Ok so even if the technologies are different we should be both at the top here wrt transparency and accuracy.
Well, cables are a bit controversial subject. I confectioned my XLR interconnects with a surplus run of aerospace/military signal cable (so shielding, for instance, is quite a thing) and speaker cable is Sommer Cable Quadra Blue used like a “big star quad”.
I built my own passive power conditioning system, and use Eupen EMC power cables. If their ability to eliminate EM/RFI noise is good for nuclear plants and particle accelerators (they are actually built for those) then they should be ok for me as well ;-)
Good.
Quite difficult. I am about 6000 miles from you.
Such as?
Well, here's the rub: The device does not aim at being able to drive absolutely everything. Some amps it may better drive directly, other ones via a preamp. There is nothing wrong with the design, in fact it has some clever ideas right from the first IV conversion. It may not drive some more demanding loads because it has a small power supply. Since the Universal Buffer by Neurochrome is very easy to drive, then this may be the reason for a total absence of graininess, and for the almost unreal smoothness of the treble.
In fact, it may as well be that the differences heard between this DAC and other ones, even significantly more expensive, all boil down to electrical interfacing with the next stages. Because, otherwise any graininess, macro-dynamic or micro-dynamic defect would be detectable as distortion.
This DAC has replaced a Soekris DAC1541 and there was not even contest. Maybe I am in a honeymoon with this DAC, but it could be a definitive product – not because it cannot be improved, I am sure it can, but almost any other improvement in any other component would be of a greater order of magnitude.
For instance, I am of the "persuasion" that such a construction would just be a waste of cycles (and energy).
hmmm...
well, I would just say that running direct results in less distortion than adding an additional preamp. And my experience suggests that the OPAs sued in the output stage are not really the right choice for driving a cable interface: the have very low current capabilities.
If you are of the belief that a distortion measurement based on a single tone tells you all you might need to know about how a component actually sounds when listening to music, so be it. My experience suggests otherwise.
"For instance, I am of the "persuasion" that such a construction would just be a waste of cycles (and energy)."
Since you have a D-90, perhaps actually testing such would give you more than just a "persuasion". And I would also note that Jussi's Lasko has measured the improvement available when inputting DSD 256 to the direct mode of the other AKM chips, and AKM's own specifications (see the 4499 data sheet) also show this improvement.
well, I would just say that running direct results in less distortion than adding an additional preamp. And my experience suggests that the OPAs sued in the output stage are not really the right choice for driving a cable interface: the have very low current capabilities.
If you are of the belief that a distortion measurement based on a single tone tells you all you might need to know about how a component actually sounds when listening to music, so be it. My experience suggests otherwise.
"For instance, I am of the "persuasion" that such a construction would just be a waste of cycles (and energy)."
Since you have a D-90, perhaps actually testing such would give you more than just a "persuasion". And I would also note that Jussi's Lasko has measured the improvement available when inputting DSD 256 to the direct mode of the other AKM chips, and AKM's own specifications (see the 4499 data sheet) also show this improvement.
It depends also on the intended usage. The LME49720 have also a better phase behaviour than, say, the OPA1612. This actually helps reduce the distortion further in a feedback circuit (if you want a no feedback single ended circuit, the OPA16xx series is better). So, if you have a long cable this output stage is not optimal. I have 4 feet interconnects. Are they already too long? When is too long for this output stage? AFAIK driving a few meters of cable is fine with the LME49720 at the output.
I am definitely not, and on ASR I have tried to explain to people why this is complete crap. Measuring DACs always at 4V may in some cases increase the distortion that one normally observes, say at 0.5V levels or less (more common when listening at home) if the output stage is a bit stressed. In other cases the distortion may increase at lower levels or, if the DAC circuit is a sign magnitude R2R Dac, the distortion might be the same... in other words this is an arbitrary value that not only does not tell the whole story, but may be even misguiding. With noise, at usual output levels the noise is of course usually higher, and may more easily drown the distortion if the latter at the same time decreases or stays equal. In other words, anything can happen and this might explain why my system sounds different if I attenuate less digitally and put attenuators in the signal...
I do not doubt that these things can be measured. In fact, I know that they can. The question is whether they are audible. And for instance, we know that they response to impulse rendition is debated, with no conclusive evidence (AFAIK), and I am not talking about primitive oversampling methods like Track and Hold, that can still be found in some chips.
This said, I am always open to update my knowledge and change opinion. But "graininess" for voices and cymbals on the D90, nah
Last edited:
One person's opinion is hardly something that proves there something 'wrong' with the D90.
Ah you have been drinking JAM brand kool aid again.
So you are basically saying 'if your ears are not good enough and your system not good enough in a different way?
As personal opinion, I don't think stationary measurements of nonlinear distortion within the audio band and maybe a little above that tell the whole story. For example, I do not believe the ESS 'hump' is a stationary distortion even though it does show up as something using standard stationary measurement techniques. I do think HF/RF noise coming out of dac analog outputs is non-trivial and it is a factor that can cause nonstationary yet audible distortion in opamps that may or may not show up very well as measured by an AP. I strongly believe clock close-in phase noise is a real issue affecting sound quality even though few have the expensive test gear to properly measure it. I do think the audible effects are in principle measurable and perfectly in keeping with standard physics. However, as a matter of common practice we do not have a good standard measurement for its nondeterministic audible effects in dacs. There is more, but I will stop here. My personal opinions, that's all.
Standing ovation!
I was temped to just try to find the "ignore" setting on the forum once I read that a phono system would be superior, esp if you have the "best of everything". In other news, people that spent inordinate amount of money on bling will never admit that cheaper stuff can be as good. The most expensive turntable can be preferred by some, but it will never reproduce the signal as well as a decent modern digital system.
Some people just do not realise they are in an intellectual cage. You detect them by the fact that if you do not belong to the same cage then you MUST be in a different cage.
If you steer from the purely objective approach on ASR you are labeled as a "pure subjectivist", whereas if you, for instance, here dare doubt that the capacitor brand on the power supply may have an effect on sound, you are labeled as a Vulcan pure objectivist that should retreat to ASR. And let us not mix SBAF of head-fi or other niches in this mix.
Signs of this insanity: that we value tiny changes more than other ones which have a much stronger impact on sound (such as tweaking headphones, changing their pads, adding a little resonator there, or modifying the room treatment). At least in a DIY forum we are supposed to have fun.
I should add that personally I do not use phono, but I know people who are not crazy who like it more than they like digital. They listen in a different way than I do. That is, they listen to sound in a way that is not focused in terms of how we measure. I know its hard to believe. Didn't used to believe it either. I understand both sides now, objectivist and subjectivist (if one wants to name them inaccurately), having spent some time with both. Neither is crazy and both are quite biased.
Agreed.
Well, I do not believe that it is relevant. It is "measurbation".
Agreed.
I think that this is a mostly solved problem in modern gear?
I tend to agree with you. The trend to reduce everything to one value, the SINAD, measured at an output level that is almost never relevant to listening is a clever marketing tactic to draw readership to a forum (and get clicks). Measurbation.
Personal preferences have to be respected – they do not imply that a system is better than another. A friend of mine still prefers phono to digital, esp in the treble. He is used to a certain smoothing and warmth that if anything corresponds to a certain treble irregular attenuation.
If anybody prefers that, fine, but when they try to claim that because of this phono is better than digital, and the latter can never be "hi-fi", well, then I think we have a problem.
The D90 may have some graininess wrt DAC X. Ok. Which one is more correct? The D90 or DAC X? Or are they both off the mark in a different way, and there are more important things?
Regarding close-in phase noise, I don't think it is as solved as it might be. I have tried Accusilicon, Crystek 957, NDK SDA, trying to see if I can get my hands on an NDK DuCOLoN. The local NDK rep is a friend of a friend, so we'll see. Ended up using Crystek. Won't go into all the details of squeezing the best I could out of those clocks. It seems to me that my very best work with them is the best I personally have heard from AK4499. Don't know if there is audibly better or not. I am working on trying to find out though, one way or another.
Last edited:
Quote from a much smarter guy than me:
"Some purists argue that this is some sort of violation of the recording, that playback should be
as accurate as possible. It's a valid opinion, but it's not my purpose to argue that point here.
My own view is that this is entertainment - not Dialysis - and the customer should use
whatever makes him happy. After all, he is the guy who pays for it."
Taken from :
http://www.firstwatt.com/pdf/art_h2.pdf
I'll add that following your conversations here is entertaining too.
Best wishes to all of you guys.
"Some purists argue that this is some sort of violation of the recording, that playback should be
as accurate as possible. It's a valid opinion, but it's not my purpose to argue that point here.
My own view is that this is entertainment - not Dialysis - and the customer should use
whatever makes him happy. After all, he is the guy who pays for it."
Taken from :
http://www.firstwatt.com/pdf/art_h2.pdf
I'll add that following your conversations here is entertaining too.
Best wishes to all of you guys.
And HE is absolutely right.
In fact, I also contend that "I bought it because I like it" is a better path to sonic bliss than "I bought it because it has better measurements".
As long as we do not mistake preference for reproduction quality and that we agree that the improvements that actually matter are those that also have an audible effect (in a proper test, not sighted), then we should be fine. We may want to bring the THD level of a DAC to -150Db – this is an engineering and scientific challenge, and it definitely has a value – it is also nice to know that one can build a device with even less close-in phase noise, and I commend the efforts to kill that as much as possible, but if it is not provably audible, I am not interested to spend money to put it in my audio chain for that reason alone.
I should note that I enjoy vinyl (have two turnables in the living room) as the whole ritual is calming and gets you in the mood to kick back, relax and enjoy the music. But I also accept that the medium has an audible noise floor some 40-50dB above digital, >2%THD and some wacky FM effects from eccentricity and mechanical resonances that make worrying about close in phase noise in digital clocks seem like comparing an earthquake to a gravitational wave.
But I have simple needs and for me once you are below the noise floor and distortion of the microphone you are into either turd polishing, or doing things for the fun of it. No issues with that. Dismissing a product as having a problem on a sample of one report is a worry to me though.
But I have simple needs and for me once you are below the noise floor and distortion of the microphone you are into either turd polishing, or doing things for the fun of it. No issues with that. Dismissing a product as having a problem on a sample of one report is a worry to me though.
Mark, it's a pity you don't own SOTA phase noise measuring gear. You would likely get a lot closer than others in correlating various jitter artifacts to what
is heard. In the past I've come close to pulling the trigger on a Timepod etc but I just don't think I would get value out of it these days.
I have mixed feelings about the "well tempered" clocks here for sale. By the time a complete SC-cut clock system is assembled with doublers, squaring,
etc etc to get the usual switchable 45 / 49MHz or even 22 / 24MHz, it's very expensive and spread over a huge amount of physical real estate.
That is another discussion, I may or may not jump into one of those at some point or maybe just procure Magicxtal to do a few, they will end up a similar
spec and probably cheaper.
TCD
That doesn't mean anything. However, room acoustics quality means a significant audible improvement or detriment.
"custom manufactured wire" sure smells like one.
Vinyl is great. The experience has a special flavour and the ritual has its fascination. It is a low fi reproduction mechanism (by today’s standards) with very good treble (if you have everything properly set up and a good pressing) and the only reason I do not have it is that I would end up ruining myself financially knowing how much I involve myself on my hobbies.
The technical points I may raise have nothing to do with personal appreciation.
Roberto