You are right. I guess calling them snide remarks is giving them too much credit:
No, I didn't. I answered. You ignored my answer.
Anyway, this thread is in The Lounge for a reason. Let's leave it here, and finally let it rest in peace!
The tests don't usually allow for that, it's usually a switch box.
Some people say the fast switch is absolutely necessary and needed, like Julf's friend Ethan Winer.
In essence you are right, for sure, blind listening per se is very flexible.
I am glad that wasn't a snide remark 🙂
So let me leave with a quote from J. Gordon Holt, the founder of Stereophile:
I listened to your answer about the higher performance.
Another way to phrase it is why not aim for more advanced design?
With a theoretically perfect design then we don't need to take listening data evidence into account.
Aim upwards not horizontal. That is my solution at least.
______
Anyway you're still avoiding my question about the evidence, if you have any.
Perhaps it's a "decline to answer"?
Oh well, one last answer. Clearly "rest in peace" is a concept that doesn't translate...
How do we specify "more advanced" or even "better".
Take 2 playing systems. One costs $50, fits in a shoe box, and uses 1 W when idling. The other costs $2500, weighs 75 kg, takes up 3 units of rack space, and uses 250 W. Both sound the same. Wouldn't you agree the first system is "better"?
There is no point in over-engineering unless it brings benefits - *audible* improvements in sound quality or increased reliability etc.
How do we specify "more advanced" or even "better".
Take 2 playing systems. One costs $50, fits in a shoe box, and uses 1 W when idling. The other costs $2500, weighs 75 kg, takes up 3 units of rack space, and uses 250 W. Both sound the same. Wouldn't you agree the first system is "better"?
There is no point in over-engineering unless it brings benefits - *audible* improvements in sound quality or increased reliability etc.
We know there are no "theoretically perfect designs". There is only "not possible to improve it in a way that would make an audible difference". And to determine that, you need listening tests.
All fine as long as you know which way is up...
Evidence that ABX works? Do you have access to the AES library?
Ah but to which one could you add bespoke expensive audiophile cables ($275/m or more) to tailor the sound and reveal that hidden detail that you know is in there hiding behind some copper atoms and which gives the best boasting rights...
Hundred dollar cables would just look silly on a $50 system!!! and other even more expensive additions would look even sillier, all those QUANTUM tuning aids only work on systems that cost a lot of money, they have no effect on cheep systems that can play undistorted music.
😀
No, not everybody does, unless they pay. Individual papers are pretty expensive unless you are a member, and many organisations have a subscription.
You could start with this one.
I also recommend reading Sensory Evaluation Techniques by Morten C. Meilgaard, B. Thomas Carr and Gail Vance Civille.
No, you got it all wrong. Make like 47Labs - take the $50 one and mark it up to $2500 and you're getting somewhere.
The first one is statistical analysis, that's not applicable to a null result due to masking effect.
The second is a book, I'm not ordering a book then continuing this discussion when it arrives in October.
No internet resources then?
The second is a book, I'm not ordering a book then continuing this discussion when it arrives in October.
No internet resources then?
If you say so...
Oh, it is all about instant gratification, isn't it?
Only Chuck Norris can lead a horse to the water and make it drink.
In brief fast switching is necessary for pitch and volume, but there is skepsis how well it works with "the rest" of sound, i.e. everything which is not pitch and volume.
It's well known that we are constantly deleting sensory input.
It's well known that we are constantly deleting sensory input.
Point taken. No more feeding of this particular one (or even trying to make it drink).
In other words you have no internet resources.
Nor book citations for us.
Then as far as this thread is currently aware, there is no evidence that ABX is "all revealing".
I want to be clear though, almost no one is against blind testing, definitely not myself, but I do "fear" the idea that it masks differences at times and I think I have reasonable data / theory to think that it may.
If you don't fear that idea, like NwAv-ist's don't, just whisk it off, that's not "completeness".
If we entertain that ABX testing is flawed it doesn't seem all that deadly to me anyway.
Put it this way, if someone really needs ABX to differentiate say MP3 versus FLAC, i.e. they can never hear the difference without ABX, isn't FLAC fairly pointless in that case? Since they don't listen to music with ABX anyway.
So if ABX theory is flawed, it's not really much of a loss to the listening community anyway.
If you can hear it sighted, you should be able to identify it blind, without any A/B at all! That is more my line of thought!
It needs practice though, selecting random university students with random equipment won't reveal anything at all. Practice and familiar equipment is necessary / key.
I think the unique noise patterns in certain equipment construct unique neuron networks over time.
Imho.
Put it this way, if someone really needs ABX to differentiate say MP3 versus FLAC, i.e. they can never hear the difference without ABX, isn't FLAC fairly pointless in that case? Since they don't listen to music with ABX anyway.
So if ABX theory is flawed, it's not really much of a loss to the listening community anyway.
If you can hear it sighted, you should be able to identify it blind, without any A/B at all! That is more my line of thought!
It needs practice though, selecting random university students with random equipment won't reveal anything at all. Practice and familiar equipment is necessary / key.
I think the unique noise patterns in certain equipment construct unique neuron networks over time.
Imho.
Last edited:
Coming in late here - and I haven't read the thread - I have direct experience of 12 bit resolution without embellishment by dithering, say, not being good enough: our 20 year old Yamaha keyboard uses samples at that level, I pretty certain, and you can hear the "blockiness" in the decay trails of the acoustic grand notes quite clearly. Okay, 20 year old plus technology; later units would use full 16 bit depth, and smart dithering everywhere - but it fairly clearly demonstrates, in a borderline case, limits of the resolution required in digital, using no extra "tricks" like dithering, to get "transparent" sound ...
Last edited:
- Status
- Not open for further replies.