LAiV Harmony R2R DAC Impression and Reviews
May 29, 2024 at 12:45 AM Post #241 of 251
Why do we care if a twenty-something can hear above 20 kHz? Do we need proof of everything to accept anything?
If one claims to appreciate harmonics beyond 20khz as appreciable to the experience, yes.

Digressing into snake oil purchase justications down that rabbit hole which lends itself to religious beliefs rather than the realm of audio reproduction is neither productive nor appropriate here.

I addressed a particular claim which boils down to ones opinion or one's own optimization of their chain; not a declatory Truth.
 
May 29, 2024 at 12:53 AM Post #242 of 251
I disagree on a few things. Or let me redefine the framework. And get some definitions straight. Broaden the horizon so to say.

I learned a lot in school, but what I learned best is that most teachers have no idea what they are teaching you. Besides the language and math you get indoctrinated with a LOT of ideology. One of the Fairy Tales is that by learning you will be introduced into the elite club of Reasonable Men, devoid of alterior motives or emotion. They say "Trust the SCIENCE". And with that you are an acolyte in the new religion of Scientism. Which has its own dogmas, gods and priests. Modern science as it is discribed has very little to do with the pursuit of knowledge but a lot more with establishing consensus. Denying their own foundation in 1 particular religion or worldview. That the universe can be studied and abides to common laws. Laws that apply always and everywhere the same. Without that faith there is no workable scientific method. So much for "science as opposing to "religion". They MUST go hand in hand.

The thing I would like to pull apart here is timing. I was talking about sound. Not timing of digital signals. Analog sound as perceived with or sense of hearing and the brain as processor of that digital input rendered by the synapses. Yes, our eyes and ears are biological digital sensors. Lots and lots of computing going on in that grey matter between our ears.
But here is my point: over 90% is not conscious. Like in a computer most things are machine instructions (lower level language) that are not visible for the user on a UI (user interface). Yet errors can very much throw a spanner in the works. Most of our hearing is subconscious. Just put one finger in your ear and notice how your 3D perception is totally distroyed. That is the same as having a hearing aid: you can't follow a conversation in a crowd because you can't locate anymore on timing differences between 2 ears. And thus you can't focus on one location you are listening to.

Back to my 2 dimensions of sound: amplitude and timing. And information and noise. Noise in amplitude is simply called noise. And you know instantly what I mean, signal to noise. That reduces the amount of information. When a sound is twice as loud as a similar sound or overtone, that's 10dB difference, the lower sound is masked. Yet when there is noise from everywhere you can still perceive 1 signal if it's 40dB (iirc that is) below the noise floor. Like a voice in the crowd. That is pretty amazing, 10⁴ difference in volume.

That is also due to timing. But what if that voice in the crowd is like 4 voices saying the same thing? Or 10? That's not so easy anymore. Heck, it's hard to understand a choir singing the same song even when the background is quiet. And how irritating is it when one eager beaver is starting to sing just ahead of the rest because he has a bad sense of timing?
Back to those 4 voices in the crowd. You can't focus on 1 pinpoint anymore. It's turned into an area, a wider spot with differences in timing and amplitude. That is noise in the timedomain.

But what to call noise in the timedomain? Distortion? Smearing? Confusing? Quantum physics? It is confusing. And most often it's not even consciously. It seems that people can believe 2 conflicting statements as true at the same time. That's called cognitive dissonance and you can often tell because they fall flat, their brain goes on tilt when asked about one or the other. It also causes headaches and irrational behaviour. That's all unconscious. The brain can't produce 1 clear conscious answer. Try and scientifically study that! You'll get a different answer every time. Like Schrödingers cat.

Besides the tangents, timing of sound is much more important than we think, because it's in our subconscious non-think area. It's the iceberg we can't see.

Finally, on NOS and oversampling.
NOS doesn't cause HF roll-off, it's the filtering that does. It's the pre- and post ringing that makes the HF sound clearer (in other words its artificial detail). Oversampling causes pre-ringing. That like the arithmic singer in the choir. A pond never starts to ripple before the stone hits the surface. That causes our brain to come up with an explanation from the natural world that makes the impossible possible. Result: brain freeze, cognitive dissonance, headache, listening fatigue... Post-ringing is much less of a problem as is is much like physical resonance, reverberation and harmonics.

And now my sense of timing tells me I've ranted for far too long. 😄
I was referring to the human perception of jitter. There are many claims of night and day preferences which can be reduced to purchase validation.

There's some OKAY and affordable DDC offerings. There's music tailored word clocks which will run you 4k to pair a DDC/clock combo. There's the Holo Red which might not have everything one is looking for, however, has the best audio analyzer results to date.

As for NOS and digital reproduction filters, there's no way around red book roll off without OS filters due to frequency dependent non-linearities. This is the attractive part of NOS; some prefer limited to no digital interaction while others use this aspect to get out the way of their high performance software US/OS filters solutions without a modulator in the way.

As for ringing in time domain, you physically/literally can not reproduce a digital signal in the analog domain without either pre-ringing or a frequency dependent phase shift to facilitate post ringing with delta sigma equipment. Psuedo-NOS/Ultra Slow DS filters which approximate a NOS impulse response and ringing characteristics have heavy roll off beginning around 16.5Khz as currently implemented.
 
May 29, 2024 at 1:19 PM Post #243 of 251
I was referring to the human perception of jitter. There are many claims of night and day preferences which can be reduced to purchase validation.

There's some OKAY and affordable DDC offerings. There's music tailored word clocks which will run you 4k to pair a DDC/clock combo. There's the Holo Red which might not have everything one is looking for, however, has the best audio analyzer results to date.

As for NOS and digital reproduction filters, there's no way around red book roll off without OS filters due to frequency dependent non-linearities. This is the attractive part of NOS; some prefer limited to no digital interaction while others use this aspect to get out the way of their high performance software US/OS filters solutions without a modulator in the way.

As for ringing in time domain, you physically/literally can not reproduce a digital signal in the analog domain without either pre-ringing or a frequency dependent phase shift to facilitate post ringing with delta sigma equipment. Psuedo-NOS/Ultra Slow DS filters which approximate a NOS impulse response and ringing characteristics have heavy roll off beginning around 16.5Khz as currently implemented.
So we agree pretty much. Just talking about different aspects. I was specifically not talking about digital time domain (jitter). And you do. You speak about ringing and sigma delta, I about it's absence in NOS R2R.
Look at the first page of the Pasithea review on 6Moons. That plot is so clean!

One thing I forgot to mention: I chose to remove all analog filtering from the TDA1543 DACs I did because it distroys the benefit of high res material if you neuter it >15kHz. The consequence it with test-tones going up to 20k you will hear aliasing tones going the other way. But in actual music there are no high treble that loud. Levels at >10k are already so low you can't hear anything of that specific noise. So it measures like crap, but sounds fantastic.

Also, I strongly prefer 96k material over standard 44k Redbook. I don't really consider 44k particularly audiophile (depending on the recording though). You are not bound to 44k to listen in NOS mode. With 24/96 flac there is no HF roll-off you can hear. So NOS=roll-off is a non sequitur if you define your parameters correctly.

I came across some really affordable DDC and clocks on AliExpress. Really not rubbish. But a bit amateurish DIY looking acrylic 'case'. Parts are premium though, Amanero+FPGA+Femtosecond clocks; €114 (incl p&p and VAT).
https://a.aliexpress.com/_EzWKZPZ

And this one with only XMOS chip is even cheaper with the standard crystal at €50 (€100 with best clocks)
https://a.aliexpress.com/_EukOYqL

And the simple but useful Douk Audio U2 at €50
https://a.aliexpress.com/_Ey9m12n
 
Last edited:
May 29, 2024 at 3:37 PM Post #244 of 251
I came across some really affordable DDC and clocks on AliExpress. Really not rubbish. But a bit amateurish DIY looking acrylic 'case'. Parts are premium though, Amanero+FPGA+Femto second clocks; €114 (incl p&p and VAT).
Sounds very interesting! Do you have a link?
 
May 29, 2024 at 6:51 PM Post #245 of 251
Sounds very interesting! Do you have a link?
I was editing... I always first finish the essay and edit the specifics in later. Links are added now.
 
May 30, 2024 at 1:43 PM Post #247 of 251
Is the LaiV Harmony generally a bright, neutral or warm sounding dac ?
Hard to tell without properly defining what you mean by those terms. They are very subjective, and can mean all sorts of things to different people. One man's neutral is another ones lifeless, one man's warm is another ones lacking detail.

Just read or watch some reviews. Most will say neutral with a touch of warmth, detailed but not fatiguing and with body to voices.

In your location it should be easy to get a demo somewhere
 
May 30, 2024 at 7:19 PM Post #248 of 251
Hard to tell without properly defining what you mean by those terms. They are very subjective, and can mean all sorts of things to different people. One man's neutral is another ones lifeless, one man's warm is another ones lacking detail.

Just read or watch some reviews. Most will say neutral with a touch of warmth, detailed but not fatiguing and with body to voices.

In your location it should be easy to get a demo somewhere
Thanks. Indeed neutral with bit of warmth seems to be the consensus. Certainly it all depends on one’s system synergy and what benchmark is used to compare.

I’ve heard it before during Camjam; together with the Immanis and Armageddon amp, it was an amazing combination in terms of the dynamics, transparency and staging ! But of course that is an expensive set up.
 
May 31, 2024 at 12:47 AM Post #249 of 251
May 31, 2024 at 3:20 AM Post #250 of 251
So we agree pretty much. Just talking about different aspects. I was specifically not talking about digital time domain (jitter). And you do. You speak about ringing and sigma delta, I about it's absence in NOS R2R.
Look at the first page of the Pasithea review on 6Moons. That plot is so clean!

One thing I forgot to mention: I chose to remove all analog filtering from the TDA1543 DACs I did because it distroys the benefit of high res material if you neuter it >15kHz. The consequence it with test-tones going up to 20k you will hear aliasing tones going the other way. But in actual music there are no high treble that loud. Levels at >10k are already so low you can't hear anything of that specific noise. So it measures like crap, but sounds fantastic.

Also, I strongly prefer 96k material over standard 44k Redbook. I don't really consider 44k particularly audiophile (depending on the recording though). You are not bound to 44k to listen in NOS mode. With 24/96 flac there is no HF roll-off you can hear. So NOS=roll-off is a non sequitur if you define your parameters correctly.

I came across some really affordable DDC and clocks on AliExpress. Really not rubbish. But a bit amateurish DIY looking acrylic 'case'. Parts are premium though, Amanero+FPGA+Femtosecond clocks; €114 (incl p&p and VAT).
https://a.aliexpress.com/_EzWKZPZ

And this one with only XMOS chip is even cheaper with the standard crystal at €50 (€100 with best clocks)
https://a.aliexpress.com/_EukOYqL

And the simple but useful Douk Audio U2 at €50
https://a.aliexpress.com/_Ey9m12n
Yes.
What I forgot to address with NOS redbook;
44.1/48khz may not roll off egregiously early depending in implementation but it will always have aliasing requiring "hi-res" files or upsampling for the worried about this to push all aliasing images well beyond audible FR as well as mitiging 16-20khz roll early roll off affecting tonality.

Only truly top tier NOS implementations will be ruler flat 20-20khz. (Yet still have red book aliasing) while providing ideal impulse response with no phase shift.

96khz is indeed all you need to academically not worry about either of the above issues. Most streaming "hi-res" is 24/192 standard with random MQA unfolded sampling at 88.1khz (tidal users with non-mqa dacs)
 
May 31, 2024 at 3:53 AM Post #251 of 251
Hi,

I own at this time a Merason Frerot + Pow1 and a Holo Audio Cyan 2.
I'd like to sell them to buy:
- Laiv Harmony (2999€ here in France)
- Merason DAC1 on the second hand market. (3700€)

I like the analog sound of the Frerot but the reviews of the Laiv make me hesitate...
And the small footprint of the Laiv make it easier to stack.

Is anyone try the 2?
 

Users who are viewing this thread

Back
Top