Tbh this is more like formant synthesis + concatenative synthesis. Well. One day I’ll hopefully accomplish…just formant synthesis
November 3, 2026 at 4:43 AM
Tbh this is more like formant synthesis + concatenative synthesis. Well. One day I’ll hopefully accomplish…just formant synthesis
Concatenative Combinators aren’t a #concatenativelanguage in the strict sense. Although it’s true that any two valid programs catenated together form a valid program, it is not true that any bisection of a program results in two valid programs.
This is because concatenative combinators includes […]
This is because concatenative combinators includes […]
Original post on functional.cafe
functional.cafe
November 3, 2026 at 9:35 PM
Concatenative Combinators aren’t a #concatenativelanguage in the strict sense. Although it’s true that any two valid programs catenated together form a valid program, it is not true that any bisection of a program results in two valid programs.
This is because concatenative combinators includes […]
This is because concatenative combinators includes […]
I have not failed to write a fix point in #concatenative combinators, I have found nearly 10,000 constructions that won't work.
Now switching my focus and examining the algorithm for rewriting programs. Perhaps I’m looking for the problem in the wrong place.
Now switching my focus and examining the algorithm for rewriting programs. Perhaps I’m looking for the problem in the wrong place.
November 6, 2026 at 6:56 PM
I have not failed to write a fix point in #concatenative combinators, I have found nearly 10,000 constructions that won't work.
Now switching my focus and examining the algorithm for rewriting programs. Perhaps I’m looking for the problem in the wrong place.
Now switching my focus and examining the algorithm for rewriting programs. Perhaps I’m looking for the problem in the wrong place.
Well, I ruled transparency out. Although the original paper doesn't dictate whether concatenative combinators have transparent or opaque quotations, I implemented transparent quotations, a/k/a "rewriting all the way down."
My attempt to write the Z […]
[Original post on functional.cafe]
My attempt to write the Z […]
[Original post on functional.cafe]
![This construction has a bit of a caveat, which should be pointed out: in order for this construction to work out, we had apply rewrite rules inside of a quotation. If a particular formalism considers this to be valid, then we say that it has transparent quotation, whereas if this is not valid, then it has opaque quotation. The standard Joy interpreter actually has opaque quotation in that "[[B] [A] [i] dip i]" is considered distinct from "[B A]", even though both quotations would do the same thing when executed (they are distinct because, for example, "size" would give "5" in the first case and perhaps "2" in the second). However, even though the two above quotations can be distinguished, in many contexts they would be considered interchangable.](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:tdu2ua5n5twpnw3w3nzqe6pq/bafkreidmah2z3okmrw6wpgyhswa6wjyshy3kity5aznlln74kgjy7wojwy)
November 6, 2026 at 10:25 PM
Well, I ruled transparency out. Although the original paper doesn't dictate whether concatenative combinators have transparent or opaque quotations, I implemented transparent quotations, a/k/a "rewriting all the way down."
My attempt to write the Z […]
[Original post on functional.cafe]
My attempt to write the Z […]
[Original post on functional.cafe]
The factorial combinator in an implementation of Concatenative Combinators. CC is a concatenative language, so operators come after their arguments, which are quoted with [ and ]. Programs, subroutines, whatever you want to call them are also quoted.
There’s […]
[Original post on functional.cafe]
There’s […]
[Original post on functional.cafe]
![export const fact = construction([ [A] ],
[
[1], swap,
[
bury2,
dup, [2],
isLessThan,
[
dup,
[[multiplier], cons, cons], dip,
[pred], cons,
dig2,
m
],
[
zap, nip
],
cond
],
m,
i
]
);](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:tdu2ua5n5twpnw3w3nzqe6pq/bafkreidyuxlypoaagx4wa5ip6mh7evdiu3fneuuxwv3xikkxptolr36kz4)
October 27, 2026 at 4:26 PM
The factorial combinator in an implementation of Concatenative Combinators. CC is a concatenative language, so operators come after their arguments, which are quoted with [ and ]. Programs, subroutines, whatever you want to call them are also quoted.
There’s […]
[Original post on functional.cafe]
There’s […]
[Original post on functional.cafe]
![import { B, A } from "../combinator-tools/placeholders";
import { render } from "../combinator-tools/render";
import { l } from "./l";
test("l combinator", () => {
expect(render([ [B], [A], l ])).toEqual(render([ [B], B, A ]))
});](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:tdu2ua5n5twpnw3w3nzqe6pq/bafkreiezvvgedtz6vqk437gwsvolftnmnuna6ixd2m6zvctjmy5umopvfy)
October 26, 2026 at 5:01 PM
Example: Here is euclidean division implemented in concatenative combinators. There are two M combinators, on lines 37 and 29.
The one on line 37 invokes the recursive program in lines 21 – 36 while putting a copy in front of the existing two parameters […]
[Original post on social.bau-ha.us]
The one on line 37 invokes the recursive program in lines 21 – 36 while putting a copy in front of the existing two parameters […]
[Original post on social.bau-ha.us]
![export const euclideanDivisor = construction([ [B], [A] ], [
[0], bury2,
[
bury3,
peek, peek,
isLessThan,
[
[[subtrahend], cons, cons], sip,
[[succ], cons], dip2,
dig3,
m
],
[
zap,
[zap], dip2
],
cond
],
m
]);](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:7lhbtvgyqc376j7eh3vkjvqf/bafkreidwpco43wrduht5cqgtukvlblqcbpkgb7gvbhzjdptdslva66wfoe)
October 25, 2026 at 6:53 PM
Example: Here is euclidean division implemented in concatenative combinators. There are two M combinators, on lines 37 and 29.
The one on line 37 invokes the recursive program in lines 21 – 36 while putting a copy in front of the existing two parameters […]
[Original post on social.bau-ha.us]
The one on line 37 invokes the recursive program in lines 21 – 36 while putting a copy in front of the existing two parameters […]
[Original post on social.bau-ha.us]
The first cut of euclidean division in concatenative combinators good enough to dismiss with scorn:
https://tunes.org/~iepos/joy.html
https://tunes.org/~iepos/joy.html
![// [B] [A] euclideanDividedBy == [B \ A] [B % A]
export const euclideanDividedBy = construction([[B], [A]], [
// [dividend] [divisor]
[zero], bury, // [quotient] [dividend] [divisor]
[
bury3, // [callback] [quotient] [dividend] [divisor]
dup, // [callback] [quotient] [dividend] [divisor] [divisor]
[swap], dip, // [callback] [quotient] [divisor] [dividend] [divisor]
[dup], dip, // [callback] [quotient] [divisor] [dividend] [dividend] [divisor]
isLessThan, // [callback] [quotient] [divisor] [dividend] [T/F: dividend < divisor]
[
// [callback] [quotient] [divisor] [dividend]
swap, dup, bury, // [callback] [quotient] [divisor] [dividend] [divisor]
[minus], cons, cons, // [callback] [quotient] [divisor] [dividend - divisor]
bury, swap, // [callback] [dividend - divisor] [divisor] [quotient]
[succ], cons, // [callback] [dividend - divisor] [divisor] [quotient++]
bury2, // [callback] [quotient++] [dividend - divisor] [divisor]
dig3, // [quotient++] [dividend - divisor] [divisor] [callback]
m
],
[
// [callback] [quotient] [divisor] [dividend]
nip, // [callback] [quotient] [dividend]
dig2, zap // [quotient] [dividend]
],
cond // [quotient] [remainder]
],
m, // [quotient] [remainder]
]);](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:tdu2ua5n5twpnw3w3nzqe6pq/bafkreifguyijsr2hbxphpsd3p5pqjszb4dpjiofqcfjodlpdzpnieparxa)
October 18, 2026 at 8:33 PM
The first cut of euclidean division in concatenative combinators good enough to dismiss with scorn:
https://tunes.org/~iepos/joy.html
https://tunes.org/~iepos/joy.html
The M combinator (aka “Mockingbird”) in action: The Mockingbird and Sage Bird (Y Combinator) are Computer Science’s primordial implementations of recursion.
https://tunes.org/~iepos/joy.html
Here we are implementing Euclidean Division in concatenative […]
[Original post on functional.cafe]
https://tunes.org/~iepos/joy.html
Here we are implementing Euclidean Division in concatenative […]
[Original post on functional.cafe]
![// n.b. returns quoted numbers, not applicable numbers,
// as it is necessary to select whether one wants the quotient, remainder, or both.
const euclideanDividedBy = construction([ [B], [A] ], [
[zero], bury, // [quotient] [dividend] [divisor]
[
bury3, // [callback] [quotient] [dividend] [divisor]
dup, // [callback] [quotient] [dividend] [divisor] [divisor]
[swap], dip, // [callback] [quotient] [divisor] [dividend] [divisor]
[dup], dip, // [callback] [quotient] [divisor] [dividend] [dividend] [divisor]
isLessThan, // [callback] [quotient] [divisor] [dividend] [T/F: dividend < divisor]
[
// [callback] [quotient] [divisor] [dividend]
[[minus], cons, cons], sip, // [callback] [quotient] [dividend - divisor] [divisor]
dig2, // [callback] [dividend - divisor] [divisor] [quotient]
[succ], cons, // [callback] [dividend - divisor] [divisor] [quotient++]
bury2, // [callback] [quotient++] [dividend - divisor] [divisor]
dig3, // [quotient++] [dividend - divisor] [divisor] [callback]
m
],
[
// [callback] [quotient] [divisor] [dividend]
nip, // [callback] [quotient] [dividend]
dig2, zap // [quotient] [dividend]
],
cond // [quotient] [remainder]
],
m, // [quotient] [remainder]
]);](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:tdu2ua5n5twpnw3w3nzqe6pq/bafkreie4fkqktvzzsuq55r4stoeojyqio5enaubjvpeoiemal6afeiwwyy)
October 16, 2026 at 2:13 PM
The M combinator (aka “Mockingbird”) in action: The Mockingbird and Sage Bird (Y Combinator) are Computer Science’s primordial implementations of recursion.
https://tunes.org/~iepos/joy.html
Here we are implementing Euclidean Division in concatenative […]
[Original post on functional.cafe]
https://tunes.org/~iepos/joy.html
Here we are implementing Euclidean Division in concatenative […]
[Original post on functional.cafe]
Neat - I was attempting something similar in 2010 using regression trees proceedings.mlr.press/v11/stowell1... so I'm pleased to see your much larger-scale work. My use-case was mapping two different situations (datasets) in same modality, so I'll point out that yr method can be within-modality too
Cross-associating unlabelled timbre distributions to create expressive musical mappings
In timbre remapping applications such as concatenative synthesis, an audio signal is used as a template, and a mapping process derives control data for some audio synthesis algorithm such that it p...
proceedings.mlr.press
October 12, 2026 at 3:55 PM
Neat - I was attempting something similar in 2010 using regression trees proceedings.mlr.press/v11/stowell1... so I'm pleased to see your much larger-scale work. My use-case was mapping two different situations (datasets) in same modality, so I'll point out that yr method can be within-modality too
If it can maintain its concatenative-esque charm with the bells and whistles of AI... we could really have something special on our hands.
I've been initially impressed and then been let down by Yamaha before though, so I'm just gonna wait and see. I like what I hear so far though.
I've been initially impressed and then been let down by Yamaha before though, so I'm just gonna wait and see. I like what I hear so far though.
October 11, 2026 at 2:15 AM
If it can maintain its concatenative-esque charm with the bells and whistles of AI... we could really have something special on our hands.
I've been initially impressed and then been let down by Yamaha before though, so I'm just gonna wait and see. I like what I hear so far though.
I've been initially impressed and then been let down by Yamaha before though, so I'm just gonna wait and see. I like what I hear so far though.
alexander panos botanica sound collage alon mor foley water sounds phritz minimeters twitter clip nurture vibes all about lily chou chou intro nascent close up of a flower harmonizer bon iver concatenative synthesis granulator max msp kontakt orchestral tools zoomer sound design
October 9, 2026 at 7:27 PM
alexander panos botanica sound collage alon mor foley water sounds phritz minimeters twitter clip nurture vibes all about lily chou chou intro nascent close up of a flower harmonizer bon iver concatenative synthesis granulator max msp kontakt orchestral tools zoomer sound design
also we were robbed of getting an english concatenative vocaloid of adachi rei because v6 crosslingual synthesis was less work to develop. that would've actually been cool unlike another engine that smooths out the harsh sonic qualities that make her voice so beautiful.
October 9, 2026 at 8:01 PM
also we were robbed of getting an english concatenative vocaloid of adachi rei because v6 crosslingual synthesis was less work to develop. that would've actually been cool unlike another engine that smooths out the harsh sonic qualities that make her voice so beautiful.
Hand-written tests for porting boolean operators from the Lambda Calculus and/or Smullyan Forest to Concatenative Combinators. Yet another reminder to...
“Beware of the Turing tar-pit in which everything is possible but nothing of interest is easy.”—Dr. Alan J. Perlis, 1982
“Beware of the Turing tar-pit in which everything is possible but nothing of interest is easy.”—Dr. Alan J. Perlis, 1982
![// [B] [A] And == [B && A]
test("boolean and", () => {
expect(display(rewrite([ [False], [False], And ]))).toEqual(display(rewrite([ [False] ])));
expect(display(rewrite([ [False], [True], And ]))).toEqual(display(rewrite([ [False] ])));
expect(display(rewrite([ [True], [False], And ]))).toEqual(display(rewrite([ [False] ])));
expect(display(rewrite([ [True], [True], And ]))).toEqual(display(rewrite([ [True] ])));
});
// [B] [A] Or == [B || A]
test("boolean or", () => {
expect(display(rewrite([ [False], [False], Or ]))).toEqual(display(rewrite([ [False] ])));
expect(display(rewrite([ [False], [True], Or ]))).toEqual(display(rewrite([ [True] ])));
expect(display(rewrite([ [True], [False], Or ]))).toEqual(display(rewrite([ [True] ])));
expect(display(rewrite([ [True], [True], Or ]))).toEqual(display(rewrite([ [True] ])));
});
// [A] Not == [¬A]
test("boolean not", () => {
expect(display(rewrite([ [False], Not ]))).toEqual(display(rewrite([ [True] ])));
expect(display(rewrite([ [True], Not ]))).toEqual(display(rewrite([ [False] ])));
});
// [B] [A] Xor == [(B || A) && ¬(B && A)]
test("boolean or", () => {
expect(display(rewrite([ [False], [False], Xor ]))).toEqual(display(rewrite([ [False] ])));
expect(display(rewrite([ [False], [True], Xor ]))).toEqual(display(rewrite([ [True] ])));
expect(display(rewrite([ [True], [False], Xor ]))).toEqual(display(rewrite([ [True] ])));
expect(display(rewrite([ [True], [True], Xor ]))).toEqual(display(rewrite([ [False] ])));
});](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:tdu2ua5n5twpnw3w3nzqe6pq/bafkreicxis6ini7ovwo4nhaw3swqtvevxrdo7ppumyuwhbjivdqaxoy64e)
October 8, 2026 at 6:36 PM
Hand-written tests for porting boolean operators from the Lambda Calculus and/or Smullyan Forest to Concatenative Combinators. Yet another reminder to...
“Beware of the Turing tar-pit in which everything is possible but nothing of interest is easy.”—Dr. Alan J. Perlis, 1982
“Beware of the Turing tar-pit in which everything is possible but nothing of interest is easy.”—Dr. Alan J. Perlis, 1982
Factor Overview https://lobste.rs/s/bwtzhy ##concatenative
Factor Overview
re.factorcode.org
October 5, 2026 at 12:15 PM
Factor Overview https://lobste.rs/s/bwtzhy ##concatenative
Keren Shao, Ayaka Kawano, Shlomo Dubnov: Relational Synthesis: Structure-Mediated Concatenative Synthesis for Foley and Retrieval-Augmented Audio Generation https://arxiv.org/abs/2610.05768 https://arxiv.org/pdf/2610.05768 https://arxiv.org/html/2610.05768
October 6, 2026 at 6:44 AM
Keren Shao, Ayaka Kawano, Shlomo Dubnov: Relational Synthesis: Structure-Mediated Concatenative Synthesis for Foley and Retrieval-Augmented Audio Generation https://arxiv.org/abs/2610.05768 https://arxiv.org/pdf/2610.05768 https://arxiv.org/html/2610.05768
why tf does cymatics have a concatenative synth lmaoo
more power to them but i do not think their target audience would appreciate it let alone know how to use it
(not tryna be elitist btw i dont think alot of people in general get concatenative synthesis including me fully lmao )
more power to them but i do not think their target audience would appreciate it let alone know how to use it
(not tryna be elitist btw i dont think alot of people in general get concatenative synthesis including me fully lmao )
October 3, 2026 at 6:46 PM
why tf does cymatics have a concatenative synth lmaoo
more power to them but i do not think their target audience would appreciate it let alone know how to use it
(not tryna be elitist btw i dont think alot of people in general get concatenative synthesis including me fully lmao )
more power to them but i do not think their target audience would appreciate it let alone know how to use it
(not tryna be elitist btw i dont think alot of people in general get concatenative synthesis including me fully lmao )
I find myself missing commercial concatenative synths. Lately, AI synths haven't been as interesting to me.
Partially because of all the problems in the last post I made, ( and that I find DiffSinger unpleasant to tune in ). But also because I feel there's kind of a push for realistic voices
Partially because of all the problems in the last post I made, ( and that I find DiffSinger unpleasant to tune in ). But also because I feel there's kind of a push for realistic voices
September 29, 2026 at 11:39 PM
I find myself missing commercial concatenative synths. Lately, AI synths haven't been as interesting to me.
Partially because of all the problems in the last post I made, ( and that I find DiffSinger unpleasant to tune in ). But also because I feel there's kind of a push for realistic voices
Partially because of all the problems in the last post I made, ( and that I find DiffSinger unpleasant to tune in ). But also because I feel there's kind of a push for realistic voices