And the the digital PLL is set to 120 seconds if I recall correctly- that way it comes away with such a small amount of memory. Clever, but not as good as Ians "all in solution"
//
Soekris chose to make due with the FPGA's internal memory. I suspect that it was a VFM-based decision.
Ian went "hi-end", and pretty much everyone is grateful for that.
What I've learned from my Soekris DAC (and from any DIY audio device I've built, for that matter) is that no two Soekris DAC implementations sound the same.
I'm pretty sure that it's easy to mess up with Ian's gear (by hooking it up to a lousy DAC), like it's easy to have a mediocre sounding Soekris. I'm not saying that your Soekris implementation it bad - I don't know the details of your implementation, and even if I did I could only guess at what it sounds like - but there may be more to get out of it with hw & sw tweaks.
Considering it's not very hard to do proper measurements of jitter and noise it is more than just strange we have a majority of bad sounding dacs made by all those engineers !
The ones knowing may open a thread about what & how to measure that around the digital stuffs.
It may also some interactions are not yet understood and there are also perhaps some others factors to measure. If a receipe would be known : everybody should be enough to make at minima a decent dac with it ! But it is not !
For instance, despite the high level of Soekris, his dac stock sounded very bad and I solded mine as I understood it will be time consuming to make it sound better ! Here it's not just about noise & Jitter as those two parameters where good enough ! Full SMT everywhere for better inductance ?? It seems some people who added bigger and less efficient caps than MLCC at some strategic places had a better result. Remember me Hiraga & la Revue de l'Audiophile about experiment and non logical big caps bank which musical result where higher an proportional inverse to theory of regulated supply....
trade offs !
The ones knowing may open a thread about what & how to measure that around the digital stuffs.
It may also some interactions are not yet understood and there are also perhaps some others factors to measure. If a receipe would be known : everybody should be enough to make at minima a decent dac with it ! But it is not !
For instance, despite the high level of Soekris, his dac stock sounded very bad and I solded mine as I understood it will be time consuming to make it sound better ! Here it's not just about noise & Jitter as those two parameters where good enough ! Full SMT everywhere for better inductance ?? It seems some people who added bigger and less efficient caps than MLCC at some strategic places had a better result. Remember me Hiraga & la Revue de l'Audiophile about experiment and non logical big caps bank which musical result where higher an proportional inverse to theory of regulated supply....
trade offs !
Last edited:
Do we?
I guess we need some convention to indicate, in the subject of a thread, that suggesting measurements or objective verification is frowned upon.
is that an excuse not to measure what we do understand?
That is your opinion. I disagree. There are lots and lots of more than decent DACs out there.
I guess this is where we get into the endless discussion about subjectively pleasing vs. objectively transparent. Yes, a lot of gear that measures badly sounds subjectively pleasing, because it adds colouration. But do you want to see everything through the same tinted glasses?
Yes the ones who know should write it !
Beware with the excuse of transparent sound, I even read some say than the Soekris sounded bad because it was in fact transparent ! Of course, not !
Even the Hypex can sound bad with the "bad" wires after it as it is "transparent"! and it's not the speaker : some changed the internal wire of the speaker and the outside one between speaker and hypex amp (especially with Ncore)...
So even the term transparent is a cultural point of view: you never have the studio sound result "exactly" in the living room : like a photograp is not transparent and IS an interpretation of a scene and how to look !
Beware with the excuse of transparent sound, I even read some say than the Soekris sounded bad because it was in fact transparent ! Of course, not !
Even the Hypex can sound bad with the "bad" wires after it as it is "transparent"! and it's not the speaker : some changed the internal wire of the speaker and the outside one between speaker and hypex amp (especially with Ncore)...
So even the term transparent is a cultural point of view: you never have the studio sound result "exactly" in the living room : like a photograp is not transparent and IS an interpretation of a scene and how to look !
Soekris chose to make due with the FPGA's internal memory. I suspect that it was a VFM-based decision.
Ian went "hi-end", and pretty much everyone is grateful for that.
I think that the Ian clock could have such a mode for lip-sync. I couldn't sync an Oppo playing blue-ray to my DAC
//
And how do you know if you don't measure?
You do realize that the plural of "anecdote" is not "data"?
Ah, that old, erroneous analogy. The original music in the studio is the scene. The recording is the photograph. And yes, the recording is not the original music, just like the photograph isn't the scene - but once you have a recording or a photograph, you don't want to have to view all your photographs through pink sunglasses...
No, the ears in this domain is a good indicator, some prefer to believe their eyes focused on one measurement only ! If more than an experienced musician says it sounds bad, the measurement don't give all the corelations !
some should give indication to how to measure but not believe they have the last word, as many dacs who measure good, give a poor experience of what is music ! It gives a good indication but it is not the only one ! Alas !
It's not anecdot : it's experienced! Alas and because you don't listen to a DATA or a scope, some amps can sounding better than an Hypex in certain conditions : the wirerinf environment after an hypex stays important, it is not as universal than a poorer Quad 909... on the paper! It is far more demanding on how to mate the devices around : interlinks, etc. The excuse of : in fact with Hypex, all the stuffs around are bad if it doesn't sound good is a bad one !
But off topic,
Marce iirc have been beginning to give documents and talk about jitter , etc, some should follow this instead dialecting around tauthology...
some should give indication to how to measure but not believe they have the last word, as many dacs who measure good, give a poor experience of what is music ! It gives a good indication but it is not the only one ! Alas !
It's not anecdot : it's experienced! Alas and because you don't listen to a DATA or a scope, some amps can sounding better than an Hypex in certain conditions : the wirerinf environment after an hypex stays important, it is not as universal than a poorer Quad 909... on the paper! It is far more demanding on how to mate the devices around : interlinks, etc. The excuse of : in fact with Hypex, all the stuffs around are bad if it doesn't sound good is a bad one !
But off topic,
Marce iirc have been beginning to give documents and talk about jitter , etc, some should follow this instead dialecting around tauthology...
Last edited:
Indeed. Let's have a fact-based discussion instead of anecdotes ("a usually short narrative of an interesting, amusing, or biographical incident").
Let's remember the wise words:
"Anecdotal evidence is an informal account of evidence in the form of an anecdote. The term is often used in contrast to scientific evidence, as evidence that cannot be investigated using the scientific method. The problem with arguing based on anecdotal evidence is that anecdotal evidence is not necessarily typical; only statistical evidence can determine how typical something is. Misuse of anecdotal evidence is an informal fallacy."
So, any factual information on the noise or jitter of the different connection methods?
not erroneous analogy ihmo :
Of course an Hypex, any amp, and all the stuffs are staying pinked sunglasses : the amp is not the eyes, or the ears either ! Talking of transparence is already a way of perceiving, it's a biased word and datas doesn't make it unbiased as far as the experience doesn't corelate the datas ! It's science : experience should confirm the datas and not the opposite ! If the datas don't corelate the experience, there is missing datas, or you don't see the whole thing with it !
Like some Photographer shoot a finished photograph, an amp and the whole sytem is a picture into the picture, you will never have the same result than the sound engineer in his studio; you even have not the same definition on how to listen to music ! Here we must believe experience and experienced acoustical musicians ! If the equation is good but the rocket crash, it was simply no the good equation to keep or something was missing !
I was saying than measuring jitter and noise shoulbe given here to allow some to also measure elswhere as the jitter and noise don't say all of why a stuff is sounding better between two good stuffs with equal measurement ! It seems now the easiest thing to solve, but most of us don't know how to solve this simple link to go further ! It was the sense of my input. You may teach us to go further and allow the majority to go farer than the jitter/noise combo !
Of course an Hypex, any amp, and all the stuffs are staying pinked sunglasses : the amp is not the eyes, or the ears either ! Talking of transparence is already a way of perceiving, it's a biased word and datas doesn't make it unbiased as far as the experience doesn't corelate the datas ! It's science : experience should confirm the datas and not the opposite ! If the datas don't corelate the experience, there is missing datas, or you don't see the whole thing with it !
Like some Photographer shoot a finished photograph, an amp and the whole sytem is a picture into the picture, you will never have the same result than the sound engineer in his studio; you even have not the same definition on how to listen to music ! Here we must believe experience and experienced acoustical musicians ! If the equation is good but the rocket crash, it was simply no the good equation to keep or something was missing !
I was saying than measuring jitter and noise shoulbe given here to allow some to also measure elswhere as the jitter and noise don't say all of why a stuff is sounding better between two good stuffs with equal measurement ! It seems now the easiest thing to solve, but most of us don't know how to solve this simple link to go further ! It was the sense of my input. You may teach us to go further and allow the majority to go farer than the jitter/noise combo !
Last edited:
Didn't we already conclude that this was really going off-topic?
It is, because:
It is, because:
You can easily have the same result as the sound engineer in the studio. And you might not agree with his/her taste, but what the mastering engineer hears is the final result of the combined artistic (and commercial) effort - just like a photograph. If you don't like it, and want it colored more pink, you are changing it to match your taste and in fact creating a derivative work - your own modified version.
OK, so would it help if we explained to you how to measure noise and jitter?
I think that the Ian clock could have such a mode for lip-sync. I couldn't sync an Oppo playing blue-ray to my DAC
//
Ian's buffer by definition will create a delay in your audio. If it is not creating a delay, it is not doing buffering. So what you need is to also delay the video. But that goes way beyond Ian's buffer's scope.
Ian's buffer by definition will create a delay in your audio. If it is not creating a delay, it is not doing buffering. So what you need is to also delay the video. But that goes way beyond Ian's buffer's scope.
I'm aware of this. However the delay was so great that even in the extreme setting, the Oppo could not "wait" enough.
//
And if we want to understand a physical phenomenon, the best tool we have is the scientific method. The first step is to verify the observations - does the I2S link really give a better result than the USB output? Is it a statistically significant number of people? How many people did *not* hear a difference, but didn't report it? Can the perceived difference be explained in other ways (such as exception bias), and if so, what experiment would isolate the effect of those other possible explanations?
If you are trying to solve a problem, the first thing to do is make sure you actually have a problem in the first place...
So.. again from all the reading makes my mind blow . I'm curious. Ian's kit comes with a couple of simple clocks to test the material. That's perfect to start. Some tips have been givesn for the Crystec clocks iirc.
I'm curious, .. I've been reading really good stories about Femto clocks. They are not cheap.. did anyone try them on the board yet?
. I'm curious. Ian's kit comes with a couple of simple clocks to test the material. That's perfect to start. Some tips have been givesn for the Crystec clocks iirc.I'm curious, .. I've been reading really good stories about Femto clocks. They are not cheap.. did anyone try them on the board yet?
If you mean in a general sense "Can different inputs into a DAC that reclocks and isolates sound different?", I think that is well established:
Barriers to Computer Audio - Page 4
This is John Swenson talking about this topic in regards to his Bottlehead DAC design which not only has reclocking & isolation, but also has unique configurations on the USB and S/PDIF inputs to minimize these differences.
If you browse through his posts on the CA and Bottlehead forums over the last couple of years, he has been relentless in searching for the causes of these differences between sources and how to eliminate them. He put his best effort based on what he knew at that time into his Bottlehead DAC. Another result of that search was the UpTone Audio Regen.
Computer Audiophile - PS Audio DirectStream DAC Review
This review recounts how the sound of the subject DAC, the PS Audio Directstream, which also included significant engineering to minimize the sonic differences between sources, performed with different sources. Here too, the reviewer found significant differences.
Searching through the PS Audio Directstream forum will net you a number of other similar experiences.
I think Ian's FIFO is a brilliant piece of kit which along with a few more recent technology bits (Acko's S03 & Twisted Pear's Hermes-Cronus combination come to mind) allow us DIY enthusiasts to use some of the tools previously only employed in fairly expensive DACs to minimize these differences.
But even these silver bullets don't eliminate those source to source differences, as much as we might hope they do.
Greg in Mississippi
Is that really the link you wanted to post? It links to a discussion about noise (as in actual sound, not electric noise) from hard disks. And I have to say I will always take anything posted to computer audiophile with a grain of salt - the site is a commercial site, living off advertising from "audiophile" vendors, and has a very pro-placebophile moderation policy.
Has he published any of his findings in peer-reviewed publications? Has he published any measurements or double-blind listening test results?
If "well established" means "according to a subjective review on a paid-advertisement web site", I think we have a slightly different standard for evidence.
Last edited: