An other comment:
When I had the old speakers with four 8" woofers covering 100 - 500 Hz, I had a lot of room resonance peaks and other nasties that needed a lot of EQ. Of some reason, the ribbons behave far better, and very little EQ is required. The ribbons also sounds a lot cleaner in the entire lower midrange, possibly because of less room resonance?
Even if we EQ away the response peaks caused by room resonance, the energy is still "ringing" in the room, and we still hear the resonance even if the frequency response is flat. This is a very strong argument for dipoles.... because they do exite the room modes a lot less than onmipole speakers. The difference is greatest in the modal range of the room, from some 200 Hz and down, but it's noticable up to at least 500 Hz.
When I had the old speakers with four 8" woofers covering 100 - 500 Hz, I had a lot of room resonance peaks and other nasties that needed a lot of EQ. Of some reason, the ribbons behave far better, and very little EQ is required. The ribbons also sounds a lot cleaner in the entire lower midrange, possibly because of less room resonance?
Even if we EQ away the response peaks caused by room resonance, the energy is still "ringing" in the room, and we still hear the resonance even if the frequency response is flat. This is a very strong argument for dipoles.... because they do exite the room modes a lot less than onmipole speakers. The difference is greatest in the modal range of the room, from some 200 Hz and down, but it's noticable up to at least 500 Hz.
WTF?
Would you care to post a file from your two current systems containing an ungated impulse from the listening position out to at least 150mS so we can ask Elias to run them in his Wavelet Transform analysis such as this:
.
http://www.diyaudio.com/forums/multi-way/161299-directivity-dipole-tweeters-18.html#post2428600
.
It is a very revealing way to look at room interaction and room tuning.
Would you care to post a file from your two current systems containing an ungated impulse from the listening position out to at least 150mS so we can ask Elias to run them in his Wavelet Transform analysis such as this:
.
http://www.diyaudio.com/forums/multi-way/161299-directivity-dipole-tweeters-18.html#post2428600
.
It is a very revealing way to look at room interaction and room tuning.
650 Hz would be about right for a pure 15" driver. 350 Hz could be the value for a 15" driver in a wider baffle or a H frame.
Hello,
I just did a sim with Edge for a speaker shaped like the B&G´s choosing three distances 1.5 m ( red ) , 2 m ( purple ), 2.5 m ( green ) :
B G RD 75 drei Dist | 75, dist, doityourself, linearray, rd | hifi-forum.de Bildergalerie
I just did a sim with Edge for a speaker shaped like the B&G´s choosing three distances 1.5 m ( red ) , 2 m ( purple ), 2.5 m ( green ) :
B G RD 75 drei Dist | 75, dist, doityourself, linearray, rd | hifi-forum.de Bildergalerie
Attachments
Baffle
Does it look any smoother if you put it in a triangular baffle, 40cm across the bottom?
Does it look any smoother if you put it in a triangular baffle, 40cm across the bottom?
Hey,
and I found some frequency and distortion data on the BG 75 in Dipole-mode. The measurement is shown in a thread at Hifi-Forum in Germany done by Tiki. It was done inside at 12,5 cm distance at aprox. 80 cm of ground an are claimed to be roughly spl-calibrated. Because it´s a number of graphics please click yourselves through them:
Index of /skvf/speaker/linearray/messungen/20060114
To me the distortion behaviour seams not to be perfect but still not critical. OK, there is a rise of THD around 1.5 - 2 kHz though it is not dramatic. And they imo show clearly that crossing the RD 75 at 180 Hz is no problem. But I am not very used to the interpretation of multi-tone measurement.
Those who are interested can find the whole tread from witch the measurements are taken here: http://www.hifi-forum.de/index.php?action=browseT&forum_id=104&thread=5195&back=&sort=&z=1
@ sendler, Edge works with an ideal source and is a little over the top sometimes, what makes a simulated line source look much wilder than it acoustical appears to the ear. Obviously the real RD´s are different than the simulation - look at the amount of EQ Stig applied to get it flat. The RD´s don´t need a baffle.
And Edge is free and easy to use: Home of the Edge
Excuse me for being lasy
,
G F
and I found some frequency and distortion data on the BG 75 in Dipole-mode. The measurement is shown in a thread at Hifi-Forum in Germany done by Tiki. It was done inside at 12,5 cm distance at aprox. 80 cm of ground an are claimed to be roughly spl-calibrated. Because it´s a number of graphics please click yourselves through them:
Index of /skvf/speaker/linearray/messungen/20060114
To me the distortion behaviour seams not to be perfect but still not critical. OK, there is a rise of THD around 1.5 - 2 kHz though it is not dramatic. And they imo show clearly that crossing the RD 75 at 180 Hz is no problem. But I am not very used to the interpretation of multi-tone measurement.
Those who are interested can find the whole tread from witch the measurements are taken here: http://www.hifi-forum.de/index.php?action=browseT&forum_id=104&thread=5195&back=&sort=&z=1
@ sendler, Edge works with an ideal source and is a little over the top sometimes, what makes a simulated line source look much wilder than it acoustical appears to the ear. Obviously the real RD´s are different than the simulation - look at the amount of EQ Stig applied to get it flat. The RD´s don´t need a baffle.
And Edge is free and easy to use: Home of the Edge
Excuse me for being lasy
, G F
It's interesting to see, how things change when we include the floor reflection. Note how the EQ necessary between 50 and 200 Hz changes from fabels 8-9 dB to just 4 dB with floor "bounce". I only did it for 1,5 m.
Attachments
Hey Rudolf,
thank you, good idea. I did it again mirroring the source.
B G RD 75 Boden drei Dist | 75, boden, dist, doityourself, linearray, rd | hifi-forum.de Bildergalerie
And I found that Tiki has done interesting intense exploration about the vertical behaviour of this source. He did a vertical directivity sono at a distance of 0.5 m going up parallel to the line in steps of one cm - wow :
28.12.2009 Messungen RD 75, Abstrahlverhalten vertikal
G F
thank you, good idea. I did it again mirroring the source.
B G RD 75 Boden drei Dist | 75, boden, dist, doityourself, linearray, rd | hifi-forum.de Bildergalerie
And I found that Tiki has done interesting intense exploration about the vertical behaviour of this source. He did a vertical directivity sono at a distance of 0.5 m going up parallel to the line in steps of one cm - wow
:28.12.2009 Messungen RD 75, Abstrahlverhalten vertikal
G F
That is an admirable piece of scientific research for sure.And I found that Tiki has done interesting intense exploration about the vertical behaviour of this source. He did a vertical directivity sono at a distance of 0.5 m going up parallel to the line in steps of one cm - wow
But I still wonder if it tells me anything that would be relevant for the listening process.
Rudolf
Hello,
the first link in post 1069 is an old one, the description not really correct, as I understand.
These rough measurements were done on an 8-array of good fairly mids the right one.
More measurements are to be found here, all in German of course.
And no, that's not science, only a bit of diligence.
One may hear the interference structure very clear, which is an intrinsic behaviour of sources, bigger than (lambda/2 to) lambda, using appropriate signals. Somebody may be satisfied with this listening normal music or speech signals. I "believe" in the advances of CBTs meanwhile, knowing, this is not the perfect implementation.
Regards,
Timo
the first link in post 1069 is an old one, the description not really correct, as I understand.
These rough measurements were done on an 8-array of good fairly mids the right one.
More measurements are to be found here, all in German of course.
And no, that's not science, only a bit of diligence.
One may hear the interference structure very clear, which is an intrinsic behaviour of sources, bigger than (lambda/2 to) lambda, using appropriate signals. Somebody may be satisfied with this listening normal music or speech signals. I "believe" in the advances of CBTs meanwhile, knowing, this is not the perfect implementation.
Regards,
Timo
ups
Hello Tiki, hello all,
ups, excuse me, it seams I got lost a little in the depth of somebody else's server. The link I wanted to give in post 1069 is this:
Index of /skvf/speaker/linearray/messungen/20060115/RD75
As I read this data here the THD reaches more than 3% in the area around 2kHz nevertheless it is measured + or - 10 dB. That could be a bit much of distortion. Hmmm ...
G F
Hello Tiki, hello all,
ups, excuse me, it seams I got lost a little in the depth of somebody else's server
. The link I wanted to give in post 1069 is this:Index of /skvf/speaker/linearray/messungen/20060115/RD75
As I read this data here the THD reaches more than 3% in the area around 2kHz nevertheless it is measured + or - 10 dB. That could be a bit much of distortion. Hmmm ...
G F
Hello,
the only meaningful measurement I can show you:
.
The horizontal behaviour is to be seen clearly. The WG was not closed and did not contain the midwoofers behind the "grills". The Aura NS4 behaved like this in a small box with a similar front mesh.
I did not realise the WG due to the estimated production effort and costs.
Hello fabel, using the skvf-server may be "risky". I told the admin to delete the data there. Most of the information is to be found on the ibtk-server.
Best regards, Timo
the only meaningful measurement I can show you:
The horizontal behaviour is to be seen clearly. The WG was not closed and did not contain the midwoofers behind the "grills". The Aura NS4 behaved like this in a small box with a similar front mesh.
I did not realise the WG due to the estimated production effort and costs.
Hello fabel, using the skvf-server may be "risky". I told the admin to delete the data there. Most of the information is to be found on the ibtk-server.
Best regards, Timo
And here's some measurements, all done in listening position with the listening chair removed...
ETC left speaker:
Waterfall left speaker (smoothed):
ETC right speaker:
Waterfall right speaker (smoothed):
MLS files from CLIO for those who are interested.
I tried to do distortion measurements as well, but my CLIO cant handle the total delay through my system, so the results was just rubbish. I wanted to measure the entire system with EQ and XO, but it seems that it's not possible because of the delay introduced by the digital filters. I did sit and listen to the sinewave-based measurement, by it did not reveal any disturbing sounding things. The ribbon does buzz a little, but so do most other drivers as well.....
ETC left speaker:
An externally hosted image should be here but it was not working when we last tested it.
Waterfall left speaker (smoothed):
An externally hosted image should be here but it was not working when we last tested it.
ETC right speaker:
An externally hosted image should be here but it was not working when we last tested it.
Waterfall right speaker (smoothed):
An externally hosted image should be here but it was not working when we last tested it.
MLS files from CLIO for those who are interested.
I tried to do distortion measurements as well, but my CLIO cant handle the total delay through my system, so the results was just rubbish. I wanted to measure the entire system with EQ and XO, but it seems that it's not possible because of the delay introduced by the digital filters. I did sit and listen to the sinewave-based measurement, by it did not reveal any disturbing sounding things. The ribbon does buzz a little, but so do most other drivers as well.....
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.