Hello Dr. Geddes,
I know you recommend the measurements to be taken at 7'5 degree's, or at least 10. Do you find the plots at the end of this kit manual to be useful? I find them quite similar and illustrative compared to your plots.
Here: 18 Sound 8" Kit
It would be interesting to compare this to the Harpers.
I know you recommend the measurements to be taken at 7'5 degree's, or at least 10. Do you find the plots at the end of this kit manual to be useful? I find them quite similar and illustrative compared to your plots.
Here: 18 Sound 8" Kit
It would be interesting to compare this to the Harpers.
I think that the difference here is that those are not selected reviews, but all the reveiws. Thats a big difference. One can always find a review that is positive, but I think that it would be rare to find that everyone had the same positive impression.
Personally, I would have been happy with just posting measured data. But people wanted reviews. So I started posting them. Now people say, well those aren't "independent" reviews. You mean like the ones in the audio press? Yea, those are real objective! (Not!) According to those reviews there are no bad speakers, only great ones that differ in the manner in which they are great. Lets face it, the only truely meaningful assesment is that done by the measurements. I'm happy to stand on those alone.
Last edited:
Those are a vast improvement over the typically useless on-axis plots, but are no where near as high resolution as mine. If I were to 1/3 octave smooth mine as well as decrese the dB resolution they would be on the same scales. That would of course smooth my significantly. But, I'd much rather see the type of plot shown there than what is usually shown, although I think that much higher detail would be better. It would begin to show the kinds of things that make an audible difference.
zilch: Do you have an actual technical argument or are you just trolling?May I repeat my question: "how do polar plots like the ones proposed by Earl correlate to perception of (phantom) sound sources, i.e. their timbral and spatial qualities?"
In addition: how much frequency smoothing can be applied without hiding important information?
Best, Markus
The constant-directivity 2-way is a Geddes innovation?
[EARL -- send this man $100 for posting that....]
Look at the timing of the 'Econowave' project...is it merely coincidence that they become popular as soon as Dr. Geddes puts out a great design?
I'm well aware of the project and Augerpros on AVS forum...the use of a 'plug,' the drivers, the look of the design, the timing, the design goals, etc....how is it not a clone?
The use of Class A, Class A/B, Class D, etc. amps are not new innovations...but last I checked, a First Watt F5 clone was a First Watt F5 clone. Not a 'Tim Rawson Dream Machine' (using his name as an example, nice guy btw)
The Gedlee waveguides are available via kitforms at attractive prices, maybe not the same as the mass-produced QSC waveguide, but for the performance...the 'value' of his kits are tremendous. You're accusing him of not crediting others' work and then deny calling the 'Econowave' projects clones.
This will be my last off-topic post and I apologize. You're not doing others a service by insulting someone providing interesting discussion.
"Phantom sources"? I am not sure how much direct correlation that there is with "imaging" phenomina and a polar plot. Imaging comes mostly from a stereo or greater setup, in a real room. The measurements are of a single source only. What needs to be understood to make the connection is how the source and room interface and what is desirable in that interface in order to improve or accentuate the "image" aspects of sound reproduction. The directivity and coloration will both impact the perception of image.
This would also be true of spatial qualities, although the room very heavily impacts this.
Coloration would be almost all the speaker and will show up in the polar map as vertical aberations.
I would think that circularly sweeping aberations as from diffraction would be more detrimental to imaging than show up as coloration since there frequency "average" is much broader.
I have played with this and I think that 1/3 is definately too wide. Smoothes away the interesting stuff. 1/10 works great, but is sometimes a bit noisy. 1/6 is OK if 1/10 is not.
Could you please give examples from your paper, from your measurements or any given suitable data (18Sound Maybe) of vertical aberrations and circularly sweeping aberations?
I'd like to learn how to spot them. It seems to me this is one of the few methods of identifying speaker coloration and most important, diffraction effects. Especially the last bugger seems to me particularly hard to nail.
Thank you!
First directivity is a form of diffraction, so the kinds of things that you are looking for in this regard are the sweeping curved lines that have nulls along them - basically the beam itself as in Fig 3. Now look at Fig 6. The additional "waves" in the response, which are not normal to the frequency axis are pure diffraction. The side that they sweep from is the side which has the diffracation. If it is sweeping equaly on both sides then the center of the diffraction is in the same plane as the source.
Fig 5 shows a pure resonance - a coloration.
Now in Fig. 7 you can see both.
Real speakers, of course, have everything going on.
In Fig 12 there are the obvious resonances as shown as straight vertical lumps. If you look carefully there are sweeping lines leading to the tweeter at about 5 kHz. The peak there is acoustic in nature, and it does not go away with EQ (on the bottom) as the mechanical resonances do. The tweeter diffraction can still be seen as the "wings in the borttom of Fig 12 at about 2.5 kHz. No electrical EQ can "fix" this problem. Fig 13 has the same thing, but its not so noticable because the whole speaker is much worse. Fig 13 - 15 show some as well. Fig 17 shows some diffraction ala HOM at above 3 kHz. This too is not correctable by any crossover or electronics. But in this example it is well controlled and minimal.
Hi Earl,
I am interested to know if you have done polar distortion measurements. If you have does the distortion track with level or do you see genuinly different levels of distortion at different angles?
Thanks for any information you might have, I am just curious about this.
Regards,
Andrew
I am interested to know if you have done polar distortion measurements. If you have does the distortion track with level or do you see genuinly different levels of distortion at different angles?
Thanks for any information you might have, I am just curious about this.
Regards,
Andrew
Thank you for your clarifications Dr. Geddes, much appreciated!
I can no see that higher resolution is preferable. The 18 Sound plot looks very clean by comparison, espeaceally in the horizontal plot. For example, there is nothing I can identify with HOM and something tells me this is rather due to the lack of resolution rather than the lack of HOM.
I can no see that higher resolution is preferable. The 18 Sound plot looks very clean by comparison, espeaceally in the horizontal plot. For example, there is nothing I can identify with HOM and something tells me this is rather due to the lack of resolution rather than the lack of HOM.
I don't do distortion measurements on loudspeakers. They have never been shown to correlate with subjective perception, so its really a waste of time. People do them because they can, not because they should.
Isn't the same true for polar maps? They are easy to measure but they don't correlate with subjective perception.
No, not at all! I find the polar maps to correlate very well with perception. This has not been quantified like the distortion stuff however.
Floyd's work is close, but he simplifies the maps down to a few lines. I prefer the higher data density.
But there have been several studies on nonlinear distortion and they have all shown the same thing - no correlation.
Why not? Is there a cheaper, better performing HOMless speaker out there?
We're talking about how polar maps correlate to perception. It is out of question that the indirect sound field does affect perception. But then, distortion affects our perception too
Best, Markus
In the "Predicting Listener Preference from Measurements," section of Toole's book and in many other places looking at pro speakers there is a "Directivity Index" given in decibels usually scaled from 0-10 dB or something like that. Is that scaling how far down the response would be at 45 degrees? I don't actually see anything else about polar response in the data. There are indication of the polar response, but nothing direct. There is measurements at the "Listening Window", "Early Reflections", "Directivity Index" and "Sound Power" given, but no polar response graphs.
Thanks again,
Dan
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.