Diffraction be gone

Hi Tvad

When you were on the McMaster site which felt did you use for your DIY diffraction pads? Also what thickness did you use?

They have the following choices:

Firm Gray F3 Felt
Fabricated from coarse wool, this material is a common choice for vibration mountings.

Ultra Soft Gray F26 Felt
Use for packing and padding applications where the felt will be held between other materials.

Very Soft Gray F13 Felt
A gray, lower wool content version of F10 felt, this material is often used for grease and oil retention.

Soft Gray F7 Felt
When fabricating grease seals and wipers, this material's lower wool content makes it an economical alternative to F5 felt.

Thanks and advise when you can.

"And it's old, but it's good". NY

Damn good post, Shad. Using the pads described above, off axis energy above the crossover also measured appreciably down in level (-4db) and therefore side wall and room reflection is made a little more benign. Audibly, I would hope. Guys with small rooms can use all the help they can get. Those measurements referred to above using the pads are on axis. A flatter on axis curve is audible I do believe in tonality and, I’ll go out on a limb here, the perception of space. Then, there is the time thing Dr. Geddes was referring to. That may be where the greatest benefit is derived. It matters not only to imaging specificity in my experience but I believe to the entire soundscape in the listening room. That's the way it seems where I sit. Literally. Cheers.

In practice either 1 alone or a combination of 1 and 3 will work best.

Oops I meant to say "In practice either 2 alone or a combination of 1 and 3 will work best."

I did not mention the trade offs but when you have no baffle at all around the tweeter then you lose LF extension. This may not see important but remember that 6" woofers tend to beam above 1Khz and tweeters distort at rather modest SPL's if driven anywhere close to that low....so often there are several conflicting compromises involved in the design choices. B&W top of the speaker mounted tweeters (which lack any baffle boost) are often crossed over quite high say at around 4 Khz - no doubt they would probably distort at all too modest SPL's if driven much lower - so this often means that the 6 inch woofer is already beaming significantly (limited off axis response) by the time the tweeter kicks in (in a sense a smoother off axis frequency response has been traded for precise imaging of a baffle-less tweeter) The huge success of B&W indicates this small compromise is well worth it to many audiophiles because of the precise imaging that it attains.

a smoothing of peaks and valleys from 2k to 5kHz can be seen from 1 to 3db resulting in a much flatter response curve.

That would be expected. linkwitz lab diffraction measurements show that anything that reduces edge diffraction (smaller baffle, square instead of circular shape, and an off-center tweeter mount) will result in a flatter response.

It is not clear if these slight frequency variations are actually audible. It seems accepted, however, that it somehow confuses the ear/brain as to the precise direction for the source of the sound and hence the "imaging" accuracy. In essence, a point source will image very well but anything reflected within 5 to 10 msec can be detrimental. (Our brain does a careful analysis of the difference in high frequencies arriving in each ear, which is how we are able to tell direction of a sound.)

The science suggests there are three options for great imaging.

1) Tweeter mounted with almost no baffle and away from any possible sharp edges. B&W tweeter on top designs achieve this 2/3 of the way round the tweeter - but undoubtedly will suffer from diffraction off the baffle box below the tweeter but this is very close to the tweeter and may not affect imaging much. Other designs with a small triangular baffle also try to keep the diffractions close to the tweeter. This means that the diffraction signals detected by our ears are very close to the point source of the tweeter and therefore one can still locate the origin of the source fairly precisely.

2) Tweeter mounted in an infinite baffle such as mounting the speaker flush in a wall. In this case, with a smooth transition from the speaker baffle edge to the smooth wall, there is no diffraction at all, PERIOD. Imaging will be as absolutely precise as the recording will allow.

3) Felt or absorptive foam surround around the tweeter. If you can reduce the strength of the diffraction by 10 db then it may become so small that the ear/brain will not detect it from the primary signal and therefore it will no longer confuse the brain as to the precise source of the sound.

In practice either 1 alone or a combination of 1 and 3 will work best. As a rule of thumb, anything that reflects the primary signal with about 7 msec can affect the image (so keeping speakers well away from sharp objects or walls is a good thing id you want a precise image). The brain ignores later arrivals in its quest to assign a location to the source of the sound. In practice there is a big improvement by placing the speakers 3 feet or more away from a wall or an object - after this you still get improvements up to 5 feet away, however, it is diminishing returns after about 3 feet. Speakers that are in a cluttered position and next to sharp corners of furniture or close to a wall will never image properly.

Atkatana- The quote from Dr. Geddes was imported from another discussion elsewhere for it's relevance to the discussion here and to what you're getting less of now, timing and phase error. Regards frequency response, one learned fellow said above that baffle reinforcement could be down 3db and while that may be true in some cases, in real in room measurements made by Danny Richie of GR Research on his production two way standmounts and a fellow in Australia using DEQX on the output of his tweeters (Dynaudio Special 25's), a smoothing of peaks and valleys from 2k to 5kHz can be seen from 1 to 3db resulting in a much flatter response curve. You shouldn't want waveforms loading to your baffles, IMO. More than one thing goes wrong. You want waveforms arriving as they are propogated without early interference.

I am using the active 20's about 3 feet from the back wall and due to the room shape size and power outlets along the long wall which opens on the right hand side to the kitchen, and yes as luck would have it a bay window on the left :) This has always pulled female voice to the right, and has never allowed anything I have put into the room to image like my old Celestion SL-600's (Loved those speakers). Again due to the room being a more or less public place in our house, I keep the grills on. (kids/pets little fingers make me nervous with almost 9k in speaker sitting in the room) Side walls are well out of the way but of different depths and material to either side. I do like the idea of flush mounting the speakers and as we are shopping for a new house it may be something I can do. At present we are in a 124 year old house in the North East and I have limited options on what I can do. (Just updating and Re-wiring the place a couple years back was more then a new car :) It also looks like we are or will shortly be declared a historical building limiting my options even more.

The deffraction begone pads do what they say at least on the mini monitor front. I have not noticed a loss of detail but will be the first to say as I age my hearing is not what it used to be either. What I have noticed is the image is more locked in, stable and more enjoyable in the room that I use. Again room issues are the heart of the question here. Simply put like me I would think many cannot just throw up a bunch of room tunes, corner traps etc. At some point you run out of hiding places and start to effect the room itself as shared space with the family this was not an option for me, or more to the point I have gone as far with it as I can. As something that mounts on the speaker there was little debate I slipped them on and got a better overall image and balance from the 20's. Again at least in my room with my setup.

I do like the idea of reaching out to ATC, love there stuff and the time I got to spend with the Atcive 100's was perhaps the best I have ever spent. (I top out at around 5k for components, and might be able to stretch to 10 on a good year) but as yet have not been able to come up with the dollars for even a used pair of the 100s. So I do wonder what they might say about this, I have to think at least some of there customers have had issues and or modified the baffle on the speakers. The one issue I have always seen around the ATC speakers is imaging is not up to the rest of the speakers performance. For the most part I would tend to agree, but they do everything so well it seems to be a minor issue and overall I could not be happier with the 20's.

As the pads I have come on and off easy enough, for 50 bucks it seemed like a no brainier to try. I as really fairly surprised by the results. Many of the classic British monitors have some surround material, and the celestions had an etched baffle that I assume made up a good deal of there imaging prowess.

Overall it sounds like people have had some luck with these types of mods, maybe I will give them a try on the Talon's. A much larger speaker and not a mini monitor by any stretch. Hope I covered the questions to date and thanks for the response.

Most people think of diffraction as simply having a small effect on the frequency response, but that is incorrect. Because diffraction is time delayed from the original sound it can be highly audible even if it only has a small measureable effect. The ear masks well in frequency but very poorly in time. In fact, there is a lot of evidence to say that diffraction audiblity increase with SPL - think about the implications of that!

Earl Geddes

Spicas, famous for their imaging, used thick felt pads on the baffles oftheir speakers. Same w/ Dunlavy,I think.

Here is a picture of the Dual Hyperdome Tweeter - so you get the idea. The faceplate was recessed so that the foam "gasket" sat flush with the baffle. When you consider that a tweeter generally costs a mere $25 in most speakers and that foam can be dented/scratched and will all to easily shows finger marks or will discolor with age - then you can understand why these extra details might be rare in commercially available designs.

Interesting.

Do you have ASLT or 20-2?

Do you listen with the grill off? (grill should definitely be ON for the towers to avoid those sharp baffle edges)

How close are the speakers to side walls? (4 feet or more away is best)

FWIW: I found a significant improvement in clarity, definition and jaw dropping imaging - all by simply soffit mounting my ATC's - similar to what they do in studios for the mains (it was an expensive project however that nearly cost more than the speakers). With teh saoffit mount my observations were that cymbals sounded more open and natural and vocal/instrument locations tightened in the soundstage. I also found mid bass or male vocals improved their articulation, as well as a more tangible bass. Some GIK tri-traps helped with lower midbass definition/clarity too - although they had little effect on soundstage. (downside is that room modes became more of an issue with the soffit mount - so treating the wall behind the listener becomes important too - and more expense).

My very old Energy Pro 22's (I love these speakers too) had a circular foam surround around the "hyperdome" tweeter (it was about 2 inches wide but fairly thin - 1/4 of an inch of foam) - these speakers imaged like there was no tomorrow - so diffraction be gone is surely on to something! (The hyperdome was great but it was not that reliable unless you fed it really clean music - perhaps this is why Energy no longer make it - sending out replacement tweeters to keep customers happy must have been costly)

One word of caution, however, by adding this pad you will change the dispersion of the tweeter and reduce some of the baffle reinforcement (3 db boost). At the crossover of 3800 Hz, where the tweeter kicks in, wavelengths are around 4 inches - so anything within 2 inches of the center of the tweeter is likely to affect the primary sound too and not just the off axis/diffraction stuff. Above about 6 Khz the pad is probably doing what you want it to (reduce edge diffraction issues coming off the sides of the speakers box). Frankly, I'd check with ATC and see what they say - they consult for pro clients regularly - so I'm sure they have come across this and have an opinion. In essence, a 3db drop in the 4 to 6 Khz range may make the speaker sound less harsh or forward - and you might even like this too, however, you could also do this with an EQ if you wanted.

As an aside, my Von Schweikert speakers have a similar foam surround on the grill. I discovered that when I had the grill off, the sound image was blurry. When I put the grill back on, the sound improved. It was then I discovered that the grill had the foam surrounds. I am a believer in these types of devices, and I assume Albert Von Schweikert is also.