Hi Earl,
I think your spec may have exceeded my excel skills. Excel can almost certainly do it but I am just a casual excel user.
I was going to look at it but I can't even work out how to get the data you described out of Holm inpuse. The only complete datasets I can get are the impulse responses. How do you get a set of frequency data without exporting each one seperately?
Regards,
Andrew
I think your spec may have exceeded my excel skills. Excel can almost certainly do it but I am just a casual excel user.
I was going to look at it but I can't even work out how to get the data you described out of Holm inpuse. The only complete datasets I can get are the impulse responses. How do you get a set of frequency data without exporting each one seperately?
Regards,
Andrew
Well to be honest, I don't and maybe thats why this can't work. Ask said that he would setup HolmImulse to output a frequency array, but perhaps he just hasn't done that yet.
I read the impulse responses into MathCad, then add in the nearfield measurement for the low end, smooth to 1/10 or 1/6 octave while I output log spaced points. The smoothing then just becomes part of the interpolation. It may well be that this later aspect is too much to ask Ask to do and it may be quite tough in Excel (i'm not an Excel user either, I don't really like it all that much except for data analysis from subjective tests).
Best bet might be the VB program, but that will take some time.
I just got Visual Studio 2008 and wanted something to try it out on.
I think I will gracefully bow out at this point as I know I can't solve that problem. If anyone wants the raw excel file to have a play with; warts and all, let me know.
However I did also want to say, thankyou to Earl for his very illuminating paper. I had struggled with the concept of constant directivity speakers as it seemed that this was a description of an impossible design and I must be missing something. Your paper made it very clear that CD does not in fact mean Constant directivity in this case and controlled directivity is a much better term.
Whilst I think I find the 3D contour maps easier to read, I do now understand your plots. I would recomend changing the plots on your site to be symetircal like the ones in your papers, if you can find time, as this was part of what helped me to visualise better what was going on.
Regards,
Andrew
However I did also want to say, thankyou to Earl for his very illuminating paper. I had struggled with the concept of constant directivity speakers as it seemed that this was a description of an impossible design and I must be missing something. Your paper made it very clear that CD does not in fact mean Constant directivity in this case and controlled directivity is a much better term.
Whilst I think I find the 3D contour maps easier to read, I do now understand your plots. I would recomend changing the plots on your site to be symetircal like the ones in your papers, if you can find time, as this was part of what helped me to visualise better what was going on.
Regards,
Andrew
Hello Andrew,
Please post the excel here. In a week or so I'll be able to look close at it and maybe I'll find a way to easily import the data.
If not, I have a colleague at work, excel wizard, and he is doing just that. Finding ways to get txt. data imported to excel.
I think this would be a usefull piece of data vizualisation...
Please post the excel here. In a week or so I'll be able to look close at it and maybe I'll find a way to easily import the data.
If not, I have a colleague at work, excel wizard, and he is doing just that. Finding ways to get txt. data imported to excel.
I think this would be a usefull piece of data vizualisation...
I would like the data, can you PM me with it.
Yes, Controlled Directivity may be a better term, but its also very easy to say that anything is "controlled directivity", so in that sense it doesn't mean very much.
Yes, I agree that the symmetrical plots make more sense. But it is redundent data, so thats why I left it off. Its easy to just fold it over and make a full 180 degree plot so I will do that, thanks.
Hi,
I have attached the excel spread sheet. It just fits the upload limit zipped.
This is a rough working sheet so its not documented well.
The CODA III data I had was only at 12.5deg steps so I averaged it to create more steps. It was also a very rough measurement as the reason I was measuring them is the base drivers are failing (the voice coil former is unwrapping - quite common with this speaker) and I was looking at them to try to find a replacement that might work with minimal modification to the crossover. (which I couldn't find, so the garage will be short of speakers for a while longer)
The graphs are on the later sheets so do look at all of them.
The last sheet has some initial prototype work on a linear interpolation algorithm and linear data set. It works to extract log steps from linear data. However the log step sizes need to be sorted to creat a larger set of data.
I did say it was warts and all and there are plenty of warts
Share and enjoy.
Regards,
Andrew
I have attached the excel spread sheet. It just fits the upload limit zipped.
This is a rough working sheet so its not documented well.
The CODA III data I had was only at 12.5deg steps so I averaged it to create more steps. It was also a very rough measurement as the reason I was measuring them is the base drivers are failing (the voice coil former is unwrapping - quite common with this speaker) and I was looking at them to try to find a replacement that might work with minimal modification to the crossover. (which I couldn't find, so the garage will be short of speakers for a while longer)
The graphs are on the later sheets so do look at all of them.
The last sheet has some initial prototype work on a linear interpolation algorithm and linear data set. It works to extract log steps from linear data. However the log step sizes need to be sorted to creat a larger set of data.
I did say it was warts and all and there are plenty of warts
Share and enjoy.
Regards,
Andrew
There are VB addins to Excel to make contour plots too. See Contour Plotting in Excel for example. It took me about 10 minutes to figure out how to use this.
Last edited:
There may be a very simple low tech solution to the data format dillemma......
I'm not an Exel expert, neither have I played with Holmimpuls yet, but a fairly standard delimited ASCII format would look something like this:
Frequency,Angle1,Angle2,Angle3, etc. etc.
1000,91,95,94, etc.etc.
Once your measurement software can spit out this format, either tab or comma delimited, for instance, you can simply import that into a blank worksheet, let Exel create the columns, then simply copy the columns, and paste them in the corresponding columns of the graphing worksheet.
One can even do any normalizing/massaging of the data in the 1st blank worksheet to make it conform to whatever the graphing worksheet will need, before copying and pasting.
This format seems to be universal, in that my Audio Precision software, and LinearX Leap 5 & LMS 4 will generate data in this format, easy for data exchange between the two, as well as for importing into Exel for whatever analysis/manipulation I may need to do.
Lukas
I'm not an Exel expert, neither have I played with Holmimpuls yet, but a fairly standard delimited ASCII format would look something like this:
Frequency,Angle1,Angle2,Angle3, etc. etc.
1000,91,95,94, etc.etc.
Once your measurement software can spit out this format, either tab or comma delimited, for instance, you can simply import that into a blank worksheet, let Exel create the columns, then simply copy the columns, and paste them in the corresponding columns of the graphing worksheet.
One can even do any normalizing/massaging of the data in the 1st blank worksheet to make it conform to whatever the graphing worksheet will need, before copying and pasting.
This format seems to be universal, in that my Audio Precision software, and LinearX Leap 5 & LMS 4 will generate data in this format, easy for data exchange between the two, as well as for importing into Exel for whatever analysis/manipulation I may need to do.
Lukas
An externally hosted image should be here but it was not working when we last tested it.
A few months ago I saw "The Watchmen", and Ozymandias TOTALLY reminded me of John K
Not to stray away from topic, but I've been wondering........ Dr Geddes, Toole's book seems to suggest the wide dispersion speakers sound better to the end user, why do you design narrow(er) directivity speakers? Can you sum it up for me in a few sentences, I know there are good reasons I don't like Bose or that line of thinking etc... Hopefully it's not just my pride disgusted with mainstream whatever. I know you said 10msec is a good idea prior to first reflection, can you explain the "why" of this idea for me and the rest of who's thinking?
Thanks,
Dan
Thanks,
Dan
What, no Summa vertical polar map, Earl?
Scroll to bottom here:
http://www.eighteensound.com/staticContent/applications/kits/18Sound_kit8.pdf
[Yah, CLIO's been generating them for years, now, in any desired resolution.... ]
Scroll to bottom here:
http://www.eighteensound.com/staticContent/applications/kits/18Sound_kit8.pdf
[Yah, CLIO's been generating them for years, now, in any desired resolution....
]
Floyd and I argue this often. Its the one thing that we disgaree on. His research is based on "preference" with piston loudspeakers and a CD loudspeaker of the type that I propose was never tried. What one must always keep in mind, and Floyd would admit this, is that his data is based on "what we know today". If someone comes along with new data then the current beliefes have to take this into consideration. I believe that my approach is "new data". I have not reduced it to the quantified status that Floyds data is in, because I don't have Floyds resources. But I firmly believe that if my speakers were put into a reverberant room along with his wide directivity speakers, that mine would be prefered. They may even be prefered in the less reverberant room as well.
Floyd and crew do not even consider image in their criteria for a loudspeaker. It is heavily weighted on spaciaousness. They don't even consider any dynamics, nothing in the amplitude aspects. There are a lot of holes in Floyds work, but its still the best there is on the subject if you can recognize where these holes are at.
To my knowledge they do (a "spatial quality questionnaire" can be found in Toole "Sound Reproduction" p. 132).
Dan, if you want to know more you can find Sean Olive as Tonmeister2008 at avsforum.com
Best, Markus
What, no Summa vertical polar map, Earl?
Scroll to bottom here:
http://www.eighteensound.com/staticContent/applications/kits/18Sound_kit8.pdf
[Yah, CLIO's been generating them for years, now, in any desired resolution....
I'd like to see that too. We can't talk about power response without knowing horizontal and vertical directivity.
By the way Zilch, could you post polar maps for the EconoWave (horizontal and vertical, ±90°, 7,5° increments)?
Best, Markus
Last edited:
It's not new data, and you know it, that having appeared in the JBL 4430 brochure, for example, 30 years ago.
And we (you and I, specifically) have previously discussed the origins of defined directivity and an enhanced image rendition zone (IRZ) here, based upon the work of Keele in other JBL products, also decades ago:
http://www.audioheritage.org/vbulletin/showthread.php?t=5671
http://audioheritage.org/vbulletin/showthread.php?t=8373
As Durwood suggested above, it'd sure be good if you provided references in your marketing "White Papers" as opposed to misleading readers into supposing the work is original.
System polars?
Nope. I've only just begun studying the verticals using the "move the mic" technique you and I developed in these pages over a year ago, which is limited to +/- 20° vertical as I implement it here. I'm doing 5° increments with that to find the influence of the vertical nulls and with inverse polarity, aiming the forward axis. I have not yet exercised CLIO's directivity sonogram capabilities with this.
Earl did the polars on the waveguide itself,* but there is no EconoWave system, per se, rather, hundreds of them, as it's an open source design. For my own part, I'm presently working with the QSC waveguide and 12" woofers with a minimum 10.5" C/C distance. Augerpro measured that waveguide's polars, and they are similar to our standard JBL/Pyle(clone) PT waveguide with respect to pattern control, basically, in a larger form factor, with integral mouth roundovers. Both of us have been able to demonstrate a +/- 10° vertical "window" (-6 dB) at 1.3 - 1.5 kHz crossover with that combination, posted on other sites.
*Since, deleted from the GedLee website....
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.