If after listening for a while, you "learn" to spot something in the sound, why would it go away just because you're not allowed to use anything but your ears?
All you have proven is that above certain thresholds noise and distortion is audible and noticed to a statistically significant degree.
What these are are by now well established for test tones over headphones in an otherwise anechoically quite environment. In real environments with ambient noise and real signals with masking etc., they are higher, exactly how much by is a lot to do with the signal.
rcw
What these are are by now well established for test tones over headphones in an otherwise anechoically quite environment. In real environments with ambient noise and real signals with masking etc., they are higher, exactly how much by is a lot to do with the signal.
rcw
As I mentioned in an early post, because one's mind likes to perceive continuity in experience. I've had this effect occur to me on numerous occasions, listening to my own and other people's systems: once a certain aspect is recognised which is seen, incorrectly often, to be a contiguous part of the whole then the brain "wants" to keep hearing that same message, and "fills in the gaps".
If the anomaly is filed, within your brain, as something that exists quite apart from the main stimulus, it's extraneous in a very strong sense from the key experience, then it's relatively easy for you, the listener, to pick its absence or presence. For example, very light rain is falling on the roof while listening to music, it stops and starts rapidly; you'll find it very easy to say at any point in time what the current status is. But, then you have the effect of very light rain as a sound effect within the music, that fits in so to speak; and deliberately, effectively stop and start that effect. I would suggest that a lot of people would find it confusing, and make many mistakes when asked whether the effect was there or not.
Frank
Yeah - have a look at Tom Nouisane's site, for example
L.
Okay, thanks. I looked at the "flying blind" paper, though, I don't agree with a lot of what he says. He says we are more sensitive to a stimulus when first exposed to it which is not true. Then he contradicts himself and explains that he'd much prefer an experienced listener doing blind testing as they do much better than inexperienced listeners.
He thinks extended listening is worse. Not true. We respond much more strongly to sounds we are familiar with, and that have importance to us.
I have complaints with most blind tests I see. I will have to do my own blind testing sometime, to satisfy myself as to what I can and can't hear.
John
Last edited:
As an example of how I get around my brain trying to "fool" me by glossing over something, is that I deliberately emphasise the characteristic, make it stand out as something apart from the "musical message". So, use very "bad" recordings which means that distortion unconnected with the music really hits you "between the ears", and listen with the ear very close to the tweeter at normal listening levels. Problems with the reproduction are bolded, underlined, increased font size, and marker penned, like all get out, this way ...
And don't worry, I don't do this for long, a couple of sec's tells me all I need to know ...
Frank
Look here.
I cannot see where and how it was beaten to death, and I'm still waiting for an answer to the last question from Jakob2.
That is a superb idea- I wish more people had your curiosity. Please don't hesitate to contact me off line if you want any help in setting it up.
John,
Looks like second stage of JC 80 ? Phono preamp stage with active RIAA bass boost. First stage likely has passive RIAA HF filter at it's output. Was the input jfet 2sk146 or coupled 2sk147 - don't think you were using 2sk389 back then ?
Is there a door prize? A bottle of Patron anjeo ? A six pack of Bud?
Looks like second stage of JC 80 ? Phono preamp stage with active RIAA bass boost. First stage likely has passive RIAA HF filter at it's output. Was the input jfet 2sk146 or coupled 2sk147 - don't think you were using 2sk389 back then ?
Is there a door prize? A bottle of Patron anjeo ? A six pack of Bud?
Last edited:
John,
Looks like the second stage of maybe the JC 80 Phono preamp stage with active RIAA bass boost. First stage if it follows your usual topology likely has passive RIAA HF filter at it's output. Was the input jfet 2sk146 or a pair of thermaly coupled 2sk147 - don't think you were using 2sk389 back then ?
Is there a door prize? A bottle of Patron anjeo ( I don't like the silver) ? A six pack of Bud? Would a double blind test reveal the difference between them or am I so burnt out I couldn't tell?
Looks like the second stage of maybe the JC 80 Phono preamp stage with active RIAA bass boost. First stage if it follows your usual topology likely has passive RIAA HF filter at it's output. Was the input jfet 2sk146 or a pair of thermaly coupled 2sk147 - don't think you were using 2sk389 back then ?
Is there a door prize? A bottle of Patron anjeo ( I don't like the silver) ? A six pack of Bud? Would a double blind test reveal the difference between them or am I so burnt out I couldn't tell?
Last edited:
Still wondering why if you can't hear sound differences in a blind test, then the test itself must be possibly faulty by definition.
L.
Of course. But, if for example, the switch box does overshadow the differences of two amps, you would have the differences without switchbox but not in the test itself.
That's one reason why it's important to insert, in this example, two amps in the setup which are known to show differences when tested blind.
If the test person can hear it, you may go forward and insert the amps of interest.
If not, either the setup or the test person (or, in other tests, the measurement gear etc.) is incapable of differentiating, despite in an earlier controlled test it was possible to do so.
Just one example.
You do that in many tests. Inserting a kind of baseline group, with known characteristics.
If the DUT's behave differently as previously known, you have to investigate further, and the test as it is is not valid anymore.
That's one reason why it's important to insert, in this example, two amps in the setup which are known to show differences when tested blind.
If the test person can hear it, you may go forward and insert the amps of interest.
If not, either the setup or the test person (or, in other tests, the measurement gear etc.) is incapable of differentiating, despite in an earlier controlled test it was possible to do so.
Just one example.
You do that in many tests. Inserting a kind of baseline group, with known characteristics.
If the DUT's behave differently as previously known, you have to investigate further, and the test as it is is not valid anymore.
Last edited:
I didn't realize that I could confuse anyone with this schematic. No wonder, that its big brother, the Para JC-2 was virtually ignored. (Shock and Awe, I guess) '-)
It comes from a NOT produced discrete phono stage, designed 18 years ago for Parasound. This happens to be the second stage, and it has part of the RIAA in the feedback loop.
The design is modeled closest from the AD825 IC video amp, but with a complementary mos fet output stage. Most parts would be Toshiba, jfet would be 2sk170, but BF862 would probably work OK.
It is here as an alternate example that has been offered to rival the truly complementary jfet input design of the Vendetta. It is not as 'good' if 'quality' or elegance is important, but it will measure and sound very good.
It comes from a NOT produced discrete phono stage, designed 18 years ago for Parasound. This happens to be the second stage, and it has part of the RIAA in the feedback loop.
The design is modeled closest from the AD825 IC video amp, but with a complementary mos fet output stage. Most parts would be Toshiba, jfet would be 2sk170, but BF862 would probably work OK.
It is here as an alternate example that has been offered to rival the truly complementary jfet input design of the Vendetta. It is not as 'good' if 'quality' or elegance is important, but it will measure and sound very good.
Attachments
I think that selecting
As far as the switchbox influences and DBT procedures are concerned, I'd like you to have a look at a real example: http://www.nousaine.com/pdfs/Wired Wisdom.pdf, and tell what in your opinion is wrong with test methodology and results.
L.
without blind testing them could be quite difficult...
As far as the switchbox influences and DBT procedures are concerned, I'd like you to have a look at a real example: http://www.nousaine.com/pdfs/Wired Wisdom.pdf, and tell what in your opinion is wrong with test methodology and results.
L.
You misunderstood.
- Switch box was an example which could influence the outcome due to the test setup. Could be somthing different too. Eg. Room, Speakers, whatever. Not really important, it was an example.
- The two amps formerly tested as being different, difference which had be detected in a earlier DBT test (by, perhaps, different testers in another test setup, but accepted to be different sonically by a scientific method)
I was merely explaining why it's important to have positive controls too.
If the test person is deaf, no difference is there between an IPhone speaker and PA speakers, when you evaluate it's test results.
You have to find out before carrying out your test, that your measurement setup is sensitive enough for a certain task.
- Switch box was an example which could influence the outcome due to the test setup. Could be somthing different too. Eg. Room, Speakers, whatever. Not really important, it was an example.
- The two amps formerly tested as being different, difference which had be detected in a earlier DBT test (by, perhaps, different testers in another test setup, but accepted to be different sonically by a scientific method)
I was merely explaining why it's important to have positive controls too.
If the test person is deaf, no difference is there between an IPhone speaker and PA speakers, when you evaluate it's test results.
You have to find out before carrying out your test, that your measurement setup is sensitive enough for a certain task.
- Status
- Not open for further replies.