Well, there seem to be only hard-core audiophiles with "esoteric" claims on the one side,
and DBT fans that believe that a "null" result proves anything (which is unscientific
as well) on the other side with no room in between.
Remember that there were DBTs which proved that a 128kB mp3 is indistinguishable from CD,
a AD-DA chain from the early eighties is sonically transparent, all amplifiers that are
not driven into clipping sound the same, etc.
Also remember that a DBT for audio relies solely on a listeners ability to dedect an effect in
a conscious way and to make a clear distinction. (I´m not aware of a medical blind test
for example where the result would reley on questions like "could you distinguish drug A from
drug B".)
So my take on this would rather be, if a DBT shows a positive result, fine, then there
is certainly something to investigate. If you get a null result, you might well say
a difference does either not exist or is so subtle that you don´t care, but you cannot say
there is proof.
Where the typical esoteric (and "snake oil") claims typically fail in my opinion, is
to present a plausible mechanism for the allegedly perceived effect. If there is (and only if)
a plausible mechanism, a set of tests can be crafted (including DBTs if needed) to check
if it exists.
An ("anecdotal" yes I know ;-)) example: I had the opportunity to listen to a pair Stax DA-50M
over some days. It´s an amplifier switchable from class A/B to A. It´s excellent either way
but I preferred A when listening over a longer (half an hour or more) period of time.
In A it has lower high order distortion products (one plausible mechanism - there might be others too)
but measures very good in A/B too.
I´m pretty sure I would _not_ be able to distinguish those two settings in a DBT.
I think we should stick to an engineering approach here, and leave the question what is
audible or not to the fields of biology and medicine.
and DBT fans that believe that a "null" result proves anything (which is unscientific
as well) on the other side with no room in between.
Remember that there were DBTs which proved that a 128kB mp3 is indistinguishable from CD,
a AD-DA chain from the early eighties is sonically transparent, all amplifiers that are
not driven into clipping sound the same, etc.
Also remember that a DBT for audio relies solely on a listeners ability to dedect an effect in
a conscious way and to make a clear distinction. (I´m not aware of a medical blind test
for example where the result would reley on questions like "could you distinguish drug A from
drug B".)
So my take on this would rather be, if a DBT shows a positive result, fine, then there
is certainly something to investigate. If you get a null result, you might well say
a difference does either not exist or is so subtle that you don´t care, but you cannot say
there is proof.
Where the typical esoteric (and "snake oil") claims typically fail in my opinion, is
to present a plausible mechanism for the allegedly perceived effect. If there is (and only if)
a plausible mechanism, a set of tests can be crafted (including DBTs if needed) to check
if it exists.
An ("anecdotal" yes I know ;-)) example: I had the opportunity to listen to a pair Stax DA-50M
over some days. It´s an amplifier switchable from class A/B to A. It´s excellent either way
but I preferred A when listening over a longer (half an hour or more) period of time.
In A it has lower high order distortion products (one plausible mechanism - there might be others too)
but measures very good in A/B too.
I´m pretty sure I would _not_ be able to distinguish those two settings in a DBT.
I think we should stick to an engineering approach here, and leave the question what is
audible or not to the fields of biology and medicine.
Last edited:
That is exactly what audiophiles do - trust their ears and have no intention/willingness to reaffirm their trust with a DBT
That's an elliptical way of saying that they don't trust their ears.
A null DBT result means that those who claim obvious/'night and day'/clear differences have failed to establish that as fact. That is useful information. If you want proof you instead become a mathematician.
A positive DBT result could still be due to chance. Hence 'golden eared' people can never prove their skill, but given sufficient tests they could establish it beyond reasonable doubt. This has not yet happened.
A positive DBT result could still be due to chance. Hence 'golden eared' people can never prove their skill, but given sufficient tests they could establish it beyond reasonable doubt. This has not yet happened.
You appear to be assuming that we will naturally assume that all these results are obviously false. This reminds me of the way trendy liberals assume that it must be wrong and evidence of unfair gender bias in recruitment if equal numbers of boys and girls don't go into mechanical engineering and midwifery. What if all correctly-used amplifiers really do sound the same? What if boys and girls really are different?gk7 said:
Really? And the previous posts in this thread arguing the null result is no proof are what?
There was no such DBT result. Negative results only mean the test assumption "a difference can be distinguished" was not confirmed by the test taker in this particular test setting on that particular test equipment. I.e. that tester results did not get above the agreed limit between random(incidental)/non-random(controlled). Nothing more, nothing less. No general rule could be drawn from either result.
However, if someone claims to hear something on his gear, he should definitely accompany his claim with DBT results testing his ears on his gear. His claim is no general scenario but a very specific phenomenon within scope of DBT validity.
The engineering approach reached the pinnacle with the earliest cd players and the Jap amps of the seventies. There has been no objective reason for development in either amps or digital ever since. Ok, maybe class D and MP3 were the final steps on the road to objective Nirvana but thankfully both are now conquered.
So, i guess we are just killing time here
Well you said that and I agree, but the "has not been proved in a DBT so it does not exist" argument is used quite often.
Sorry, yes it was 256k, some musically trained persons could dedect 128k:
In german unfortunately: Der c't-Leser-Hörtest: MP3 gegen CD | c't
Yes, exactly. (Can´t see where we disagree about that BTW)
I have already posted that example in another thread where you asked the same question some weeks ago. But ok (to big to attach):
http://www.hal9000.at/tmp/all_amps_sound_the_same.pdf
Last edited:
Old stuff with a limited sample universe. The structure of the question being asked is different as well. Sorry, that's not even close. I would have no trouble setting up a test between two amps that aren't clipping and having you be able to tell one from the other in a DBT. I couldn't do the same with two functioning USB cables.
No.gk7 said:
It is known that all amplifiers measure differently. This does not mean they will sound differently because the parameter may be too small to be heard. That is the whole point of tests: to find out which measured differences are important and which are not.
It is easy to show that certain measurements are different for boys and girls, especially after puberty! This doesn't tell you much about career choice, but it does dismiss any claim that they are essentially the same. If minor differences in body chemistry can create significant differences in body shape then it is quite possible that these also create differences in preferences. Anyone who has known any boys and girls (or men and women) will know that this is the case. It is merely silly for people to assume the opposite and require sensible people to do the same.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.