advertisement


Anyone tried blind testing DACs?

h.g. you've done it again, a bit, but I think the gist of what you are saying is that it is straightforward to test whether something is inaudible, rather than whether it is audible.

Would you have any suggestions as to the form such a test might take, in the sort of context we're dealing with here?
 
Sue Pertwee-Tyr said:
What you are saying here, I think, is that if you do a blind test, and you don't hear a difference, that is not proof that no audible difference exists. Which does somewhat invalidate Keith's mantra.
Click to expand...
Yes, correct.
Sue Pertwee-Tyr said:
I agree with you about the discussion of stats. I didn't do stats at school (we did pure maths) so most of this goes over my head, but also I presume h.g. is a scientist, because he writes in that style that scientists use, that seems almost designed to repel non-scientists. Technical scientific writing is not alone in this, of course, many disciplines (especially the law, in which I have a little experience) also use jargon and a written style and terminology which excludes the lay reader, but on a forum such as this, it is arguably more appropriate (and more polite) to write for a lay audience, or take your discussion off to pink petri-dish media, or something.
Click to expand...
I think that most people who have deep knowledge of a certain subject can explain it simply and in a way that a layman can understand.
 
"Most people wouldn't know music if it came up and bit them on the ass." - Frank Zappa.

If this discussion is anything to go by then you could justifiably substitute "audiophiles" for "people".

I've just been comparing two DACs (both solid-state) and they couldn't sound more different, truly chalk and cheese: one makes me want to continue listening, the other doesn't, and I'm confident I could distinguish between them in a blind test.

The sonic attributes that so many "objectivist" audiophiles obsess over are completely irrelevant to me.
 
Thetiminator said:
There’s absolutely no doubt that taking your eyes out of play enhances your hearing. Only a fool could argue otherwise. Of course this enhancement is not linear over time, or like an on/off switch.
Click to expand...

? are you confusing listening in the dark with blind testing, where blind means "not knowing what is playing", nothing to do with having eyes closed or masked or room darkened.

I agree BTW that listening in the dark does enhance hearing, so blind blind tests might be a good idea.

Tim
 
Sue Pertwee-Tyr said:
h.g. you've done it again, a bit, but I think the gist of what you are saying is that it is straightforward to test whether something is inaudible, rather than whether it is audible.

Would you have any suggestions as to the form such a test might take, in the sort of context we're dealing with here?
Click to expand...

I have already done this several times in this thread. Joe has contributed. What more is there to add? It is up to you to decide what is a fact and what is marketing.

If you haven't twigged my main interests are a technical one in sound reproduction and how and why the audiophile phenomenon works. I am not an audiophile objectivist. I have little interest in persuading audiophiles to respect technical/scientific knowledge as opposed to being interested in why they don't. As mentioned earlier I have no interest in audibility experiments to demonstrate what is already established for the technically literate. The ones I might be interested in performing would relate to trade-offs in what is perceived due to reflections in small rooms (i.e. the stuff that actually matters in terms of technical performance) and even here there is a fair chance the existing literature is sufficient for my needs.

If you want to be guided to the path of truth and light you will almost certainly be better off approaching someone like Jim Lesurf who is technically literate and with a much higher tolerance for audiophile nonsense than me. Perhaps Darren will continue his interest.
 
h.g. thanks, and so sorry to trouble you. It seems my layman's reading of what you wrote hasn't given me any idea that you'd actually described the sort of tests one might carry out. I saw lots about how and school kid could design such an experiment, and lots about statistical analysis, but nothing about the experiment itself. Perhaps we could invite a schoolkid to come on and explain what they'd do?
 
onlyconnect said:
? are you confusing listening in the dark with blind testing, where blind means "not knowing what is playing", nothing to do with having eyes closed or masked or room darkened.

I agree BTW that listening in the dark does enhance hearing, so blind blind tests might be a good idea.

Tim
Click to expand...

Yes sorry taking your eyes out of the equation completely. I was replying to something earlier...
The difference your sight makes to what you hear is huge...but not a realistic to test for picking a product..
 
Thetiminator said:
Yes sorry taking your eyes out of the equation completely. I was replying to something earlier...
The difference your sight makes to what you hear is huge...but not a realistic to test for picking a product..
Click to expand...
Not least because if you've got your eyes shut, how do you know you picked the right one up when you left the shop?
 
With regard to removing bias- is there really any point? If someone is determined to prefer a particular product, will them walking out of the shop with something different lead to long term satisfaction?
Maybe convincd themselves that it didnt sound better in the shop and buy their original preference? Hey if it makes you happy.....
 
Blind testing has its place, but my motto is buy what you like, like what you buy.

Well, that and live long and prosper, especially during a pandemic.

Joe
 
Sue Pertwee-Tyr said:
Perhaps we could invite a schoolkid to come on and explain what they'd do?
Click to expand...
If you are comfortable with the possible result. Alternatively as originally suggested you could quietly try to design a scientifically valid audibility experiment yourself in order to bring out what is an association (the key to how audiophile nonsense gets accepted) and what is sufficiently real to be precisely identified, predicted and measured as required by the scientific method. No need to perform the experiment given the process of designing one is likely to be rewarding for those with a genuine interest in what is true in a scientific sense but start off with a bunch of conflicting audiophile beliefs.

Of course people that are inclined to work things out for themselves are rather unlikely to have picked up audiophile beliefs in the first place and so it is not a particularly realistic thing to ask but it does make a point.
 
h.g. said:
If you are comfortable with the possible result. Alternatively as originally suggested you could quietly try to design a scientifically valid audibility experiment yourself in order to bring out what is an association (the key to how audiophile nonsense gets accepted) and what is sufficiently real to be precisely identified, predicted and measured as required by the scientific method. No need to perform the experiment given the process of designing one is likely to be rewarding for those with a genuine interest in what is true in a scientific sense but start off with a bunch of conflicting audiophile beliefs.

Of course people that are inclined to work things out for themselves are rather unlikely to have picked up audiophile beliefs in the first place and so it is not a particularly realistic thing to ask but it does make a point.
Click to expand...
No, sorry, you're doing it again, so I'd probably best leave it there.

You're telling me to design an experiment, despite me telling you that I have no such experience to draw on. And if this is what you mean by 'I have already done this several times in this thread' then we have a fundamental failure to communicate, you and I, because I inferred from that that you'd already explained what such an experiment would look like and I'd missed it.
 
Steve,

Sue Pertwee-Tyr said:
How about 'live long enough, and prosper enough to be able to enjoy some decent hifi'...
Click to expand...

Aye. So many good Trek lines to choose from, such as...

412cgi.jpg


Go ahead. Do it because you're Worf it.

Joe
 
Rigorous blind testing has a place in audio design and development. It's how we learned, for example, how shitty an mp3 file can be before people notice. In a sense, it's like this science fair experiment.

90f.jpg


Joe
 
gingermrkettle said:
Just a little bit of cart before horse in that statement.
Click to expand...
Quite the opposite. One cannot design an experiment to measure something without first knowing precisely what that something to be measured is. Not the amount which is obviously part of the result of the experiment but the something that is being measured. This is what will push an audiophile towards accepting a scientific definition of sound perception in order to make measurements. Vaguer audiophile notions and associations concerning sound perception will start to fall apart when faced with trying to measure them.

As pointed out earlier in this thread one of the marketing tricks that supports audiophile nonsense is putting forward unscientific rather than scientific hypotheses. That is, ones where the quantity cannot be defined in a form that can be measured. People with a scientific/technical outlook tend to see such hypothesis for what they are but audiophiles don't which is why the marketing works.
 
Anyone ever done a blind test where you think the difference is gross, only to find out you're just guessing?
 
I've done two or three blind tests with DACs a few years ago.

I could spot the difference between an Auralic Vega and the DAC I had at the time, an Audiolab 8200CD. But it was much tougher than I had expected.

Perhaps the incentive for you to get it right makes it more difficult to spot differences...
 
You must log in or register to reply here.


advertisement


Back
Top