Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Tip: Randomly selecting alternating patterns
Or a table from the FX column, overruled by a range so it chooses between like 3-4 tables with different settings and this column is also additionally happening probably... just to mention some options which could do some of the mentioned.
PLEASE use the search function if something have been asked or discussed before.
Every (unnessesary) forum support means less time to develop! But of course, i am here to help!  Smile
(01-22-2021, 05:39 PM)mgd Wrote: Maybe I misunderstand your setup. Your P(01) is my P(04) and vice versa. Same for P(02) and P(03).

However when the first decision made is "Pattern 01 or not" then that's the 25% chance for 01. When I understand your setup correctly, there is no other possible path to pattern 01 and therefor that's it. A similar argument goes for the rest.

But since you basically solved your problem (nice one actually Smile ) this is not really important

Kind regards,

Nice thread and idea!
Not that it is very important but I think I see grib’s point:
-There is indeed a 25% chance for 01 to win on the first row
-But for 01 to happen you then don’t want neither the 2nd, 3rd or 4th row to win otherwise 01 will be overwritten (by 02, 03 or 04). The last winner takes it all!

  • the probability for 01 is 25% * 75% * 75% * 75% (win, lose, lose, lose)
  • the probability for 02 is 100% * 25% * 75% * 75% (does not matter, win, lose, lose)
  • the probability for 03 is 100% * 100% * 25% * 75% (does not matter, does not matter, win, lose)
  • the probability for 04 is 100% * 100% * 100% * 25% (does not matter, does not matter, does not matter, win)
  • the probability for 00 is 75% * 75% * 75% * 75% (lose, lose, lose, lose)
Fun idea, really!
I'm not quite sure I understand what you write or imply.

Events aren't concurrent but happen one after another. Therefor the relevant probabilities form a decision tree (Entscheidungsbaum would be the german word). In the current example each branch is essentially a YES/NO decision and each branch gets a probability Pn (for the n-th node) with the other branch getting 1-Pn [stochastically you could have more than two branches but that's not what we have here].

The first node in our tree is "play pattern 01 or not". Likewise the other nodes would be "play pattern 02 or not", "play pattern 03 or not" and "play pattern o4 or not".

Here comes an ascii graph depicting that:

 |  (1-P1)        (1-P2)        (1-P3)        (1-P4)
 +-------------+-------------+-------------+-------------> Pattern 00
 |  P1         |  P2         |  P3         |  P4
 V             V             V             V
Pattern 01   Pattern 02   Pattern 03   Pattern 04

The '+' sign denotes a node where a decision is made. The vertices of the graph have their respective probabilities annotated. The probabilities for each single event are determined by walking the path from start to the event and multiplying the probabilities that are on each vertice along the path. I'm not repeating them here, because that's exactly what I wrote in a previous post.

For those not understanding the above I'm happy to try to explain it differently (if you ask). Otherwise I'd suggest one of the many excellent sources on stochastics on the internet.

Kind regards,

PS: The spacing is completely off...I'll fix that later
PPS: Spacing is now fixed.
(02-24-2021, 04:06 PM)mvdirty Wrote: While the release candidate NerdSEQ firmware is gaining more control over its existing randomization functionality, it won’t (that I know of) yet be gaining the kind of a) generated repeating variation or b) slow evolution of its randomization, either of which you’d be anticipating from Bloom. On the upside, the NerdSEQ isn’t a hot mess of UX issues and bugs. Big Grin

Amen to that. Actually Bloom is ending up mostly for percussion. I find it most useful for that. The warts are less apparent.

You guys are giving me a headache, but I might try to figure out your random thoughts on the weekend  Rolleyes

You see what I did there?  Big Grin
(02-26-2021, 06:13 AM)mntbighker Wrote: Actually Bloom is ending up mostly for percussion.

Thanks for mentioning that. Before I became too pissed off with it overall to even touch it, I had experimented with using it for modulation but admittedly hadn’t tried finding a use for it in percussion. That is definitely worth a try.
If I just want to loop through a selected list of patterns (tracks) while omitting the rest, what is the simplest way? So at the simplest level play 3, then 4 and loop. Do not play them at the same time. Can I include 7 and 8 in the loop?
Make them an island and jump to them.
(03-04-2021, 10:09 PM)mgd Wrote: Make them an island and jump to them.

Just in terms of searching the manual for jump and island you would not find the way to do that. At least it wasn't obvious to me. Yet it seems like I have stumbled into that before. Right now I'm just not seeing it. Maybe I saw that somewhere on YT.
Page 17 (first page of explanation what the sequencer screen looks like).

An island (on a track) is any isolated block/chain of patterns. Isolated means there is no pattern directly before or directly after the block/chain (-- before and after). Similar for several tracks.

You jump to a pattern with FX STSP trr where t is the track number aka sequencer screen column (or 0 for all) and rr is the sequencer screen row. You'll find that explained on page 27, last list item.

Does that help?

Kind regards,

PS: All page numbers refer to Version 1.23 of the manual.
(03-04-2021, 11:04 PM)mntbighker Wrote:
(03-04-2021, 10:09 PM)mgd Wrote: Make them an island and jump to them.

Just in terms of searching the manual for jump and island you would not find the way to do that. At least it wasn't obvious to me. Yet it seems like I have stumbled into that before. Right now I'm just not seeing it. Maybe I saw that somewhere on YT.

Just to quickly add to what mgd said: And you need not use FX for this necessary, for example if you are performing with the song screen, as you can instead just set up the islands and then either start the track with the cursor on the first pattern of the island or you can start the whole row.

(02-24-2021, 04:11 PM)XORadmin Wrote:
(02-24-2021, 04:06 PM)mvdirty Wrote: Perhaps, as an example, an FX command that permanently perturbs the note value in the row where it appears, within a controlled range, each time the FX command is executed. That could then be driven probabilistically using NerdSEQs existing probability FX.

You mean like a 1 shot table with a random value within a selected range on it's first row and a hop 0 so it happens only once and this happens probabilistic (does this word happen anyway??? you understand me).'s there Big Grin

(I somehow missed your reply that day and just spotted it now.)

Not quite, because the other person is not looking for randomization of the note every time the pattern passes that row. They are instead looking for the pattern’s note value on that row to be permanently altered, at least in terms of my layman’s attempt to map the workflow of their module of reference onto what is likely reasonable with NerdSEQ.

Constrained by FX probability, this would randomize a given row’s note on average every N passes through the pattern, likely with low percentage and therefore a high value of N, but would continue to play that specific randomized note on subsequent passes until FX probability causes another randomization to occur, at which point that row would continue playing _that_ note for a while, and so on.

Forum Jump:

Users browsing this thread: 1 Guest(s)