Is there a way to refine a old piano sample library?
-
- KVRer
- 4 posts since 16 Apr, 2017
Some musicians may know that there was a piano library called GIGAPIANO presented by nemesys. I love that rich and beautiful sound, but the samples of it are too few, so anybody know how to refine it? I mean the way to simulate more samples. 
-
- KVRist
- 33 posts since 16 Jan, 2006
I've often had similar thoughts. I've got lots of old sample libraries with some beautiful sonic material but too few velocity layers (eg 1, 2 or 4) to be satisfyingly playable. What I'd love would be some kind of 'sample magic' software that could take any set of such samples and smooth velocity transitions. Maybe it would have different modes for different kinds of instruments, piano, drum, bass, etc so as to simulate the different characteristic attack and harmonic changes as you go from low to high velocity. Even maybe able to add sympathetic resonance and release effects etc in the case of pianos. I'd pay good money for this opportunity to breathe new life into old samples. I'm aware Kontakt has its AET filter but I was disappointed by its results when I tried it. Is anything better on the market?
-
- KVRian
- 1367 posts since 30 Jul, 2013
To some extent it depends on what velocity samples you have. I think quite a number of sampled instruments were limited in allowable memory and may have only had one velocity layer. In this case all you can really do is manipulate the sound to insinuate velocity. Likely the hardest notes were recorded. In theory I think that delaying the attack and trying to lessen the decay would make a note sound softer. Also a higher velocity note would have more 'ring' to it I believe so EQing that down should make it sound like a softer note.
Another thing to watch for is that samples were frequently quite short and the decays looped and faded down to simulated a note fading. Older samples can cut off quite abruptly if you don't have this loop info in the wave.
If you do have a loud and soft note available, then AET can be useful, but it will do nothing for single velocity layers. Also there are crossfade scripts that might help. Fake round robin scripts can help make it seem like an instrument was more fully sampled.
In general it is easier to degrade a heavily sampled library than to polish a lightly sampled library, trust me I've tried it many times.
Another thing to watch for is that samples were frequently quite short and the decays looped and faded down to simulated a note fading. Older samples can cut off quite abruptly if you don't have this loop info in the wave.
If you do have a loud and soft note available, then AET can be useful, but it will do nothing for single velocity layers. Also there are crossfade scripts that might help. Fake round robin scripts can help make it seem like an instrument was more fully sampled.
In general it is easier to degrade a heavily sampled library than to polish a lightly sampled library, trust me I've tried it many times.
- KVRAF
- 26033 posts since 20 Oct, 2007 from gonesville
I only remember the version which came with GigaStudio.
I think what you got is what you get.
I found some of Gigaharp a couple years back, which was still fantastic. This was one case I have no recollection of 'shit, do I need more dynamics layers here or what.'.
I think what you got is what you get.
I found some of Gigaharp a couple years back, which was still fantastic. This was one case I have no recollection of 'shit, do I need more dynamics layers here or what.'.
-
- KVRer
- Topic Starter
- 4 posts since 16 Apr, 2017
Thanks you for your reply, actually Gigapiano have 3 velocity layers(recorded in pedal down and up), even real staccato samples. So I will try my best to refine it.bigcat1969 wrote:To some extent it depends on what velocity samples you have. I think quite a number of sampled instruments were limited in allowable memory and may have only had one velocity layer. In this case all you can really do is manipulate the sound to insinuate velocity. Likely the hardest notes were recorded. In theory I think that delaying the attack and trying to lessen the decay would make a note sound softer. Also a higher velocity note would have more 'ring' to it I believe so EQing that down should make it sound like a softer note.
Another thing to watch for is that samples were frequently quite short and the decays looped and faded down to simulated a note fading. Older samples can cut off quite abruptly if you don't have this loop info in the wave.
If you do have a loud and soft note available, then AET can be useful, but it will do nothing for single velocity layers. Also there are crossfade scripts that might help. Fake round robin scripts can help make it seem like an instrument was more fully sampled.
In general it is easier to degrade a heavily sampled library than to polish a lightly sampled library, trust me I've tried it many times.
BTW, I've already got fake RR scripts, but how can I get that X-Fade script?
-
- KVRer
- Topic Starter
- 4 posts since 16 Apr, 2017
I never used AET filter before, I will try it. I've been using the equalizer in Adobe Audition before, but the effect is not satisfactory. The characteristics of the sample can not be changed, only the frequency of loudness changes, it is difficult to control the transitions between velocity layers.stevebard wrote:I've often had similar thoughts. I've got lots of old sample libraries with some beautiful sonic material but too few velocity layers (eg 1, 2 or 4) to be satisfyingly playable. What I'd love would be some kind of 'sample magic' software that could take any set of such samples and smooth velocity transitions. Maybe it would have different modes for different kinds of instruments, piano, drum, bass, etc so as to simulate the different characteristic attack and harmonic changes as you go from low to high velocity. Even maybe able to add sympathetic resonance and release effects etc in the case of pianos. I'd pay good money for this opportunity to breathe new life into old samples. I'm aware Kontakt has its AET filter but I was disappointed by its results when I tried it. Is anything better on the market?
-
- KVRer
- Topic Starter
- 4 posts since 16 Apr, 2017
Kontakt can load .gig formatted samples. Where did you find GigaHarp?jancivil wrote:I only remember the version which came with GigaStudio.
I think what you got is what you get.
I found some of Gigaharp a couple years back, which was still fantastic. This was one case I have no recollection of 'shit, do I need more dynamics layers here or what.'.
-
- KVRian
- 1367 posts since 30 Jul, 2013
Here is the crossfade script...
http://www.nilsliberg.se/ksp/scripts/sc ... #crossfade
This might be totally unneeded, but if you want a simple Kontakt GUI you are welcome to use mine...
http://bigcatinstruments.blogspot.com/2 ... b-pad.html
http://www.nilsliberg.se/ksp/scripts/sc ... #crossfade
This might be totally unneeded, but if you want a simple Kontakt GUI you are welcome to use mine...
http://bigcatinstruments.blogspot.com/2 ... b-pad.html
-
- KVRAF
- 2312 posts since 9 Jun, 2002 from East of Santa Monica
I'm not jancivil, but I used to see it pop up on Ebay occasionally a few years back. Basically some stores selling their leftover old stock.verares wrote:Kontakt can load .gig formatted samples. Where did you find GigaHarp?jancivil wrote:...
I found some of Gigaharp a couple years back, which was still fantastic. This was one case I have no recollection of 'shit, do I need more dynamics layers here or what.'.
Garritan re-did it for their Aria sound engine into "Garritan Harps" which is still available at the Garritan website.
- KVRAF
- 26033 posts since 20 Oct, 2007 from gonesville
in an old box. then it converted pretty well. I don't remember about its 'Performance Tool'.verares wrote:Kontakt can load .gig formatted samples. Where did you find GigaHarp?jancivil wrote:I only remember the version which came with GigaStudio.
I think what you got is what you get.
I found some of Gigaharp a couple years back, which was still fantastic. This was one case I have no recollection of 'shit, do I need more dynamics layers here or what.'.

