Frequency bands

[mojoyup1528]

Allow for tighter/narrowbanded frequency bands, such as higher sampling from 1-20kHz. Possible to accommodate each frequency as a band? Requesting precise measurements :) Will financially support this overhaul if requested!

[mojoyup1528]

I found that this may be accomplished in the appconst.xml file under currentRelease if that helps.

[Picaut]

We are testing a simple solution in order to consider narrow bands for you own use. We come back to you as soon as we have found something.

[mojoyup1528]

Thank you so much! I really want to use this software, it fits my needs! I hope to hear from you soon and let me know if I can help. I've been working with acoustics for 15 years :)

On Tue, Jul 9, 2019, 4:04 AM PICAUT Judicael [email protected] wrote:

We are testing a simple solution in order to consider narrow bands for you own use. We come back to you as soon as we have found something.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Ifsttar/I-Simpa/issues/231?email_source=notifications&email_token=AKKRJZESWCAXDH2KVQZOFB3P6RBAPA5CNFSM4H6WW3FKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZPORWY#issuecomment-509536475, or mute the thread https://github.com/notifications/unsubscribe-auth/AKKRJZD4U5GDVYITIKXUMCLP6RBAPANCNFSM4H6WW3FA .

[nicolas-f]

Hi

I've tested.

Good news is, there is few things to do:

Set good number of frequencies and add frequency bands and A weighting here: https://github.com/Ifsttar/I-Simpa/blob/v1.3.3/src/isimpa/data_manager/appconfig.cpp#L126

Update internal database of I-Simpa here or do not use predefined spectrum in your project: https://github.com/Ifsttar/I-Simpa/blob/v1.3.3/currentRelease/appconst.xml

You want to use your own computation code to be able to compute under 50 Hz ?

[mojoyup1528]

Awesome! Thank you, I will check that out today :) you are amazing :)

Yes, I would like to try to calculate below 50Hz, even though I know that is an LF cutoff standard for concert loudspeakers. I want to try to use the same computations if possible, I just am not sure which formulas the apps uses yet.

I would assume that the computation takes the name of the frequency array index as a variable rather than a constant (i.e. 50, 63, 80, etc), which would be good! If not, then I will have to see what I can accomplish otherwise which is ok :)

Thank you so much for the help, and I hope to keep in contact :) If I find things useable in the near future, may I pay for your time and effort?

Thank you again!

On Tue, Jul 9, 2019, 5:02 AM nicolas-f [email protected] wrote:

Hi

I've tested.

Good news is, there is few things to do:

Set good number of frequencies and add frequency bands and A weighting here:

https://github.com/Ifsttar/I-Simpa/blob/v1.3.3/src/isimpa/data_manager/appconfig.cpp#L126

Update internal database of I-Simpa here or do not use predefined spectrum in your project: https://github.com/Ifsttar/I-Simpa/blob/v1.3.3/currentRelease/appconst.xml

You want to use your own computation code to be able to compute under 50 Hz ?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Ifsttar/I-Simpa/issues/231?email_source=notifications&email_token=AKKRJZFAUO3YOOMCI6S4LUTP6RH3LA5CNFSM4H6WW3FKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZPTVZY#issuecomment-509557479, or mute the thread https://github.com/notifications/unsubscribe-auth/AKKRJZBAUJRHSJ52KT3OTCDP6RH3LANCNFSM4H6WW3FA .

[nicolas-f]

Yes computation codes read the frequency values.

The SPPS code does not model all acoustic emission and propagation conditions and returns a limited number of acoustic indicators. In particular, it is important to specify that the SPPS code does not take into account:

• Interference phenomena : by construction, given the "high frequency" hypothesis, interference phenomena are not modelled.
• No diffraction phenomenon (by a panel, by an edge... . ) is modelled.

For such low frequency results, a finite element method should be appropriate.

Maybe take a look at http://acousto.sourceforge.net

[mojoyup1528]

Good stuff :) I def noticed the diffraction issue right from the start but I may also be able to implement that if I use the same code you have, which based on the "for" loops, looks like something I can def work with :)

That may be a defining factor for me too:) I am excited to see what I can do!

On Tue, Jul 9, 2019, 10:22 AM nicolas-f [email protected] wrote:

Yes computation codes read the frequency values.

The SPPS code does not model all acoustic emission and propagation conditions and returns a limited number of acoustic indicators. In particular, it is important to specify that the SPPS code does not take into account:

• Interference phenomena : by construction, given the "high frequency" hypothesis, interference phenomena are not modelled.
• No diffraction diffraction phenomenon (by a panel, by an edge... . ) is modelled.

For such low frequency results, a finite element method should be appropriate.

Maybe take a look at http://acousto.sourceforge.net

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Ifsttar/I-Simpa/issues/231?email_source=notifications&email_token=AKKRJZHKA3Y3MHHTS3ASXOTP6SNKBA5CNFSM4H6WW3FKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZQNQDA#issuecomment-509663244, or mute the thread https://github.com/notifications/unsubscribe-auth/AKKRJZAIKAWKEAD7LGTDEZLP6SNKBANCNFSM4H6WW3FA .

[mojoyup1528]

I'll check out acouSTO also :) have yet to run across it. Thank you again!

On Tue, Jul 9, 2019, 10:28 AM JDesignz [email protected] wrote:

Good stuff :) I def noticed the diffraction issue right from the start but I may also be able to implement that if I use the same code you have, which based on the "for" loops, looks like something I can def work with :)

That may be a defining factor for me too:) I am excited to see what I can do!

On Tue, Jul 9, 2019, 10:22 AM nicolas-f [email protected] wrote:

Yes computation codes read the frequency values.

The SPPS code does not model all acoustic emission and propagation conditions and returns a limited number of acoustic indicators. In particular, it is important to specify that the SPPS code does not take into account:

• Interference phenomena : by construction, given the "high frequency" hypothesis, interference phenomena are not modelled.
• No diffraction diffraction phenomenon (by a panel, by an edge... . ) is modelled.

For such low frequency results, a finite element method should be appropriate.

Maybe take a look at http://acousto.sourceforge.net

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Ifsttar/I-Simpa/issues/231?email_source=notifications&email_token=AKKRJZHKA3Y3MHHTS3ASXOTP6SNKBA5CNFSM4H6WW3FKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZQNQDA#issuecomment-509663244, or mute the thread https://github.com/notifications/unsubscribe-auth/AKKRJZAIKAWKEAD7LGTDEZLP6SNKBANCNFSM4H6WW3FA .

[mojoyup1528]

I am going to use visual studio 2019 for recompiling the 1.3.3 with the changes to the cpp and xml and let you know if they work and it executes on windows 10 platform :) Thank you for the help so far! Very appreciated, no idea how much!

[nicolas-f]

You can also use AppVeyor it will do the compilation for you. You just have to fork this repository, apply the modifications and trigger AppVeyor build.

It will give you standalone zip and setup file.

[mojoyup1528]

Thank you :) That may be what Ill do because boost 1.6.6 build even using 7zip took forever :P

[mojoyup1528]

UPDATE: I ran a test build and it was satisfactory on the current 1.3.4. Now, I am going to try to modify where I needed to see how it effects the propagation and calculations for all particles, rays, etc from the original and compare time steps, etc.

Quick Question, where I make the modifications, will it be to this repo since I have it forked or is it through the downloaded folder of the repo for AppVeyor? Thank you for the awesome help so far and I will be committed a lot to making this work for the initial request as well as implementing diffraction when possible to help conclude the less common acoustical parameters that programs need for results :)

I am grateful to have found this and you guys!

EDIT UPDATE: I ebelieve I figured it out an ddid a pull request. I am also editing the appconfig.xml at the moment to reflect an idspectre "8" and name as ELC_Full with the passband configuration for the cpp file. Let me know if I need to adjust anything as far as my workflow :)

[nicolas-f]

Hi,

You don't have to make a pull request. You can activate AppVeyor by login in AppVeyor website and pointing to your fork at https://github.com/mojoyup1528/I-Simpa

[mojoyup1528]

Thank you :) yeah, I kind of figured that out afterwards lol. But all set now :) I got the app to recognize the response range using a new user spectrum I just have to change the materials and also an error for source/ recover spectrum missing so I'll be working on that tomorrow :)

On Thu, Jul 11, 2019, 2:56 AM nicolas-f [email protected] wrote:

Hi,

You don't have to make a pull request. You can activate AppVeyor by login in AppVeyor website and pointing to your fork at https://github.com/mojoyup1528/I-Simpa

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Ifsttar/I-Simpa/issues/231?email_source=notifications&email_token=AKKRJZHKNDKL7QDIWJPCOMLP63KSJA5CNFSM4H6WW3FKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODZVWSTA#issuecomment-510355788, or mute the thread https://github.com/notifications/unsubscribe-auth/AKKRJZD6PTUSWSOKB7GLTL3P63KSJANCNFSM4H6WW3FA .

[mojoyup1528]

I've been playing around with the responses and just added a test material for all freq and its interesting :) I am getting a few errors but they seem to not phase the calcs yet. They exist in one file as scripted warnings, but not sure how to rid them yet. Its the errors saying source and receivers spectrums dont exists, even if I select one.

It also just recently started crashing when going into the receiver background noise profile, but Im quite sure its an xml typo on my end, or a missing connection between spectrum values.

Other than that Ive just been playing with it and it works, just not quite as expected for resonances but close! Geting usable info for calibration testing :) very fun so far :P

Might close this soon if I can fix the crash and spectrum warnings :) Then I can test more and soon after go into the calc files to check diffraction coding if possible. Great stuff so far!

[mojoyup1528]

So, I concluded the input for the full frequency band, and the results are satisfactory so far :) I did a comparison of the LF response from the spps calculations with a popular response curve calculator software for the same enclosure test/room resonance test. They both show similar peaks, and nulls in the response as well as resonances above the focus area.

For example, I purposely made the room a certain size that had a high focus at 60Hz, 125Hz, 160Hz, and 400Hz. In the curves attached (which came from your software), you can see the 60Hz, 160Hz, and 2nd order of 400Hz(800Hz). These curves are just zoomed in from the same full range curve single calculation. Also, the curve is very similar to the other software, even the gain by about 1-3dB.

I am happy so far with the results. Going to test more, but I will work on diffraction next, and then see about calculating impedance curves as well if that is ok :)

focused Response psychoacoustic: https://imgur.com/oVh394x lower Response 1: https://imgur.com/By5gxzl lower Response 2: https://imgur.com/ja73UOu upper Response: https://imgur.com/dUnfaRD

default Response: https://imgur.com/1ouLlz0

[mojoyup1528]

UPDATE: So, I ran more tests on different (pretty much opposing) resonances and I am noticing something I need to check, the dB values, where I might have the cumulative values wrong. I see patterns no matter the shape, size/volume, positioning of the receiver etc which Id like to change. So, a step back but worth it for me :) Will update again shortly! Hope this becomes helpful one day to someone.

[mojoyup1528]

UPDATE: So, I noticed even using the default calcs and responses I am getting results that somewhat seam to not differ for each different type of mesh imported. Here is my flow:

Create an stl file through sketchup. Make sure face normals are correct and export to desktop. Import into ISimpa as repair model(default import) Create a sound source, giving it a spectrum and 80 for the Lw. Make a group for the material and use 0% absorption (or default has 0 also). Create a receiver and not change anything there. Position source and receiver in solid room mesh.

NOTE: I have not created surface receivers for this test, though I know they might be different, just focusing on punctual receivers response curve.

Calculate the SPPS (mainly so I can see if particle timing is correct with default specs). I also do change some parameters such as particles for display to 1500. And leave calc particles to 150000.

The response curve for any mesh with these parameters are the same. They are barely different than +/- 0.5-1.0dB no matter the frequency. This, I know through real time testing and acoustical calculations cannot be true because of resonances and gains from combined in phase particle collisions.

I do know that these calculations are only for impulses but maybe a continous calcualtion could improve realtime results because simulation time does not change the calcs obviously.

Any thoughts?

[nicolas-f]

Hi,

The issue here is that you must set a realistic material with absorption coefficient > 0 .

[mojoyup1528]

Ok Thank you. I will do so with a 0.4 and see how that works and then change to different Abscos for each frequency afterwards :) You are very very helpful I appreciate it very much

On Wed, Jul 17, 2019 at 4:40 AM nicolas-f [email protected] wrote:

Hi,

The issue here is that you must set a realistic material with absorption coefficient > 0 .

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Ifsttar/I-Simpa/issues/231?email_source=notifications&email_token=AKKRJZBWKRPJ67ACC6BQWMDP73LJNA5CNFSM4H6WW3FKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD2DPIDI#issuecomment-512160781, or mute the thread https://github.com/notifications/unsubscribe-auth/AKKRJZFCU2V32G7X6UOHSYLP73LJNANCNFSM4H6WW3FA .

-- Jason K. Evans, Entrepreneur

• GKADT Design Leader :* www.gorillakdesigns.com http://www.gorillakdesigns.com*

• *Developer, Android Games-*MOJOGO: https://play.google.com/store/apps/details?id=com.MojoGo.blockDODGE https://play.google.com/store/apps/details?id=com.MojoGo.blockDODGE

• *Artist, Graphite professional: *https://www.facebook.com/dvisionist023/?ref=settings https://www.facebook.com/dvisionist023/?ref=settings

[mojoyup1528]

I haven't had time yet this season to do any work on my changes, but I have an interest of doing so now with some time. Will let you know if I find anything useful for updates.