He's not saying the same thing.
People selling you stuff say you need things that cost more and therefore considered subjectively to be better... because of the theory of "Surely it must be?? Costs so much! "
He's saying you neeed objectively, measurably, better equipment to test. Cost not mentiioned at all anywhere (anything relating to costs is oneself reading between lines, listening to one's internal thought over and above the actual external communication).
If you conduct a scientific experiment to measure gear, all the equipment performing the tests has to be of an order of magnitude more accurate than the thing being tested.. otherwise any results could be the test equipment's margin of error or indeed masking the results completely.
This is audio subjective listening test but still, the surround equipment has to be an order of magnitude (ok, that's more of a figure of speech here, than anything too meaningful) more "accurate" (again, choose the critieria for judging that..) than the equipment being tested.
It's just logic isn't it?
I am only explaining what Guho is saying - I'm not relating it to the details of the original test. I can't be bothered to go back and check what gear was used and how good or bad it was.
Is there anyone else that totally disagrees with the above? The question should be can you hear a difference under normal listening conditions. Where do you draw the line? What is the point if it only sounds better when used in a padded soundproof room while standing on one leg.
Last edited:
I assume your post is directly related to NATDBERG´s post.
What do you consider as "normal listening conditions"? Is listening in a so-called "ABX test" part of normal listening?
Further it is a good basic rule that any generalization of experimental results should be dependent on the conditions of the experiment _and_the conditions of the situation(s) that are covered by the generalization.
You mentioned the "normal listening conditions" and that are the normal conditions in which the practical relevance is important.
If the conditions are different for the individual listeners any categorical conclusion would be questionable.
Guha imo only questioned (for good reasons) the broad generalization that some posters asserted.
@ scottjoplin,
why should a thread like this be closed?
The topic(s) are around for ~ 400 years - the so-called great debate about golden ears and "blind" tests really started in the 1970s - and it will most likely stay for another while. New members will join the forum, not evey member will already noticed this thread, so is it an advantage to close this thread and force opening of new threads whenever this or a similar topic is will pop up?
And then there is the question about digital sound, and perfect sound forever.
An analog direct-to-disc recording is only limited by the size of vinyle molecules to reproduce what is recorded on the disc.
Any digital recording is a sampled process recording, where no attempt at all is made to preserve the original sound.
And the question "what is really there in the recording" has totally disappeared.
There has been some debate lately about single versus multi bit D/A converters, but what about the type of A/D converters used?
I my self has built many D/A converters, and you quickly come to the conclusion that the digital sound can be quite nasty, and you have to spend a lot of time and effort to make it listenable and enjoyable.
And the whole art of making a good D/A ist to mask all the digital drawbacks.
But the art of extracting more good sound from an all analog recording is the opposite to a digital one, to get better sound from analog you strive to open up and reveal more of what there is, not to cover up something.
An analog direct-to-disc recording is only limited by the size of vinyle molecules to reproduce what is recorded on the disc.
Any digital recording is a sampled process recording, where no attempt at all is made to preserve the original sound.
And the question "what is really there in the recording" has totally disappeared.
There has been some debate lately about single versus multi bit D/A converters, but what about the type of A/D converters used?
I my self has built many D/A converters, and you quickly come to the conclusion that the digital sound can be quite nasty, and you have to spend a lot of time and effort to make it listenable and enjoyable.
And the whole art of making a good D/A ist to mask all the digital drawbacks.
But the art of extracting more good sound from an all analog recording is the opposite to a digital one, to get better sound from analog you strive to open up and reveal more of what there is, not to cover up something.
I understand that, but isn´t it true in every direction?
To follow your reasoning, what is the advantage to know about being unable (more precisely that other individuals were unable) to distinguish between two pieces of equipment under extreme conditions?
Conditions that are never going to apply. (means to the normal listening situation at home)
And as said before, Guho imo related to the very categorical (and broad) conclusions that were drawn from the experimental results. (OP´s experiment(s))
Imo one of the reasons for the wide ranging discussion is that this original experiment didn´t have a clearly expressed objective, therefore the conditions were almost randomly changed within which is a "deadly sin" in these kind of tests.
Last edited:
if swapping btwn high/low end dac makes no audible diff, this is a strong signal suggesting the presence of bottlenecks
again this is a diy forum talking expensive gears is not too useful to me, as many cheap diy projects are good enough for the job
Maybe you've forgotten the beginnings of this thread and are introducing a different topic of your own.
This thread is really about testing gear, not about what you have at home.
Everyone is free to do what they want and select gear on any critieria they want! Personally aesthetics, build quality and general 'feel' of equipment features highly in my criteria (often why I have trouble with DIY builds - I'm only happy if it's got a high quality feel to it)
No-one's talked about extreme conditions.
Basic logic of test set-up is what Guho was talking about..
To your knowledge.... ? Have we ever met?
I enjoy it more, yes, and that mostly to do with practical environment - I live in a busy city, in a block of flats. I don't see what relevence it has, other than some daft ad hominem? I didn't consider that my comment was even worth ad hominem, but I guess these things can be very important to some?
I listen mostly to single-speaker radios if you were to log all music listening. As I am now - going to change to a good radio show on Mixcloud now, via ipad/hifi if you're interested. The music is more important than being too anal about sound quality.
All this discussion about the playback system not being sufficient to reveal the type of differences expected to be found is only possible to resolve when some positive controls are used i.e audio files of known audible differences are used in the test to evaluate the sensitivity of the playback system/listener's sensitivity
How many such controls have you ever seen used or even suggested by the ABX proponents? Ever wonder why such simple test calibration is never mentioned by these proponents?
Without this self-checking control, the test is of unknown quality
seemed odd to mention it other than to signal to people to discount my opinion for some reason - clearly not only were yours twisted but probably riding up to a wedgy ...
You've got enough self righteousness to go around
There are no reasonable people on the planet. This is a tidy "no true Scotsman" logical fallacy. Not to mention the extreme power of influence. You hear someone call a capacitor "dark" and you buy it in the hopes it does the same in your own system, then the likelihood is you're going to hear it as "dark" too.
I'm sorry, your rant is a caricature of the testing space entirely. Double blind testing is the gold standard, and listening impressions are not for a good reason. Since you want to take the side of science, I invite you to pick one forum, bias it as heavily as you want to your side, and count the number of people influenced by badly designed DBTs versus the number of members influenced by listening impressions that are stated as fact. Pick a month or two so you don't lose your mind. Pre-describe your selection criteria and test protocols like many journals and funding agencies are demanding (best practices). I want to see the implications of "many people".
So individuals reports on audible characteristics and differences is reliable and yet DBT is not? Scientifically persuasive veneer indeed. You're painting yourself into a corner as an anti-DBT troll, not someone remotely interested in the truth and science. And please go ahead and find me a test where they showed that impressions are durable across a volume change.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.