[buonassi]

None of what you describe can be attributed to noise or jitter, and the fact that you've identified those qualities as related to them, AND are making fully-sighted and biased listening tests is why you're getting the results you are.

I'm definitely not closed off to the idea that expectation bias is at play, wanting to justify my investment in the USB cleaner. But I'll be damned if I'm making all this up. There is a definite jump up in the quality of the waveform and the bandpass noise (lack thereof?). Theoretically, there shouldn't be. Current J tests and distortion measurements would suggest that DACs with Asynch USB are already good enough and any jitter/noise distortion is so far below the audible spectrum (or masked by music) that it can't be audible. Yet audible differences are there.

Same for my newest investment. A $30 IEM cable made by monoprice (good old oxygen free copper). I bought this on a whim for my Campfire Jupiters because it was 1 foot longer than the stock cable (silver plated copper litz) and I wanted a bit more room to route the cable. I never expected to hear a difference because I didn't believe cable differences were audible. There shouldn't be, after all - they all measure perfectly flat from 20-20kHz, right? Yet there it was - a much rounder, less treble intensive sound. I really wanted to admit there was a superiority to the SPC litz cable, and maybe there is. But it's just a little too zingy and live sounding for me.

Anyway, my point is that I agree with you about sighted tests, expectation bias, etc. But I also think that the hardcore objectivists (not suggesting you fall into this camp - but I definitely did until recently) need to balance their skepticism with some humility. It is quite possible that we just can't measure everything that the human auditory system can pick up on. Or that we're measuring the wrong things. IE how are one or two perfectly oscillating sine waves anywhere near a good approximation for real music?

 

[pinnahertz]

I'm definitely not closed off to the idea that expectation bias is at play, wanting to justify my investment in the USB cleaner. But I'll be damned if I'm making all this up. There is a definite jump up in the quality of the waveform and the bandpass noise (lack thereof?). Theoretically, there shouldn't be. Current J tests and distortion measurements would suggest that DACs with Asynch USB are already good enough and any jitter/noise distortion is so far below the audible spectrum (or masked by music) that it can't be audible. Yet audible differences are there.

Well, you can't have it both ways. If you acknowledge the presence of expectation bias, then you must also allow for it's effects on your test results. Or, ignore expectation bias.

Same for my newest investment. A $30 IEM cable made by monoprice (good old oxygen free copper). I bought this on a whim for my Campfire Jupiters because it was 1 foot longer than the stock cable (silver plated copper litz) and I wanted a bit more room to route the cable. I never expected to hear a difference because I didn't believe cable differences were audible. There shouldn't be, after all - they all measure perfectly flat from 20-20kHz, right? Yet there it was - a much rounder, less treble intensive sound. I really wanted to admit there was a superiority to the SPC litz cable, and maybe there is. But it's just a little too zingy and live sounding for me.

Cables can have an audible impact, but it's also a measurable impact. There are real scientific reasons, given certain conditions. But I see no measurements being done here, so there's no point in analyzing the conditions either, especially since you still have bias present.

Anyway, my point is that I agree with you about sighted tests, expectation bias, etc. But I also think that the hardcore objectivists (not suggesting you fall into this camp - but I definitely did until recently) need to balance their skepticism with some humility. It is quite possible that we just can't measure everything that the human auditory system can pick up on. Or that we're measuring the wrong things.

We certainly can measure everything that's audible, but we do often look at the wrong things or misinterpret what we measure. The classic is single-figure THD. Does 1% sound better than .1%? Yes, sometimes, no sometimes. There's a lot to know about THD and a single figure tells you basically nothing.

IE how are one or two perfectly oscillating sine waves anywhere near a good approximation for real music?

That's an easy one. What we're trying to measure is a change caused by a system or component. To do that we need to know what the original was then measure the difference. We can't do that with music, but we can with tones. The selection of type of tones, frequency, level, and number of tones is very important, as is the method of analysis. But music cannot be measured, it's too time-variant, and the original is not known. Tones are an approximation of certain aspects of music, but you have to use quite a variety, and an equal variety of analysis methods, to get an answer that correlates with what you may (or may not) hear with music.

 

[buonassi]

Based on standard measurements alone, I guess everyone should just buy a topping d30 and get on with their lives. No point in going down the audio rabbit hole?

 

[buonassi]

result of the second test, iUSB3.0 nano vs Schiit Eitr.....

The Eitr wins by a noticeable amount. Pretty apparent, didn't even have to do too much A/B or search for differences. It is that good with my particular DAC. Not saying everyone is going to have the same experience, but this is my best source solution so far for my setup. I don't care how the inputs measure and that we "shouldn't" hear a difference, nor am I interested in theories as to why I my emotions are tricking my brain into believing there is a difference. While these replies may convince others, it will not convince me. I know damn well what I'm hearing, and it's not subtle to me.

• less grain in the shimmer of cymbals - they are less "ringy/splashy" sounding and more focused, polite strikes where the decay is much more nuanced, not there then gone, but trails off nicer
• less sibilance in the 5-9khz region - smoother and more tame here. more comfortable to listen to
• depth of soundstage increase (very slight)
• more precise and focused imaging - easier instrument localization.

Not expectation bias either. I'm not really a schiit fan after my experience with Asgard 2. I bought this fully intending (or expecting) to send it back and incur the 5% restock fee. It's why I bought it at regular price and opted to not save $30 by buying the B stock units (can't return B stock). My goal was to test all the inputs of this DAC and stick with the best sounding one. Coax via Eitr - case closed.

 

[pinnahertz]

result of the second test, iUSB3.0 nano vs Schiit Eitr.....

The Eitr wins by a noticeable amount. Pretty apparent, didn't even have to do too much A/B or search for differences. It is that good with my particular DAC. Not saying everyone is going to have the same experience, but this is my best source solution so far for my setup. I don't care how the inputs measure and that we "shouldn't" hear a difference, nor am I interested in theories as to why I my emotions are tricking my brain into believing there is a difference. While these replies may convince others, it will not convince me. I know damn well what I'm hearing, and it's not subtle to me.

• less grain in the shimmer of cymbals - they are less "ringy/splashy" sounding and more focused, polite strikes where the decay is much more nuanced, not there then gone, but trails off nicer
• less sibilance in the 5-9khz region - smoother and more tame here. more comfortable to listen to
• depth of soundstage increase (very slight)
• more precise and focused imaging - easier instrument localization.

Not expectation bias either. I'm not really a schiit fan after my experience with Asgard 2. I bought this fully intending (or expecting) to send it back and incur the 5% restock fee. It's why I bought it at regular price and opted to not save $30 by buying the B stock units (can't return B stock). My goal was to test all the inputs of this DAC and stick with the best sounding one. Coax via Eitr - case closed.

And exactly how did you remove bias?

 

[ThinkerMakerDIYer]

well there are (principally) a way to improve jitter - replacing the OC of MAC to high grade one like TXCO or even OCXO... Saw some one offering PCI-E based OXCO board to give SCard a extremely low jitter (<5ppm) reference.