That’s interesting to see as we’ve used a bunch of Akotherm in the past.
It’s very reasonably priced if you’re buying a box of it. I think that’s partly because it falls under industrial products, rather than being focused at audiophile, studio or hifi markets.
Another interesting material that popped into my radar recently is this stuff:
https://technetics.com/products/feltmetal-acoustic-media/
I doubt it’s anywhere near so cheap, but the ability to ask them to create it to specific acoustic parameters might come in handy for certain situations or projects. The consistency of response under high sound pressure levels is also pretty cool.
Thanks for the hints @fluid and @kyleneuron. I was not unable to find the flow resistance and absorption of polyester wool but the following is helping:
https://www.stretchgroup.net/webshop/acoustic/polyester-wool-acoustic-insulation.htmlhttps://www.researchgate.net/figure...-m-3-and-thickness-40-mm-Sound_fig3_228933963(check the other pics for more interesting graphs)
Maybe a trick is to use a combination of 2 densities to find the right absorption/delay combo.
Try the melamine "magic" sponges, should be available at grocery store cleaning department. Easy quick and cheap test while waiting the exotic mail order stuff to arrive 
I'm still using it, and it is working fine see post #327 and details of the implementation if you scan the thread backwards from there. It is a bit brittle as material so perhaps not the most optimal as such. It would be nice to see if it worked for other boxes than mine.
I'm still using it, and it is working fine see post #327 and details of the implementation if you scan the thread backwards from there. It is a bit brittle as material so perhaps not the most optimal as such. It would be nice to see if it worked for other boxes than mine.
Last edited:
Here is some information on Tontine Acoustisorb polyester batting, you could use it to compare relative densities to gas flow resistivity numbers.
TONTINE
Acoustisorb 1 (22kg/m3 100mm)......4350 mks Rayls/m
Acoustisorb 2 (32kg/m3 100mm)......8070 mks Rayls/m
Acoustisorb 3 (48kg/m3 100mm).....14230 mks Rayls/m
Magic melamine sponges seem to be made of Basotect https://agro.basf.ca/East/Innovation/ProductsAndTechnology/item1894.html
Whose absorption coefficients attached (captured from product PDF) as well as here https://www.cmfacoustics.com.au/Basotect1.pdf
Flow resistivity I cannot find other than from Gearspace "According to BASF: The flow resistence of Basotect G as listed in DIN EN 29052 averages 12,4 kPa*s/m²"
What is great about the melamine sponges is that they are very uniform in size and easy to cut with sharp knife / wire. But, the material is also brittle, breaks by hand quite easily. I'm pretty sure the melamine sponges I have, sold under local brand name, are made in china and not by BASF so perhaps not exactly Basotect G
Whose absorption coefficients attached (captured from product PDF) as well as here https://www.cmfacoustics.com.au/Basotect1.pdf
Flow resistivity I cannot find other than from Gearspace "According to BASF: The flow resistence of Basotect G as listed in DIN EN 29052 averages 12,4 kPa*s/m²"
What is great about the melamine sponges is that they are very uniform in size and easy to cut with sharp knife / wire. But, the material is also brittle, breaks by hand quite easily. I'm pretty sure the melamine sponges I have, sold under local brand name, are made in china and not by BASF so perhaps not exactly Basotect G
Attachments
the problem is you need a specific attenuation and delay profile vs frequency. you can find that profile by modelling the slot back wave ports by arrays of small drivers, a low pass filter and dsp delay but I don't know a way to relate that profile to the kind of data in the previous post. There is no escaping good old fashioned cut and try. Best just to pick a material with stable acoustic characteristics in the right ball park that is easy to cut.
Even if you go active and actually include the arrays of small drivers in your speaker you are in for multiple rounds of measurements as you tweak the filter to optimize the response. I'm hoping that with vituix that one could measure polars of the side array and combine them with measurements of the main driver, optimize cardioid response in simulation and go directly to a good measured cardioid response.
Even if you go active and actually include the arrays of small drivers in your speaker you are in for multiple rounds of measurements as you tweak the filter to optimize the response. I'm hoping that with vituix that one could measure polars of the side array and combine them with measurements of the main driver, optimize cardioid response in simulation and go directly to a good measured cardioid response.
One big problem in designing a cardioid speaker, either passive or active, is the measurement difficulties in lower frequencies. To keep the room reflections at bay and avoid the room modes contamination of the sound, one will need to do gated measurement down to ~100hz. It would be a 3.4meters tall tower, acoustically transparent on a turntable. Soon we will receive a new turntable from a Dutch company with a 210 cm tower, and to my estimations it will provide gated resolution down to ~140-150hz.
We may soon add a Klippel NFS to our lab, and if that happens I'm gonna optimize the design of the passive cardioid to see how far we can go. Hopefully there will be a well documented cardioid design available for the DIY community in years to come.
We may soon add a Klippel NFS to our lab, and if that happens I'm gonna optimize the design of the passive cardioid to see how far we can go. Hopefully there will be a well documented cardioid design available for the DIY community in years to come.
Off topic but the Turntable, what brand?One big problem in designing a cardioid speaker, either passive or active, is the measurement difficulties in lower frequencies. To keep the room reflections at bay and avoid the room modes contamination of the sound, one will need to do gated measurement down to ~100hz. It would be a 3.4meters tall tower, acoustically transparent on a turntable. Soon we will receive a new turntable from a Dutch company with a 210 cm tower, and to my estimations it will provide gated resolution down to ~140-150hz.
We may soon add a Klippel NFS to our lab, and if that happens I'm gonna optimize the design of the passive cardioid to see how far we can go. Hopefully there will be a well documented cardioid design available for the DIY community in years to come.
Martijn's DIY cardiod project was extremely well documented years back.One big problem in designing a cardioid speaker, either passive or active, is the measurement difficulties in lower frequencies. To keep the room reflections at bay and avoid the room modes contamination of the sound, one will need to do gated measurement down to ~100hz. It would be a 3.4meters tall tower, acoustically transparent on a turntable. Soon we will receive a new turntable from a Dutch company with a 210 cm tower, and to my estimations it will provide gated resolution down to ~140-150hz.
We may soon add a Klippel NFS to our lab, and if that happens I'm gonna optimize the design of the passive cardioid to see how far we can go. Hopefully there will be a well documented cardioid design available for the DIY community in years to come.
There is also another one very well documented (forgot the name atm)
Why not using the near field-far field stitching method?
Or this can also be a ground plane - far fiels stitching method well.
No need for fancy measuring towers.
Klippel NFS basically does the same thing, just automated.
Automatic turntable is not that expensive, specially when you have to do this on a regular basis.
There is a full size gantry type CNC machine available. It's much much cheaper than $100K NFS. In fact, four times more expensive.
I can't remember the name. I'll ask and drop you a PM. We studied all the available options and this vendor turntable seems the best if a gated measurement is needed, and it works flawlessly with all the measurement softwares including Clio, ARTA. Though REW has no turntable control capability, but the seller provided a 3rd party solution for it that works like a charm with REW and other possible softwares. And it's cheaper than Outline and Four Audio and others. I think we bought it 2450 Euros.
Oh, sure. I didn't mean to belittle their endeavors. D&D is great, but it's a commercial product, so there is the least documentation about its design, and that's understandable. The Speaker Martijn designed at first place finally lead to 8C, but these have few things in common.
Near-field + far-field gated measurement won't tell a lot of information about speaker directivity in lower frequencies in off-axis. Lower frequencies couldn't be measured in near-filed.
Klippel NFS can give us a lot of useful information about speaker like 3D propagation in space in any distance. Klippel NFS is a revolution in speaker measurement. I wish it was more affordable so the world would gradually become a better-sounding place.
The lower part of frequencies is rather predictable, so that's really a combination of simulation, near-field measurements as well as other partial measurements.
Combine those with like ground plane measurements and you will get extremely good results.
There is even a youtube video out there that Martijn even explains that's the route they used.
Klippel NFS on itself is mostly extremely convenient, with a bit of creativity and time one can do most things manually.
So I wouldn't call it revolutionary, but it's a very nice system to have, I totally agree with that.
Although over 100k is even a lot of money for most companies I worked for.
If you have any idea about how can we convert horizontal and vertical measurements into a holographic sound propagation in 3D space, just tell us about it.
How is that a necessity for making a great speaker design?
But besides that, 3D graphs can be done on multiple ways, it just takes an awful lot of time and most people don't even bother looking at it anyway (that's probably a reason that Erin doesn't show them as well).
2D contour plots (normalized) plus off-axis freq resp will give you most of the information you need.
Yes. I am able to get resolution down to 20 Hz, as long as the ambient noise floor is good... i.e. waiting for no vehicles or other noise sources.
In theory, a ground plane measurement is equivalent to anechoic. In practice, we are limited by environmental considerations. But the rotation of the device under test is no obstacle at all.
Excellent thread on ground plane testing by Erin:
https://www.diyaudio.com/community/...how-close-is-close-enough-to-anechoic.353347/