Ett sista och förmodligen onödigt inlägg i denna fråga suboptimerande fråga men Faktiskt är ju ibland hårklyveriernas forum
:
TLDR version: Se
ovan inlägg av I-orEdit, det verkar som att många DACar tystar utgången när insignal saknas, vilket gör att testet med lyssna på bruset kanske kräver någon form av insignal?
Längre version:Tell, du har i princip ganska (men nog inte helt) rätt ang punkt 1, kvantiseringsbruset verkar till stor del vara just (vitt) brus. Men inte bara enligt nedan snabba gurglande:
Från
Reconstruct a DAC Transfer Function from its Harmonic Spectral Content"The output spectrum will contain harmonic content due to the DAC’s non-ideal transient and static behavior."
I denna gamla artikel
Analog Devices, Evaluating High Speed DAC Performance pratar man om
"Correlated (A) and Uncorrelated (B) Quantization Noise Spectrum"
Här sägs det att (förvisso rörande ADC):
Additive noise behavior is not always a valid assumption. Quantization error (for quantizers defined as described here) is deterministically related to the signal and not entirely independent of it. Thus, periodic signals can create periodic quantization noise. And in some cases it can even cause limit cycles to appear in digital signal processing systems. One way to ensure effective independence of the quantization error from the source signal is to perform dithered quantization (sometimes with noise shaping), which involves adding random (or pseudo-random) noise to the signal prior to quantization.
...
At lower amplitudes the quantization error becomes dependent on the input signal, resulting in distortion. This distortion is created after the anti-aliasing filter, and if these distortions are above 1/2 the sample rate they will alias back into the band of interest. In order to make the quantization error independent of the input signal, the signal is dithered by adding noise to the signal. This slightly reduces signal to noise ratio, but can completely eliminate the distortion...
Quantization noise is a model of quantization error introduced by quantization in the ADC. It is a rounding error between the analog input voltage to the ADC and the output digitized value. The noise is non-linear and signal-dependent. It can be modelled in several different ways.
...
For complex signals in high-resolution ADCs this is an accurate model. For low-resolution ADCs, low-level signals in high-resolution ADCs, and for simple waveforms the quantization noise is not uniformly distributed, making this model inaccurate.[17] In these cases the quantization noise distribution is strongly affected by the exact amplitude of the signal.
Från Taking the Mystery out of the Infamous Formula, "SNR = 6.02N + 1.76dB," and Why You Should Care
Although the rms value of the noise is accurately approximated by q/√12, its frequency domain content may be highly correlated to the ac-input signal under certain conditions. For instance, there is greater correlation for low amplitude periodic signals than for large amplitude random signals. Quite often, the assumption is made that the theoretical quantization noise appears as white noise, spread uniformly over the Nyquist bandwidth dc to fs/2. Unfortunately, this is not true in all cases. In the case of strong correlation, the quantization noise appears concentrated at the various harmonics of the input signal, just where you don't want them.
Här,
https://www.audiosciencereview.com/foru ... rom.53876/You don't have to have the original waveform to compare against, to objectively identify quantization noise as noise, because the noise is correlated to the input signal, and our brains are very good at picking out such noise.
We can mask that correlated noise by injecting uncorrelated noise, which is much easier for the brain to ignore.
Från:
Understand SINAD, ENOB, SNR, THD, THD + N, and SFDR so You Don't Get Lost in the Noise Floor som gäller ADC:
"If the bandwidth of the measurement is dc to fs /2 (the Nyquist bandwidth), THD + N is equal to SINAD.
SINAD is a good indication of the overall dynamic performance of an ADC because it includes all components which make up noise and distortion....
.. SNR is often misused in defining audio DAC performance. ... A more proper term to describe the noise floor for an audio DAC is idle channel noise"
Ovan var bara en snabb gurgling.
I de mätningar från ASR jag snabbt kollat syns att spektrumet inte är jämnt fördelat, varför man inte kan betrakta DACar endast utifrån kvantiseringsbrus. Och som vi redan diskuterat flera gånger och verkar vara överens om, så minskar marginalen (SINAD) om man använder den digitala volymkontrollen.
******
Ska man testa detta ämne på ett bra sätt behöver man bästa förutsättning, dvs bra lyssnare, bra anläggningar och bra programmaterial.
Jag har haft svårt att hitta bra tester men den som I-or länkade till är mer intressant än många andra:
http://archimago.blogspot.com/2020/06/b ... -high.htmlOm man exkluderar de sämre lyssnarna kan det vara möjligt att det går att särskilja distorsion om 0,02% eller -75 dB från t ex 0,3% (-50 dB). Snubben kör chi-två-test för att signifikanspröva så studien är verkligen inte dålig trots att den är gjort på 'fritiden''. Det hade varit fint med större svarsfrekvens för att få högre signifikans men annars verkar den vid snabb koll vara välgjord.
Under more controlled "lab" conditions with very high-fidelity systems, it's possible (likely) that the absolute threshold of audibility for harmonic distortion would be even lower. I also expect audibility of THD would be easier with test tones rather than real music.
Even though we can see correlations between distortion level and preference, we don't need to be neurotic about insisting on owning DACs and amplifiers with vanishingly low THD either. Just as "extreme subjectivists" might erroneously claim measurements make no difference, we need not be "extreme objectivists" either thinking that a -120dB THD DAC "sounds better" than one at -100dB beyond just being able to appreciate the engineering efforts to achieve excellent performance!
Jag har själv erfarenhet av att det är tydlig hörbar skillnad på högtalare som distorderar ca 0,1% jämfört mot ca 0,5%. Skulle jag sätta det blint? Ja, jag tror det. Har jag testat blint? Nä, det kräver noggrann ekvalisering. Hur kan jag då tro att det är distorsion som ger skillnaden? Därför att det låter mer grumligt, grusigt etc med högre (uppmätt) distorsion och inte som tonkurveförändringar.
Dessa är intressanta i sammanhanget:
https://www.audiosciencereview.com/foru ... les.36296/https://www.audiosciencereview.com/foru ... ents.5734/
.
