Hornresp

Hornresp Update 3140-130403

Hi Everyone,

CHANGE 1

The horn throat sound pressure and throat particle velocity charts that were recently removed, have now been reinstated. The diaphragm pressure and diaphragm velocity charts introduced in Version 31.30 have been retained, and a new diaphragm acceleration chart has been added.

The throat sound pressure and particle velocity charts can be selected from the Tools menu when the SPL response window is displayed.

The diaphragm pressure, velocity and acceleration charts can be selected from the Tools menu when the diaphragm displacement window is displayed.

Charts showing particle velocity along the horn axis at a given frequency and particle velocity at all frequencies for a given position on the axis, as suggested by Mårten (more10), have not been included because of the amount of work that would have been required to fully integrate them into the program. Also, rather than including a diaphragm "force" chart, a diaphragm acceleration chart has been provided instead.

CHANGE 2

The sampling functionality associated with the impulse spectrogram and directivity polar map has been enhanced. Moving the mouse pointer over the chart window still displays the value at a particular point, but now the point can be set or moved by clicking or dragging the mouse, or by using the Up, Down, Left and Right arrow keys. Pressing the Esc key releases a set point.

Kind regards,

David
 

Attachments

  • Acceln.png
    Acceln.png
    53 KB · Views: 137
  • Sample.png
    Sample.png
    157.2 KB · Views: 138
Well, never tought I'd jump in to the rescue for Mr McBean, but here it goes: :p


Just a quick remark regarding filters:

You can actually export SPL and Impendance data from the resp. HornResp graphs and by slightly modifying the first line of the files, you can create SpeakerWorkshop compatible .spl and .zma files.
If you import this into SpeakerWorkshop you can have a serious swing at it.

I remember I think Patrick Bateman who figured this out, and incidently he was also trying to scale various drivers into a too small Tapped Horn. (Much like I'm trying to figure out the bees knees of passive QB5 alignments.)
I haven't told him yet how to smooth/filter away his Quasi Tapped Horn bump, but I think he dropped the idea. Mr. Bateman, if you read this: try the import+optimization in SW! (And stiffen your cars suspention for the large passive components at your desired power level and low frequencies! :D )

AkAbak should also be able to calculate what most want, but for some reason SpeakerWorkshop and AkAbak give different results with me?


That said, there are various spreadsheets around who simulate the response based on the tables that HornResp already seems to (internally???) generate. I don't think it would be that difficult to drag a filter algorithm over it, but I'm not a programmer nor do I know how HornResp uses that data to calculate the other goodies. If - for example - the throat pressure is lineair to the SPL, these things could be re-calculated after filtering.
Maybe there can be an "externally filtered import" feature with limited graphs? Then you would still hold your ground on the "I'm not programming filters!" thing... ;)

kind regards,

Cordraconis
 
I don´t want to spoil your enthusiasm, but this of course is an old "trick". Exporting the SPL and IMP data from hornresp into other tools is possible for many years, most professionel tools in trandsucer design are capable of importing or exporting data from external tools. Audiocad was one of the first we used, that was some 20 years ago, the good old MLSSA did of course the same, LSP-Cad is doing this since 20 years and the list could go on for long.
Nevertheless, this is a very good idea. The different results are due to the fact that the models to simulate crossovers are not identical in every peace of software and that some import the raw txt-file to SPL but internaly, they put this virtual "speaker" in an enclosure, convert the response to freeair, given a certain measuerment baffle, etc.... One needs to look closely on how the software handles the data...
When dealing with subs, I wouldn´t look tooo close on response, the room does a lot more change to the sound than some minor changes in your cross-over design.

One tool which is very good in doing realistic simulations while designing x-overs is Boxsim. It takes some time to get to know everything possible, but then it gives veery good results. And: it´s free.
 
I don´t want to spoil your enthusiasm, but this of course is an old "trick". Exporting the SPL and IMP data from hornresp into other tools is possible for many years, most professionel tools in trandsucer design are capable of importing or exporting data from external tools.

No problem! I figured it wasn't something new, but I first had to read it somewhere before I realized it. That is why I put it here. :)

Thanks for the tips. So based on your "virtual speaker in an internal enclosure", I can assume AkAbak is the best sim, since it already has the correct enclosure defined in it, or I should find a way to define it as accurately as possible in SpeakerWorkshop.

This should be discussed in my QB5 thread, but hang on for a while as I have no time now to work on it. (In fact I am "working myself in" again after a few months of dust collecting.) My apologies for the short Off Topic.
 
Akabak is able to simulate the horn itself, so it gives you the freedom to simulate what the mechanical parameters in conjunction to the electrical side of the Xover does. Other tools are only capable to import the SPL/IMP Data - if you want to change your tapped horn, you have to go back to hornresp. If you are in Akabak, you don´t.
When importing it in Akabak, you export the whole project, not just the SPL/IMP Data. Akabak does it´s own hornsimulation then. Only in cases where Akabak is not able to do what hornresp does, you are limited to the "import SPL/IMP" way of doing things. "Exotic" hornflares would be one example of a case where akabak is not capable of doing the same as hornresp. but when dealing with subs, there is no problem.

I would go for Akabak, if you have the time. It ist capable of so much more and on learns a lot when dealing with it. Try playing around with things like: "my Box is standing on a floor with a wall 1m behind the box and I am standing 3m away from it with my ears in 1,65cm height". :)

Speakerworkshop probably is easier for a quick shot of trying to do a x-over.... (i´d still go for an active x-over instead of a passive one :) ) .
 
Hello David,

About introducing filters in Hornresp at least we should discuss about some requirements and priorities.

1) I think a high pass filter is sufficient.
2) Butterworth filters from 1 to 3rd order and LR filters LR2 and LR4 should be considered.
3) The output of the HP filter could come in replacment of Eg assuming Eg becomes a table (versus frequency) of complex numbers.

Then for the Hornresp code is it possible to replace Eg by such a table I don't know.

Best regards from Paris, France

Jean-Michel Le Cléac'h
 
But, there are other forum members who use Hornresp to actually build speakers, for me it is just an area of interest and learning, so I'd rather you take input from e.g.: mwmkravchenko, jbell or NEO Dan (I know, there is a long list) than from me.

It's a very much chicken then egg scenario.

Being an egg is easy once there is a chicken around.

Without Hornresp we would be forced to use the evil AkaBak! Just kidding.

But without us David would get bored with programming. So it is a mutually inclusive feedback loop.

We just need to feed the back.




Oh yeah ...

I'm finally number one in a list!

Whoo hoooo!
 
David is old enough to speak for himself, but didn't he clearly say that he doesn't want to include filtering in hornresp? I think we should respect that.

Or did I miss something and did David open the door to implementing filters ?

Hello Sabbelbacke,

My message was intended to propose to David if needed, my help to integrate such filters in a same way David and me used to collaborate to introduce the "Impulse Response" module and the "Spectrogram" module, that surely many like to see now in Hornresp.

Best regards from Paris, France

Jean-Michel Le Cléac'h
 
Don´t get me wrong, I´d love filters in hornresp, too :) I just got the impression that people ask a lot about something where david said "no" already.

I also know that once you start to put passive filters in hornresp, there is no "this is enough"...

If hornresp were to get filter capability, I´d strongly recommend to implement something like in BoxSim, so ALL possibilities are there (important impedance corrections, notches, etc..). Maybe the Boxsim guys want to share their work?
 
About introducing filters in Hornresp at least we should discuss about some requirements and priorities.

Hi Everyone,

With regard to Jean-Michel's comment, would it be sufficient to be able to specify a passive filter in accordance with the attached network template, with only those components actually required being included?

If so, then a real time 'Filter Wizard' could be quite a nice thing to have... :).

R = resistor, C =capacitor, L = inductor.

Kind regards,

David
 

Attachments

  • Filter.png
    Filter.png
    5.2 KB · Views: 158
My message was intended to propose to David if needed, my help to integrate such filters in a same way David and me used to collaborate to introduce the "Impulse Response" module and the "Spectrogram" module, that surely many like to see now in Hornresp.

Hi Jean-Michel,

Thanks for your kind offer, but I should be okay. The circumstances are a bit different this time - I better understand the underlying theory :).

Kind regards,

David
 
Hi Everyone,


If so, then a real time 'Filter Wizard' could be quite a nice thing to have... :).

R = resistor, C =capacitor, L = inductor.

Hello David,

If we limit the choice to a high pass filter, with a few more components than in your schematics we can also provide impedance equalization and HF uplifting.

(I put 2 RLC impedance equalizors but one could eventually be enough, even if with horns there is often more than 2 peaks of impedance).

Best regards from Paris, France


Jean-Michel Le Cléac'h
 

Attachments

  • filter_2.png
    filter_2.png
    8.7 KB · Views: 154
I love the filtering in Boxsim, it looks like this:
Weicheneditor1.PNG

It is possible to put in whatever one wants, no limits. I could ask Uwe, the author, if he would share his modul of the function...
Another idea in the same direction would be to take some parts of SPICE and integrate them.
SPICE - Wikipedia, the free encyclopedia

Either way, both active and passive filtering are interessting, if Ih had to choce, I´d take the passive one (after all, aktive Filtering is "boring" in this regard, since there is no dependancy on the impedance of the speaker system... The real interesting things happen when you take the real impedance of the speaker into account).

At least, we´d need:

- Coils with editable R and mH
- Capacitors
- Resistors
- Shunts

The passive filtering should be right before "the box", not behind the amp, in front of the cable:

amp - cable - x-over - speaker/box

AjHorn does it like this:

DateneingabeFW1.JPG


nod as flexible as BoxSim, but sufficent.