Once again, polarmap will not run on a local machine, it is not setup to do that. And the setup program uses a "call" that is non-standard and not all browsers handle it properly. So if you want use polarmap right now you should run it from within Explorer. This will likely change in the future, but for right now that's the way it is. (It is old code and has some outdated calls. I am rewriting it now, but that may take a little while.)
Also, the zip file can only be opened in HolmImpulse, and the txt file can only be used in my analysis program. Neither of these files is used by Polarmap.
As a user, you would create a project in Holm, like the example, export the data as a text file, like the example, and send that to me. I would analyze it and post the results to the database. You could then view them there.
Also, the zip file can only be opened in HolmImpulse, and the txt file can only be used in my analysis program. Neither of these files is used by Polarmap.
As a user, you would create a project in Holm, like the example, export the data as a text file, like the example, and send that to me. I would analyze it and post the results to the database. You could then view them there.
Last edited:
All in one enclosure? Otherwise, just do the midrange and up. That's really the only place where free field measurements mean anything. Low frequencies always need the room involved. But yes, 200 lbs is extreme. I don't think that anyone can handle that, not even Harman.
I guess you will just have to "hope" that the response is OK.
Of course I am sure that they sound good to you so who wants to see those pesky measurements that may not agree.
I guess you will just have to "hope" that the response is OK.
Of course I am sure that they sound good to you so who wants to see those pesky measurements that may not agree.
On another thought, it would be interesting to know what the least amount of local control in the software would be necessary to get a users attention. Do you have to have complete control or you won't use it? What if you could do anything except that all data has to be saved to the web database? You could not save or access it locally.
Of course you can say that only full control would interest you, but given the amount of time I have in this and how critical it is to my proprietary designs, that is very unlikely to happen. Definitely not in the short term. But I have not decided what I will allow users to do themselves with this capability.
Clearly competitors would not use it if they had to post their results for all to see, unless their results were very good, in which case I'd love to see them. I think we would all love to see them. We have to stop going on about how such and such a speaker is great with nothing more than our (or someone else's) personal opinion to support it.
I also have a crossover simulation that uses the same techniques - and then some - and someday this could be available as well, but all of this is proprietary development software that I use extensively in my own designs. It's one thing for DIY usage, its quite another for commercial usage.
Of course you can say that only full control would interest you, but given the amount of time I have in this and how critical it is to my proprietary designs, that is very unlikely to happen. Definitely not in the short term. But I have not decided what I will allow users to do themselves with this capability.
Clearly competitors would not use it if they had to post their results for all to see, unless their results were very good, in which case I'd love to see them. I think we would all love to see them. We have to stop going on about how such and such a speaker is great with nothing more than our (or someone else's) personal opinion to support it.
I also have a crossover simulation that uses the same techniques - and then some - and someday this could be available as well, but all of this is proprietary development software that I use extensively in my own designs. It's one thing for DIY usage, its quite another for commercial usage.
Last edited:
The measurements will not conform well to the Harman paradigm- Kantor believed in asymmetric polar patterns and specific baffle angles. The measurements I've done on them show a pretty flat response at the listening position (I moved the mike around a 30cm square window where my head would be), but it would be really interesting to see how they show on a high angular resolution measurement set. If I can get some sturdy assistants...
I had those discussion with Floyd Toole once when he wanted to measure my speakers. He did not seem to think that using a non-symmetric, or in my case, off-axis, listening position would be an issue. I would tend to agree.
I also agree that a non-symmetric polar response about the "listening axis" is highly desirable. But it is not so easy to create a pattern that is not symmetric about some axis. To me just toeing in the speakers gives the right results in the forward direction. It would be nice if the response fell off faster as the angle went away from the listening axis towards the back. This is certainly possible, but then you have to make a left and right speaker that are unique. And then if there is a center channel this is different as well. From a practical standpoint it is just not feasible.
I also agree that a non-symmetric polar response about the "listening axis" is highly desirable. But it is not so easy to create a pattern that is not symmetric about some axis. To me just toeing in the speakers gives the right results in the forward direction. It would be nice if the response fell off faster as the angle went away from the listening axis towards the back. This is certainly possible, but then you have to make a left and right speaker that are unique. And then if there is a center channel this is different as well. From a practical standpoint it is just not feasible.
Well, it definitely changes placement options a smidge in a rectangular room. I'm sure you're familiar with Duke LeJeune's tilted baffle.
It essentially forces the correct angle to the head in order to (in Kantor's view) optimize crosstalk. At the same time, it forms a pi/8 corner for the woofer, and forces the correct distance between woofer and midrange to eliminate the so-called Allison effect. There's an absorber on the outside edge of the midrange/tweeter array but not the inside.
Subjectively, when I'm in the sweet spot, the image and sense of the acoustic space are better than just about any other speaker I've built or owned. I can hear all sorts of engineering tricks in recordings.
(photo shamelessly stolen from Home Page | Stereophile.com)
Attachments
So these are forced into the corners? That's pretty wide in many rooms. No certainly not the way I would do things.
I would agree that if that was the intended placement that this would be the way that they should be measured. But I do not think that the results would come out quite like you hope. They won't be constant directivity and the angular polars would likely be pretty rough. That situation could be done with my approach by mirroring the response about the plane of the wall side of the speaker. The gap may be an issue as that would be "bad" but hard to account for.
We all have different subjective opinions about things. This thread is trying to get away from that.
I would agree that if that was the intended placement that this would be the way that they should be measured. But I do not think that the results would come out quite like you hope. They won't be constant directivity and the angular polars would likely be pretty rough. That situation could be done with my approach by mirroring the response about the plane of the wall side of the speaker. The gap may be an issue as that would be "bad" but hard to account for.
We all have different subjective opinions about things. This thread is trying to get away from that.
No, they're actually spaced well away from the corners, but placed against the back wall. The back wall and the speaker cabinet side form a virtual corner.
The low resolution angular measurements are actually pretty smooth:
http://www.stereophile.com/content/nht-33-loudspeaker-measurements
My versions are somewhat different than the production version, but not so much that this measurement would be significantly affected. And as you might gather, I love measurements and have no particular religious attachments to commercial products. The measurements are what they are, and I think it's always better to have them. If you've improved the genus, I'm all for it.
The low resolution angular measurements are actually pretty smooth:
http://www.stereophile.com/content/nht-33-loudspeaker-measurements
My versions are somewhat different than the production version, but not so much that this measurement would be significantly affected. And as you might gather, I love measurements and have no particular religious attachments to commercial products. The measurements are what they are, and I think it's always better to have them. If you've improved the genus, I'm all for it.
No not really. I have not seen Duke in many years. He got too esoteric for me.
To make the software very interesting, I would need to get the end result fast (iow during the measurement session itself so I can continue to make adjustments). When I measure speakers, I like to do incremental changes and see the results (to learn the effect of individual changes, and also sometimes just trail and error...). This could mean that I make at least 20 polars before I am ready (or before my "measurement time" is up). When having to send data to you for processing, this would take way to much time. Saving all this data on your database would be no problem for me, but if all users to this, it would turn in one big heap of useless data I am afraid...
So I do not see many choices between allowing full control, and what you are doing now. You could of course display the result only as a picture with some embedded "Gedlee watermark"...
Fast is not possible when one has to rotate the speaker between sweeps and a change to the design makes all the previous runs invalid.
This was never intended as a development tool for users. My development software is quite different and does exactly what you want, but that is not in the cards at this point.
My intent is for people to show off their end result, not work-in-progress. Once you think that you have it right - "sounds good to me" and all that, then measure it and let us all see it. I think that would be very useful.
As to the database, I am already looking at how to do this. Of course its easy on the web and I know database programming so there is nothing beyond me. But for now the txt approach works just fine and this can easily handle several hundred speakers. So there is no hurry to changing that aspect.
I tend to agree that what is there now is fine until the full capability were made available. That, of course, will never happen if no one uses what is there now. If I get several dozen new speaker posts then I'll look at releasing more flexibility. At that point I won't want to do the analysis myself anymore - it will be too much work.
Fast is possible with a DIY automated table and some programming (takes roughly 4 minutes for a polar in ARTA...)
I understand you do not make your software publically available (yet), I also didn't do that with mine (partly due to licencing issues)
What I do see as a problem, is that the user does not see the end result of your software before it is publically available. When people send the result to you of their hard work, I can imagine they are afraid of seeing bad results (they did not know of, because they did not have the correct tools).
Also, when your software is not available during the development phase, it seems not very fair to compare measurement results with speakers (yours) that had the software available during the entire development phase. So even if we were an as good devolloper as yourself (unlikely), we now have a serious disadvantage in trying to get equally good (or better) results.
Last edited:
Fair points, except that ARTA and an automated table is not a low cost DIY solution.
I would have no problem giving people the option of exclusion if the results are very bad. But the point of this is to show things the way they really are. If everyone withdraws their results because they aren't perfect then that kind of defeats the point doesn't it?
I would certainly always allow someone to replace one set of data with another of the same system. That's only fair if a mistake was made.
As to my having a proprietary position because I am good at what I do is not something that I am going to apologize for.
I would have no problem giving people the option of exclusion if the results are very bad. But the point of this is to show things the way they really are. If everyone withdraws their results because they aren't perfect then that kind of defeats the point doesn't it?
I would certainly always allow someone to replace one set of data with another of the same system. That's only fair if a mistake was made.
As to my having a proprietary position because I am good at what I do is not something that I am going to apologize for.
Last edited:
Arta is 79€, and I spend about the some money building the automated table. For the time it has saved my, money well spent.
I will try to make some measurements in Holm, and send them to you. It could take a couple of weeks until I have a chance to do so, but it really does interest me to see the differences in polars between what you are doing and what is done by ARTA.
I have to get some sleep now.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.