What is the difference between harmonics being scaled linearly or logarithmically, I've been looking at the manual section on Spectroblend and Spectromorph, and it states that Spectromorph harmonics are scaled logarithmically while, Spectroblend harmonics are scaled linearly.
I always thought that a harmonic was a whole number integer of the fundamental frequency, and that was it. So does this logarithmic scaling apply to the way the harmonics are spread across the display. Meaning that the linear display spreads them evenly, while the logarithmic display groups more and more harmonics in smaller and smaller spaces, the further you move along the display.
Thanks in advance.
Linear vs Logarithmic?
-
- KVRist
- 78 posts since 12 May, 2005 from Finger Lakes, NY, US
X is the harmonic number; Y is the amplitude. 
hakey wrote:Logarithmic
The Y axis is the harmonic number and the X axis is amplitude.
Most people wouldn't know good music if it bit them in the ass. –FZ
-
- u-he
- 30246 posts since 8 Aug, 2002 from Berlin
Hehe, ok... as for the original question...
In Spectroblend you have one bar for each of 128 harmonics. The distance between each harmonic is always the same frequency, which is the fundamental frequency of the note played by the oscillator. In terms of pitch, the first 8 bars span 4 octaves, the first 64 bars span 6 octaves and the remaining 64 bars span another octave. Thus for Spectroblend the editor is laid out as linear frequency, or logarithmic pitch.
In Spectromorph there are no individual bars. There's a curve that spans the full spectrum of 10 octaves. But this time those 10 octaves of harmonics are evenly distributed across the editor's x-axis. Thus the leftmost 50 or so pixels deal with just the fundamental, the next 50 or so deal with harmonics 1 and 2, then 3-7, and so on until the rightmost 50 or so pixels deal with 512 harmonics at once. So here we got a layout of linear pitch, or exponential frequency.
Latter is a more natural way of dealing with perception of frequency, whereas former is a more technical approach due to the constraints of the method.
In Spectroblend you have one bar for each of 128 harmonics. The distance between each harmonic is always the same frequency, which is the fundamental frequency of the note played by the oscillator. In terms of pitch, the first 8 bars span 4 octaves, the first 64 bars span 6 octaves and the remaining 64 bars span another octave. Thus for Spectroblend the editor is laid out as linear frequency, or logarithmic pitch.
In Spectromorph there are no individual bars. There's a curve that spans the full spectrum of 10 octaves. But this time those 10 octaves of harmonics are evenly distributed across the editor's x-axis. Thus the leftmost 50 or so pixels deal with just the fundamental, the next 50 or so deal with harmonics 1 and 2, then 3-7, and so on until the rightmost 50 or so pixels deal with 512 harmonics at once. So here we got a layout of linear pitch, or exponential frequency.
Latter is a more natural way of dealing with perception of frequency, whereas former is a more technical approach due to the constraints of the method.
-
- KVRist
- Topic Starter
- 48 posts since 20 Sep, 2012
Thanks for the answer Urs, I think you meant to say that the first 8 bars of the SpectroBlend span 3 octaves, right?
I've reread your post quite a few times, and now completely understand how the two methods lay out the harmonics. Thanks.
I've reread your post quite a few times, and now completely understand how the two methods lay out the harmonics. Thanks.