Jump to content
Please note: You can easily log in to MPN using your Facebook account!

Soundfonts question


analogman1

Recommended Posts

What are they? How do they compare with .wav or .aiff files? Are there limitations to how they can be used in a sampler...and, what types of sampler(s) would you use them in?

Thanks in advance!

Tom

Nord Electro 5D, Modal Cobalt 8, Yamaha upright piano, numerous plug-ins...

Link to comment
Share on other sites

  • Replies 13
  • Created
  • Last Reply

I don't have too much experience with SoundFonts so I'm not the best person to answer your question, but here goes...

 

AFAIK it's another sample format, there are others such as EXS (Logic) and Giga. The difference is that an AIFF or WAV file is just a sound file, there's no parameter/mapping information for the sampler.

 

More information on SoundFonts at:

 

http://emusician.com/mag/emusic_downloadable_sounds/

 

The Wikipedia entry is also quite useful:

 

http://en.wikipedia.org/wiki/Soundfont

Link to comment
Share on other sites

Right -- it's just a sampler format. The only thing that sets it apart from all the others is that it was standardized and published by Creative/EMU, who created it. This makes it kind of the "lingua franca" of sample formats.

 

There are differences in sophistication of the parameters for different sample formats that soundfonts don't support. However, for basic keyboard-style instruments without release samples, it's fine. (Frankly, release samples for things like piano and Rhodes are for folks whose purpose is to make it sound like the real thing, rather than for folks who focus on the MUSIC. If anyone could have made a real piano without release sounds, they would have!)

 

If you want to use soundfonts, the easiest way is to use your computer. You don't need a special or hot or fancy computer -- it just has to be close enough to your MIDI keyboard (or vice versa).

 

For Windows PCs, to get started you need line or headphone outputs (built-in on most computers), and a MIDI interface. A simple MIDI interface, one I use live, is the MidiSport UNO. Some soundcards have MIDI ports though you might need an adaptor plug.

 

For software, try "sfz", which is a plugin .dll file intended to be plugged into other programs (like digital audio workstations). However, sfz also comes with a "standalone" .exe program. Run that.

 

With a built-in soundcard, you'll have too much latency -- the time between playing a key and hearing the sound. There's a fix for that called "Asio4all". Google, download, install. When you run sfz, set it up to use the asio4all device and the latency goes away.

 

Way fun. To use it live, you'd need a transportable computer. I use a laptop, which is handy.

Link to comment
Share on other sites

Originally posted by learjeff:

There are differences in sophistication of the parameters for different sample formats that soundfonts don't support. However, for basic keyboard-style instruments without release samples, it's fine. (Frankly, release samples for things like piano and Rhodes are for folks whose purpose is to make it sound like the real thing, rather than for folks who focus on the MUSIC. If anyone could have made a real piano without release sounds, they would have!

Interesting. Any idea on the differences/limitations of other formats, such as EXS and Giga?

 

(And I quite agree that it's absurd to replicate the unwanted sounds of musical instruments.)

Link to comment
Share on other sites

From Learjeff's suggestion, I downloaded an sfz player (rcg:audio sfz) and I used a free piano soundfont sample called "Splendid Grand". 72Meg sample from Akai I think.

 

So now I have this free piano VST plugin which is not bad actually. However, it does suck a lot of juice from my computer. If I play a lot of notes and quickly, the computer cannot keep up and bump into 100% cpu.

 

In any case, it was a nice find to learn about soundfonts and have a nice application of it to use.

Hamburg Steinway O, Crumar Mojo, Nord Electro 4 HP 73, EV ZXA1

 

Link to comment
Share on other sites

Originally posted by soundscape:

Interesting. Any idea on the differences/limitations of other formats, such as EXS and Giga?

 

(And I quite agree that it's absurd to replicate the unwanted sounds of musical instruments.)

Advanced software formats, Giga, EXS, HAL, Kontak, SampleTank II pose few real limits for the developer. All support 128 levels of velocity splits, thousands of zones/groups. All (except SampleTank) support disk streaming. Kontakt 2.0 even includes a scripting language.

 

Busch.

Link to comment
Share on other sites

Originally posted by burningbusch:

Advanced software formats, Giga, EXS, HAL, Kontak, SampleTank II pose few real limits for the developer. All support 128 levels of velocity splits, thousands of zones/groups. All (except SampleTank) support disk streaming. Kontakt 2.0 even includes a scripting language.

Why, then, does GigaStudio claim to be "The World's Most Power Sampler. Period."?
Link to comment
Share on other sites

Originally posted by learjeff:

Right -- it's just a sampler format. The only thing that sets it apart from all the others is that it was standardized and published by Creative/EMU, who created it. This makes it kind of the "lingua franca" of sample formats.

Exactly. :thu:

 

In my case, I only use .SF2 files which are mainly build for sound cards that use static cache memory, like the SoundBlaster Live! and so on.

 

The advantage is there's no resource hog anywhere. Heck, my old PC runs only at 400 MHz and I've never experienced lagging with .SF2 files. :) It's a totally different story when I run a resource hog like B4 or other software.

 

Of course, I'm limited to 160 Mb of cache, but IMHO that's way enough to build a good sound font. And if I want to modify existing files, I just open another software.

Link to comment
Share on other sites

Originally posted by soundscape:

Originally posted by burningbusch:

Advanced software formats, Giga, EXS, HAL, Kontak, SampleTank II pose few real limits for the developer. All support 128 levels of velocity splits, thousands of zones/groups. All (except SampleTank) support disk streaming. Kontakt 2.0 even includes a scripting language.

Why, then, does GigaStudio claim to be "The World's Most Power Sampler. Period."?
Well we're talking about sample formats which is different than samplers. Tascam can claim it and words like powerful are so nebulis that they're tough to prove one way or they other. Maybe they base the claim on the fact that they can reach higher polyphony than the others on the same machine, or that they include convolution. Gig 3.0 is advanced no doubt and there are some things that are easier to do in Giga. I own GigaStudio 3.0, Kontakt 2.0, Mach Five, EXS24, Halion. They are all, roughly, in the same league.

 

Busch.

Link to comment
Share on other sites

Originally posted by burningbusch:

Well we're talking about sample formats which is different than samplers.

Indeed. I thought the question worth asking, though.

 

Tascam can claim it and words like powerful are so nebulis that they're tough to prove one way or they other. Maybe they base the claim on the fact that they can reach higher polyphony than the others on the same machine, or that they include convolution. Gig 3.0 is advanced no doubt and there are some things that are easier to do in Giga. I own GigaStudio 3.0, Kontakt 2.0, Mach Five, EXS24, Halion. They are all, roughly, in the same league.
It seems to also have something of a reputation for being top of the heap. So much so that I'd assumed it had some capabilities far beyond anything else. I guess there's no reason for me to get GigaStudio then, I'll just stick to EXS.
Link to comment
Share on other sites

Originally posted by burningbusch:

Originally posted by soundscape:

Interesting. Any idea on the differences/limitations of other formats, such as EXS and Giga?

 

(And I quite agree that it's absurd to replicate the unwanted sounds of musical instruments.)

Advanced software formats, Giga, EXS, HAL, Kontak, SampleTank II pose few real limits for the developer. All support 128 levels of velocity splits, thousands of zones/groups. All (except SampleTank) support disk streaming. Kontakt 2.0 even includes a scripting language.

 

Busch.

Soundfont format puts no limitation on those things. Some soundfont players may, but I don't think I've run into it yet.

 

Some things that most of these sample formats offer that Soundfonts don't:

 

- Convolutions, for modeling sympathetic resonance (for imitating playing notes with the pedal up)

 

- Different samples depending on controller values, such as pedal up/down, mod wheel position, etc. Thinks like this allow you to choose different articulations, such as a legato vs. staccatto attack for violin or sax, or playing the same note in different neck positions on guitar.

 

- Cycles of samples so that when repeatedly playing the same note at the same velocity, you don't play the same sample twice in a row. This is a big help to make GM drum format sound less mechanical, so that alternately hitting snare cycles between a left-hand and right-hand strike. (A better keyboard map would be a better solution for this specific problem, but groups of samples for the same zone solve the same problem in other cases too.)

 

The biggest limitation to any sound designer is the limited input from the keyboard. A breath controller helps a lot for wind instruments and strings, but still is very limited compared to the nuance of articulation real instruments are capable of. But what the heck, we still have lots of fun with it, and we can do a lot better job of imitating other instruments in the band than they can!

 

It's fun sometimes to play a hot sound that imitates an instrument someone in the band is playing. The keyboard player generally comes off as the loser in tone, articulation, and fluidity (not to mention authenticity) but when the other player is willing to have fun with it you can support and play off the other player and it's great fun.

 

Finally, the key to imitating any instrument (esp. live) is not the sound but the way you use it. You have to play the licks they play on that instrument, work it like they do.

 

And it's amazing how much better horns sound when there's one real horn in the arrangment. A poor sound well played can work surprisingly well (which I've learned more from listening to others than doing myself).

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...