Guitar is an Ibanez RG350EX with a Seymour Duncan bridge pickup instead of the original one.
Strings are Ibanez super light gauge nickel wound (.009/.011/.016/.024/.032/.042).
All strings are tuned one half tone down (e.g. e strings on d#).
A capo was used to hold down the strings on the respective frets.
All non-picked strings and the springs of the tremolo were dampened with some tissue, making the recordings as dry as possible.
The guitar was recorded directly using a Behringer UPhoria UMC202 HD device.
Sampling rate is: 44.1kHz, 24bit
The sound font contains two presets:
000:000: Ibanez neck D#2
using the neck pickup only
strings were picked right above the neck pickup
recommended velocity range: 75-95
000:001: Ibanez bridge D#2
using the bridge pickup only
strings were picked right above the middle pickup
recommended velocity range: 70-90
I did not record any samples using the middle pickup
Ardour: for recording (one wav file per tone, each containing some samples with a different picking strength)
sox and some bash script magic to:
cut out the samples
apply a 2ms fade-in using a half sine wave
apply a 1s linear fade-out
detect the peak level and rename the sample accordingly
Polyphone: to create this sound font
How to use properly:
Try to remain within the recommended velocity range, because in this area the sample density is high.
Use a little velocity randomization so a different sample will be used for the same tone most likely. This will make your track sound a bit more realistic.
Timing randomization should not be required. Due to the fully automated cutting of the samples during recording/preparation there already is some randomness in the attack delay.
Use a separate FluidSynth instance (e.g.) where you can boost up the gain/volume without boosting it also for your other MIDI tracks. The loudness of this soundfont is quite low compared to standard General MIDI soundfonts. This is because making it equally loud just was no priority for me.
Apply plugins (e.g. Guitarix, Calf Vintage Delay, Calf Equalizer, etc.) to shape the sound like you want. Remember that the output of the soundfont is just the dry notes directly recorded from the pickups.