Thank you for posting about your experience. Why do you suppose the diffraction panels propagated a "uniform" waveform back into the room? Wouldn't diffraction material break up that waveform and make it the opposite of uniform?
Does a Stepped Front Wall Adulterate the Rear Wave From a Dipole Speaker?
- Thread starter Ron Resnick
- Start date
Thank you for posting about your experience. Why do you suppose the diffraction panels propagated a "uniform" waveform back into the room? Wouldn't diffraction material break up that waveform and make it the opposite of uniform?
Thank you very much Dude, Bruce, Tao, kcin, Mark! I really appreciate the analyses!
I am inclined to forfeit the 4" of room length (which I was trying to keep) and "fir out" the entire rest of the front wall so it is one flat plane.
I am inclined to forfeit the 4" of room length (which I was trying to keep) and "fir out" the entire rest of the front wall so it is one flat plane.
You could also put a matching 4" of depth on the opposite side so your wall is totally symmetrical. Space is at a real premium in places like LA and London, so totally understand where you are coming from.
Last edited:
It’s far from a perfect analogy, but look at the images on pages 30 and 31 of the Legacy Audio Wavelet product manual. The “rear wave” is far less “symmetrical and uniform” than most picture in their head. The diffusers “break up” the “irregular waves” thus creating a far more “random reflection”. I’d instead say “more random” rather than “more uniform” but they’re pretty similar conceptually.
Last edited:
Thank you, LL21, but my concern with that idea is that, while it would create visual symmetry, I might also be adding to the right channel the same rear wave problem (if it is even a problem) I have on the left channel.
You could also create an entire wall diffusor when filling in the 4". I have seen that done to great effect.
Based on my experience with MartinLogan speakers I believe in starting with a front wall which is plain drywall. I do not believe in absorbing or diffusing the rear wave.
Gryphon Audio recommends either an untreated front wall or, if anything, only diffusion on the front wall.
Gary's big Genesis speakers use a ribbon driver (for the midrange) which is very similar to the ribbon driver in the Pendragons. Gary, in his Genesis Technologies Prime Operating Manual, agrees with not absorbing or diffusing the rear wave:
Gryphon Audio recommends either an untreated front wall or, if anything, only diffusion on the front wall.
Gary's big Genesis speakers use a ribbon driver (for the midrange) which is very similar to the ribbon driver in the Pendragons. Gary, in his Genesis Technologies Prime Operating Manual, agrees with not absorbing or diffusing the rear wave:
4.6 Room Treatment
No room is perfect. To optimize your sonic presentation it may be helpful to treat your room. Here are some guidelines:
1. Frontwalls. The Genesis Prime loudspeaker is a dipole and therefore there is sound coming from both the front and the back. How the front wall is treated or not treated is important. Do not use either diffusion or absorption on the front wall.
No room is perfect. To optimize your sonic presentation it may be helpful to treat your room. Here are some guidelines:
1. Frontwalls. The Genesis Prime loudspeaker is a dipole and therefore there is sound coming from both the front and the back. How the front wall is treated or not treated is important. Do not use either diffusion or absorption on the front wall.
Wilson Audio also writes the same in their manuals. However, please note that many people advise in general against diffusion in the front wall for one of two reasons - most diffusers absorb a lot, sometimes more than 60% depending on frequency, and if used close to speakers diffusors can create comb filtering.
Not all diffusors are created equal - there is a large difference in the performance of different types of acoustical treatments carrying similar names. Very few have complete specifications and trusted data sheets - RPG and KineticsNoiseControl are known for providing excellent data.
Ron,
I am going to put the JL Audio F113/2 subs in place next week, so today I re-called my REW set-up. To check operation I used my usual calibrator - a Quad ESL63 dipole panel. I attach a spectra with the speaker at 40" from a frontwall, microphone at 80" from speaker (magenta) and the speaker at 44" from a very reflective frontwall (drywall and wood), microphone also at 80" from speaker (5dB per division). My Quad ESL63 pairs are really matched at 2 dB, I would not like to compromise their factory matching having one 4" farther from the reflective front wall. Please note that this is not an optimal position for ESL63 - it just an easy position for testing.
In some aspects I am an instrument dependent audiophile - when adjusting tilt with the curved Soundlab's or Quad's I make the final adjustment for equal response of frequency in the high treble for left and right speaker, it is much more accurate than visual or laser measurement.
Please note that the ESL 63 is a point like dipole, your line array dipole has a different pattern of dispersion.
Attachments
Francisco, thank you very much for performing this experiment and posting it!
How do you interpret the results?
Are we seeing differences between the blue line and the red line from the 2dB difference between the speakers, the 4" front wall distance difference, or possibly some of both?
How do you interpret the results?
Are we seeing differences between the blue line and the red line from the 2dB difference between the speakers, the 4" front wall distance difference, or possibly some of both?
The two spectra were taken with the same speaker - so the only difference is the slight difference in position. Unfortunately I did not have any RPG Abfusor panels at hand, otherwise I would have checked the effect of absorbers. It would be a nice help to interpret what is happening.
I understand. Thank you for performing this experiment!
I think this shows that removing the 4" difference and equalising the channels is worth sacrificing the 4" length of floor space. Do you agree, Francisco?
This also solves the problem of Jack being slightly disturbed by the visual asymmetry when he visits me someday!
I think this shows that removing the 4" difference and equalising the channels is worth sacrificing the 4" length of floor space. Do you agree, Francisco?
This also solves the problem of Jack being slightly disturbed by the visual asymmetry when he visits me someday!
Francisco, no apology needed for that comment. Instruments are very useful tools in helping set up a system even with your ears as the final arbiter. However, I'm curious why you didn't take microphone measurements from the listening position. Isn't that the most relevant position? Finally, if I may, you might consider looking carefully at the horizontal front/back level of the XLFs as assessed by a digital level. Differences of less than 1 degree can significantly effect floor bounce and if your XLFs are similar to the measurements seen with the ESL63, possibly help tame some of the large deviations you have at 180Hz, (this is something relatively common in most set-ups with XLFs or Alexandrias).
As a point of reference back in 2011, Nyal Mellor (Acoustic Frontiers) and Jeff Hedback (Hd Acoustics) suggested an acoustical standard for a high performance audio system of +/- 3dB difference in the midrange between L & R speakers measured at 1/3 octave.
Having taken many REW measurements, I know I'd have a hard time hearing the differences between Francisco's 2 plots on casual listening. The difference could be more audible on back to back comparison, but might still be swamped by left to right speaker positioning. This plot doesn't look like 1/3 octave, but less smoothed.
Having said that, I still wouldn't want a 4" difference to the front wall.
Having taken many REW measurements, I know I'd have a hard time hearing the differences between Francisco's 2 plots on casual listening. The difference could be more audible on back to back comparison, but might still be swamped by left to right speaker positioning. This plot doesn't look like 1/3 octave, but less smoothed.
Having said that, I still wouldn't want a 4" difference to the front wall.
Last edited:
Marty,
This was just a quick experiment with ELS63 that does not represent my listening conditions at all - the only requirement were the distances and a large free reflective wall - something it is not easy to find in my listening room! My front wall uses RPG Abfussors, bass traps and diffractals and I did not want to move the XLFs from their places. The XLF's measure very differently from the ESL63. Please note that considering that no optimization was carried at all - the speaker was just dropped and measured - the general trend of the plots is excellent, with symmetrical peaks and dips.
Also, in order to show the differences in detail, I did not carry the usual flattering 1/3 octave smoothing - just 1/24 octave.
I must say that I never measured the XLF's in detail - my main interest in measurements has been the bass frequencies under 100 Hz.
Similar threads
