http://ift.tt/2bnxouK
I gave it another go, this time with FFT (for analysis), which figures out the predominant frequency (hopefully) and saturates according to the shape of a parametric EQ. The EQ has unlimited bands but it slows down after a while because as it draws the graph it is also storing the magnitudes in a table. You can adjust the 'plotsize' slider to give it more 'resolution', or lower it to allow for more bands, or simply lower it while adding bands for a kind of draft mode, and then crank it up when you're finished. There is a global drive slider as well in case you add too much gain and things start crackling.
Frequencies/gain may not correspond to real life/exact values because I'm just kind of guessing as I go, but it should work as I intended, as an exciter without the need for crossovers
Edit: the parametric eq interface is just out of simplicity, you could probably implement it in some freeform way with bezier curves/drawing interface but I'm not going to mess with it. I might try to figure something out with the phase plot vs. FFT phase which in that case would require it to use a parametric/biquad filter gui.
Edit2: also don't use highpass filters as far as I can tell, lowpass seems ok
Edit3: the plot will look funny because it is LINEAR not logarithmic in this case, because the lookup table has to be linear in order to match with FFT
Edit4: mousewheel up adds/moves to the next band, failed to mention that. Also left click moves gain/freq and right mouse click and drag changes bandwidth
I gave it another go, this time with FFT (for analysis), which figures out the predominant frequency (hopefully) and saturates according to the shape of a parametric EQ. The EQ has unlimited bands but it slows down after a while because as it draws the graph it is also storing the magnitudes in a table. You can adjust the 'plotsize' slider to give it more 'resolution', or lower it to allow for more bands, or simply lower it while adding bands for a kind of draft mode, and then crank it up when you're finished. There is a global drive slider as well in case you add too much gain and things start crackling.
Frequencies/gain may not correspond to real life/exact values because I'm just kind of guessing as I go, but it should work as I intended, as an exciter without the need for crossovers
Edit: the parametric eq interface is just out of simplicity, you could probably implement it in some freeform way with bezier curves/drawing interface but I'm not going to mess with it. I might try to figure something out with the phase plot vs. FFT phase which in that case would require it to use a parametric/biquad filter gui.
Edit2: also don't use highpass filters as far as I can tell, lowpass seems ok
Edit3: the plot will look funny because it is LINEAR not logarithmic in this case, because the lookup table has to be linear in order to match with FFT
Edit4: mousewheel up adds/moves to the next band, failed to mention that. Also left click moves gain/freq and right mouse click and drag changes bandwidth
parametric exciter again (FFT version)
Aucun commentaire:
Enregistrer un commentaire