Sound Designs

Home
Up
About
News
Products
Ordering
SoundFonts
Emu EOS
Emulator X
SY77
F.A.Q.
Synthesis
CS6x


Send Email

Samples Vienna 2.3 Importing Looping Instruments Presets Tips
Last Changed 01/19/2005

SoundFont® Bank Tips

  •  Don't stick to only using your Global Loop points. By changing the Local Loop point of a sample in the Instrument Pool, you can transform a single waveform into several variations simply by changing the start or end loop point. For example, if you are using a perfectly looped Sine wave that sounds smooth and flowing, you could move the Local Start or End loop points (not both) away from the perfect loop point, and this will add some grunginess to the Sine wave. This way you can use a single Sine wave sample to create a small palette of waveforms which vary between Sine and something similar to Sawtooth. Watch your tuning because once you move away from the perfect loop point, the tuning of the sound changes and you have to manually retune it. This is relatively easy to counter ... if you make the loop smaller (Start and End move closer together), than the pitch of the sample will increase and you need to lower the tuning. Conversely if you move the Start and End points further apart, the pitch of the sound will lower and you will have to increase the tuning of your instrument.
     
  •  When you sample your sounds, don't normalize (maximize volume) them to 100%. Rather keep them between 85% and 95%, so that you leave a little headroom for the hardware or software which is playing them back. Some soundcards distort easily if the sound overburdens the card, so if you keep the level a little lower, this always helps prevent harsh clipping and distortion.
     
  •  There are some samples you just cannot loop well. Don't make it your life's ambition to loop them if it seems impossible to do so. If you can possibly generate the sound again and force it into a pattern which promotes easy looping, then do this. If you have no other recourse, decide if the quality of the loop is acceptable and make a decision to keep or throw it away. If you can, fire up an editor and attempt to create the sample yourself, with better looping possibilities, otherwise you create a sub-standard SoundFont and pollute the already-cluttered SoundFont arena.
     
  •  When sampling sounds off other machines, it is always a good idea to try and keep modulation properties (LFO, vibrato, tremolo) to a minimum if you want to be able to loop easily and save RAM. Some sounds don't work without keeping the original LFO settings, and in these cases you really need to do some heavy multisampling. However, if the sound relies on properties you can replicate in Vienna, then turn them off on the original machine. It makes the sound a lot easier to loop and masks weird sounding changes as you move up and down the keyboard.
     
  •  What about some general tips and information using the SoundFont format to it's maximum ...

    1
    . First thing to do is create multi-sampled waveforms, like sawtooth, square, triangle, sine, noise, and various pulse-width waveforms (10%, 20% 30% ... etc). This gives you a basis to work with. You can use some freeware waveform editors for this task, such as Cool Edit or Sound Forge.

    2. Next, take the samples and invert them, so you now have a palette of normal waveforms as well as their inverted counterparts.

    3
    . Load these into your sampler and loop them. Looping the inverted waveforms should be easy, since the loop points are identical to the original samples.

    4
    . When layering your waveforms into a preset, always try to use at least 2 layers and detune them slightly, or apply different pitch LFO's to each. This gives that chorussy effect so prevalent in multi-oscillator analog synths. If possible, apply different settings for each layer, so that their LFO's are slightly different, their tuning is slightly different, the filtering is a bit different ... etc. This allows for a more evolving and unpredictable sound that simulates an analog synth.

    5
    . Sawtooth and Square waves are always your best choices when emulating analog sounds because they are at the heart of analog synths.

    6
    . Layer a normal sample with it's inverted counterpart. Under normal circumstances this will actually output no sound as the opposing waveforms cancel each other out (not on the inaccurate SB Live! though). However, detune them slightly and you'll get yourself a pulse-width waveform . Set your Modulation wheel to modulate the pitch of one of the waveforms, or the pitch LFO of one of them (very slowly), and you can control the pulse-width in real time. Actually, setting a controller to modulate amplitude, filter cutoff, or filter rez also has interesting results.

    7
    . Take a waveform and set the Local Loop point off by a few samples. This works well with Sine waves because you get a completely new sounding waveform when the sample loop is inaccurate. The more you offset the loop point , the more the Sine wave approaches the sound of a Sawtooth waveform. However, the more you offset the loop-point, the more the waveform plays back out of tune so you'll have to adjust for this.

    8. When layering samples, try layering one an octave lower than the other. This can really phatten up the sound and this is particularly good for bass sounds.

    9. Another trick related to point 8 is to layer a Sine wave one octave lower than whatever sound you are creating. You can get a super stacked, phatt sound by doing this because it adds a lot of low-end ballz to the sound. It can be pretty roof-shaking if you work the sounds a little.
     
  •  I have written a very simplified article on emulating analog synths, and you can find this right HERE - the article is called Vintage Synth Emulation on a Sampler. It's old, but still useful for giving one a simple idea of how things work, and from which direction to approach analog emulation of samplers.

    There's also an article on creating waveforms in Visual Basic which, although not quite what we're looking at, can still make interesting reading if you are so inclined. You can also find this
    HERE under Wave Modulation Synthesis for Programmers.

    One thing I can add to the 6 basic waveforms previously is the use of more complex waves. For example, if you have access to software which can generate additive, AM, FM, or other types of waveforms (Cool Edit and Sound Forge are good candidates), this enhances your sound palette much further, allowing you to create some of those old ring-mod sounds as well as DX and other FM sounds. It helps create some crispy bell-like textures, and clangourous noises, clangs, and clashes. Since samplers often don't have access to real time synthesizing (like ring-mod, FM ... etc), the only way to get this into your sound is to include these types of waveforms. With Additive, AM, and FM, it is relatively easy to create simple waveforms which can be looped easily, but result in an overtonic (yes, I know, no such word!) sound. It's great for pads if you dub this over your 6 staple waveforms. Also, nothing creates bass sounds like FM can. I know, analog synths are supposed to be bass kings, but it's just not true - FM synths are capable of bass sounds that a Moog only dreams of being able to create!
     
  •  The SoundFont 2.0 standard is quite limited when it comes to getting real time interaction out of your SoundFonts. The lack of Portamento, multiple filters, LFO shapes, and controller routings forces you to be very creative when trying to get unconventional sounds, or when trying to create something that hasn't been done before. Soundfont format 2.1 is much more capable, but is not available on anything but the Emu ASP and Creative Labs Audigy soundcards.

    One of the methods previously mentioned, and one we have used for quite some time, is to layer inverted versions of the same sample on top of each other. You will be surprised at how much this can phatten up a sound, or how much livelier you can make it.

    Since the samples essentially cancel each other out in a perfect soundcard (the SB Live! does not), you have no sound until you start messing with the parameters of one of the inverted samples. The avenues you can take are as follows:

    1. Change the Resonance setting of one of the samples. This brings in a filtering effect similar to a high-pass filter, where certain frequencies of the untouched waveforms begin to show through. It's not an epiphany, but it does widen your sound creation palette by giving you another way to drag yet another sound from a single sample.

    2. Change the filter cutoff of one of the samples. This again brings out various frequencies as the sound-cancellation properties of the inverted samples are messed with. It works in a similar manner to changing the resonance of one sample (point 1 above), but the sound is different - just another way of changing the tonal characteristic of a sample and getting more use out of it.

    3. Detune the samples. This allows you to employ a pulse-width of sorts - the more the detune, the bigger the width. You can use this method to change layered Sawtooth-Inverted samples into a pulse width of between 0% and 100%, so once again you stretch the sonic potential of a single sample into many different timbres.

    4. Change the volume of one of the samples. Again, a similar effect to the above, but just different enough to give you a few more timbres from a single sample.

    5. Use different ADSR settings. If your inverted samples have different Attack, Decay, Sustain, and Release settings, you can get various pulse-width effects based on whether you're applying different Amplitude, Pitch, or Filter ADSR envelopes. You can get some quite interesting effects this way.

    6. Use the LFOs to give you a more consistent variation to point 5 above. You can have a consistent rate of pulse-width change, for example, but applying a pitch LFO to one of the inverted waveforms.

    7. If you have access to the true Soundfont 2.1 standard, try routing velocity to, say, the pitch of one of your inverted waveforms. This will increase the pulse-width the harder you press your keys. Aftertouch is even better in this regard because the harder you press down a key, the more pronounced the PW effect. Once you have access to the full spectrum of controller routings, Soundfonts take on a new life and can be extremely expressive because you can vary all sorts of parameters based on controllers such as a modulation wheel, expression pedal, all manner of knobs and sliders, velocity, aftertouch ... etc. The possibilities are mind-boggling.


     

Back Home

_____________________________________________________
All contents Copyright © 2000-2003 Analoguesque Sound Designs