|
Post by oliv928 on Jul 13, 2011 23:15:34 GMT
thanks kpr. i will use your advice. Did You, kpr or Jess, try the program ? Do You like it ? except this clip noise problem.
|
|
|
Post by Jesse on Jul 14, 2011 0:18:47 GMT
thanks kpr. i will use your advice. Did You, kpr or Jess, try the program ? Do You like it ? except this clip noise problem. Yes, I tried it, it is interesting, I think I will try layering it with some other sounds once the click noise is eliminated, I tried to examine the Fusion samples on my computer, but I guess you need the wav files for that, I will try to examine them on the Fusion if that is possible, it will be a good learning experience for me. The volume gets real loud fast if you turn the volume control up on he fusion to around a 1/4 +/- of the way up.
|
|
|
Post by kpr on Jul 14, 2011 5:46:43 GMT
thanks kpr. i will use your advice. Did You, kpr or Jess, try the program ? Do You like it ? except this clip noise problem. No, I haven't try it, I already have enough samples in my Fusions :-) But it's great, that you did it and most important, that you like it yourself.
|
|
|
Post by oliv928 on Jul 14, 2011 12:33:15 GMT
Thank You Jess. Yes, why not layering it with another sound (strings, pad ?). It is easy to do, as the program use only one Multisample. You can use the other Multisample (MS2) with another sample. I dont think You need the wav files to examine them. Just use the sample/edit function in sampler mode and you can then edit them (crop, silence, loop points,...). good luck and talk soon.
|
|
|
Post by oliv928 on Aug 28, 2011 12:21:08 GMT
"I haven't tried the Preset, but clicking when release the keys sounds like Envelope problem. If the Release amount is too low it might happen. I read that the samples are unlooped, so perhaps there is not any Envelope setting? You can view also the Modulation Matrix for wrong Env Mod settings. " I think I am close to the solution. Thank You Klauss for Your usefull help. As the samples are unlooped, I did not look at any envelop settings. I dont know if one envelop is "by default" active or not. I did not put any env in the mod list. When I press "Env" I see an ASR default curve. I will try your advise, ie, to change the release time. What I ear is : When i play short note, the click appears. When I play long note, it doesnt. I think the problem is not in the sample. The problem is : If the note stops before a certain time, we can certainly have a high level harmonic due to the "dirac" function : ie : stop the wave from its actual value to zero. This should be the case with any unlooped sample, isn't it ? How can we avoid that ? Certainly with a release time isn't it ? This is common sampling technique. Sorry for this ignorance. I am "quite" beginner. Any way thank You so much. I see that you are composer and performer. I will have a ear at your tracks. You can find my "old album" here : www.jamendo.com/fr/album/1323
|
|
|
Post by bluebear on Oct 1, 2011 13:13:27 GMT
Hi, tried both links, no success to download Cheers!
|
|
|
Post by parametric on Oct 8, 2011 13:16:06 GMT
I copied your pianoss files to the hard drive and it did not work due to one reference to the CF card in the program file and eleven in the multisample file, I changed these references to hard drive using a hex editor (based on info I read from mibrue fusioneer.proboards.com/index.cgi?action=display&board=hints&thread=671&page=1#4323) I copied the modified files to my Fusion hard drive and your pianoss acoustic piano program worked just fine running from the hard drive, except for the click noise you mentioned after you release the keys. I have been playing around with various settings, but have had no luck so far in eliminating this key release noise. I've been trying to load olive928's Steinway v12 program to the Fusion HD and have not been able to get it to appear. I've modded the .afp as described, and all the instances in the Multisamples file - but still, no joy . . . I even allocated the program to category :A in the Bank.txt file to see if would appear in "Pianos" . . . . Any ideas what I could be doing wrong? Best parametric
|
|
|
Post by kpr on Oct 10, 2011 12:30:18 GMT
Instead of changing the reference data I recommend that the original sample instrument contributor just moves the data from his computer to his Fusion's harddisk. Then he can change the OSC assignments of the already created Program Presets, all this is done in minutes, and then upload the data package to the file hoster. This makes sure that everything is flawless and no problems will ever occur on anyone else's Fusion when loaded with the data.
|
|
|
Post by parametric on Oct 10, 2011 14:36:19 GMT
"I haven't tried the Preset, but clicking when release the keys sounds like Envelope problem. If the Release amount is too low it might happen. I read that the samples are unlooped, Maybe "unlooped" is the problem? Perhaps this is some help: Check Page 18 of Joeri Peeters' excellent document here www.vstream.be/pdf/vstream_manual.pdfMight be relevant . . . . Best parametric
|
|
|
Post by oliv928 on Oct 24, 2011 18:44:38 GMT
Yes the samples are unlooped. I will try to look at that.
|
|
|
Post by oliv928 on Oct 24, 2011 21:55:43 GMT
OK . Here it is ! New One without release noise, and some improvements on the mod matrix. As kpr said even if the samples are unlooped I have put a Volume envelop and now it is OK. Feel free to adjust the release time at Your taste. I don't now exactly what a standard piano should have as release time. www.mediafire.com/?zfukmbwvf30ue34
|
|
|
Post by kpr on Oct 25, 2011 10:56:10 GMT
The release times of real piano tones are longer than the most might expect. Consider even about 30 seconds in the lowest ranges, and it will be shorter towards the upper ranges. But it is ok the program just 15 seconds decay time for the lowest ranges as the sound behaviour of a sample is different to the original tone. When using samples you will "stack" audio signals when playing several notes at the same time or pressing the damper pedal and hold it. In result there is not "real" interaction between those notes and very long release times will sound "unnatural".
Using samples unlooped is ok, using the tone shaping parameters will bring the same results for looped and unlooped samples. But a looped sample saves memory space and allows to adjust the decay time to a value even higher than the original sample. So you could turn a Piano sample into a pad, just for instance. Using a second Envelope for the Filter Cutoff shaping is a must when programming a Piano based on only f samples. f=forte contains a lot of harmonics that don't appear when playing piano or mezzoforte, so the Filter Envelope can simulate this. Assign Velocity to Cutoff and set the original Filter Cutoff amount to a rather low value to control the harmonics when hitting the keys harder or softer. Also the decay time of a softly played Piano tone is shorter compared to heavy hitted tones.
The mid range of a Piano sample is usually the problem zone. First you need many samples for a decent mid range and second the dynamic range is different compared to the lower and upper range. To make the instrument playable with a overall keyboard consistence the keyboard level scaling options are helpful. If the mid range is a bit weak you can reduce the dynamic range of the upper and lower to create a overall 88 keys consistence. It also happens to huge Piano sample libraries that here and there is a single tone or dynamic (V Switch) that will sound "unnatural". The best idea is to delete this sample. Perhaps the original string was improper or the tone hit not exactly the same compared to a neighbourhood tone hit. If nothing helps you can always try to "design" this single tone with a decent audio editor where you can enhance a specific harmonic element of the tone or even shape it with a good Equalizer.
Making a small size Piano sample sound good is more difficult and time consuming than making a huge GB sized Piano sample sound good. It's a matter of fact that the original Fusion HG Piano is a very good one although it's only 3-V-Switch and the difference between the ff and the mf is very hearable. But all the other details are greatly done and it is still one of my favourites, even compared to other instruments like Kronos, Kurzweil, Clavia.
|
|
|
Post by oliv928 on Oct 25, 2011 12:32:40 GMT
Thank You kpr for all this details. Concerning my program I have the following problem : I wanted to have a softer sound at low velocity and brighter and higher. As the samples are quite ff themselves, I wanted to ad the following mod : Velocity>Sample start. ( As sample start positive makes the sound softer, bypassing some of the First attack sound sample ). But at that point I cannot make it work. Even with additive, multiplicative, new curve with a negative slope... No way. I don't know why. If You have any idea ...
|
|
|
Post by kpr on Oct 25, 2011 14:17:10 GMT
Thank You kpr for all this details. Concerning my program I have the following problem : I wanted to have a softer sound at low velocity and brighter and higher. As the samples are quite ff themselves, I wanted to ad the following mod : Velocity>Sample start. ( As sample start positive makes the sound softer, bypassing some of the First attack sound sample ). But at that point I cannot make it work. Even with additive, multiplicative, new curve with a negative slope... No way. I don't know why. If You have any idea ... To find out how Sample Start modulation works you can assign it to a Control Knob just for test purposes. The maximum range is rather small, so you can just control the main attack element and as this is the hammer noise the result will be a rather gentle sound with a tiny little attack Using Velocity for Sample Start modulation is a good idea though, but to keep the sound natural a really little amount might be right and also only as part of a multi-modulated sound shaping. To get control over the harmonic structure time duration after hitting a tone the typical "old school" sound shaping as described above is the most helpful.
|
|