Why do we want to add resistor ladders and analog junk to a DAC in the 1st place?
What is going on with the R2R movement? What is their end target they hunt for?
What is going on with the R2R movement? What is their end target they hunt for?
A warm and full sounding lower midrange. OTOH, some DSD dacs have sounded like they are a little recessed in that part of the frequency spectrum. Other than that difference, DSD dacs tend to sound more accurate and better in every way as opposed to R2R. It seems that some people really like that warm sound, and they don't like a recessed lower midrange.
Also, it occurred to me that a warm lower midrange sound might be more like what age related hearing loss sounds like. IOW, warm low midrange may tend to sound more natural/real to old men.
However IME, Marcel's RTZ dac with SC Pure clocks does not have that subjective recessed characteristic. Rather, IME it is perceptually very well balanced across the frequency spectrum.
Also, removing the ferrites and nonlinear caps from Andrea's DSD dacs seemed to help a lot with the fixing the slightly recessed low midrange effect. Thus, I have some suspicion correlated noise could be a factor involving the perception.
Last edited:
Appeal to authority is a logical fallacy.
Those standards you call out for are ENGINEERING standards, not scientific standards.
From a scientific perspective, we do not have a model that maps psychoacoustics to engineering.
To make the claim that those engineering standards that you call out are sufficient to ensure "good sound" - which is what I believe you are implying - requires a good scientific model that ties psychoacoustics to engineering.
Since the scientific model does not exist, then your claim is unproven.
Means nothing to me, this dac stockhave a great margin to sound better and can sounds different according the TDA1541A dac chip you have and passive parts as well as active and power supplies not saying about the front yet !
Ah yes a same DAC can sound very different whatever are saying the inexperienced guys of ASR whom most of them are just able to unpack two box to listen to dacs side by side and then draw conclusions suited by numbers to looks like science.
Presently, I am listen two DACS same dacs with just playing with op amps, power supplies and parts on a same layout ! Change just a resistor with same value but different tech and it sounds no the same ! It asks a lot of listening and benchmarks to make it sounds better. And it is not in an expensive design, although some parts can cost: I have some SMD resistors at 6 euros a 1206 smd size ! Better sommetimes, sometimes not at the same aera !
Guys at ASR don't listen to music, the listen to SINAD screens and as scientologist say that you who is biased if you don't glue to the mantra...
Attachments
We know from your many posts that you are a realtime DAC builder. You build and listen, rebuild and listen again, and the same again and again.
That's the kind of DIY I respect: not posting a lot of abstract thoughts but actually building stuff and telling us what happens. That's useful and informative for other builders, and what this forum is about.
That's the kind of DIY I respect: not posting a lot of abstract thoughts but actually building stuff and telling us what happens. That's useful and informative for other builders, and what this forum is about.
So far, many of the claims I have seen herein are unproven; there are some hypotheses, some theories and some speculation. I'm not saying that they all are incorrect, it is that there is no proof, at least that I have seen, that all of it is correct. Without proof, what we are left is belief/faith and subjective listening impressions.
Nonetheless, I enjoy the discussions and debates, as well as reading different perspectives, especially those that are contrary to mine. It helps me to continue learning.
If you look at the first pic, you can see I also layouted an output near the dac output for tubes (wires). You can also use Tht caps & resistor for the I/V and use the RCA output. Initially this 4 layers is made to benchmark I/V alone or I/V + buffer/gain follower for fast op amps that are demanding about the layout.
All the guys saying the opa861 is bad don't know it sounds very good with a goodpower supply and chosen cap & resitor for the I/V of a current output dac chip. How strange a same dac sounds not the same with just 2 differents passive parts and any trained guy or musician can immediatlt hear it whatever he is not aware of the mod... SINAD number, the same, one sounds more acurate the other so-so... 😉 .
Here I test the opa891 you will not see in a cheap chi-fi dac, but the layout was also made to benchmark on a same DAC smd parts or Tht I/V parts for the special opa861. Ad811 curent operational amplifier was tested too. No one sings the same, and no one reacts the same to the same passive parts. There are not ultimatly best passive parts you can put everywhere. It simply doesn't work like that. It asks time and hours of experiments !
What chi-fi low cost doesn't is to voice their DACs : you need to play with everything for correct tonal, it's really not about SINAD !
All the guys saying the opa861 is bad don't know it sounds very good with a goodpower supply and chosen cap & resitor for the I/V of a current output dac chip. How strange a same dac sounds not the same with just 2 differents passive parts and any trained guy or musician can immediatlt hear it whatever he is not aware of the mod... SINAD number, the same, one sounds more acurate the other so-so... 😉 .
Here I test the opa891 you will not see in a cheap chi-fi dac, but the layout was also made to benchmark on a same DAC smd parts or Tht I/V parts for the special opa861. Ad811 curent operational amplifier was tested too. No one sings the same, and no one reacts the same to the same passive parts. There are not ultimatly best passive parts you can put everywhere. It simply doesn't work like that. It asks time and hours of experiments !
What chi-fi low cost doesn't is to voice their DACs : you need to play with everything for correct tonal, it's really not about SINAD !
Last edited:
I think we reach the point where we loop back in a // universe of 10 years ago, then 10 years ago, then 10 years ago. then this dac chip used to, 10 years ago
So DACs have replaced thermionic valves in the audio debate. Not that valve debates will even end...
If the D to A conversion was done well and good, wouldn't the rest be up to the surrounding components then?
Clock jitter is for extreme money purists?
The bread and butter and ears on is in the board components supporting the D to A?
A human can not hear clock jitter?
Its the way the clock purrs the sample of bits to the DEM.
If it has some delay or overshot the bits samples are not sent to the R2R ladders in time, and then the switches don't operate in synchronization with the signal, which cause a loss of fidelity in complex signals...
I guess it is important.
Like I just forgot to reconnect the ground to the pre-amp tube and cathode follower tubes to GROUND.
When I powered the power amplifier it destroyed in a flash the resistor in series in the amp OUT pins...
I hope the output transformer survived , it is a 500$ one.
I just also removed all the power tubes 200mA fuses, very pyrotechnic
If it has some delay or overshot the bits samples are not sent to the R2R ladders in time, and then the switches don't operate in synchronization with the signal, which cause a loss of fidelity in complex signals...
I guess it is important.
Like I just forgot to reconnect the ground to the pre-amp tube and cathode follower tubes to GROUND.
When I powered the power amplifier it destroyed in a flash the resistor in series in the amp OUT pins...
I hope the output transformer survived , it is a 500$ one.
I just also removed all the power tubes 200mA fuses, very pyrotechnic
Maybe the D to A conversion isn't all so well and good as one might hope.
Once again, every DAC and every ADC needs two analog references. An analog time reference is needed, and an analog voltage reference. The latter can affect dac sound by adding AN, while the former can affect dac sound by adding PN (where AN = amplitude noise, and PN = phase noise).
For small errors of either type, they are considered approximately interchangeable. That is to say, the right signal amplitude at the wrong time is just as bad as the wrong amplitude at the right time. Problem is LT3042 is cheap, but SC Pure costs more to make because its not a simple IC with lots of them made all at once on one wafer. SC-cut quartz crystals have to be precision lapped and polished to the right frequency and on the correct crystalline lattice plane (which is what SC-cut refers to). More info at: https://www.electronics-notes.com/a...l-xtal/crystal-resonator-cuts-at-bt-sc-ct.php
Last edited:
I didn't say that. It's a whole and the basement. Now it's all about bits and speed needed on the boards. My library is Red Book and enough to play better than 24/192 K hz on a cheap Chi-Fi Dac. A DAC that measure good can sound bad. Now they often measure all good, but say little between two dacs.
I have dedicated a great deal of time to DAC in the past year & a half and I have gained a great deal of knowlege and perspective but it seems like a big pile of audio mud. I have the feeling it will be an expensive journey for me.
All you folks are the best. It was sad to see buddy walk away earlier from the thread today, he has a good perspective and knowledge.
I want to come back and hash it out some more as I probe deeper into it.
I may also pull some moves and come back with my own real world perspective.
IDK
The waters still look muddy to me and that is not what I expect in D to A
I was always an analog guy, never digital.
Rock on.
All you folks are the best. It was sad to see buddy walk away earlier from the thread today, he has a good perspective and knowledge.
I want to come back and hash it out some more as I probe deeper into it.
I may also pull some moves and come back with my own real world perspective.
IDK
The waters still look muddy to me and that is not what I expect in D to A
I was always an analog guy, never digital.
Rock on.
Thanks for the explanation, that is beyond my simple diy knowledge and very interesting.
When I studied TDA it was interesting to see how L/R always have a delay and how the splitting works, very fascinating indeed. What is your take on PCM53? why did it had all the bass? I wonder is my new pcm58 I will have soon will give me the same results?