Oscillators, Pitch variations and other parameters

[Sugizo]

Hi folks,

Although I am mainly a guitarrist, I have been using the KORG TR61 for a couple of years now studio and live and have become quite familiar with playing around with it. I have come to the point where I feel I should take a deeper dive into all of its wonders. I'm OK with creating combis however I've never attempted to play around with program editing beyond the basics (editing efects, basic parameters, filters, MIDI). I've read the manuals and roamed these forums, however a few concepts still remain slightly blurry and I would much appreciate it if someone could give me an earthly explanation of the following:

- The nature and relationship of Oscillators and Multisamples (are we talking a similar relationship to programs and a combis?). What are the benefits of using two oscillators?

- The differences between and the varying of pitch with pitch shift, pitch EG intensity, transpose/tune, LFOs.

Once again thank you for sharing your superior knowlodge!

[kanthos]

The advantages of using two oscillators instead of two separate programs in a combi are that a) you use only one slot in your combi and can thus add more programs and b) Things like LFOs and envelopes can operate uniformly on the oscillators in your program - while you could theoretically accomplish this over two separate programs, you'd have to make sure the settings were identical.

Pitch shift and transpose/tune are two ways of doing a static pitch change. If I detune an oscillator or transpose my program, whether saved in the program or done in a combi, those are global, constant changes - all keys are transposed or detuned by precisely the same amount for that program (i.e. if you transpose by -7, middle C is a G, D is an A, etc.) The pitch is constant and is affected by a constant amount.

The pitch envelope varies pitch over time in a well-known way. Pitch starts at some level below the actual pitch of the key when you first press the key. Over the attack time, the pitch rises to the attack level, then over the decay time, it decays to the normal level (i.e. if you play the A below middle C, after the decay phase, the pitch is at 440 Hz). Pitch stays here until you release a key, then it decays further. See page 11 in the parameter guide for diagrams and indications of how this works. The point of this is that most physical instruments don't always have a constant pitch; the envelope describes how it the pitch changes over time in a predictable way.

LFOs, on the other hand, if they're set to modify pitch, modify pitch always and at a constant rate. You set the frequency of the LFO (i.e. the speed at which the LFO's waveform repeats itself - a higher speed means faster repetitions, of course), and the pitch is constantly varied according to the waveform. Say you've got a sine wave (google for a picture): the pitch will rise and fall over a smooth curve; increase the frequency and the pitch will vary much faster; increase the amplitude (the height of the peaks in the wave) or, in the case of the TR, you'll be increasing the amount that the LFO effects pitch, and the variations will be more extreme.

[Sugizo]

Cheers Kanthos,

I understand the grand majority of your definitions, I just think I am mixing concepts. Let's go part by part.

1) Constant pitch variations: take any sound you want, and go to the edit basic -> OSC1 menu, where you can find Transpose and Tune. Now if I set transpose to -12 and tune to 1200 and then viceversa (always compensated values, I believe the "Stretch" parameter in the P.Edit menu is doing exactly this) the pitch does not vary, however the sound is different. What phenomena are acting behind this sound change?

2) With respect to the envelope gradient, does an increase in intensity (menus 3.1 osc and EG) mean and increase in pitch? (I think this is the part where I'm mixing concepts). Acoustically speaking, upon increasing the intensity, I hear that something is winding up and it does feel as if this something is increasing its energy but also its pitch (now do these correlate to each other?). However, once again, this sound is totally different when compared to increasing pitch with tune or transpose.

Still concerning the EG diagram, would this mean that, if I hit any "C" key, the pitch rises slightly over C when attacking, is exactly C when sustaining and falls below C when decaying or decays to C?

3) Say an LFO is oscillating considerably (considerable waveform amplitude), would the correct representation of the EG graph hence include a waveform instead of straight lines (god I wish I could draw this out but hopefully you can get what I mean). Or am I again mixing concepts?

Thank you for having patience with me!

I'm sure most of you are thinking: just play the damn keyboard and listen! yea, you are right, but hey it's nice to have some musical physics (for dummies - hey I should google that) in the forum evry now and again

[X-Trade]

The pitch EG has a number of parameters and thus can be configured in a number of ways. It is unique in the programs though that the pitch EG is shared between the two oscillators, whilst the amp and filter EGs are independent for each oscillator.

Anyway, if you play with the pitch EG and look at the graph, you will see you can move the start, peak (attack peak), and end point.

roughly speaking, the pitch will rise from the start point, to the peak point (but that could be lower than the start point), over the times specified. Then decay to the middle (zero). Then for release it will go (over release time) to the endpoint (release) level you set. Again that could be above or below zero.

The pitch EG is also unique compared to the other EGs in that it has zero as its middle point in the graph - you can go above and below zero. I believe the other EGs range from 0 to +1, whilst the pitch EG goes from -1 to +1. The pitch EG intensity for each oscillator sets where that + and - 1 actually are. if you set the intensity to 0.10 then when the EG reaches the top of the graph (if it ever does - again it depends on your EG settings), that will add on 0.1 of a semitone, or 10 cents. again if your eg hits the bottom of the graph, that point will be minus 10 cents in pitch.

Of course if you are crazy you could set the intensity to 12.0, where +1 on the EG relates to an additional octave of pitch!

generally the pitch EG isn't used much.




As for pitch stretch and the transpose/detune parameters, what you are hearing is an effect which is a byproduct of the sampling.
A multisample is typically made up of a number of samples of different keys, spread across the keyboard to produce the best consistency in tone/attack over the range of the keyboard. Otherwise with just a sample of (for example middle) C, higher notes would have a rediculously 'sped up' sound, whilst lower notes would be 'slowed down'.

now, what happens is, that the Transpose parameter takes effect before the sample is selected for the keyboard, but detune is used after the sample has been selected. So you will notice with a transpose of -1 and +100, you will often not hear any difference, whilst with -12 and +1200 it is selecting the sample from an octave below, and tuning it up to the new pitch.

[Sugizo]

Fantastic clarification X-Trade. I have played around with the EG intensity and graph levels and now fully understand the behaviour of the sound. Ta!