Hornresp

Shrub0, in reply to your post #11313 from simulations I have run I would say yes you are correct, a model having a longer impulse spectrogram also has a longer group delay - shorter is better. The scales on your 2 screenshots are different, so take that into account. The left one is complete at the red/yellow border around 20milliseconds, while the right one is complete around 25 milliseconds. For the light/dark blue band the left model is done by about 30ms or under, while the right model is done at/past 40ms. The right hand model also contains a group of sharp spikes at the top which from other user comments I understand to be HOM's(higher order modes), undesirable. I would say the left hand model is cleaner, settles quicker, and is more desirable of the two.
 
I think this is a sort of wavelet analysis, there's a thread on that here: Help on making MULTIRESOLUTION WAVELET ANALYSIS available ? and an interesting piece of software for Gnu Octave was developed by DIYaudio member Elias. His website may have some useful info: http://elias.altervista.org/html/Easy_Start_on_Wavelets.html

Thanks Bjørn.

There are still things that I do not fully understand. And this will help. One of the things I enjoy most about working in this field. The more you learn, the less you really know.
 
Speaking of Akabak, I take it you didn't care for the different TH functions I suggested?

I must confess that I didn't pay particularly close attention to them at the time :).

I had assumed (mistakenly it seems) that you were just "spitballing" with Booger weldz.

Needless to say, AkAbak (not Obi-Wan Kenobi or Hornresp) is your only hope...
 
Just to inform, as i was trying re-entrant stub to damp port resonnance, i played with akabak. A port with re-entrant stub in the form of a deviation from of the center segment of a 3 segment port works fine with akabak. But without hornresp tweaking power, i don't manage to see if it has advantage from a 2 stub port not connected each others. I cannot optimise all the parameter intuitively as hornresp do. For damping, i found a old thread in wich X explain to use waveguide instead of duct, and use acouResistance from RA=5000 to 10000.
 

Attachments

  • pouet.PNG
    pouet.PNG
    48.6 KB · Views: 152