it's worth at least remembering that for the most part what makes a Strad really sound like a Strad is the player, not the wood. The tonal advantage is latent and must be coaxed out of the instrument by an expert violinist.
A violin isn't a complicated piece of electronics, but any physical phenomenon depending on so many variables (the interrelationship between the player, his bow, the violin's strings & soundboard, the acoustic and the human ear) would produce an unbelievably complex prospect for "white paper" analysis.
This is absolute fact. The player
is the largest variable. The way each player attacks the strings is very different. Where he bows the string along it's length, how much pressure he applies with the bow while he draws it, how much pressure he hits the string with as he bows are very important events in the timbre of the note the violin will produce.
A direct analogy is a drum. A snare drum for example is a simple instrument. There are many different sounds that a snared drum can make. They are very sensitive to velocity, stick weight, the stick tip material, angle of strike, location of strike, duration of strike. All these different elements will affect the attack, timber, resonance, fundamental pitch, pitch modulation, overtone, amplitude and duration of the note a snare drum produces. Despite the fact that it is tuned to a fixed note, it's a very complex sounding instrument.
If you
listen to an experienced drummer hit 4 evenly spaced consecutive hits in as similar a fashion and location as he can, it will appear to sound the same at the time you are listening to it. If you were to take those 4 hits and record them (in a digital sampler for example) edit and isolate each hit, then trigger them closely side by side and compare the strikes, they would in fact be all different. Often it's the first and last strikes that are the most different.
I learned this over 15 years ago when I first began using digital samplers. There are a complex set of parameters to edit recorded sounds with after a sound has been sampled, in order to make them sound more as if a
human were playing the sounds, and not a machine simply triggering recorded samples.
There are amplitude, attack, envelope, slope, filters, pitch, layers, crossfades, mulitsamples, modulation parameters to name a few that must be used in order to add the randomness of different elements a human player makes in the sound of a performance and timbre of a certain instrument. Drums are very sensitive to these parameters. As are stringed instruments (they are probably the most difficult actually). A very simple instrument is a harpsichord. It is a very simple instrument to reproduce because of the fact it has very little in the way of human variation. It is a plucked instrument, as opposed to a hammered instrument like the piano. It has basically two events. Note on, and note off. You can't play a note softly (per se) on a harpsichord. It's attack is very consistant, and therefore variations of all the elements that shape the timbre of a note event are all but eliminated.
What the elements are that shape how sounds are made from instruments is a fascinating and complex topic. Much too complex to try and summarize here, but very, very real, and very, very complex. Even a simple drum.
The best and easiest description so that people might get a sense of what kind of difference these elements might make in terms of sound is this example.
If one were to digitally sample a single snare drum shot, trim the sample, edit it for proper velocity sensitivity and amplitude, even add a certain amount of LFO (Low Frequency Oscillator) to add randomness to the playback pitch, you would have a certain randomness being played back. The triggered sample is different than the stored sample. The triggered one has all the parameters acting upon it according to the way it is triggered. You assign parameters to elements of velocity, duration, aftertouch of a triggering of a sample.
If you were to trigger the sound numerous times, people would definitely hear differences between the (triggered) samples.
If how ever you were to try and play the sample very close together to simulate a double stroke roll on a snare drum, it would sound like a series of very closely spaced echoes. It would sound that same way machine gun fire sounds like. Tat,tat,tat,tat,tat,tat.....If you were to play back the samples very,very close together to try a simulate a flam, you would get a huge phase cancellation and the sound would be full of comb filtering and flanging.
The only way to try and have it sound anywhere close to what it sounds like in reality would be to have 3 or 4 different samples (like the 4 shots we had our pro drummer play earlier) edited like I described earlier. If you would then play back the samples in random patterns in order to try and simulate sixteenth note figures, rolls, and flams, it would come very, very close (for all intents and purposes, virtually real)to sounding as if they were really being played in actuality by a real drummer and not a machine triggering a series of samples. Snare rolls would sound more like ta,ti,ta,to,tit,ta,tat,to....more randomness and very human.
Cheers
