Soundfont editor?

[No name]

I need one to put together some soundfonts. Other than Vienna (didn't work with my soundcard) is there another?

[kevink]

VSampler3 will do the job - I've never actually used the feature for real work, but I did at least verify that I could do a save-as-SF2.

[beltrom]

http://www.synthfont.com/ look for Viena (NOT Vienna).

[dougsyo]

VSampler3 will do the job - I've never actually used the feature for real work, but I did at least verify that I could do a save-as-SF2.

I've built soundfounts with VSampler3. It depends mostly on the quality of the samples going in... you get some wav files that are looped right and keyed right, and it makes as good a soundfont as any for me.

Because of its somewhat obtuse interface, it's not my choice for playing soundfonts (I do that with sfz+).

Doug

[MidiMan]

Because of its somewhat obtuse interface, it's not my choice for playing soundfonts (I do that with sfz+).

Doug

Agreed! sfz+ is THE sf-player. I use Awave Studio to make the sf's, though.

[learjeff]

I have some "nerdware", if you're interested, that streamlines the process of creating big layered soundfonts -- I wrote it to help build my jRhodes3 soundfont.

You can find a link to the tools from that page. By nerdware, I mean you have to load Python and run the scripts from a DOS shell.

For each velocity layer, I sampled all the notes for that layer into a single wave file, and then did any processing (EQ, noise reduction, etc.) on the layer files by hand in a wave editor. Then I feed the layer files into a program that chops them up and names them by note and layer. Another program takes the list of files (samples) and some user input from a text file and creates a keyboard/zone map. A third program takes the above and knits together an sf2 format file.

I then use Extreme Sample Converter (well worth the small price, under $50 IIRC) to tweak the filter paramters, etc.

HTH
Jeff

[No name]

I have some "nerdware", if you're interested, that streamlines the process of creating big layered soundfonts -- I wrote it to help build my jRhodes3 soundfont.

You can find a link to the tools from that page. By nerdware, I mean you have to load Python and run the scripts from a DOS shell.

For each velocity layer, I sampled all the notes for that layer into a single wave file, and then did any processing (EQ, noise reduction, etc.) on the layer files by hand in a wave editor. Then I feed the layer files into a program that chops them up and names them by note and layer. Another program takes the list of files (samples) and some user input from a text file and creates a keyboard/zone map. A third program takes the above and knits together an sf2 format file.

I then use Extreme Sample Converter (well worth the small price, under $50 IIRC) to tweak the filter paramters, etc.

HTH
Jeff

Sounds pretty complicated. I might have quite a few samples, and learning how to do that might be quite time consuming. I'm just looking for something that is rather easy and intuitive. Vienna never worked with my soundcard though. I'm still gonna give Viena a try though. I really appreciate your help! Thank you very much.

[M'Snah]

VSampler3 will do the job - I've never actually used the feature for real work, but I did at least verify that I could do a save-as-SF2.

I second that. It takes some time to understand how VSampler works, but after that it's very easy to build banks, multi-layered, etc, and convert them to SF2 (or just keep them in the native format). Included is import from most other formats.

If you don't want to edit, but only want tot convert: ESC.

[learjeff]

Sounds pretty complicated. I might have quite a few samples, and learning how to do that might be quite time consuming.

It would be a waste of time if you only had a few samples -- in that case, any good soundfont creator would work just fine. (Or any sample format editor, and then convert to sf2, for that matter. I did my first attempt using GigaStudio.)

The more samples you have, the more time my nerdware will save you. In either case (my tools or a wysiwig program), you have to do any hand-editiing first (normalization, de-noising, etc., if you plan to do those kinds of things). With a wysiwig program, for each sample, you have to know what note it is and build a zone for it and map it by hand. I didn't want to have to do that manually 65 times (once for each of my 65 samples), which is why I wrote the programs.

I would definitely advise finding a soundfont editor you like and can use and do some experimenting first, building some simple soundfonts to get familiarized with the tools and the process. Only then should you attempt to record the large number of samples for your ultimate soundfont. (Unless sampling is easy, like when you're sampling a MIDI synth.) But if you're sampling an acoustic or electromechanical instrument (like acoustic or electric guitar or piano), the recording process is rather painstaking.

That process is a bit easier if you don't have to write down for each sample what the file name is, what note it is, and what velocity layer it is.

You can think of my tools as being a "soundfont compiler". It uses the information inherent in the sample sets, plus a simple layer file naming convention, and a few parameters in a text file to build a soundfont file without any user interaction.

BTW, EXSC is good for editing certain kinds of things, like loops and filter parameters.