[bigshot]

If you are going to base audibility on measurements and not actual hearing tests, then you should do as much research into the limits of human hearing as you do into electronic engineering. Most audiophiles think their ears hear an order of magnitude better than they actually do. They invest bias into their hearing being superior to other humans, The truth is that human hearing at its absolute best performance is remarkably consistent, and quite humble. The variations in hearing are all related to degradation below that level of peak performance, not being able to hear better than the norm.

But the simplest way to know if DACs sound different or alike is to just grab a range of different kinds and price ranges and do some controlled listening tests. When you take the time to do that, then you'll know. Audibility is where the rubber hits the road. When you sit down to play some music, it's all that really matters.

 

[SoundAndMotion]

If you are going to base audibility on measurements and not actual hearing tests, then you should do as much research into the limits of human hearing as you do into electronic engineering. Most audiophiles think their ears hear an order of magnitude better than they actually do. They invest bias into their hearing being superior to other humans, The truth is that human hearing at its absolute best performance is remarkably consistent, and quite humble. The variations in hearing are all related to degradation below that level of peak performance, not being able to hear better than the norm.

But the simplest way to know if DACs sound different or alike is to just grab a range of different kinds and price ranges and do some controlled listening tests. When you take the time to do that, then you'll know. Audibility is where the rubber hits the road. When you sit down to play some music, it's all that really matters.

You talkin' to me? (adjusts DeNiro mask) You talkin' to ME? (hard to tell, since you often don't quote people)

I'm not an electronic engineer. I'm primarily a neuroscientist, among other things, focusing on multi-sensory perception. I understand audibility and auditory thresholds quite well, thank you. I've done many, many perception tests, including listening tests. I can say unequivocally that on the issues of audibility, human hearing and listening tests, it is best to steer clear of most of your advice. There are too many errors to correct.

Please don't do your usual real-world vs. lab, and golly-gosh-gee-whiz test vs. publishable test set of excuses. A listening test designed to produce significant, useful results, (publishable or not) has methods that your hunches don't cover.

I do believe when it comes to all things animation, especially preservation of the history (great job there! I'm considering a donation.), you would be the go-to-guy.

Yeah IMUs are great (we use them a ton at work). I haven't written on my blog in a while, maybe I should start doing that again! How did you integrate the IMU data? Or did you just use the accelerometer to detect if it was moved?

I used the InvenSense chips (6050 and 9250), which do sensor fusion on chip. There's an arduino library to handle all the register calls. I "played" with the orientation calcs at my desk and it was really nice. In the lab we had moving ferrous metal which screwed the magnetometers. I wanted high-frequency linear position and started to use a Kalman filter with GPS data to kill the double integral drift, but the lab again... no GPS signals inside. Settled on the most important for me: raw 6DOF linacc and angvel.

 

[bigshot]

No, I don’t reply to you.

 

[SoundAndMotion]

I understand. In your position, I might not either. Have a nice day.

 

[gregorio]

[1] But yes, you are 100% correct - if you sample above the Nyquist rate, the Shannon-Nyquist Sampling Theorem states that you can perfectly reconstruct the signal. This is absolutely a true statement and you are right, I just don't think it can be implemented in real-time.
[2] This is an excellent point. The thermal noise can definitely be higher than the DAC quantization noise.
[3] I'm interested in how one can go about comparing the DACs. As we said, if we use a rubbish amplifier, obviously the difference in the DACs won't be as noticeable. When people test these DACs, do they use a ridiculously nice amplifier?
[4] I appreciate that we can have a civil debate here

1. I would agree. So the question effectively becomes, how close to the ideal of the Shannon-Nyquist Sampling Theorem can we implement in practice and, is the difference between the ideal and what is implementable audible? Not only can we get remarkably close but modern technology can achieve it for peanuts and reliable/controlled testing indicates that it's not audible, by quite a margin.

2. If you think that even accounting for TDPF dither, the digital noise floor of 24bit would be around -138dB, about the same as the thermal noise produced by a single 10k resistor, so even the output of a DAC, let alone the output of an amp and transducers is always going to be way higher than the 24bit digital noise floor.

3. Standard practice for listening tests when testing ADC/DACs is to loopback x10. Take the DAC output, feed it into the ADC input and record it, then take that recording and repeat the process another 9 times. This multiplication of artefacts makes audible comparisons between ADC/DACs possible. Doing this with Pro ADC/DACs doesn't require an amp because the output is balanced line level, as is the input. Putting an amp in the chain makes the process effectively impossible because the amp will have a higher noise floor than the ADC/DAC, even a good amp and of course the amp is amplifying the ADC/DACs noise floor. For a side by side comparison of DACS it doesn't matter so much because if the amp (and everything else) is unchanged between the two DACs, then the only difference is the DAC. A null test can be illustrative but at these levels and the relative amount of thermal noise, which of course is random (with a Gaussian distribution, if I remember correctly), it's likely to be inconclusive unless there is some specific distortion/artefact. Specialist measuring equipment like the APx555 would probably be required.

4. It's only when when get ridiculous absolute assertions/claims (that have already been debunked numerous times over many years/decades), based on personal perceptions and repeated without any reliable supporting evidence, that the discussion typically turns uncivil. You haven't done that, your assertions are not contrary to the facts/science, they're not based on purely personal hearing perception and you have presented them more as questions than absolute assertions of fact. So even though some of your assertions don't necessarily represent the entire story, a civil debate is entirely possible, and of course preferable!!

G