c / a | H Y P E R _ H O L O G R A M [presenting our work using Magenta]

[c.sla...@gmail.com]

hi all,

we are c / a [pronounced as cslasha], an anonymous electronic music duo. we just wanted to introduce our work to you.

we work with ML and AI and performed a major body of work called HYPER_HOLOGRAM for the first time last month. it was at Sonar Festival, Istanbul. 

here is a two minute preview of the live show. 

https://vimeo.com/272264858

also a bonus; on the attachment you can find a track called "global_step_sec=0.065243". this was made with melody_rnn with a model trained on an amalgam of various artists ranging from Aphex Twin to Ravel. surely, here, there are lot of people who are familiar to the work of Aphex Twin - when we first heard the MIDI our model generated we were quite fascinated cause it had a distinct Drukqs vibe to it. also, it was also one of the most beautiful melodies we could create so far. 

now, we want to improve our live setups to utilize more custom code, NSynths and so on. if you would like to collaborate, have any idea to share, have any questions or comments we would be beyond happy to hear them.

we just wanted to share our work with you, get in touch and thank you all

thanks for reading and best wishes.

c / a

[Leandro Garber]

very cool performance ! hope to hear more from you soon

--

leandrogarber.info

--
Magenta project: magenta.tensorflow.org
To post to this group, send email to magenta...@tensorflow.org
To unsubscribe from this group, send email to magenta-discuss+unsubscribe@tensorflow.org
---
You received this message because you are subscribed to the Google Groups "Magenta Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to magenta-discuss+unsubscribe@tensorflow.org.

[whoisc whoisa]

thanks a lot Leandro, we hope to perform and get the word out there more hopefully!

best wishes,

c / a

x

[Adam Roberts]

Very cool track and performance. Thanks for sharing and keep it coming!

On Fri, Jun 1, 2018, 5:29 AM whoisc whoisa <c.sla...@gmail.com> wrote:

thanks a lot Leandro, we hope to perform and get the word out there more hopefully!

best wishes,

c / a

x

On Tue, May 29, 2018 at 9:13 PM, Leandro Garber <leandr...@gmail.com> wrote:

very cool performance ! hope to hear more from you soon

--

leandrogarber.info

2018-05-28 19:10 GMT-03:00 <c.sla...@gmail.com>:

hi all,

we are c / a [pronounced as cslasha], an anonymous electronic music duo. we just wanted to introduce our work to you.

we work with ML and AI and performed a major body of work called HYPER_HOLOGRAM for the first time last month. it was at Sonar Festival, Istanbul. 

here is a two minute preview of the live show. 

https://vimeo.com/272264858

also a bonus; on the attachment you can find a track called "global_step_sec=0.065243". this was made with melody_rnn with a model trained on an amalgam of various artists ranging from Aphex Twin to Ravel. surely, here, there are lot of people who are familiar to the work of Aphex Twin - when we first heard the MIDI our model generated we were quite fascinated cause it had a distinct Drukqs vibe to it. also, it was also one of the most beautiful melodies we could create so far. 

now, we want to improve our live setups to utilize more custom code, NSynths and so on. if you would like to collaborate, have any idea to share, have any questions or comments we would be beyond happy to hear them.

we just wanted to share our work with you, get in touch and thank you all

thanks for reading and best wishes.

c / a

--
Magenta project: magenta.tensorflow.org
To post to this group, send email to magenta...@tensorflow.org

To unsubscribe from this group, send email to magenta-discu...@tensorflow.org
---

You received this message because you are subscribed to the Google Groups "Magenta Discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to magenta-discu...@tensorflow.org.

--
Magenta project: magenta.tensorflow.org
To post to this group, send email to magenta...@tensorflow.org

To unsubscribe from this group, send email to magenta-discu...@tensorflow.org

---
You received this message because you are subscribed to the Google Groups "Magenta Discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to magenta-discu...@tensorflow.org.

[Sean Farrell]

Very cool guys!  The performance was so visually engaging and the music sounds awesome too!

I have also been using RNNs to make music of my own!

For the first collection I made about a year ago now I used this RNN https://github.com/hexahedria/biaxial-rnn-music-composition from Daniel Johnson.

https://soundcloud.com/sean-farrell-4/sets/bot-beats-vol-1

But recently I have been working on another go round using some of the magenta RNNs.  I have been digging the performance RNN personally!  Here is one made with that model.

https://soundcloud.com/sean-farrell-4/feelzgoodruff-mix

Basically I take the output from the RNNs and "sample" parts that catch my ear, then break them out into different synths.  From there I add a little of my own flair by playing on top or by adding some samples.

I am working on performing these pieces in more of a live context with the addition of the AI Jam setup pioneered by the Magenta team, so hopefully I will be sharing some video soon.

Keep up the good work, and feel free to reach out if you want to collaborate!

Sean

--

leandrogarber.info

To unsubscribe from this group, send email to magenta-discuss+unsubscribe@tensorflow.org
---

You received this message because you are subscribed to the Google Groups "Magenta Discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to magenta-discuss+unsubscribe@tensorflow.org.

--
Magenta project: magenta.tensorflow.org
To post to this group, send email to magenta...@tensorflow.org

To unsubscribe from this group, send email to magenta-discuss+unsubscribe@tensorflow.org

---
You received this message because you are subscribed to the Google Groups "Magenta Discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to magenta-discuss+unsubscribe@tensorflow.org.

--
Magenta project: magenta.tensorflow.org
To post to this group, send email to magenta...@tensorflow.org

To unsubscribe from this group, send email to magenta-discuss+unsubscribe@tensorflow.org

---
You received this message because you are subscribed to the Google Groups "Magenta Discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to magenta-discuss+unsubscribe@tensorflow.org.

[whoisc whoisa]

thanks a lot Adam!

cool, thanks for sharing Sean! all of that doesn't sound much different from our modus operandi though a bit in reverse. we started off experimenting with magenta's RNN models then got haunted by the idea of writing our RNN models from scratch.

for example there is one techno-ish track in the teaser [about in the middle], it was born from a few hours spent with AI Jam setup as well. we wanted to utilize these methods to live situations but then gave up on that for various reasons. we just wanted to maintain the idea that the project and the audio/visual show would remain as a work of its own rather than a presentation or experimentation if that makes any sense - not that we were afraid that the model would produce 'bad music' or would not work at all or something.

though, we are really curious of the possible outcomes. thanks a lot and keep in touch!

we spent the most time on the aphex twin models so both the breakbeat drums and synth melody parts could match and produce something really in the vibe of Richard.