All I can say is meet the flockers Linon New Flokati Black Area Rug | eBay
Btw I think Jim was trolled pretty bad as if he was some snake oil salesman when in fact he was offering a good product at a fair price, too bad.
Stereophile lifted an old test of speakers with felt srips around the tweeter, very much like in David Ralph's tests. Sadly no measurements. Dick Olsher Dahlia speaker | Stereophile.com
Wilson is perhaps the best known user of felt - star shape! Their Alexx was in test bench and tweeter response shows nothing special in my eyes, specially compared to eg. Magico Q series and Vivid Audios which have round baffles. Notice that only CSD measurement is of single on-axis measurement!
Wilson is perhaps the best known user of felt - star shape! Their Alexx was in test bench and tweeter response shows nothing special in my eyes, specially compared to eg. Magico Q series and Vivid Audios which have round baffles. Notice that only CSD measurement is of single on-axis measurement!
Yeah I went and read a lot of those old threads... people were not nice to him at all. But feedback from those who actually used his pads was overwhelmingly positive. I think another issue he had was he was not too adept in dealing with forum moderators.
The wool rug is not a bad idea although I think the 1.5" pile might be a bit overkill. Have you tried something like this?
My speakers now have a synthetic fleece material on them. They seem to image better with it. They image better than I ever thought was possible with a speaker. The larger the baffle the better the effect, I probably wouldn't bother putting material on a minimal baffle.
The wool shag rug is the best choice I've come across online, a bit pricey though. I believe wool will absorb better at lower frequencies than other materials.
The nice part is that you don't have to countersink anything. Just use the baffle (off the cabinet) as a template to cut the material. Then after the fur is on the baffle you can trim the fur so that the hairs are flush with the edges of the drivers.
It works nice if the baffle sinks into the cabinet and sits on the bracing sort of like harbeth style construction. I believe in a lot bracing
Last edited:
I have some f10 (12" x 12" cost $24) that I'm close to cutting.
I was inspired by dunlavy loudspeakers and my results playing with foam either in a horn or at the mouth edges.
Currently I'm using some foam around dome tweet (cat 308 morel).
It sharpens things up.
I couldn't tell you the exact density of my grey foam.
maybe the comi-grey like here
Foam Types | Cutfoam Ltd
I was inspired by dunlavy loudspeakers and my results playing with foam either in a horn or at the mouth edges.
Currently I'm using some foam around dome tweet (cat 308 morel).
It sharpens things up.
I couldn't tell you the exact density of my grey foam.
maybe the comi-grey like here
Foam Types | Cutfoam Ltd
I got my foam from Acefoam via ebay who are also UK based - they will make whatever you want. I got a sheet of 15mm adhesive backed foam that they made for me. it's easier to work with then the felt IMHO and appears to have similar results. At least it did in my limited testing which I posted here earlier. Cutting circles/holes is very tricky if you're doing it freehand though.
Nice to see all that deep soul-searching comparing my felt to your foam albeit based on unsubstantiated impressions of image "sharpness".
Funny almost nobody seems to address a primary intervention: avoid symmetry when you mount your tweeter. Just why is symmetry a cardinal imperative among the engineering-minded in instances when heterogeneity is better.
Pity that few manufacturers want to put themselves to the trouble of making left and right pairs. But for DIYers, seems basic smart strategy for issues such as diffraction, if it is any issue at all.
B.
Funny almost nobody seems to address a primary intervention: avoid symmetry when you mount your tweeter. Just why is symmetry a cardinal imperative among the engineering-minded in instances when heterogeneity is better.
Pity that few manufacturers want to put themselves to the trouble of making left and right pairs. But for DIYers, seems basic smart strategy for issues such as diffraction, if it is any issue at all.
B.
Last edited:
I don't know what this is supposed to mean or who it's addressed to, but earlier I provided measured before and after FRs using felt with a tweeter, and did the same with foam which gave similar results. Both work much the same and the measurements proved that.
Asymmetrical designs don't solve diffraction problems at all, which is why they're not favoured, given the added complexity of having two opposite cabinet designs.
Asymmetrical positioning of a driver can make the on axis response measure a little bit flatter because the diffraction from equidistant left and right sides no longer stacks at the same frequencies. By spreading the peaks and dips more the overall response is a bit flatter, and that is what most people are focusing on when they try asymmetric driver positioning, especially when playing around with sims like the edge.
However it completely fails to address the main problem with diffraction at high frequencies - large frequency response changes over quite small vertical, and in particular, horizontal angular movements away from on axis. All it does is alter the starting response but doesn't prevent these rapid response changes with small angular changes.
These rapid changes just off axis make for a speaker that is very picky about the amount of toe in/out and exactly where the listener sits.
Driver directivity and/or absorption before the signal reaches the edges addresses this problem allowing for a very smooth, gradual change in frequency response as you go off axis.
And once you have sufficient directivity and/or absorption to make the polar response smooth and uniform you no longer gain any advantage in the on axis response by using an asymmetric layout, so you might as well just make it symmetric.
I'm also against designs that offset different drivers by different amounts for example tweeter to one side and woofer still centrally positioned. (perhaps out of necessity due to its size if the baffle isn't much bigger than the woofer)
As soon as the listener goes off the horizontal axis on such a speaker you are changing the time of flight between woofer and tweeter unnecessarily, which will have undesirable effects on the summing of the drivers that differs depending on which side you go to as the tweeter will be more ahead to one side and the woofer to the other.
Such changes in time of flight are unavoidable in the vertical plane in a non-coaxial design but are completely avoidable in the horizontal plane by simply lining up all the drivers vertically.
If you are going to offset the drivers to one side at least offset all drivers by the same amount, but personally I wouldn't bother, I'd focus on directivity and/or absorption to combat diffraction.
I've built and listened to asymmetrical designs and also symmetric designs that take care of diffraction with directivity and I much prefer the latter as not only is the change in frequency response moving off axis a lot less, even the on axis response practically speaking can be a lot flatter when you're not having to deal with high amounts of diffraction.
Last edited:
This is good advice. Symmetry in loudspeakers is not a good thing. I once did an internal study of making parts of a loudspeaker non-symmetric and it showed good results as the natural resonances were suppressed - smeared if you will. An acoustician knows that "degenerate" modes, as symmetric modes are called, are always worse. That's why symmetry in rooms is such an issue.
Asymmetry may not "solve" the problem, but as your second paragraph states, it does make it less offensive.
If you're not going to use absorption or directivity then what makes sense for a baffle is to do two completely different positions for the tweeter.
Choose the tweeter position for the right and left speakers such that the peaks and nulls in the frequency response offset eachother at the listening position. A commercial manufacturer would never do this though because people would think they made a mistake.
All it really does is moves the "least offensive" response from being somewhat off axis to on axis.
But now the off axis response that was previously least offensive is now more offensive. So you're only making the response on one axis better by making other axes worse.
Also most people do not listen directly on axis but listen about 10-15 degrees off axis anyway when they toe their speakers to cross in front or behind the listener.
And they would be right, as deliberately introducing differences in frequency response between left and right speakers is a terrible, terrible idea.
If you're going to use an asymmetrical layout it should at least be mirror images for left and right channels so that near the listener position the frequency response from both speakers is identical at higher frequencies.
Having very different frequency responses due to completely different diffraction effects will pull the phantom centre channel image apart and considerably worsen the imaging.
Part of getting good imaging from a pair of speakers is to ensure close pair matching of the frequency and phase responses of the two speakers, so anything that deliberately makes their responses different is a bad idea.
Last edited:
Well idk, because with absorption small peaks and dips are easier to absorb than the larger ones that symmetrical positioning creates. However with equalization symmetrical maybe better (more consistent off axis)
The problem with a seriously offset tweeter is it creates a strange off axis lobe depending on the crossover.
Yea it might mess with imaging. I should probably delete that. But then again its not that much different than two identical baffles that are offset (not mirrored)And they would be right, as deliberately introducing differences in frequency response between left and right speakers is a terrible, terrible idea.
If you're going to use an asymmetrical layout it should at least be mirror images for left and right channels so that near the listener position the frequency response from both speakers is identical at higher frequencies.
Having very different frequency responses due to completely different diffraction effects will pull the phantom centre channel image apart and considerably worsen the imaging.
Part of getting good imaging from a pair of speakers is to ensure close pair matching of the frequency and phase responses of the two speakers, so anything that deliberately makes their responses different is a bad idea.
Well yeah... and if it looked aesthetically pleasing there would probably be less debate over whether or not it is worth doing.
I apologize if this has been added before, but this site has interesting graphs involving the freq response manipulated by use of felt.
Diffraction Doesn't Have to be a Problem
Diffraction Doesn't Have to be a Problem
You may be misleading with this statement. Edge rounding will also 'smear' diffraction in time and space, and it makes best use of the baffle area. It is always about having enough to reduce the effects below perception - either way. Centring the source may focus resonances on axis, but if they are below perception? When the speaker has been deliberately designed for a centred source and isn't trying to make some sort of compromise, then...
By worse, I might accept specifically 'stronger', but there can be other reasons for accepting this. I know that speaker/room placement works when modes are spread via placement, but I think this is good advice for speakers that were not made to work specifically with a room at lower frequencies rather than a blanket statement.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.