Magazine Archive

Home -> Magazines -> Issues -> Articles in this issue -> View

Technically Speaking

Article from Making Music, July 1987

Andy Honeybone, brave man that he is, tries to swap programs on the DX-series synths.

CAN DX7 voices be transferred to lesser four-operator eight algorithm synthesisers? That's the question behind this article, and a short series to follow.

Both the DX21 and TX81Z allow dual mode whereby all but seven of the DX7's 32 algorithms can be concocted by stacking. The problem areas are largely down to fixed frequency operators and envelope complexity. It's with the latter subject that I'll kick off this monograph, and hope that by explaining aspects of the two systems we'll end up wise.

The first envelope generator that I ever built was of the analogue AR variety. Just attack and release rates were variable and the sustain level was always at maximum. The DX7's envelope generator might well be more complex but at least the terminology has stayed the same — we're still talking about rates and levels.

The controls on my analogue EG were variable resistors which determined the charge and discharge times of a capacitor. The laws of physics being as they are, a capacitor charges quickly initially and thereafter slows more and more as it approaches full charge. The discharge rate similarly falls quickly and then drops to a trickle.

These voltage/time profiles are described as exponential and although the discharge curve is a good approximation to natural sound decay, the attack shape is far from perfect. The purpose of this aside is to remind you that the straight lines depicted joining the various sections of multi-stage envelopes on the front panels of synthesisers are nothing like what you would see if you fed the audio output to an oscilloscope. Although the envelope generators are digital, there's more than a hint of exponentiation in their software.

While digressing on non-linearity, the parameter values on Yamaha X-series equipment give the greatest amount of variation toward the high end of their range. They too can be said to have an exponential response. The audible effect of a change from 99 to 98 will be much greater than that of a change from 19 to 18.

There can be few synthesists not familiar with the ADSR envelope model. This EG has become standard even though some early circuits were rather touchy about what they'd do when the key was released before the attack phase had completed. The ability to set the sustain level to zero allowed very staccato sounds to be produced and, in order to avoid embarrassing silences, the multiple trigger facility became mandatory on mono synths.

The ADSR was fine for click organs and brass, but was incapable of realising an effective piano envelope with a rapid attack, the fall-back decay after the hammer stike, and then the final decay of the vibrating string.

The introduction of the DX7's four-rate and four-level EG simultaneously gave the flexibility to create naturally occurring profiles and acute culture shock to those who had to use them. It's easy to panic when confronted with the eight envelope parameters — but on closer examination, several of these controls are only used in specific situations.

Looking at the apparently more simple EG of the DX21/27/100/TX81Z, the number of parameters is reduced to five: four rates and one level. The missing three level parameters can be imagined to be preset.

The first level to which the attack climbs is fixed at maximum. The second level (D1L) is a parameter which can be controlled and is either the break point between two decays or, if the second decay rate (D2R) is set to zero, the sustain level. The third level is set equal to the second level when D2R is zero, but otherwise is set to zero. The fourth level is, of course, the starting level of the next envelope to be generated. No flexibility is allowed here though — its value is firmly set to zero.

The parameter economies of the DX21/etc EG enforce single stage attacks. It is not possible to create double-strike attacks or three-slope attack curves. Most restricting is the fixed final level which wipes out any chance of adding all those rattles and buzzes on key release which are so distinctive of the votings on the DX7II. The DX21 EG is essentially ADSR with an option on ADDR for piano type envelopes.



In showing that the DX21 EG is cut down from the DX7, I hope I've shown that the extra flexibility of the full blown thing is well worth mastering.

I'll finish then with some thoughts on rates and levels. Each parameter on the DX7 EG has a range of zero (lowest level/slowest rate) to 99 (highest level/fastest rate). This seems a little more orderly than the DX21's assortment of 15 and 31 maxima.

Thinking graphically now, level moves us in the vertical plane while progress from left to right is proportional to rate. As a rate of 99 is almost instantaneous, the actual amount of horizontal movement can be estimated from 99 - rate.

The main rule to remember is that a rate between two equal levels is meaningless. A ball will only roll if it is on a slope. Changing the R2 or R3 rates on the INIT VOICE envelope will have no affect as the L1, L2 and L3 levels are all set to 99. If you need to bypass one stage of the envelope, it is usual to set the levels either side of the stage to be equal. Even though the corresponding rate will be ignored, Yamaha say it should be set to 99 (it's good style).

To further demonstrate the above point and to get you out of the arm chair and over to the DX7, you can easily create an ADSR type envelope by setting levels L1 to 99 and L4 to 0 and then setting levels L2 and L3 equal to some sustain setting (around 50). This is what is already done for you on the DX21. The rates R1, R2 and R4 become the A, D and R controls. It's a terrible straight jacket to put on the EG but it will get you started on familiar territory.

The level settings are predictable as they are exactly what you imagine them to be. The rates however, are what are known as relative which means that their affect is totally dependent on where they've come from and where they're going. The book of words cheerfully says that the same rate will be faster for an attack than a decay. The rate will also depend on the degree of slope between adjacent levels. In other words try it and see.

Just to close, a quick point on the graphics. The level settings have been represented as their exponential values. It seems only right that inaudible levels around 30 are squashed down to the base line while small changes at the top of the ranges are magnified. To misquote Tina, "What you see is what you hear."

Previous Article in this issue

Tokai Guitar

Next article in this issue


Publisher: Making Music - Track Record Publishing Ltd, Nexus Media Ltd.

The current copyright owner/s of this content may differ from the originally published copyright notice.
More details on copyright ownership...


Making Music - Jul 1987

Feedback by Andy Honeybone

Previous article in this issue:

> Tokai Guitar

Next article in this issue:

> Demology

Help Support The Things You Love

mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.

If you value this resource, you can support this project - it really helps!

Donations for July 2024
Issues donated this month: 14

New issues that have been donated or scanned for us this month.

Funds donated this month: £20.00

All donations and support are gratefully appreciated - thank you.

Magazines Needed - Can You Help?

Do you have any of these magazine issues?

> See all issues we need

If so, and you can donate, lend or scan them to help complete our archive, please get in touch via the Contribute page - thanks!

Please Contribute to mu:zines by supplying magazines, scanning or donating funds. Thanks!

Monetary donations go towards site running costs, and the occasional coffee for me if there's anything left over!

Small Print

Terms of usePrivacy