Aggregate Music System 1.0.1 - Update available!
This thing looks absolutely brilliant! Congrats to the release, Rob!
This brings up the whole interesting question of what a generative player is intended to do in the first place. At first I was very sold on the idea that they might be able to create something human-sounding, like a good piano solo. I made some progress in this direction with Evolution, but ultimately I gave it away because it took so much work to program & automate that in the end it was way quicker just to compose & play or draw the notes into the sequencer myself (and the results were better)guitfnky wrote: ↑25 Aug 2021looks really impressive, but if the demo is any indication, it suffers from the same problem every other generative player suffers from--it sounds like a generative player. until these types of things are able to create believable human-sounding performances, they'll probably continue to be best used for background/augmentative purposes.
Then I started making generative setups, but again it was a lot of work for results that were strangely sterile and boring. They sounded like algorithms, not like something I was making myself. And that's the problem - the more I hand the compositional work over to the player, the less character I end up with in the music. Doesn't matter how many hours I spend inputting & modulating parameters, the results always sound weirdly uninteresting. For some reason I don't get this problem when working with CV or playing manually
In the end I still find generative players useful, but only for getting started with ideas. Evolution and Inspiral are a great combination, I just do a bit of tweaking, print the MIDI to the sequencer, and then comb through it looking for interesting little runs & progressions. But that's just the beginning of the composition process, and from that point onward I don't use the players at all
Of course I don't rule out the possibility that this is all my problem and it still might be possible to make really cool, distinctive music with generative players. If anyone's doing it, I'd love to hear it. Maybe we need an Aggregate Music System song challenge?
Could be worse - We’ve made lots of progress since the 1980s!dvdrtldg wrote: ↑25 Aug 2021I've got it on trial, been playing with it for a few hours
First thoughts: amazing bit of gear, and while I'm still way down at the bottom of the learning curve, I can see that a few more hours flipping between the device and the manual will really open things up
That said, it seems way too complicated for a useful composition tool. Unless what you want to compose is a noodly, open-ended sort of plink-plonk-scape. I love generative setups and spend a lot of time creating them, but I'm way more comfortable doing them with CV, where the link between input and result is relatively easy to follow. With this thing, there's just too much math & mystery
I guess I might change my mind, and will certainly get more psyched if I hear people start creating interesting music with Aggregate, or if I manage to come up with some myself. But everything I've heard so far - including the examples on the promo video - is boring, sounds like a software demonstration more than it sounds like something I'd choose to listen to
I love Lectric Panda REs and I know everyone else does too, so I feel I can say all this without it seeming like I'm shitting on some poor struggling developer's hard work. LP rules and will continue to rule. But I dunno, I'm not feeling this one, not yet anyway
Selig Audio, LLC
that’s pretty much my experience too, although I gave up on using them as a primary source of creating music much quicker. I do still use them for auxiliary parts (like some cool background bell or pluck sounds), but never been able to get them to work as a main focus.dvdrtldg wrote: ↑25 Aug 2021This brings up the whole interesting question of what a generative player is intended to do in the first place. At first I was very sold on the idea that they might be able to create something human-sounding, like a good piano solo. I made some progress in this direction with Evolution, but ultimately I gave it away because it took so much work to program & automate that in the end it was way quicker just to compose & play or draw the notes into the sequencer myself (and the results were better)
Then I started making generative setups, but again it was a lot of work for results that were strangely sterile and boring. They sounded like algorithms, not like something I was making myself. And that's the problem - the more I hand the compositional work over to the player, the less character I end up with in the music. Doesn't matter how many hours I spend inputting & modulating parameters, the results always sound weirdly uninteresting. For some reason I don't get this problem when working with CV or playing manually
In the end I still find generative players useful, but only for getting started with ideas. Evolution and Inspiral are a great combination, I just do a bit of tweaking, print the MIDI to the sequencer, and then comb through it looking for interesting little runs & progressions. But that's just the beginning of the composition process, and from that point onward I don't use the players at all
Of course I don't rule out the possibility that this is all my problem and it still might be possible to make really cool, distinctive music with generative players. If anyone's doing it, I'd love to hear it. Maybe we need an Aggregate Music System song challenge?
I’d love for someone (maybe Lectric Panda) to work on algorithms for enhancing the human-ness of the performative side (e.g. velocities, timing and phrasing considerations), as opposed to note selection, and simple randomization parameters for the rest. even the “humanizers” I’ve heard so far are pretty lacking in terms of making stuff sound believable—presumably because they rely on basic randomization.
- StephenHutchinson
- Posts: 464
- Joined: 27 Sep 2019
- Contact:
Agreed, it's pretty much out of whack. As of this writing (Wednesday August 25th, 9:35 a.m. EST/DST). $69 USD = around $84.50 CDN.doze wrote: ↑25 Aug 2021$69 USD = $99CAN on RS... prob is they have an arbitrary rate that isn't based on current conversion. Would be better if they just charged us in USD and let our payment gateway deal with it.dvdrtldg wrote: ↑25 Aug 2021I'm not an unusually stupid person and I'm pretty comfortable with Evolution, PSQ 1684 and some of the other more complicated REs. But the manual for this one is just berserk
I'l probably buy it, but I've already given up any hope of figuring it out
Is it a Reason thing? I'm in Australia so I get punished by the exchange rate ($69 USD = $95 AUD at the moment), but I can't see how that's Reason Studio's fault
Reason User Since Version 1.0
Sound Design & Music Content Creation
Sound Dimension :: https://www.sounddimension.io
Sound Design & Music Content Creation
Sound Dimension :: https://www.sounddimension.io
-
- Posts: 149
- Joined: 11 Oct 2018
Man, I used to love M! I had it on the first computer I got, around '99. I spent hours and hours playing around on that thing.selig wrote: ↑25 Aug 2021Could be worse - We’ve made lots of progress since the 1980s!dvdrtldg wrote: ↑25 Aug 2021I've got it on trial, been playing with it for a few hours
First thoughts: amazing bit of gear, and while I'm still way down at the bottom of the learning curve, I can see that a few more hours flipping between the device and the manual will really open things up
That said, it seems way too complicated for a useful composition tool. Unless what you want to compose is a noodly, open-ended sort of plink-plonk-scape. I love generative setups and spend a lot of time creating them, but I'm way more comfortable doing them with CV, where the link between input and result is relatively easy to follow. With this thing, there's just too much math & mystery
I guess I might change my mind, and will certainly get more psyched if I hear people start creating interesting music with Aggregate, or if I manage to come up with some myself. But everything I've heard so far - including the examples on the promo video - is boring, sounds like a software demonstration more than it sounds like something I'd choose to listen to
I love Lectric Panda REs and I know everyone else does too, so I feel I can say all this without it seeming like I'm shitting on some poor struggling developer's hard work. LP rules and will continue to rule. But I dunno, I'm not feeling this one, not yet anyway
I acctually don't want or need a self generating mozart player. This one just adds more control over a sequence and over much greater amount of time. I think this is the strongest part of this player. Which is awesome for long sequences. I don't expect it to play a perfect piano tune. Just feels like if I'm going to make such long sequence, maybe I should just enter notes in a piano roll until it sits right.guitfnky wrote: ↑25 Aug 2021that’s pretty much my experience too, although I gave up on using them as a primary source of creating music much quicker. I do still use them for auxiliary parts (like some cool background bell or pluck sounds), but never been able to get them to work as a main focus.dvdrtldg wrote: ↑25 Aug 2021This brings up the whole interesting question of what a generative player is intended to do in the first place. At first I was very sold on the idea that they might be able to create something human-sounding, like a good piano solo. I made some progress in this direction with Evolution, but ultimately I gave it away because it took so much work to program & automate that in the end it was way quicker just to compose & play or draw the notes into the sequencer myself (and the results were better)
Then I started making generative setups, but again it was a lot of work for results that were strangely sterile and boring. They sounded like algorithms, not like something I was making myself. And that's the problem - the more I hand the compositional work over to the player, the less character I end up with in the music. Doesn't matter how many hours I spend inputting & modulating parameters, the results always sound weirdly uninteresting. For some reason I don't get this problem when working with CV or playing manually
In the end I still find generative players useful, but only for getting started with ideas. Evolution and Inspiral are a great combination, I just do a bit of tweaking, print the MIDI to the sequencer, and then comb through it looking for interesting little runs & progressions. But that's just the beginning of the composition process, and from that point onward I don't use the players at all
Of course I don't rule out the possibility that this is all my problem and it still might be possible to make really cool, distinctive music with generative players. If anyone's doing it, I'd love to hear it. Maybe we need an Aggregate Music System song challenge?
I’d love for someone (maybe Lectric Panda) to work on algorithms for enhancing the human-ness of the performative side (e.g. velocities, timing and phrasing considerations), as opposed to note selection, and simple randomization parameters for the rest. even the “humanizers” I’ve heard so far are pretty lacking in terms of making stuff sound believable—presumably because they rely on basic randomization.
Some weird symmetries can be achieved here and I think this is another strong point of this player.
Don't know, I'll probably end up buying it, when I learn what Panda wanted me to learn with this player.
M
Last edited by Mataya on 25 Aug 2021, edited 1 time in total.
Anybody else having this issue? Mine is all covered with hearts.
(insert recycled joke from the beta test here) -
Here's a project I started in the beta. It's got AMS running pretty much all the instruments, with the help of MIDI CV Converter and a couple other players. This player (AMS) is brilliant! I love it so much! Congrats on the release, Rob.
-
Also... the manual... nice nod to Mario here! ha ha, love it
(insert recycled joke from the beta test here) -
Here's a project I started in the beta. It's got AMS running pretty much all the instruments, with the help of MIDI CV Converter and a couple other players. This player (AMS) is brilliant! I love it so much! Congrats on the release, Rob.
-
Also... the manual... nice nod to Mario here! ha ha, love it
That's exactly what I was talking about. I'm pretty sure that there will be people praising this little rather annoying sounding project of yours. But to me this really sounds awful and the fact that AMS is doing all of that is ... ok I've lost interst in writing any further...
You're right, it's brilliant.
You're right, it's brilliant.
Last edited by Mataya on 25 Aug 2021, edited 1 time in total.
That's the problem though - "humanisation" ultimately doesn't make a lot of musical sense unless it's meant in the very weak sense of "introducing imperfections". Which is easy do to, but as you say, these imperfections still sound machine generatedguitfnky wrote: ↑25 Aug 2021I’d love for someone (maybe Lectric Panda) to work on algorithms for enhancing the human-ness of the performative side (e.g. velocities, timing and phrasing considerations), as opposed to note selection, and simple randomization parameters for the rest. even the “humanizers” I’ve heard so far are pretty lacking in terms of making stuff sound believable—presumably because they rely on basic randomization.
Humanisation in the sense of "sounding like it was made by a particular human" is something else altogether. It implies a certain quality of character, and elements of surprise that aren't just random quirkiness. That character is what these players lack, although I guess theoretically it wouldn't be impossible to create it - some amazing algorithm could analyse the complete recorded works of Thelonius Monk and then generate distinctively Monkified phrasing, timing, chord selection etc. That would be interesting
I wasn't kidding about the AMS song challenge by the way. Are we still doing these? I'd really like to hear what people are coming up with
-
- Posts: 439
- Joined: 22 Apr 2020
Obviously its horses for courses - and I dont deny I love putting a bass guitar on QNG and just jamming a long with a Radical Keys or whatever for a warm up - but for me these types of players are for extracting sections of midi thats kind of working and then making them your own as idea starters; or to run these devices in semi random loops as triggers for other things eg see the Adam Fielding stream where he triggers Slice Edit 2 using Evolution, or to humanise hats or perc or whatever throughout sections. Seeing them as just a thing you tweak until it makes a 'human sounding performance' out of the box misses the point imho.guitfnky wrote: ↑25 Aug 2021looks really impressive, but if the demo is any indication, it suffers from the same problem every other generative player suffers from--it sounds like a generative player. until these types of things are able to create believable human-sounding performances, they'll probably continue to be best used for background/augmentative purposes.
To each his own, of course. I think it's pretty amazing that this player is doing all of that. I didn't work very hard on this "annoying sounding project" other than some time put into the instrumentation and arrangements. And that's good, otherwise I might have taken offense to your rather rude comment.Mataya wrote: ↑25 Aug 2021That's exactly what I was talking about. I'm pretty sure that there will be people praising this little rather annoying sounding project of yours. But to me this really sounds awful and the fact that AMS is doing all of that is ... ok I've lost interst in writing any further...
You're right, it's brilliant.
Question/challenge for you: Could you possibly word your comment any ruder than you did?
In any case, I hope that you will someday find a warm spot in your heart for Evolution and AMS. They aren't the sorts of players to just set and forget. I often find myself sending the notes to track and editing in the sequencer (and I think that is where they shine); they can help to inspire or even move a project along when creativity might be low... annoying sounding or not.
it is a very specific compositional sculpturing tool. you get out of it what you put into it, which is unlike most generative tools i've had experience with.
i suggest turning off the mutate and melody sections and start intentionally creating chord progressions with the accompaniment section and rhythm section, maybe learn whether you prefer using scale degrees or interval down progression types. then try the variety of other accompaniment features. i reckon entire tracks await creation from just this one section
i suggest turning off the mutate and melody sections and start intentionally creating chord progressions with the accompaniment section and rhythm section, maybe learn whether you prefer using scale degrees or interval down progression types. then try the variety of other accompaniment features. i reckon entire tracks await creation from just this one section
- Certified Reason expert
True. I expect most people open in it and bang the random button a few times and say, this is shit noise... But really, besides that action, there is very little in the way of random about this Rack Extension. There is way less randomization going on here than Evolution. Evolution was about 10% of the features of Aggregate, so the rest had to be filled in with 'randomization'. Aggregate gives you the control over the rest. It took a long time go sort all that out. Aggregate will only create what is specified.
Most of this came from me reading these books:
Schillinger-System-Musical-Composition
I started out with some python code to generate midi files.. and then a standalone real-time app to kick out MIDI. But I really wanted to play around with these composition techniques in reason. This software lets me do that. Its the deepest thing I've ever done. I don't expect many people to try to get it, the shit is basically a composition spreadsheet. It's quite the rabbit hole.
- StephenHutchinson
- Posts: 464
- Joined: 27 Sep 2019
- Contact:
I like this... alot. Don't let the naysayers get to ya!. Very well done indeed!challism wrote: ↑25 Aug 2021Anybody else having this issue? Mine is all covered with hearts.
(insert recycled joke from the beta test here)
Untitled.jpg
-
Here's a project I started in the beta. It's got AMS running pretty much all the instruments, with the help of MIDI CV Converter and a couple other players. This player (AMS) is brilliant! I love it so much! Congrats on the release, Rob.
-
Also... the manual... nice nod to Mario here! ha ha, love it
Capture.JPG
Reason User Since Version 1.0
Sound Design & Music Content Creation
Sound Dimension :: https://www.sounddimension.io
Sound Design & Music Content Creation
Sound Dimension :: https://www.sounddimension.io
This player gets very deep once you dig into it. One key thing that's easy to miss at first glance is at the bottom of page 6 in the manual:
Here's a patch
If you miss that you may not notice the two different types, and they should probably be a little more visually distinct.An Event type of duration is drawn darker in
the center, making it appear like a ring around the
numeric value. A Rate type of duration is drawn
with a filled circle looking like the values.
Here's a patch
If you ain't hip to the rare Housequake, shut up already.
Damn.
Damn.
agreed—I think we’re saying much the same. and there are already such algorithms out there. for example, the Jamstix VST. it does a really good job of creating realistic MIDI drum performances. shouldn’t be too far a leap from that to other types of performances—keys, strings, horns, etc. just need someone to do it.dvdrtldg wrote: ↑25 Aug 2021That's the problem though - "humanisation" ultimately doesn't make a lot of musical sense unless it's meant in the very weak sense of "introducing imperfections". Which is easy do to, but as you say, these imperfections still sound machine generatedguitfnky wrote: ↑25 Aug 2021I’d love for someone (maybe Lectric Panda) to work on algorithms for enhancing the human-ness of the performative side (e.g. velocities, timing and phrasing considerations), as opposed to note selection, and simple randomization parameters for the rest. even the “humanizers” I’ve heard so far are pretty lacking in terms of making stuff sound believable—presumably because they rely on basic randomization.
Humanisation in the sense of "sounding like it was made by a particular human" is something else altogether. It implies a certain quality of character, and elements of surprise that aren't just random quirkiness. That character is what these players lack, although I guess theoretically it wouldn't be impossible to create it - some amazing algorithm could analyse the complete recorded works of Thelonius Monk and then generate distinctively Monkified phrasing, timing, chord selection etc. That would be interesting
I wasn't kidding about the AMS song challenge by the way. Are we still doing these? I'd really like to hear what people are coming up with
- StephenHutchinson
- Posts: 464
- Joined: 27 Sep 2019
- Contact:
Something I put together with just hitting the random button at the top of the Aggregate Music System RE for two devices.. Kong & Monotone Bass... 105 bpm (As a .WAV or .AIFF file it loops perfectly but I've uploaded an .mp3 file which won't perfectly loop unless cut in some audio editor)
Reason User Since Version 1.0
Sound Design & Music Content Creation
Sound Dimension :: https://www.sounddimension.io
Sound Design & Music Content Creation
Sound Dimension :: https://www.sounddimension.io
challism wrote: ↑25 Aug 2021To each his own, of course. I think it's pretty amazing that this player is doing all of that. I didn't work very hard on this "annoying sounding project" other than some time put into the instrumentation and arrangements. And that's good, otherwise I might have taken offense to your rather rude comment.Mataya wrote: ↑25 Aug 2021That's exactly what I was talking about. I'm pretty sure that there will be people praising this little rather annoying sounding project of yours. But to me this really sounds awful and the fact that AMS is doing all of that is ... ok I've lost interst in writing any further...
You're right, it's brilliant.
Question/challenge for you: Could you possibly word your comment any ruder than you did?
In any case, I hope that you will someday find a warm spot in your heart for Evolution and AMS. They aren't the sorts of players to just set and forget. I often find myself sending the notes to track and editing in the sequencer (and I think that is where they shine); they can help to inspire or even move a project along when creativity might be low... annoying sounding or not.
Yes. Please don't hop on that offended train. I really think "music" in that video is awful. Feel free to listen some of my music and call it what you want. Unless there are some rainbows and hearts involved, everything is rude theese days. I really think it's important to point out when something is not that good in art, because I expect then that you will think about it except just get offended by it. I just don't think that being rude is that bad sometimes.
As for the player, again, not for me right now. I own evolution and it brings me joy each time I use it. Maybe AMS will do that some day. But it seems it goes where I don't need to go with my, let's call it music.
tx
M
-
- Reason Studios
- Posts: 1594
- Joined: 05 Feb 2020
Really enjoying playing with this! Can't say I understand much of it yet, but playing with that RAND button is mighty fun Great work!
-
- Information
-
Who is online
Users browsing this forum: No registered users and 95 guests