Does an 'old piano' sampleset exist??
-
- KVRist
- 467 posts since 21 Feb, 2005
Have you tried this one? http://audioshots.com/auditorium/viewtopic.php?t=369
Sounds just like Granny's old Yamaha upright.
Sounds just like Granny's old Yamaha upright.
-
- KVRist
- Topic Starter
- 156 posts since 13 Dec, 2004 from houston texas usa
I'll hold out for 93143's sample set. How's that coming?
-
- KVRist
- 261 posts since 19 Jan, 2005
I've been a bit busy lately, and that's not going to change until late November. However, I am proceeding on sample preparation; I recently spliced out a cable bump (hopefully the last one of those) from the pianissimo low A.
I have a suspicion that this is going to be too big for sfz. Since the only other sampler I own is Emulator X, I'll probably have to map and program it in that and have someone else convert it to Kontakt, Giga, etc. A smaller sf2 without the pedal crossfade (assuming I can get it working in the first place) might be good too.
Gigastudio can do ramp crossfades on pedal trigger, right?
I have a suspicion that this is going to be too big for sfz. Since the only other sampler I own is Emulator X, I'll probably have to map and program it in that and have someone else convert it to Kontakt, Giga, etc. A smaller sf2 without the pedal crossfade (assuming I can get it working in the first place) might be good too.
Gigastudio can do ramp crossfades on pedal trigger, right?
-
- KVRAF
- 6596 posts since 21 Jun, 2004 from Secret Underground Hideout
if i use soundfont enhancer, i can get close. the sf is missing twangegarrard wrote:Have you tried this one? http://audioshots.com/auditorium/viewtopic.php?t=369
Sounds just like Granny's old Yamaha upright.
-
- KVRist
- 261 posts since 19 Jan, 2005
Since I've got both pedal-up and pedal-down samples, and I want to crossfade between them, the samples for a single zone should be more or less in phase with one another. I had a couple of hours last night, so I threw together a phase-alignment code in MATLAB.
It takes two files and checks to see how many leading zeroes are necessary on each file to bring it into phase with the other one (that is, to minimize the sum of the absolute per-sample differences). It then uses the smaller of the two calculated pad lengths on the corresponding file.
It's still an ugly monolithic prototype, but it gives results. Here's a segment of the A#4 ff (blue=no pedal, red=sustain) before:
and after:
Just so you know I'm actually working on it...
It takes two files and checks to see how many leading zeroes are necessary on each file to bring it into phase with the other one (that is, to minimize the sum of the absolute per-sample differences). It then uses the smaller of the two calculated pad lengths on the corresponding file.
It's still an ugly monolithic prototype, but it gives results. Here's a segment of the A#4 ff (blue=no pedal, red=sustain) before:
and after:
Just so you know I'm actually working on it...
-
- KVRist
- 52 posts since 12 Sep, 2005
Very clever!93143 wrote:so I threw together a phase-alignment code in MATLAB.
It takes two files and checks to see how many leading zeroes are necessary on each file to bring it into phase with the other one (that is, to minimize the sum of the absolute per-sample differences). It then uses the smaller of the two calculated pad lengths on the corresponding file.
It's still an ugly monolithic prototype, but it gives results. Here's a segment of the A#4 ff (blue=no pedal, red=sustain) before:
Just so you know I'm actually working on it...
Rob