He did for me... on special request.
Back to the roots: 14 bits (of TDA1541A) - Audial
There is some communication on the bottom of this page...
Back to the roots: 14 bits (of TDA1541A) - Audial
There is some communication on the bottom of this page...
Yes I knew that, so you talk about the standalone USB to Sim board, not the complete DAC.
OK, understand now, it was a bit confusing as I also asked him by PM a TDA1540 complete DAC version but he has no time for that and it will need 40 customers imho and surely a no Go as it's a lot of works for almost no monney. But I should try to make a pcb maybe...not sure it works better at the end than the venerable TDA1541 in S1 or late Taiwan version !
sorry for the off topic Ken.
OK, understand now, it was a bit confusing as I also asked him by PM a TDA1540 complete DAC version but he has no time for that and it will need 40 customers imho and surely a no Go as it's a lot of works for almost no monney. But I should try to make a pcb maybe...not sure it works better at the end than the venerable TDA1541 in S1 or late Taiwan version !
sorry for the off topic Ken.
To the difference between NOS and OS:
A NOS analog output is a stepped signal (assuming 1st order S/H). One step is practically a square pulse. A square wave consists of the base and odd harmonics. A periodical signal can be described as a Fourier series, but the DAC output is seldom periodical. Nevertheless, even a single pulse can be dissected to harmonic components. For example a pulse having sampling period width, repeated at a rate of 300 Hz will have components @900 Hz, 1500 Hz, 2100 Hz, etc. In a music, the next sample will have different amplitude, but it can be similarly expressed as a series of harmonics.
I believe these magically appearing harmonics (which were not present in the original signal, and they are not present in OS mode) make NOS sounding different than OS.
This also means the OS is theoretically closer to the original signal than NOS.
Strangely, I still prefer NOS
Could it be because of the enriched harmonic content that is somehow pleasing to the ear?
A NOS analog output is a stepped signal (assuming 1st order S/H). One step is practically a square pulse. A square wave consists of the base and odd harmonics. A periodical signal can be described as a Fourier series, but the DAC output is seldom periodical. Nevertheless, even a single pulse can be dissected to harmonic components. For example a pulse having sampling period width, repeated at a rate of 300 Hz will have components @900 Hz, 1500 Hz, 2100 Hz, etc. In a music, the next sample will have different amplitude, but it can be similarly expressed as a series of harmonics.
I believe these magically appearing harmonics (which were not present in the original signal, and they are not present in OS mode) make NOS sounding different than OS.
This also means the OS is theoretically closer to the original signal than NOS.
Strangely, I still prefer NOS
Could it be because of the enriched harmonic content that is somehow pleasing to the ear?
Staying with the title of this thread, there is a (strong) suspicion towards a sound difference between NOS and OS, a difference that cannot be simply explained.
Einstein once mentioned: not everything that can be measured is important and not everything that’s important can be measured.
Seen in that light it’s better to listen first and then try to analyse as a second step instead of lengthy discussions that do not solve anything.
And after all, it’s very simply to excecute the proposed tests.
My two cents.
Hans
Einstein once mentioned: not everything that can be measured is important and not everything that’s important can be measured.
Seen in that light it’s better to listen first and then try to analyse as a second step instead of lengthy discussions that do not solve anything.
And after all, it’s very simply to excecute the proposed tests.
My two cents.
Hans
Agree with Hans. Sometimes the ears can provide clues in a more direct way than hypothesizing as the first step. Also, experiments can be done that change the sound in ways that may help guide further investigation. Not every experiment has to be done solely to validate or invalidate a preexisting hypothesis. There is a difference between applied research and more basic research.
Also, IME there are typically layers of problems with dacs that have to be peeled away one by one. More or less, removing one layer tends to expose the next.
Also, IME there are typically layers of problems with dacs that have to be peeled away one by one. More or less, removing one layer tends to expose the next.
Last edited:
What is good with this thread is it is not about opinion and experience only. I say experience because people know too many factors have a coloration influence : so the algorythm or suimply the upsampnling made by multiplication.
As there are many way to make this last one, Ken Newton proposed a protocol to investigate a scientific way. Which should stay the main purpose with of course the casual tchitchat that are always welcome.
I don't know if it has be on the table before : but what a bout the most simple experience possible in order to erase the multiple sounding factors that can enter in a final conclusion - FGPA, etc :
As pointed out above : try to find iso reccording of the same material both NOS and OS : simplier to find a native OS that had been down sampled that the opposite - remastering of old OS reccording have also an engineer remastering process with all the added artifact, etc
Just make the less obstructive layout but still sota layout foir the upsampling process : layout with respect of speed clock, impedance trace, the shortest as possible, LiPo cell feeded ,with just shift registers to increase the speed (even if we know it's never the ideal but perhaps better than FGPA or noisy computers ?) say to X2, X4... Measure, do a blind test of people without saying what it's about... Then... well then, rediscuss
OK, just a non tech proposal of a basic diyer, forgive me if silly !
As there are many way to make this last one, Ken Newton proposed a protocol to investigate a scientific way. Which should stay the main purpose with of course the casual tchitchat that are always welcome.
I don't know if it has be on the table before : but what a bout the most simple experience possible in order to erase the multiple sounding factors that can enter in a final conclusion - FGPA, etc :
As pointed out above : try to find iso reccording of the same material both NOS and OS : simplier to find a native OS that had been down sampled that the opposite - remastering of old OS reccording have also an engineer remastering process with all the added artifact, etc
Just make the less obstructive layout but still sota layout foir the upsampling process : layout with respect of speed clock, impedance trace, the shortest as possible, LiPo cell feeded ,with just shift registers to increase the speed (even if we know it's never the ideal but perhaps better than FGPA or noisy computers ?) say to X2, X4... Measure, do a blind test of people without saying what it's about... Then... well then, rediscuss
OK, just a non tech proposal of a basic diyer, forgive me if silly !
I understood that, the experience being about opinion, sometimes bias (and I am in the NOS camp), math being... math (we need a/several medium between math and ears°, I beleive science is a theory tested by experiments that are repeatable. The best equation is to minimize the influence of what is in between the best we can and not mixing every way to do it between the math and the listening test (plus measurement side to side to aesthetic judgement)
That's all what everybody know I simply wanted to repeat.
That's all what everybody know I simply wanted to repeat.
The funny thing about all this is that math brought us to digital sound reproduction in the first place, so it can not be that bad.... manipulation signals analog or digital is yet another story alltogether....
I was also agains OS until i was convince otherwise. The issue is that NOS is the most affordable way to a "stress free" almost analog experience...
I was also agains OS until i was convince otherwise. The issue is that NOS is the most affordable way to a "stress free" almost analog experience...
math is needed for the theory. Experiments are needed to paradigm the theory...untill a better.
The funny thing is above the math that are beyond my head even if I made it about networks computing. I ever prefered vynils untill I can find digital units tweaked to death than sounded better to me (call that analog but for me it's a non sense, all is analog and firstly what run on the pcb traces. ok there are glitch than comming from the math but prove me it's a math problem and not a hardware one, etc. A reasonement can be true from A to Z but it may be not the good one to apply (brain wants to find always an explain and humans are good to that, science proposes to experiment to corelate = better)
But remember about math, theories are often dicuseed without experiments : how much time we suffer of the "bit perfect" or "jitter" only to involve each time a bad sound !
And radiation are often discussed to know what until a bomb makes equal all the participants : experiments are mandatory and I liked Ken Newton road map for that.
The funny thing is above the math that are beyond my head even if I made it about networks computing. I ever prefered vynils untill I can find digital units tweaked to death than sounded better to me (call that analog but for me it's a non sense, all is analog and firstly what run on the pcb traces. ok there are glitch than comming from the math but prove me it's a math problem and not a hardware one, etc. A reasonement can be true from A to Z but it may be not the good one to apply (brain wants to find always an explain and humans are good to that, science proposes to experiment to corelate = better)
But remember about math, theories are often dicuseed without experiments : how much time we suffer of the "bit perfect" or "jitter" only to involve each time a bad sound !
And radiation are often discussed to know what until a bomb makes equal all the participants : experiments are mandatory and I liked Ken Newton road map for that.
Last edited:
Blindly experimenting is also not helping here. we need to have some kind of fundament to start with and only then we can set a goal...
A lot of math and formulars are also beyond me so i will not anticipate in that debate.. but i do understand the theory and can work with that (sofar i could get away with it ....lol)...
A lot of math and formulars are also beyond me so i will not anticipate in that debate.. but i do understand the theory and can work with that (sofar i could get away with it ....lol)...
Has it occurred to anyone that which dac sounds best often depends on the rest of the reproduction system. Sometimes one system component is chosen to compensate for the shortcomings of another component. Until that problem is sorted out among the various systems used by listening test participants, its probably going to be difficult extract useful findings.
Last edited:
... the dust is to vynil what the jitter is to digital : Plato, The Republic, book IV ! Me to I can
Yes and that's why a methodology is discussed. We always experiment at iso perimeter the best we should. But hey you find alsways a guy that say a lythic cap is as good as the best film cap when it comes to nuance for passive filter in a loudspeaker, transparency, etc... What can I answer to that ????
There are a lot of good flat in room system enough for that (active, lol ! no I joke)
Yes and that's why a methodology is discussed. We always experiment at iso perimeter the best we should. But hey you find alsways a guy that say a lythic cap is as good as the best film cap when it comes to nuance for passive filter in a loudspeaker, transparency, etc... What can I answer to that ????
There are a lot of good flat in room system enough for that (active, lol ! no I joke)
Last edited:
Unfortunately we are not close so we can not listen to each others system...
Yesterday i had visitor from Italy, we listened all day (we forgot the time) and after the listening session he was surprised that even after many hours of listening he did not get tired at all. Even the conversations we had during playback on a substential level did not bother him at all (no shouting was required though ;-)). I do not want to speak for other people but that is what i can say...the rest you will have to ask him if he is willing to share it...
Best wishes,
Frank
Yesterday i had visitor from Italy, we listened all day (we forgot the time) and after the listening session he was surprised that even after many hours of listening he did not get tired at all. Even the conversations we had during playback on a substential level did not bother him at all (no shouting was required though ;-)). I do not want to speak for other people but that is what i can say...the rest you will have to ask him if he is willing to share it...
Best wishes,
Frank
Not to me... LOL
I have supplied some dacs to my closest friends (they all have different set-ups)....this is what they state jokingly:
"Frank for President" LOL
I think this is meant as a compliment...
Last edited: