Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Software Tools SPICE, PCB CAD, speaker design and measurement software, calculators

VituixCAD
VituixCAD
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 3rd May 2019, 07:55 PM   #1181
mbrennwa is offline mbrennwa  Switzerland
diyAudio Member
 
mbrennwa's Avatar
 
Join Date: Mar 2005
Location: Zurich, Switzerland
VituixCAD
Quote:
Originally Posted by kimmosto View Post
Easier for me to do everything for 180 deg twice in this big puzzle and just divide two weighting factors by 2.
Just to be sure I get this right:
- your coefficient for 0/0 is (approximately) 0.0004758 / 2 = 0.0002379
- your coefficient for 180/180 is (approximately) 0.0004758 / 2 / 2 = 0.00011895

This means that the sum in your code includes the 0/0 point twice, and the 180/180 point is included four times for some reason.

Correct?
__________________
------
Group buy for augerpro waveguides for Scan Speak tweeters -- elliptical version -- circular version
  Reply With Quote
Old 3rd May 2019, 08:35 PM   #1182
kimmosto is offline kimmosto  Finland
diyAudio Member
 
kimmosto's Avatar
 
Join Date: Oct 2005
Location: Kuopio
Correct.
  Reply With Quote
Old 3rd May 2019, 10:03 PM   #1183
mbrennwa is offline mbrennwa  Switzerland
diyAudio Member
 
mbrennwa's Avatar
 
Join Date: Mar 2005
Location: Zurich, Switzerland
VituixCAD
Ok, so I tried the Tylka calculation using your coefficient values from post 1177. To this end I first tried with 0.000237... for the last coefficient for the 180 point, and then I also repeated with 0.000118... . Both results were very close to the one with the original Tylka calculation (less than 0.1 dB difference everywhere). This is not surprising, as your coefficient values are essentially identical to the Tylka values.

This means that although my Tylka code and Vituix use the same coefficients, the result is not the same. Therefore Vituix and my Tylka code therefore don't implement the same calculation. I attached the formula I used for the Tylka calculation. How is this different from the Vituix calculation?

Note: the w_n correspond to the coefficients as discussed in post 1177.
Attached Images
File Type: png power_response_formula.png (199.4 KB, 163 views)
__________________
------
Group buy for augerpro waveguides for Scan Speak tweeters -- elliptical version -- circular version
  Reply With Quote
Old 4th May 2019, 05:24 AM   #1184
kimmosto is offline kimmosto  Finland
diyAudio Member
 
kimmosto's Avatar
 
Join Date: Oct 2005
Location: Kuopio
Quote:
Originally Posted by mbrennwa View Post
How is this different from the Vituix calculation?
No difference - except 10log(4pi) offset in total.
  Reply With Quote
Old 4th May 2019, 05:35 AM   #1185
kimmosto is offline kimmosto  Finland
diyAudio Member
 
kimmosto's Avatar
 
Join Date: Oct 2005
Location: Kuopio
PS1. Just want to make sure that you calculate VCAD results with the main program - not with Calculator tool which uses the simplest weighting with sin() function i.e. 0 deg and 180 are totally ignored.

PS2. The latest built of 2.0.15.3 (downloadable since yesterday evening) uses Archimedes' hat-box theorem cos(a)-cos(b) for area calculation. Difference to old versions such as obsolete 1.1 is insignificant, visually undetectable with default dB scale.
  Reply With Quote
Old 4th May 2019, 06:21 AM   #1186
kimmosto is offline kimmosto  Finland
diyAudio Member
 
kimmosto's Avatar
 
Join Date: Oct 2005
Location: Kuopio
Difference between simple sin() and area (or integral of sin) weighting is quite small if angle interval is only 5 deg:
Click the image to open in full size.

In practice, amount of off-axis angle measurements is the most significant factor in addition to limit of dual planes only. Curve fitting would work better than weighting by sector factors but that would reduce calculation performance.
  Reply With Quote
Old 4th May 2019, 05:33 PM   #1187
mbrennwa is offline mbrennwa  Switzerland
diyAudio Member
 
mbrennwa's Avatar
 
Join Date: Mar 2005
Location: Zurich, Switzerland
VituixCAD
Quote:
Originally Posted by kimmosto View Post
No difference - except 10log(4pi) offset in total.
Alright, so now at least I know the formula used in Vituix. So the "How does it work?" question is now settled.

With respect to the question "Is the result correct?", here are some updates:

Quote:
Originally Posted by kimmosto View Post
PS1. Just want to make sure that you calculate VCAD results with the main program - not with Calculator tool
Yes, I used the main program.

There is some progress on my side. In my code, I screwed up the allocation of the data files to the corresponding H_m,n terms where the angles wrapped around the hemispheres (i.e., at the rear of the speaker). The Vituix convention is slightly different to what Tylka uses, and I got confused by this. While the w_n coefficients were right all the time, they were not always multiplied with the correct SPL curves (H_m,n terms). I fixed this now.

I have also tested my code by artificially setting all H_m,n to zero with the exception of one (n=k), which was set to 1:
- H_m,n = 1 for n = k
- H_m,n = 0 for all other n
By looking at the summation (see post 1183) you will find that whatever the choice of k is, the resulting power response (before conversion to dB) will be equal to 2 x w_k. Implementing this in the code is easy, and I have done this test for all possible k values (with a 5 angle step, there are 72 possibilities). The code always gave the correct result. I am therefore quite confident that it works correctly now.

The power response curve from the fixed code is now closer to the Vituix curve than before, but there is still some discrepancy of almost 1.5 dB above 10 kHz (see attached plot). I don't know if this is just a difference related to the numerical techniques, or if it's something more conceptual.
Attached Images
File Type: png power_response_Tylka_vs_Vituix_20190504.png (32.9 KB, 131 views)
__________________
------
Group buy for augerpro waveguides for Scan Speak tweeters -- elliptical version -- circular version
  Reply With Quote
Old 4th May 2019, 06:26 PM   #1188
kimmosto is offline kimmosto  Finland
diyAudio Member
 
kimmosto's Avatar
 
Join Date: Oct 2005
Location: Kuopio
Quote:
Originally Posted by mbrennwa View Post
Alright, so now at least I know the formula used in Vituix. So the "How does it work?" question is now settled.
That basic part was already explained 3 days ago and exception logic for weighting factors 2 days ago. But I guess Greek hieroglyphs suit better than my bad English flavored with minor amnesia. I'm done with this item.
  Reply With Quote
Old 5th May 2019, 05:34 AM   #1189
kimmosto is offline kimmosto  Finland
diyAudio Member
 
kimmosto's Avatar
 
Join Date: Oct 2005
Location: Kuopio
Rev. 2.0.15.4 (2019-05-05)

Diffraction
* Effect of Floor and Wall boundary reflections reprogrammed.
New calculation method enables easy adding of new boundaries such as front wall and ceiling, which might be added later.

Enclosure
* Added name and Vd[cm3] read-only text boxes to Driver configuration group. Name saved to user settings and enclosure project file.
* Added name and Vd[cm3] read-only text boxes to Passive group. Name saved to user settings and enclosure project file.
  Reply With Quote
Old 5th May 2019, 02:43 PM   #1190
nc535 is offline nc535
diyAudio Member
 
Join Date: Apr 2011
Hi Kimmosto:
Regarding front wall and ceiling reflections, I would love to see them. I immediately tried out the new revision but the changes apparently aren't apparent to the user.

As you may recall I tried simulating them for a line array via image method in the main program but there were multiple issues that you noted causing me to abandon that approach. But floor and ceiling reflections are critical to line array performance in a home environment so I have gone as far as I can with them in the current program and would appreciate your help going further. The goals are not just to understand line array behavior but to be able to develop shading coefficients without endless cut and try or audible in-living-room tweaking.

Any improvement to the accuracy of these sims is welcome but especially if it would shed light on why the comb filtering seen in simulations of line arrays is usually worse than seen in measurements.

Looking at geometry of a 2m tall line array, one needs high resolution vertical directivity only within 0 to 27 degrees for the main array and its ground reflection or out to 37 degrees if including a ceiling reflection. Diffraction tool only goes down to 5 degrees but I am able to import piston directivity from HornResp down to 1 degree. It would be nice if diffraction tool would do this for me as that takes some time. Its not practical for me to do 1 degree measurements but next time out I will get down to 5 degrees in this critical region. I wonder if there is a way to interpolate between 5 or 10 degree measurement data in some external tool then import or if that feature could be added to calculator.

Now I'm adding ground reflection in main program by the simple expedient of adding 2x number of drivers. This puts the vertical axis on the floor, which is not ideal, but it does give me answers. I can't add ceiling reflection this way because baffle height seems to be limited to 2m. If you would raise that limit (3m would be sufficient) then I could add the ceiling reflection. Simulation performance is not yet an issue. There is a lag to recompute but only a few seconds.

Looking at reflections in diffraction tool is certainly easier. In the main program there is more flexibility to add frequency dependent reflection coefficients in an attempt to correlate with measurements. I will shortly be able to do measurements of a full line array so this will be interesting.

Thanks,
Jack
__________________
My Synergy Corner Horns and Bass Bins
http://www.diyaudio.com/forums/multi...bass-bins.html
  Reply With Quote

Reply


VituixCADHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


New To Site? Need Help?

All times are GMT. The time now is 02:39 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Copyright ©1999-2020 diyAudio
Wiki