RapidComposer v5 beta feedback and discussion
-
- KVRAF
- 2150 posts since 15 May, 2017
The updated user manual is now online. I suggest looking at this, if you have any questions, before asking here..
- KVRian
- 1177 posts since 13 Mar, 2017
Hi.
When I double-click in a track header area on the left side of the screen, that track zoom-expands to fill the screen. However, the notes remain small and unlabeled until/unless I ctrl-drag on the keyboard to biggen and label them and then drag to center them vertically.
Then when we double-click the track header to go back to normal track vertical size, the gigantic notes (which were desired in the fully expanded view) remain (the original size/location is not restored).
I think it would be nicely helpful if each track could remember 2 separate zoom/location settings - one for the "normal" un-double-clicked situation, and one for the expanded double-clicked situation.
If such is currently somehow settable, please advise. Thanks!
When I double-click in a track header area on the left side of the screen, that track zoom-expands to fill the screen. However, the notes remain small and unlabeled until/unless I ctrl-drag on the keyboard to biggen and label them and then drag to center them vertically.
Then when we double-click the track header to go back to normal track vertical size, the gigantic notes (which were desired in the fully expanded view) remain (the original size/location is not restored).
I think it would be nicely helpful if each track could remember 2 separate zoom/location settings - one for the "normal" un-double-clicked situation, and one for the expanded double-clicked situation.
If such is currently somehow settable, please advise. Thanks!
-
- KVRAF
- 2150 posts since 15 May, 2017
Can you post a screen shot of this?sj1 wrote: Thu Oct 19, 2023 1:27 pm Hi.
When I double-click in a track header area on the left side of the screen, that track zoom-expands to fill the screen. However, the notes remain small and unlabeled until/unless I ctrl-drag on the keyboard to biggen and label them and then drag to center them vertically.
- KVRian
- 1177 posts since 13 Mar, 2017
Here is a short video which basically follows the description steps above.
https://www.screencast.com/t/f2WklPsswP
There is no sound in this one.
https://www.screencast.com/t/f2WklPsswP
There is no sound in this one.
-
- KVRAF
- 2150 posts since 15 May, 2017
- KVRian
- 1177 posts since 13 Mar, 2017
-- Below is some AI-related commentary copied forward from the V4 Forum . Cheers! --
In RC-type terms, I could see genre-oriented "intelligence" being applied to chord progression generation. This could be a direct-to-chord-progression sort of thing, or it could be a genre-to-chord-rule-set sort of thing. I think the question here becomes "Where does the genre information come from?". Should Attila himself pursue the mastery and cataloging of genre-oriented rule info to build such into RC, or should he build RC in a way that the users can select and catalog genre-oriented rules on their own?
Offhand, I tend to think the latter makes more sense. The issue then becomes (in current RC terms) how can the user ingest chord progression info efficiently prior to converting it into chord rules sets. For this I advocate for MusicXML file import into the Master Track. From there, we already have what we need to build custom specific or hybrid CRSs (very cool!).
The other primary high-level compositional need IMO is for structure generation. In RC-terms, this would be a meta-level generator of Lines and Parts. I'd love it if the user could supply a tempo range and a running-length range and then have RC "dice" create structures automatically. I would like to see these displayed in a high-level visual allowing the whole structure to be seen and then accepted (or dismissed) by the user for instantiation into the timeline.
Workflow would then be:
0. setup up VSTs, tracks, etc. for desired instruments
1. choose approximate length and tempo of the piece
2. generate candidate structures and choose one
3. generate chord progressions into the structure until satisfied
4. everything else, combining generation and manual additions at any desired granularity
Then, an even higher meta-level would be to have RC do 1-4 by itself all with a single roll of the dice.
IMO, working toward this (or some similar) high-level vision is what will keep RC relevant and important going forward into the age of AI.
Once there is a "completed efficient pipeline" for compositional workflow, then any amount or form of AI assistance can be grafted in at any level to keep the product moving into the future with panache!
In RC-type terms, I could see genre-oriented "intelligence" being applied to chord progression generation. This could be a direct-to-chord-progression sort of thing, or it could be a genre-to-chord-rule-set sort of thing. I think the question here becomes "Where does the genre information come from?". Should Attila himself pursue the mastery and cataloging of genre-oriented rule info to build such into RC, or should he build RC in a way that the users can select and catalog genre-oriented rules on their own?
Offhand, I tend to think the latter makes more sense. The issue then becomes (in current RC terms) how can the user ingest chord progression info efficiently prior to converting it into chord rules sets. For this I advocate for MusicXML file import into the Master Track. From there, we already have what we need to build custom specific or hybrid CRSs (very cool!).
The other primary high-level compositional need IMO is for structure generation. In RC-terms, this would be a meta-level generator of Lines and Parts. I'd love it if the user could supply a tempo range and a running-length range and then have RC "dice" create structures automatically. I would like to see these displayed in a high-level visual allowing the whole structure to be seen and then accepted (or dismissed) by the user for instantiation into the timeline.
Workflow would then be:
0. setup up VSTs, tracks, etc. for desired instruments
1. choose approximate length and tempo of the piece
2. generate candidate structures and choose one
3. generate chord progressions into the structure until satisfied
4. everything else, combining generation and manual additions at any desired granularity
Then, an even higher meta-level would be to have RC do 1-4 by itself all with a single roll of the dice.
IMO, working toward this (or some similar) high-level vision is what will keep RC relevant and important going forward into the age of AI.
Once there is a "completed efficient pipeline" for compositional workflow, then any amount or form of AI assistance can be grafted in at any level to keep the product moving into the future with panache!
- KVRian
- 1177 posts since 13 Mar, 2017
I'd like to plug the idea of "structure presets" here as potentially very useful for RC as it evolves to higher meta levels of Composition creation.
Here is a picture of perhaps the best implementation I've yet seen of the "structure preset" concept, Jamstix 4:
Notice how the each preset in the list of presets is a simple, but powerful specification of song structure, using specific single letters for "V"erse, "C"horus, "B"ridge, etc.
Then notice how a chosen preset gets translated into an actual song structure (where the lengths are defined, and edited if desired) on the left.
It is dead-simple to define a new structure and instantiate it.
Especially since you don't even have to use the UI if you don't want to. The structure preset definition are read in from a text file.
Brilliant! (IMO)
RC gives us thousands of phrases to work with (and allows for an infinite number of definitions) . Why not do the same thing for Structures?
Heck, I don't think anything would be wrong with adopting the exact convention shown above, then we could have the same structures defined for both RC and Jamstix, perfectly matched up and ready to go.
Synergy! I love it. Cheers!
Here is a picture of perhaps the best implementation I've yet seen of the "structure preset" concept, Jamstix 4:
Notice how the each preset in the list of presets is a simple, but powerful specification of song structure, using specific single letters for "V"erse, "C"horus, "B"ridge, etc.
Then notice how a chosen preset gets translated into an actual song structure (where the lengths are defined, and edited if desired) on the left.
It is dead-simple to define a new structure and instantiate it.
Especially since you don't even have to use the UI if you don't want to. The structure preset definition are read in from a text file.
Brilliant! (IMO)
RC gives us thousands of phrases to work with (and allows for an infinite number of definitions) . Why not do the same thing for Structures?
Heck, I don't think anything would be wrong with adopting the exact convention shown above, then we could have the same structures defined for both RC and Jamstix, perfectly matched up and ready to go.
Synergy! I love it. Cheers!
You do not have the required permissions to view the files attached to this post.
Last edited by sj1 on Wed Oct 25, 2023 8:22 pm, edited 2 times in total.
- KVRian
- 1177 posts since 13 Mar, 2017
Then, just add structure-affecting Dice and one song leads to another at the touch of a button.
The creativity and power would reside in the implementation of the structure-affecting Dice.
Are we completely changing structure? Or just changing the lengths/repetitions of the sections. Uniformly, or not uniformly. Tossing in a tempo or time sig change? A key modulation? A genre/style change (once RC is "smart enough" for that)?
A big reason why I think this could make alot of sense for RC is that no paradigm change is required. RC already has all the concepts in place, and since most of what I'm describing is basic list-walking, recombination and parameter tweaking of elements, doing this would be a relatively straightforward programming effort compared to some other ways of going.
IMO, we are not far from "new songs at the push of a button", and the ripe field that awaits harvesting is the field of Song Structures!
The creativity and power would reside in the implementation of the structure-affecting Dice.
Are we completely changing structure? Or just changing the lengths/repetitions of the sections. Uniformly, or not uniformly. Tossing in a tempo or time sig change? A key modulation? A genre/style change (once RC is "smart enough" for that)?
A big reason why I think this could make alot of sense for RC is that no paradigm change is required. RC already has all the concepts in place, and since most of what I'm describing is basic list-walking, recombination and parameter tweaking of elements, doing this would be a relatively straightforward programming effort compared to some other ways of going.
IMO, we are not far from "new songs at the push of a button", and the ripe field that awaits harvesting is the field of Song Structures!
- KVRian
- 1177 posts since 13 Mar, 2017
AI/Dice could also judge what fully-specified structures (list of section types, lengths, reps) would comply with a specified target total time and tempo.
The user could lock the time and let RC calculate structures and tempo Or the user could let the let the total time be elastic within a range while specifying either an exact tempo or an acceptable tempo range.
Lots for the computer (programmer!) to do right here ...
The user could lock the time and let RC calculate structures and tempo Or the user could let the let the total time be elastic within a range while specifying either an exact tempo or an acceptable tempo range.
Lots for the computer (programmer!) to do right here ...
- KVRian
- 1177 posts since 13 Mar, 2017
We probably also need the concept of what I'll call "absolute tracks". These would be tracks in which the events (phrases) are locked to absolute times. This is so a gunshot or gong-hit to match picture would always take place at the right spot, even if everything else is recomposed.
-
- KVRAF
- 1752 posts since 30 Nov, 2013
Now that version 5 has already been released, I still want to ask everyone who is related to the RC. Why when I select scales in the master track, I see all the scales available in the program and cannot immediately select the scale that corresponds to the chord? I can select a scale from the list, but 99 percent of the time it will not match the chord selected in the master track. I have options.. either select another chord that matches the scale or open the master track inspector and select the command to count the corresponding scales. Why can’t you immediately see the scales corresponding to the chord in the general list, but you can only see all the existing scales in the program? After all, I don’t want to change the chord to another. Don’t the capabilities of the RC allow this? Otherwise, the RC works mainly as a harmonizer of melodies (phrases).. And indeed, in this, the RC has enormous capabilities. But it is simply impossible to immediately select the appropriate scale for any chord. Isn't this logical? Either I'm bad at logic or I don't know the capabilities of the RC well.Or maybe I'm just crazy? Thank you .
-
- KVRist
- 259 posts since 11 Dec, 2018
Which rhythm generator allows the melody generator to start the bar with a rest (when regenerating, not manually edited)? What about inserting rests between some random notes? Most seem not to do it (is this a bug or is there a setting somewhere?)
I tried using the remove note variation but regenerating the melody generator does not randomize the seed or how many notes are being removed.
How do you properly generate a melody with rests that matches the chords and scale?
I tried using the remove note variation but regenerating the melody generator does not randomize the seed or how many notes are being removed.
How do you properly generate a melody with rests that matches the chords and scale?
-
- KVRist
- 410 posts since 9 Oct, 2013 from uk
Hmm.. yes please..BluGenes wrote: Fri Oct 13, 2023 1:40 pm And for everyone reading this, I did start work on a video about how to use the Rhythm Lab..![]()
System 1 - Win11; i9 13900HK miniPC; 64Gb; Iris XE graphics; Cubase 15.0.10; Studio Pro v8.0.3;UR44 i/o
System 2 - Win10; i7 4790; 16Gb; GTX750Ti; Cubase v14.0.41; WLab Pro v12.0.51; StudioOne v6.6.4
System 2 - Win10; i7 4790; 16Gb; GTX750Ti; Cubase v14.0.41; WLab Pro v12.0.51; StudioOne v6.6.4
-
- KVRAF
- 2150 posts since 15 May, 2017
Yea, you can do that with the Rhythm Lab.. I just finished a video about using it, should have it uploaded in a little while..mdstudio wrote: Tue Oct 24, 2023 1:53 pm Which rhythm generator allows the melody generator to start the bar with a rest (when regenerating, not manually edited)? What about inserting rests between some random notes? Most seem not to do it (is this a bug or is there a setting somewhere?)
I tried using the remove note variation but regenerating the melody generator does not randomize the seed or how many notes are being removed.
How do you properly generate a melody with rests that matches the chords and scale?
-
- KVRAF
- 5710 posts since 24 May, 2004 from []1
^I would like to view the Rhythm Lab video when ready.
Are there any other RC5 videos available?
Are there any other RC5 videos available?
