I think what you're both saying is the effects on FR go hand in hand with the diffracted phase cancellations/waveform aberrations. One causes the other so really which one is more audible is somewhat irrelevant. I think it would be more significant to see the effects of diffraction on different axis and frequencies at the waveform level.
At your website are links to your peer-reviewed publications and to some other write-ups but nothing that had "diffraction" in the description. ???
B.
To answer that question I'd need to know what methodology was used when establishing the audibility of "delay" in your testing.
What is meant by "delay" in this context, and how does it relate to diffraction when diffraction doesn't have one specific delay but is a smeared impulse ?
Is there a public link for your paper I could read ? If there is, I think I may have read it before but if I have it's a long time ago and I don't remember the exact details. But if it's available I'll re-read it to form a better response to your question.
I'd need to read your paper to be sure we're talking about the same thing, but if the testing was based on a synthetic signal where a reduced amplitude delayed version of the signal was electronically added, (a delayed impulse) then I can think of an obvious reason why an increased delay would sound worse than a shorter delay.
While The "mean effects on frequency response" might be similar if considering something like a 1/3rd octave averaged response, the specific frequency response will obviously differ, (and is therefore not "the same") and the longer the delay is the higher the density of the peaks and dips that will result from comb filtering effects, and therefore the steeper the slopes of these peaks and dips in the amplitude response for a given peak amplitude variation.
This higher density of peaks and dips is analogous to a larger number of more densely packed higher Q resonances - which also sounds worse than fewer lower Q resonances.
When people try to analyse perceived sound quality based on frequency response they give attention to how far the response deviates from an ideal flat response, (+/- X dB) but don't give much attention to the rate of change of that deviation, eg the steepness of the slope as graphed. Is it a very rapid zig zag or is a gradual fluctuation ?
This is why frequency responses that follow the same general 1/3rd octave averaged response but differ in the slopes of the individual variations when not smoothed don't sound the same, and it's one reason why I never use 1/3rd octave averaging in my measurements and always look at the raw, full high res responses, as that critical slope information that shows whether a response is jaggy or smooth is lost when applying averaging.
Eliminating high rates of change of the amplitude response with respect to frequency (jaggyness) so that the response is "smooth" and that any frequency response variations are smooth and gradual is key.
So if your testing did in fact add a time delayed amplitude reduced copy of the original signal and then vary the time delay independently of any other variables, my response would be that it's the increased density of comb filtering that makes the sound more objectionable.
If your testing did something else, then I'll need to read the paper first to comment.
Last edited:
What do you hear when you listen to a mono signal played through stereo speakers?
You have to ask scottjoplin that question. He says the sound of a mono speaker is different (and superior) to the sound of a virtual image.
(I am somewhere near completing a long write-up on perception of stereo music. Hint: Helmholtz' concept of "unbewusste Schluss" and conflicting perceptual cues rather than Blumlein's math.)
B.
Last edited:
Well I attempted to address this issue and bought some of the 3/8" F3 felt sheets from Grainger. They were every bit as hard to work with as I anticipated but I figured out you can cut them precisely with a jigsaw if you stick them down to something like a thin utility board, MDF or pegboard first.
Try as I might I could not hear any significant difference with the pads on or off the mids & tweeters. I'm not claiming that there wasn't an improvement, just not enough for me to justify moving forward with the modifications. It should be noted that my drivers are mounted flush, asymmetrically with baffle edges rounded over so diffraction effects are already somewhat minimized. And my ears are decent because I can easily hear the different sonic characteristics of crossover caps.
Try as I might I could not hear any significant difference with the pads on or off the mids & tweeters. I'm not claiming that there wasn't an improvement, just not enough for me to justify moving forward with the modifications. It should be noted that my drivers are mounted flush, asymmetrically with baffle edges rounded over so diffraction effects are already somewhat minimized. And my ears are decent because I can easily hear the different sonic characteristics of crossover caps.
Yes, like with beaming, you can't hear diffraction.
If beaming or diffraction were truly substantial, then you might say, "Umm, yesterday before adding the felt it sounded different, maybe, I think, sort of, in a way I can't quite put my finger on... but it sounds perfectly fine to me today anyway".
That's the analysis from the point of view of listener's perception rather than polar plot.
B.
Well I don't know if I agree with all that. You can definitely hear beaming, particularly with guitar amps, which is why you always mic them off axis.
Yes, 1 1/8" fabric dome
Yes, likely makes really a big difference where you plunk your mic (which has its own polar pattern and room pick-up) in front of a good speaker, and more so if it is acoustic garbage like a "guitar amp"*.
But that's quite different than putting a binaural human head at a distance listening to a few drivers playing in a reverberant room... and trying to remember what it sounded like before the felt was added.
You know, a lot of this thread comes from people who think they can listen to a music passage and then form a frequency response plot in their head, like REW does.
B.
*the "guitar amp" might be making delightful music (to some ears) but still acoustic garbage in the terms of builders on this forum
Last edited:
I've attached a typical dome tweeter.
There are two types of diffraction with this type of tweeter. One is from the front panel itself. The next one is at the edge of the baffle.
I'm afraid flush mounting and rounded baffle edges won't help much. Neither would attaching absorbent material around the tweeter.
There are two types of diffraction with this type of tweeter. One is from the front panel itself. The next one is at the edge of the baffle.
I'm afraid flush mounting and rounded baffle edges won't help much. Neither would attaching absorbent material around the tweeter.
Attachments
In your own words from a few posts back, what a load of wordy Nonsense.
I guess we all should just go back to not bothering to flush mount drivers, round over baffle edges, use drivers with directivity etc because according to you we can't hear the effects of diffraction...
Beg to differ.
The "guitar amp" (presumably meaning the amp and speaker(s)) is a musical instrument (some might say) and designed to meet the needs of the musician. It has to be delightful and functional as an instrument. I hope I am clear on that point.
But we are mostly enthusiasts for clean reproduction of recordings. I think it is fair to say most "guitar amps" would have trouble playing back any recording nice (likely including recordings of electric guitars)*. And the sound would cause abdominal distress to some us in reproducing a well-recorded acoustic guitar.
Moreover, even if most HiFi speakers have imperceptible diffraction irregularities in listening to music at home, all bets are off when sticking a mic in front of a "guitar amp".
I think few posters would find those thoughts off-base.
B.
*and a whole other discussion about the role of contact mics compared to mic'ing speakers compared to playing amplified instruments compared to mic'ing acoustic instruments compared to hearing acoustic instruments
Last edited:
Hi Simon
I have been out of town.
The above sounds like a strong rational for the point that I was trying to make - that the delay as well as the amplitude of the diffraction is a factor, which is what we showed.
Once again, can you please tell us which among the downloads at your website is the relevant research (hopefully from a peer-reviewed journal). Sure isn't apparent to me.
B.
http://www.gedlee.com/Papers/AES06Gedlee_ll.pdf
This paper was written with HOM in waveguides in mind. However, the techniques resemble diffraction in many ways and my position is based on these results.
I will not argue about the paper as anyone can criticize research (just look at global warming). I will entertain discussions of other papers that draw different conclusions based on valid psychoacoustic tests. Hence, while this paper may seem to be of little connection to this discussion, it is a data point that will stand unless other data points of equal validity refute them.
This paper was written with HOM in waveguides in mind. However, the techniques resemble diffraction in many ways and my position is based on these results.
I will not argue about the paper as anyone can criticize research (just look at global warming). I will entertain discussions of other papers that draw different conclusions based on valid psychoacoustic tests. Hence, while this paper may seem to be of little connection to this discussion, it is a data point that will stand unless other data points of equal validity refute them.
Had a read. Thanks.
But could you please tell us in what way that paper and its conclusions are meaningful here. Finding the connection is beyond my grasp.
B.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.