Hi Earl,
You may be thinking that the scale on the impulse response is linear, but it is a dB scale. I use the dB scale to make the first reflection easier to identify. Attached is the linear response. Sorry if this threw you off track. A question for you, how much more delay would you expect the foam would add to the start of the impulse response compared to an impulse response without foam ?
Regards
Peter
You may be thinking that the scale on the impulse response is linear, but it is a dB scale. I use the dB scale to make the first reflection easier to identify. Attached is the linear response. Sorry if this threw you off track. A question for you, how much more delay would you expect the foam would add to the start of the impulse response compared to an impulse response without foam ?
Regards
Peter
Attachments
Yes, I didn't catch that - and didn't even know you could do that in Holm. Its makes the signal look very strange. I don;t think that I would use such wide windows as beyond a few ms, its just noise.
I have no experience with foam plug but collected measurements from those who use Autotech OS waveguide. Enclosed please find data I've used for wavelet. This is not Dr. Geddes speaker but similar:
LoudspeakerSoft Forum / Geddes Loudspeakers clone
LoudspeakerSoft Forum / Geddes Loudspeakers clone
Attachments
jzagaja
Jack, I don't know what you've done with the data but you have wrong levels and maybe gating. This is more representative picture directly from Holm on my computer (raw gated data). The foam fills almost all of the waveguide volume. Honestly I can't remember if I had the time offset locked or not (I've forgotten to check that), so maybe the time scales are not comparable also.
I don't think you can see or say much from these measurements.
Marcel
Jack, I don't know what you've done with the data but you have wrong levels and maybe gating. This is more representative picture directly from Holm on my computer (raw gated data). The foam fills almost all of the waveguide volume. Honestly I can't remember if I had the time offset locked or not (I've forgotten to check that), so maybe the time scales are not comparable also.
I don't think you can see or say much from these measurements.
Marcel
Attachments
Last edited:
These are responses for "with foam" and "without foam" that I've measured on my attempt of OSWG. You can easily switch between them. It has all been measured from one point in space with and without foam for each angle. The only effect you can probably see is a slight attenuation, progressively increasing with frequency. It doesn't mean that there is nothing more - only that it's all these pictures can tell you. You can see 1/12oct smoothed version of this same data on my web.
Attachments
Last edited:
Data from Holmimpulse exported (1024 samples) to wav file and imported into ARTA.
As promised - here you possibly find some first answers:
http://www.diyaudio.com/forums/ever...-giant-leap-maturizing-audio.html#post2565176
🙂
Michael
PS
anybody be "warned" :
non of my findings are "peer reviewed" yet
LOL
Last edited:
It looks like the on axis notch between 5khz and 6khz is less deep with foam, so the foam has some good effect on the mouth/throat reflexions.
Hello !
Here I compared the posted IR with foam and without foam, using my new wavelet analysis package:
http://dl.dropbox.com/u/2400456/html/Elias_Pekonen/Wavelet_software.html
Here's a constant Q wavelet comparison:
And here's the same with amplitude normalised to better compare the temporal effects:
It seems that the foam is making the waveguide impulse response worse !
Am I the only one to see this !? 😀
- Elias
Here I compared the posted IR with foam and without foam, using my new wavelet analysis package:
http://dl.dropbox.com/u/2400456/html/Elias_Pekonen/Wavelet_software.html
Here's a constant Q wavelet comparison:
An externally hosted image should be here but it was not working when we last tested it.
And here's the same with amplitude normalised to better compare the temporal effects:
An externally hosted image should be here but it was not working when we last tested it.
It seems that the foam is making the waveguide impulse response worse !
Am I the only one to see this !? 😀
- Elias
Last edited:
Is that Earl's specific foam application or a fixed foam density? That's a significant difference since Earl specifically found it necessary to use a specific foam application. In my use of felt and to some degree foam, the difference in response can be significant with different felt/foam.
Dave
I haven't had a chance to follow all of this carefully. Jack asked for a measurement of his OS12 waveguide with and without the foam. I had one of his waveguides at the time, along with some foam I had cut, 30 ppi reticulated foam as seen here:
I was in a hurry and about to give the foam + waveguide back to its owner, so took a quick measurement with REW. I used REW as it could export the impulse response as a wave file. It was ungated and IIRC it was a nearfield, very close to the foam itself. It may have been an inch or two away. I had no idea what he intended to do with it and I also used some material not shown to hold the foam in place temporarily. I had to dispense with the first measurement of the foam, in which the material obscured the measurement and gave an uneven response which looked suspect. Having used the foam with 3 different waveguides, I knew something wasn't quite right. In each case, with 3 different waveguides of various sizes, the loss from the foam is 2db and quite consistent, and increases to 3 db above 10k.
I would be wary of basing too many conclusions on this measurement and to be honest I hope no one is basing too much on it. It's important to be sure that you have correct data before drawing conclusions. Had I known that this measurement would be analysed and scrutinised in this way, I would have taken it with a bit more care, in particular with a better method of holding the foam.
I have noticed an improvement with the foam, a significant reduction in harshness.
I currenly have running a 6" oblate spheroid waveguide with the same compression driver (DE250) and can take more measurements.
I was in a hurry and about to give the foam + waveguide back to its owner, so took a quick measurement with REW. I used REW as it could export the impulse response as a wave file. It was ungated and IIRC it was a nearfield, very close to the foam itself. It may have been an inch or two away. I had no idea what he intended to do with it and I also used some material not shown to hold the foam in place temporarily. I had to dispense with the first measurement of the foam, in which the material obscured the measurement and gave an uneven response which looked suspect. Having used the foam with 3 different waveguides, I knew something wasn't quite right. In each case, with 3 different waveguides of various sizes, the loss from the foam is 2db and quite consistent, and increases to 3 db above 10k.
I would be wary of basing too many conclusions on this measurement and to be honest I hope no one is basing too much on it. It's important to be sure that you have correct data before drawing conclusions. Had I known that this measurement would be analysed and scrutinised in this way, I would have taken it with a bit more care, in particular with a better method of holding the foam.
I have noticed an improvement with the foam, a significant reduction in harshness.
I currenly have running a 6" oblate spheroid waveguide with the same compression driver (DE250) and can take more measurements.
If you could share your new (foam non-foam) IR files here I would very much appreciate.
Please do not only take a measurements at the "mouth" but also one roughly at 1m distance on axis and at roughly 30deg (Earls target FR).
If you can do outdoor measurements, 2-3m distance would even be better - but its not really necessarily.
🙂
Michael
