Documentations of TECHMANIA projects, hosted in a repo so they can be localized.
Applies to version: 1.4
TECHMANIA defines and handles volume and pan differently from whatever game using .pt, so these parameters are not directly convertible. I have experimented and found formulas that I think sound close enough to the original, but you can customize the formulas to your liking if needed.
The input is velocity in .pt, ranging from 0 to 127. Output is volume in .tech, ranging from 0 to 1. For convenience we first normalize input, so it ranges from 0 to 1:
normalized = input / 127
Then, choose 1 of the 2 following formulas:
output = normalized ^ exponentoutput = log(base, normalized) + 1 (output is clamped between 0 and 1)The default is exponential with exponent equal to 1.5.
The input is pan in .pt, ranging from 0 to 127. Output is pan in .tech, ranging from -1 to 1. For convenience we first normalize input, so it ranges from 0 to 1:
normalized = 1 - input / 64normalized = (input - 64) / 63Then, choose 1 of the 2 following formulas:
output = -normalized ^ exponentoutput = normalized ^ exponentoutput = -log(base, normalized) - 1 (output is clamped between -1 and 0)output = log(base, normalized) + 1 (output is clamped between 0 and 1)The default is exponential with exponent equal to 0.25.
.pt has volume notes, which modifies the velocity of all notes after it in the same track. There is no equivalent in TECHMANIA, so by default, the converter applies the same formula to both volume and velocity, then multiplies the results to get the final output volume.
However, if you need to modify patterns after conversion, especially move notes between tracks with different volumes, it may be better to tell the converter to ignore volume notes, in order to treat each track equally. Note that music volume from TECHMANIA’s options menu applies to all notes in hidden lanes.
You can use WolframAlpha to visualize exponential and logarithmic curves. Type into WolframAlpha:
plot y=x^0.25, 0<=x<=1 (replace 0.25 with your exponential)plot y=min(max(log(10, x)+1, 0), 1), 0<=x<=1 (replace 10 with your base)Scroll speed determines:
Therefore it’s critical for the converter to find the correct scroll speed for each pattern. The converter will search the following 3 sources for a scroll speed and use the first one found: