Sorry but before I waste my time & to prove you are not using politician speak "perfectly acceptable evidence", please detail what you mean by this.
I mean what you meant when you said "perfectly acceptable evidence" but please don't waste your time, neither me nor anyone else here is worth it
I was politely asking for evidence at the standard you claimed for yourself - that what you say is (quote) "verifiable in the perceptual testing research". You haven't offered that, as far as I can see.
Examples of ABX tests that fail to identify an "obvious" difference are not the same as evidence for the claim that there are effects in human audio perception / physiology where differences reliably detected via "longer term" listening cannot be reliably detected via short term listening.
If you're saying that you have evidence that there exist "obvious" audible differences that no ABX test can detect that would be interesting to me, assuming the definition of "obvious" at least includes repeatably detectable unsighted.
OK. You already know I am thick, but perhaps you can help me here. Two different but otherwise very similar ABX tests cannot be compared because everything might be different - I think you told me. Yet two different types of test (short ABX, long preference) can be set up so they provide results which can be compared. Have I missed something?mmerrill99 said:
Yes, you have missed the different sensitivity of each ABX test.
Because of this you cannot judge the threshold audibility of ABX testing in general by using tests of unknown sensitivity, no matter how many such tests you use. All that can be done & it is mandatory, is to use internal controls within each ABX test which can give an indication of it's sensitivity.
Jakob2 already mentioned this but the standards for blind testing state that hidden internal controls (hidden reference) are used in order to provide a measure of the sensitivity of the test & the listeners.
From here The Well-Tempered Computer
Last edited:
Hmm, it's interesting how you change what I said (& you quoted) into the above words ""obvious" audible differences that no ABX test can detect"
Note: Underlining is mine so you don't miss it
Which is now a different request to your first request "Do you have evidence of that specific scenario? I.e. (presumably blind) long-term listening tests with (presumably) repeatable preference expressed, but where there is no consciously detectable difference? It'd be a very interesting result."
Funny, really!! I said I can provide evidence of ABX test results where obvious differences weren't detectable but which you rejected & you now try changing my original statement into something I didn't say!!
My original statement as you quoted it - my underlining
Last edited:
If you only meant "some ABX tests fail to identify an obvious difference" then I don't doubt it (although a definition of obvious would be needed). But regardless, that's a much less strong claim than what you said upthread:
Again, where is the "perceptual research" to verify this (strong, and counterintuitive, IMO) claim about the nature of human audio perception?mmerrill99 said:
Last edited:
Here is the “logic” with which you are establishing the validity of the ABX test as applied by JonBocani:
Newton's Laws of Motions are correct – therefore JonBocani’s ABX approach is valid.
Can you not see that this is a completely vacuous rationale?
Are you intentionally ignoring the fact that the criticisms (at least from me and Mmerrill) are based on a disagreement with the approach and not a disagreement with the results?
Since you seem unwilling to do any background reading on ABX testing, and find evidence of an agenda to be more informative, stay tuned.
Regarding long term ABX or other long term blind comparisons, seems to me what happens is whatever we are accustomed to listening to becomes the new normal, and if it is abruptly changed, it can be very easy to notice at that point, especially if it changes for the worse such as becoming more distorted.
Noticing changes in the other direction for the better can be harder, especially if they are small.
For example, its probably easier to notice if your car develops a new little rattle over bumps, than it is to notice if the rattle stops once you have become used to it.
Noticing changes in the other direction for the better can be harder, especially if they are small.
For example, its probably easier to notice if your car develops a new little rattle over bumps, than it is to notice if the rattle stops once you have become used to it.
Last edited:
Sorry to disappoint you but it is exactly what I said as anybody can read for themselves. Your failure to understand this is not my responsibility.
What strong & counter-intuitive claim are you talking about?
Originally Posted by mmerrill99
"..we perceive differences because we are physiologically affected by the devices but this isn't so available to consciousness which is what is being focused on in such forced choice blind testing. Over longer term listening we become more clued into how this device is affecting us"
This is quite spooky, it seems as though the device is trying to sell itself, could it have become possessed by the marketing department?
"..we perceive differences because we are physiologically affected by the devices but this isn't so available to consciousness which is what is being focused on in such forced choice blind testing. Over longer term listening we become more clued into how this device is affecting us"
This is quite spooky, it seems as though the device is trying to sell itself, could it have become possessed by the marketing department?
The one directly quoted.
Where is the perceptual research to support this? It's an objective claim about human audio perception.mmerill99 said:
If an approach is invalid, per se, doing it over and over doesn't resuscitate it.
Using the term "statistical" doesn't change the fact that the results are still invalid.
There is no threshold of repetition where the validity curve inverts.
You simply make a deeper pile.
I told you blind ABX test results can show obviously audible differences not being differentiated. But you seem to suggest this is not perceptual research? Can you say why?
Mmm, I think after normalisation, most changes are likely to be perceived as improvements. In the longer term, isn't it probable that the most distorted sound is also the most tiring?
Still, wouldn't be a bad idea to find out (calibrate) what people *can* hear under test conditions, room, fixed equipment, group of listeners, etc., things that won't be changing during the test. There should be some threshold of sensitivity you can identify.
Of course, that could take some of the fun out of it if you know all the equipment to be tested that day differs below the threshold of sensitivity. I guess you could hide the calibration results from everybody including yourself until after the testing is over.
A more interesting test would be to take people convinced that they are hearing "significant" differences sighted, and then go through an ABX. In the above test, sighted vs. non-sighted results are more or less identical ("extremely thin"), for whatever reasons (could be that there are no real differences, or some other explanation). This ABX test does not really demonstrate anything.
Another instance where ABX testing is interesting is when you compare perception versus measurement. In this case, also, there were no measurements.
I don't know. When I replaced the Benchmark DAC-1 with a DAC-3 it sounded a little different including a little more clear and detailed. My son who usually doesn't hear small differences did hear something small to him at the time. He was able to describe it to me, based on that it seemed he was hearing it about the same as I did. Now we are both used to the DAC-3. But, getting back to the issue, I wouldn't say it was tiring at all before the change, but it was already pretty good.
On the other hand, there have been systems I have never been happy with no matter how long I had to get used to them.
Last edited:
Yea, it's interesting that something doesn't have to sound 'bad' for another device to sound 'better'