difficulties uprading the Holo May (and likely also the Spring) to the latest firmware, make sure it's not active as the primary audio device (unlike in the USBFirmwareUpdateREADME.pdf guide, where the picture shows it as being active and selected, didn't work for me, I had to use my laptop for this, because my server has no alternative speakers so when connected directly the May always is default and active). This did the trick for me.

Thank you for the info. I was going to contact dealer, when I saw this and everything finally worked.

[Hensema]

@GoldenOne Have you measured the linearity of both the May L2 and May KTE? I’m wondering what the difference is because of the hand picked ladders from the KTE version.