There is often a knob that apparently some folks don't know how to use that will reduce the volume!
As a pro I am sure that you know how to turn it up, but what is really cool is that if you move it in the other direction, the volume will actually go down. Kind of like getting rid of feedback!
(Will do lower levels next time, thanks...)
Since my laptop gave no output at 150Hz sort of makes sense.
It is a poor craftsman who blames his tools!
I found rapid switching (using foobar) was best for spotting differences. Jan tells me that my guesses and yours were the same, so we can be equally embarrassed when the truth is unveiled. I still don't know what Hawksford did...
What I know is that some of the files have been passed through a 4th order Butterworth (or was it LR?) xover at 3kHz. The two bands then have been combined again.
That produces a result that has the same (ruler flat) amplitude response as the original but the phase response has been messed up, especcially around the xover frequency.
The idea was to explore audibility of in-band phase response anomalies.
jan didden
That produces a result that has the same (ruler flat) amplitude response as the original but the phase response has been messed up, especcially around the xover frequency.
The idea was to explore audibility of in-band phase response anomalies.
jan didden
Care must be taken in differentiating between flat small signal frequency response and time domain response. The crest factor of music can be changed by all pass networks.
I have no problems sorting the 6 files into two groups, but in my ears all of them sounds "compressed", maybe it's an idea to post uncompressed files in a listening test?
S
Stinius can you pm me with your results?
jan didden
Jan, tnx for interesting test.
I did the test yesterday and, unlike Pano and SY, I mentioned a null result to Jan. Since I was sure three of the files had been processed because Jan told us, and because I know I can't register absolute phase, I floated the suggestion that something might have been done with the phase. Which it was.
This being said, others might be able to pick it up. So, I hope more of of the participants in this thread will do the same test, and report back to Jan. With a null result, or with a division in two groups, whatever your ears tell you.
The results would be not only be interesting from an academic point of view. For someone developing loudspeakers it is directly relevant. Flat FR and phase shift are difficult to optimize at the same time. So I like to get as much leeway in the latter dimension to get the first one right.
vac
SY, I could rapidly switch between the files, each starting at the beginning each time. I have developed some aural memory though. The thing is I had to blind test myself. Only after I found I gave inconsistent answers, I concluded that I could not hear a difference. At the same time, you get under the impression that you may hear a little something here or there. As soon as you know which number you are listening to, I found that because of expectations, patterns emerged. However, as I said, blind testing destroyed those.
So, I reported a null result, but at times I also heard differences. Only testing discipline showed me that these were random.
vac
vac
So, I reported a null result, but at times I also heard differences. Only testing discipline showed me that these were random.
vac
vac
Hi,
Tis then, the crux of the matter.
Did this test destroy the false illusion differences where present or did it obscure real differences that where present?
If I find the time, I may prepare some files of my own to test the gullible as to what they can hear and what not, if sufficiently blinded...
Ciao T
Tis then, the crux of the matter.
Did this test destroy the false illusion differences where present or did it obscure real differences that where present?
If I find the time, I may prepare some files of my own to test the gullible as to what they can hear and what not, if sufficiently blinded...
Ciao T
Thorsten, why do you refurse to participate in the test Jan linked, and now come with something like this? I'll boycot.
Sorry Pano, but I read your post only today after coming back from a week of holiday in a Greek island (no PC and internet connection at hand
).I started this discussion on pulse test of electronics (not only amps) some days ago, but my point was different: I stated that matching impulse (transient) response could match the sound of amps.
I used a pulse that I had at hand; the focus should not be on the pulse used for the test, but on the results. I never said that "the pulse could reveal things that a swept sinewave or MLS noise signals would not".
What I suggest you to try, since you said that it was interesting to look into, it's the following experiment:
1. take two amps that in your system sound differents;
2. compare their pulse response (using whatever signal you prefer);
3. try to modify the pulse response of one of the two amps to exactly replicate the pulse response of the other one;
4. make a new listening session of the two different amps now with the same pulse response and check if they sound the same.
If no, my assumptions are completely wrong; if yes, we can continue to discuss about these and other interesting results.
One suggestion: in case you will use a dynamic loudspeaker for the listening, try to match the pulse response of the amp output current driving the loudspeaker, not the voltage output (in my experiments I modified the voltage pulse response since I used an electrostatic speaker).
I am amazed and confused. ABX tests on a computer? What is the point? Are you attempting to prove that you can not trust your own ears, as they would behave in an open test? Are you trying to show that even slightly modified test tones are essentially the same for all practical purposes? What does this have in common with hi fi reproduction? Would you not agree that it is mid-fi that is now being discussed here?
Thorsten, on your substantive points -quo-
A. Did this test destroy the false illusion differences where present or did it obscure real
differences that where present?
B. If I find the time, I may prepare some files of my own to test the gullible as to what they can hear and what not, if sufficiently blinded... -unquo-
Ad A. There was a real difference present, which was inaudible to me. Yet, when challenged to look for differences, the human mind will try to construct even completely inaudible differences into apparent ones. With non-blind testing, these perceived differences may develop into a pattern because of the way expectations screw around with your brain. Even trained ears falll into that trap. Only by blind listening it is possible to falsify or confirm these perceived differences.
ad B. do you see the other participants to this thread as 'the gullible'? If so, you are right, because everybody's perception is primed by expectations. It is a human quality, which you have to deal with in order to arrive at significant results in any kind of testing environment. In connection with the term 'if sufficiently blinded', however, the word 'gullible' assumes an even more pejorative connotation. The question is, why you are doing this? You refuse to participate in a well done test, one you can perform any way you deem best, and subsequently you deride those taking part in it. For me it is about gaining knowledge and understanding. Refuseniks are not very helpful in that respect.
It would seem more correct to describe 'the gullible' as those who believe they perceive differences between e.g. two capacitors of similar construction but of different make, without subjecting themselves to the rigour of measurements and controlled testing. Or perhaps 'purposely gullible' would be an even more appropriate term.
vac
A. Did this test destroy the false illusion differences where present or did it obscure real
differences that where present?
B. If I find the time, I may prepare some files of my own to test the gullible as to what they can hear and what not, if sufficiently blinded... -unquo-
Ad A. There was a real difference present, which was inaudible to me. Yet, when challenged to look for differences, the human mind will try to construct even completely inaudible differences into apparent ones. With non-blind testing, these perceived differences may develop into a pattern because of the way expectations screw around with your brain. Even trained ears falll into that trap. Only by blind listening it is possible to falsify or confirm these perceived differences.
ad B. do you see the other participants to this thread as 'the gullible'? If so, you are right, because everybody's perception is primed by expectations. It is a human quality, which you have to deal with in order to arrive at significant results in any kind of testing environment. In connection with the term 'if sufficiently blinded', however, the word 'gullible' assumes an even more pejorative connotation. The question is, why you are doing this? You refuse to participate in a well done test, one you can perform any way you deem best, and subsequently you deride those taking part in it. For me it is about gaining knowledge and understanding. Refuseniks are not very helpful in that respect.
It would seem more correct to describe 'the gullible' as those who believe they perceive differences between e.g. two capacitors of similar construction but of different make, without subjecting themselves to the rigour of measurements and controlled testing. Or perhaps 'purposely gullible' would be an even more appropriate term.
vac
- Status
- Not open for further replies.