I have recently posted a white paper that I wrote highlighting directivity issue in loudspeakers. I hope you enjoy it and I invite discussion.
I should state however, that requests for "what does x-speaker look like" will be pointless since I have already put virtually all of the competitive products that I have in the paper. So there is no more competitive data to be had. I have much more data on my own products as you can imagine.
The paper is at Geddes Directivity White Paper
I should state however, that requests for "what does x-speaker look like" will be pointless since I have already put virtually all of the competitive products that I have in the paper. So there is no more competitive data to be had. I have much more data on my own products as you can imagine.
The paper is at Geddes Directivity White Paper
Thanks,
I caught a couple of typo's, if you're interested:
- fig 11 the color coding is off
- fig 12 the horizontal scale values are missing
- in the text fig 13 is mentioned as the normalized version of fig 12, and later on it's mentioned as a different speaker. It's probably fig 12 A&B?
- page 13: "This speaker is widely praised for its natural and natural sound."
I caught a couple of typo's, if you're interested:
- fig 11 the color coding is off
- fig 12 the horizontal scale values are missing
- in the text fig 13 is mentioned as the normalized version of fig 12, and later on it's mentioned as a different speaker. It's probably fig 12 A&B?
- page 13: "This speaker is widely praised for its natural and natural sound."
Thanks
- fig 11 - oops, that will be hard to fix. I'll see what I can do
- fig 12 - this sometimes happened. Once I generated a figure, it is hard to go back and recreate it. That is particularly true of all the measured data. The frequency scale never changed however so its easy enough to figure out.
Corrected Fig 13 reference and typo on Page 13.
Thanks for those!
Member
Joined 2003
AwEeee Sheet! Its a real paint to recreate those curves!! Proably won't happen. I never even looked at the legend. I may take it out altogether since THAT I can do. Its really not necessary.
Just grab the figures in most any image editor and draw new lines for the legends with the correct colors. The dipstick tool will give you an exact color match to the line in the graph if you need it.
If it's vector data, well, that's harder.
Off to actually read the paper. Looks like lots of pretty pictures. 😛
Wow, just after skimming it seems I've learned a great deal in the last 2 months. I can't wait to get through this. It's good to see that some of the data mirrors what I've done at home though your graphs look much better. Makes me feel like I'm not totally crazy.
Thanks!
Dan
Thanks!
Dan
I'd like to create directivity plots like that for my own measurements. Any software out there that can do it?
More to the point: How did Mr. Geddes create those graphics? Don't recall seeing it explained in the white paper.
My software is commercial, called Sigma-Plot, but publication quality graphics software like this is very expensive.
If you can send me the data in the correct format, I will plot it as along as I can post it as well. I'd like to create a library of real polar data measurements. But I don't want to have to clean up data and reformat it etc. It has to be in the exact format that I need.
Three *.txt files:
1) list of frequencies (suggest no more than 512 points)
2) list of angles (suggest every 7.5 degrees)
3) data of angles seperated by comma's with all angle data on one line
I can then plot this in exactly the same form as I show in the paper.
The data should be smoothed in frequency, but not more than 1/3 octave, but 1/10-1/6 works best. Unsmoothed data makes a messy plot because of the random noise in the data.
Last edited:
Will ARTA do that kind of resolution? Many of these programs don't do interpolation, but Sigma-plot does. And it does a good job. Some others that I have tried have not been very good.
Last edited:
thanks for the offer Earl, much appreciated. I'll be out of the country for the next three weeks, but after that I'd like to take you up on that offer.
I'd love to see a library of designs. That way, with real data at hand, we can get to what really matters and stop all this hand waving and guessing about audio performance. I just don;t have the resources to do any more than I have already done.
Thats not bad. Certainly cheaper than SigmaPlot. Are the scales arbitrary? -90 > 90, 0dB > -36 dB?
Is this data available? I'd love to add this example to the paper if I could get the data set. And I'd like to compare my plotting with this.
Do you have any more?
Is this data available? I'd love to add this example to the paper if I could get the data set. And I'd like to compare my plotting with this.
Do you have any more?
Not exactly. Angles are arbitrary (depends only on measurement angles done, up to 360deg). But, amplitude iso-contours seems to be fixed on 1/10 of range. OTOH Color palette can be "jet" (as it is) or B-W, contoured or none. Range is up to 40dB, IIRC.
Data is available, but this is first atempt of xo for my budget surround speaker. I can send them. If you want final version, you have to wait because I am short of coils.
Data is available, but this is first atempt of xo for my budget surround speaker. I can send them. If you want final version, you have to wait because I am short of coils.
Hey Earl,
I like the paper. I think you have some good stuff in there. Some of it that you touched on I would be more tempted to make the whole paper about but that's just what I am fixated on I guess. This point kind of stuck out to me as something I don't think works.
"Recall now our discussion of image shift in a stereo situation where we
hypothesized that if the farther speaker could get louder as one moved
laterally while the closer speakers level decreased, that we might be able
to offset the time delay differences and maintain a fairly stable image
with listening position."
Thats a big "might". You can compensate with a delay for one side of the room but then you compromise the center and opposite side of the room. And I am pretty sure the delay (arrival time) is more important than the volume in terms of what will effect the stereo image. Also subjective reviews on other polar patterns than your own are very similar in this respect. Basically a report of a lack of shifting compared to the average loudspeaker. I am not sure it is exclusive to your design perceptually. I see why you could think that it is based on that physical alignment though. And yeah you could be right and that this sort of directivity is self compensating when moving around a room to a certain extent. I guess I just suspect there is something else going on.
I like the paper. I think you have some good stuff in there. Some of it that you touched on I would be more tempted to make the whole paper about but that's just what I am fixated on I guess. This point kind of stuck out to me as something I don't think works.
"Recall now our discussion of image shift in a stereo situation where we
hypothesized that if the farther speaker could get louder as one moved
laterally while the closer speakers level decreased, that we might be able
to offset the time delay differences and maintain a fairly stable image
with listening position."
Thats a big "might". You can compensate with a delay for one side of the room but then you compromise the center and opposite side of the room. And I am pretty sure the delay (arrival time) is more important than the volume in terms of what will effect the stereo image. Also subjective reviews on other polar patterns than your own are very similar in this respect. Basically a report of a lack of shifting compared to the average loudspeaker. I am not sure it is exclusive to your design perceptually. I see why you could think that it is based on that physical alignment though. And yeah you could be right and that this sort of directivity is self compensating when moving around a room to a certain extent. I guess I just suspect there is something else going on.
Last edited:
- Status
- Not open for further replies.