Not so easy when the Beaucoup D'argent types are proclaiming anything that isn't on a par with dCS, Meitner or Mola Mola is rubbish.
And one day You will learn how to handle a black and white box view and what properties are valid in their respectively interfaces. It's the next level of architectural approach but I'm sure you could convolve into a greater insight if you just tried.
//
One have to have the strength to form an own opinion. As I see it, DIY gives a satisfaction of having accomplished something myself. Is it the "best" or equally good compared to what exists.... perhaps, perhaps not (there is really no easy way to tell, is there?), is it good enough that in the end on system level, the results could match anything on the market - I think so. And to 1/10th of the cost so thats good. If you like the result, if it starts to sound what you hear at the visit of the concert hall or church, it's good. It's even great. If one judge SQ on mostly multi mic studio recordings it can only be about "nice sound" which is of course very subjective. To listen to voices that talks mostly about "nice" sound is completely meaningless as these voices are the result of the owners love of it, expressing how they experience their gear and system and room and recordings and mood and electricity and ears and music experience, and..... see? We are just being influenced by a number of influencers. The influencers motivation is most often money but self promotion in general is not uncommon.
Hang in there and try not to be so "influenced" by the D'argent.... if one seeks truthful reproduction - go to a non amplified concert - do you get the same thrill at home? It's all about music in the end, isn't it? If one wants "nice sound" - go right ahead and buy their view of what nice sound is if you can't form your own opinion - which I'm sure most of us can - and if not, chess is also fun.
//
Do you think that the concept of the "black box" holds some merits?
What properties do you believe represent the characteristics of an analog signal?
//
What properties do you believe represent the characteristics of an analog signal?
//
I don't mean that it dosen't effect the output.
But you obviously didn't read what I wrote. Here, an other chance...
"There isn't anything on a DACs analog output that is called Vref noise. Vref noise is noise present on vRef pins on a DAC chip."
If the semantics in this sentence is not clear, don't be afraid to ask for support.
//
This is what PN and AN look like in a spectrum analyzer view of a test tone played back though a dac:
You can easily see those effects in measurements presented in a diyaudio thread: https://www.diyaudio.com/community/threads/phase-noise-in-ds-dacs.387862/post-7063038
The time domain view is different. There can be errors in dac output sample point amplitudes, and or errors in dac sample point output timings. Either one leads to an error in the reproduced waveform.
To claim they are not specifically evidenced in the analog output of a black box dac would be incorrect. Thus they should be included a more or less comprehensive black box model.
You can easily see those effects in measurements presented in a diyaudio thread: https://www.diyaudio.com/community/threads/phase-noise-in-ds-dacs.387862/post-7063038
The time domain view is different. There can be errors in dac output sample point amplitudes, and or errors in dac sample point output timings. Either one leads to an error in the reproduced waveform.
To claim they are not specifically evidenced in the analog output of a black box dac would be incorrect. Thus they should be included a more or less comprehensive black box model.
What there is in the frequency domain view is the convolution of Vref noise with the audio signal. If we know what the audio signal is, then the applying inverse process should leave Vref noise observable at the output.
Moreover, since in the time domain Vref is multiplied by the dac digital output, it means the height of the spectral line at the dac output is moving up and down with Vref noise. So the amplitude noise is still there. You don't see that up and down spectral line motion on a FFT because it averages the motion out.
Last edited:
Evidenced yes, but not present particularly.
If you look into a pair of these:
Tell me how you measure jitter / PN and Vref noise by sticking someting in there with something in the end. And I mean an actual indicator saying "Phase noise" with the scale db/Hz and an other saying "Vref noise" with a scale dBv.
No, exactly, it wont happen.
//
If you look into a pair of these:
Tell me how you measure jitter / PN and Vref noise by sticking someting in there with something in the end. And I mean an actual indicator saying "Phase noise" with the scale db/Hz and an other saying "Vref noise" with a scale dBv.
No, exactly, it wont happen.
//
Now back to the topic of this thread, the OP has been presented a number of suitable inexpensive Chinese DACs. There are more information on the objective performances of these DACs over at ASR. I reckon these are sufficient to make an informed choice. Retailers like Apos have returns/ refunds so fairly low risk trying out. Would like to hear how these sound like in comparison… all the best 👍
Why don't you describe how you perform a measurement on the analog outputs that isolate the artefacts that is due to Vref noise?
//
//
You already saw that Bohrok2610 did it in the frequency domain, or did you not understand the discussion in that thread?
No no, you describe how to make a measurement in the black box interfaces that identifies from where different white box noise sources originate.
//
//
The thread is at: https://www.diyaudio.com/community/threads/phase-noise-in-ds-dacs.387862/post-7063038 IMHO, the whole thread is worth reading. In one post @peufeu explains how to differentiate phase noise from amplitude noise in the measurements.
EDIT: Not my fault if some people don't bother to read, can't understand engineering talk, etc.
EDIT: Not my fault if some people don't bother to read, can't understand engineering talk, etc.
I wont because I'm so very sure that my question above is not answered as it is simply not possible. Not if you have a closed black box i.e. you cant manipulate the test object ... hence the "black box". I also most probably followed that thread.
//
//