Blind tests are much less subjective than sighted tests, so I don't see how you can really object to them on this basis. You can always do ABX tests instead, which take subjective impressions out of the picture even more and asks the listener to only compare X to A and B.
Blind tests are subjected to the same expectation bias as sighted tests as I mentioned somewhere in that post. When someone does not see the actual switching, the expectation is that... there was no switching that actually occurred, and they may tend to think it should be the same as before. I'm not necessarily objecting to it as much as raising some potential issues that I think would affect the test result and causes the test to be inconclusive, as it has for the past while. ABX test also has its own issues, because the subject is under pressure rather than relaxing and just giving what they think. They have to strain to compare now, which is not ideal.
A blind test doesn't require every listener to be in the room at the same time, so volume can be adjusted on a per-user basis. Overall volume can be adjusted digitally to a listener's comfort level once level matched. As for the headphone issue, we have to establish that there's an audible roll-off first and we can do this with an ABX test which again asks the listener to identify rather than describe the amps. Why are you so hung up on finding subjective preferences? We're interested in finding audible differences.
Digital volume should not be adjusted because you may cause dynamic range to be compressed or some other oddities depending on the software setup. It is best for this to be done in the analog domain... But that's not the real issue I wanted to raise. The issue is that people simply hear things differently. Whether a roll-off is audible or not can be determined using an objective measurement of the frequency response, and some amps on the higher-end do intentionally EQ high frequencies down. If the amp does not do this then some DACs do. Case in point: look up measurements for the iQube V3 and AMB Gamma2.
I'm not hung up on finding subjective preferences, but I'm trying to explain how personal preference can affect the test results of a comparison test. Again, this is purely subjective, and is not along the same line as audibility. It's not like we're asking listeners to determine whether or not they can hear something playing in a completely silent room.
You can in an ABX test. Because, once again, in the world of amps when most everything actually sounds the same, we have to establish that there are audible differences at all before we worry about which one people like more.
In reality, though, amplifiers should not sound the same. They have measurable differences. Whether the differences are audible or not are tested using studies that rely on subjective impressions, which... as I have mentioned already, is not reliable. Objectively, there should be a difference. Unless you're saying measurable differences somehow don't count.
Most differences between DACs are smaller than the ones between amps, even audibly transparent amps, and their distortion isn't affected by the load itself like an amp's. If there are differences to reveal, virtually any DAC will reveal them. The headphone is important, but the HD800 isn't the one to use. It has high impedance, making it an easy load for the amp, and it's open back which will allow ambient noise to bury the tiny noise and distortion differences in amps. Probably a better choice would be a high isolation low impedance IEM, which will present a more difficult load and increase the amp's distortion. A multi-BA IEM can be used to test for differences caused by output impedance as well if desired.
People have claimed to hear the differences in DACs, though, so... I don't think it's as clear cut as saying the differences are small. Also some DACs intentionally shape the impulse a certain way. Please look up the AMB Gamma2 and "Minimum Phase" filter.
On that note, high impedance does not necessarily mean "easy load". It only means "low current" needed, and not as much interplay with output impedance. Amps that do not have the voltage swing necessary for a certain loudness level may still audibly clip or distort. Solid-state amplifiers relying on op-amps will also run into this if power supply quality is not taken into consideration. When all else are equal, though, the differences may still exist due to noise level, as well as the op-amp's inability to cope with capacitive load, which most headphones would present due to their structure (cable has capacitance, and also the diaphragm construction has capacitance). You'd be surprised to know that most op amps don't really play well with capacitance in the output, even when said capacitance is extremely small.
Back on the discussion, I don't think IEM is the way to go, because not everyone can fit IEM. Also because not all IEM can stand up to the technicality of something like the HD800, so they may not be able to resolve as much as the HD800 can. Stax SR-009 would be my pick, but then it doesn't work with many amplifiers.
On a personal level, because amp performance is so dependent on the load, the headphone to use is one you will be listening with on a daily basis. It doesn't matter how an amp performs with a 300 ohm load when your headphones are 32 ohms, for example.
Agreed. The test should be conducted with a headphone the user is familiar with. However, you wouldn't want to use a tube amp with 120-Ohm output impedance on a 32-Ohm headphone. I have tried. The distortion goes way beyond "more bass and warm sound".
Expectation bias is why we ask the ones who do hear a differences to test themselves, because there's no point testing Sound Science regulars if they don't think there will be a difference. Again, ABX can help with that slightly and reveal differences we might not have expected, but the results will still be more conclusive coming from someone who's actively trying to prove the differences exist. But anyway, I think the test you're looking for is an ABX test, not a regular AB blind test. Or we can do it the easy way and record the output of two amps and compare the results, but this bypasses the headphone load which will greatly affect the amp's output.
But Sound Science regulars are the ones who think there is no difference.
At least that is the vibe I'm getting from here.
When you say something like "all amps sound the same or the differences are not audible", and someone say "I can hear amp X is warmer than amp Y" and then the actual objective measurements tend to show minute differences, as a scientist, who am I inclined to believe more, right?
It makes no sense from an electrical standpoint that there should be audible differences. Exactly how revealing do you think our ears are?
It makes sense from an electrical standpoint. In fact, allow me to show you an article on measurements of different types of capacitors...
I think our ears are as revealing as they can be. Tests that ask people to replay their subjective impressions should be taken as a yard stick rather than as facts because... again, people are really unreliable.
When I can measure a difference, and my ears cannot pick those up, that means the problem lies with me (or perhaps my headphone/speaker), and not with the amp or DAC itself.
The O2 isn't a giant killer. It's so far past audible noise and distortion with most loads that it's pointless. Probably one of the $80 Fiio amp/DACs is the giant killer, they don't measure as well but will be audibly transparent most of the time anyway. Probably more linear than most kilobuck amps.
You would think it's past audible noise and distortion, but I can show you otherwise. It would still interact with IEMs... which should be easy loads, and even when its output impedance is at 0.5 Ohm. That should tell you
Most amps are perfectly flat from 20 Hz to 20 kHz, and most competent solid state amps have distortion specs too low to hear at any frequency. With a low output impedance, there's no EQ effect at all. There's exceptions of course, but bringing it up in a discussion of EQ is more of a distraction than anything.
Most amps are actually only "relatively" flat from 20Hz to 20KHz. Zoom in on the measurements and you'll see bumps. Notice they usually present frequency response graphs in massive scales. This is actually more due to limitations of the measuring equipment.
But anyway, even when frequency response is not taken into account, distortions and noise are still different, so the amps should, theoretically, be different. Saying they don't sound different is just a subjective impression, and not necessarily science IMO.
Unless I am mistaken, science is the pursuit to ultimately understand and comprehend a phenomenon, and not a pursuit to dismiss something based on subjective perception.
On that note, I only brought it up to further expand because some may tend to think high-end amps do EQ, but that's not the whole truth.