Advanced Delay Time Cheat-Sheet

Rusty · 2015-08-05T10:31:28.000Z

Nice one

Nymphomation · 2015-08-05T12:34:15.000Z

I was looking for exactly this. Much appreciated

dumbledog · 2015-08-05T14:09:55.000Z

Here’s a chart I whipped up for filter 1’s cutoff frequency:

Screen_Shot_2015_08_05_at_7_07_40_AM.png1157×789 48.8 KB

PeterHanes · 2015-08-05T14:27:22.000Z

""dumbledog"":

Here’s a chart I whipped up for filter 1’s cutoff frequency:

Screen_Shot_2015_08_05_at_7_07_40_AM.png1157×789 48.8 KB

Why not just say that it’s a semitone per knob increment?

dumbledog · 2015-08-05T14:54:42.000Z

Because it’s not. I don’t know why but 440Hz is at 59.17 and 880Hz is 70.74 – which would imply a semitone every… 0.96 knob steps or so.

But this probably varies on the device. My Track 3 voice for example is noticeably flatter than this – 70.74 on Track 3 is more like 860Hz. And yeah that sucks if you’re trying to get it to track to the keyboard.

I made that chart before Overbridge as well. Now that we have precise two-decimal values for the filter I can probably tighten that curve up a bit.

PeterHanes · 2015-08-05T15:33:58.000Z

""dumbledog"":

Because it’s not. I don’t know why but 440Hz is at 59.17 and 880Hz is 70.74 – which would imply a semitone every… 0.96 knob steps or so.

…

Would be nice if you used a log scale for the frequency, like a musician would.

avantronica · 2015-08-05T15:34:57.000Z

""dumbledog"":

Now that we have precise two-decimal values for the filter I can probably tighten that curve up a bit.

That’s imprecise actually - the underlying ‘raw’ data is 14bit, that’s (in OB) a clumsy way to represent it - if you have a midi monitor you’ll see the fine detail coming through on the LSB part of the NRPN string

Manoeuvre · 2015-08-12T02:21:57.000Z

Thanks for the delay time cheat sheet! Do these figures apply in any way to the quantise values? Why is there no musical translation built into the Elektron Machines? Is this a mad mathematician’s ego trip? Watching all of the musicians counting on their fingers and laughing sinisterly in the background? For me the maths dominated approach is the weak point, in Elektron’s otherwise, very creative sound machines.

roboldham · 2015-08-17T18:04:30.000Z

""snowcrash"":

I did some math on the delay times and how they relate to measures, metrics and rhythm… also helps which numbers could be useful etc… I don’t know if I did any mistakes (I’m unsure about the parameter-range since I’m not with my A4 right now). Also I didn’t check if this values work out on the actual machine. If someone finds the time, to test this, I’d be glad to update the link…

Ah great - always wondered why “40” sounded so nice
thanks for the efforts!

Sammelkorn · 2016-01-01T16:02:55.000Z

[quote=“” snowcrash""]
I did some math on the delay times and how they relate to measures, metrics and rhythm… also helps which numbers could be useful etc… I don’t know if I did any mistakes (I’m unsure about the parameter-range since I’m not with my A4 right now). Also I didn’t check if this values work out on the actual machine. If someone finds the time, to test this, I’d be glad to update the link…

A4DelayTimeMath_v0_1.png1200×1957 120 KB

PDF Version:
https://dl.dropboxusercontent.com/u/6245022/Gnankn/A4DelayTimeMath.v0.1.pdf
[/quote]

First of all thanks for creating this chart, really cleared up the delay for me! The colour coding is awesome too.

I think I found something inaccurate in here, though. Maybe it’s me that is looking at this the wrong way, I guess we’ll see…
What i mean concerns the the triplets in the chart. I think that the red lines, the supposed triplets, are actually dotted notes. Let’s put this hypothesis to the test:

1 dotted 1/8th-note = 1/8th-note*1,5 = 3/16

or numerically:

0,5 [beats] *1,5 = 0,75

The resulting values (0,75; 3/16) show up in the same line in the chart. That line is labelled “1/4tri[plet]”. Unless I have taken a wrong corner in my thought process here, I think that label should be “dotted 1/8”.

Maybe the reason why this got mixed up was that, if you look at the chart, the column “fraction”, which shows the most reduced fraction, does have numbers with periodical ,3333 ,66666 (e.g. 3/32 = 1/10,666666…). Of course this looks like we’ve got a triplet, but not really actually. The ,333 or ,6666 needs to be the in the column “beats” to make it a triplet

I was bummed when I discovered this because i quickly realised there are no triplets at all in the delay on the A4.
An 1/8 triplet would have to be 1/12, which simply doesn’t appear between 1/128 (the increments of the delay) an 128/128. (That’s basically saying 1/12 and 1/128 don’t have a common denominator).

Oh well, doted notes sound great too

Oh and just an afterthought:
I haven’t looked at the other -plets…

EDIT: Dunno what these findings means for the quintuplets etc. in the chart.
EDIT: It was Lars Erik that got me on this idea with this post: Delay & Tempo Sync - #7 by matic

mhek · 2017-11-06T00:26:10.469Z

Hi @Sammelkorn its possible to re-upload the Delay time pdf?
File not found : /
Thanks!

snowcrash · 2017-11-06T08:44:01.636Z

Dropbox changed their sharing api, I updated the post above, here’s the new link for the original pdf:

dropbox.com

A4DelayTimeMath.v0.1.pdf

snowcrash · 2017-11-06T08:56:16.000Z

@Sammelkorn thanks for the input on triplets vs. dotted. I will look into this, might be I was wrong. I didn’t measure the actual delay times. One can do proper triplets on the note grid though with micro-timing.

I don’t have my A4 with me right now, so I don’t know the exact value for the three steps (there are 2 solutions for that) but they must be somewhere there:

snowcrash · 2017-11-06T09:13:42.348Z

Btw. still very useful:
http://www.sengpielaudio.com/calculator-bpmtempotime.htm

sezare56 · 2017-11-06T12:28:20.457Z

I like to use this sometimes :

When you play a fast arp, lfo (or retrig on other lachines), or 32th with scale 2x, you can have a note depending on tempo.

A = 440 Hz. 440 x 60 = 26 400
Divide by 2 until you get common tempo :
A > 103.125 bpm

To find tempo one semi-tone above, you have to multiply by 2^(1/12)=1,0594630944

That number can be used to find frequencies too, starting from 440 Hz. Ex with 5 semi tones above :
440 (A4) x 2^(1/12)^5 = 587.3 (D5)

Silmae · 2020-10-15T12:07:06.494Z

Very nice! I was thinking of creating something like that, but search brought me to your post and you have done a detailed job! Thank you!

JeffCrossSF · 2024-03-11T02:43:46.883Z

I just picked up the Analog Heat + FX. yay!

Setting delays is a confusing mess. Why??? I’d love to get a modern answer to this issue. Elektron, why make it difficult to get simple musical values for my delay time?

Tchu · 2024-03-11T02:47:08.996Z

If you push and turn the Time Encoder, you will get the musical values.

JeffCrossSF · 2024-04-04T03:51:42.415Z

wow, thank you! I wish this was default and push and hold was whatever that is…

Andreas_R · 2024-04-04T04:27:27.555Z

Very handy and inviting to test out more values other then the usual suspects. 🩶🤘