Korg Forums

[CharlesFerraro]

An administrator might want to sticky this. Naturally these times apply to other keyboard engines that use the same envelopes such as the Krome, Kross, and MicroSTATION workstations. These numbers are actually more accurate than what you find in the manual. The information was calculated from the actual C code written by Andy Leary.


EG VALUE > ACTUAL TIME (milliseconds)

0 > 0.667
1 > 0.667
2 > 1.3
3 > 2
4 > 2.667
5 > 3.33
6 > 4
7 > 5.33
8 > 6.667
9 > 8
10 > 10
11 > 16
12 > 18
13 > 20
14 > 23.33
15 > 26
16 > 29.33
17 > 32
18 > 35.33
19 > 38
20 > 44
21 > 48
22 > 52
23 > 58
24 > 64
25 > 72
26 > 80
27 > 86
28 > 91.33
29 > 96
30 > 104
31 > 112
32 > 128
33 > 136
34 > 144
35 > 160
36 > 176
37 > 184
38 > 192
39 > 208
40 > 224
41 > 240
42 > 272
43 > 288
44 > 304
45 > 336
46 > 352
47 > 384
48 > 400
49 > 432
50 > 464
51 > 496
52 > 544
53 > 576
54 > 624
55 > 672
56 > 720
57 > 768
58 > 816
59 > 880
60 > 944

EG VALUE > ACTUAL TIME (seconds)

61 > 1.008
62 > 1.088
63 > 1.168
64 > 1.248
65 > 1.344
66 > 1.44
67 > 1.536
68 > 1.648
69 > 1.776
70 > 1.888
71 > 2.032
72 > 2.176
73 > 2.336
74 > 2.496
75 > 2.688
76 > 2.88
77 > 3.072
78 > 3.296
79 > 3.536
80 > 3.808
81 > 4.112
82 > 4.448
83 > 4.848
84 > 5.312
85 > 5.896
86 > 6.512
87 > 7.28
88 > 8.256
89 > 9.44
90 > 10.912
91 > 12.784
92 > 15.184
93 > 18.384
94 > 22.784
95 > 29.12
96 > 37.44
97 > 49.92
98 > 65.536
99 > 87.376


Practical application: Knowing the actual time output of an envelope stage can help you fine tune your envelopes to be a note division of tempo. For instance if you would like the decay of a tom to span an 8th note at 120bpm, all you do is take 60,000 (the amount of milliseconds in a minute) and divide by the beats per minute to find a quarter note (one beat in a 4/4 time signature). Then simply divide for 8th, 16th etc. In our example 60,000/120 = 500ms / 2 = 250ms for an 8th note. If we then look at our chart we can see that with no sustain, a decay time value of 42 will last 272ms. I suggest overshooting and then using the curve setting to further fashion the decay. There are of course many other musical applications beyond this simple example.

If anyone would like to know, the EG Time Value knob references a table where each position operates in multiples of 32 samples. So a time value knob setting of 13 will reference the 13th slot in the table which holds the number 30 (the 12th slot holds the number 27, the 14th the number 35 etc). That EG time value will last the length of 30*32 samples at a 48kHz sample rate. Thus all you need to do is multiply 30*32=960samples / 48,000(the number of samples per second) to find that 960 samples = 0.02 seconds or 20 milliseconds. Took me a minute to calculate all 100 EG values in this way so I hope you make good use of it!

PS. Yes the list will be amended after I do a test for EG time value 0 and write Andy about that anomaly.
_________________
paypal.me/CharlesFerraro

[timbukktwo]

I'll be the first to say thank you for this! Interesting stuff not commonly known by most of us!!

[psionic311]

Music is math, and music is emotion, and music is...

[Bachus]

Very interesting..

But in general, when changing an envelope, i changed it by what i hear, i dont even look at the specific numbers that much..

But nice to know, just need to remmeber that the real thing is aproximately 2/3 th the vallue of the screenvalue..
_________________
Korg Kronos 2/88 , Genos, Mainstage3 +VSTsu, ipad pro, GSi Gemini, Roland Integra 7, Jupiter Xm, Yamaha motif XS rack, Ketron SD90.
www.keyszone.boards.net

[CharlesFerraro]

Yes a time value setting of both zero and one both last 0.667ms or 32 samples at a 48kHz sample rate. That's about as long as a single cycle at 1500Hz so any sound below that frequency would suffer an audible click if say the attack or decay times were 0 or 1.


Attack time 0



Attack time 1



I was able to make those by throwing noise through the amp and setting the amp envelope so the start and attack were at full level with zero sustain. Then I could simply set the attack time to how long I wanted the noise to sound while recording.
_________________
paypal.me/CharlesFerraro