Is it possible to commission TroikaTronix to upgrade an Isadora plug-in to GPU video?
Back in 2011 my team developed a couple of plug-ins for Isadora. They are live particle 'curtains' that also have up to three resizable shapes that are interactive with the particle system and can be linked to blob tracking data. These plugins were developed using the Isadora SDK and are CPU based. I guess there are two questions now - is a new SDK for GPU based plugins in the pipeline? or can some arrangement be made to commission TroikaTronix to upgrade these actors? I think we would be happy to donate the plug-ins to the community. It is just we have been unable to upgrade them to GPU and therefore have not reached their full potential.
I think the only person who can answer this properly is @mark himself.
@bonemap a new, possibly similar, plugin was released at the werkstatt.
I'm unsure where to find the download. Perhaps mark will have it.
Hi Ryan, I understand my plugins might be redundant with new versions of Isadora. I am happy to wait for the next release. However, as you know some of us have been waiting for the gpu based sdk for some years now. For example I invested a lot into an iOS plugin that is now languishing. I have another Navier-stokes plugin that is stalled in development. The Isadora team and Mark have done an amazing job with this software and the total cost of liscence ownership for the software is very good compared to other commercial apps.
I don't want to be disrespectful to the hard work and effort that has gone into Isadora's development, but at what point do I give up on anticipating an sdk release?
Sorry if I came across as dismissive. That was not my intention. Replying on my phone I tend to be short. (As I am again now).
I know how much time and energy development takes. The gpu sdk is still on the Roadmap, unfortunately there have been unanticipated road blocks.
The new plugin may address a need for you. I simply hoped that letting you know about it may be helpful.
@mark Is there a sdk beta available?
Thank you. It is always good to hear about innovative updates to Isadora...
I would love to have been able to leverage the GPU sdk potential along with the rapid prototyping and proof of concept attributes of using Isadora in our development team.
Particle based systems have become the standard and often the starting point for a lot of the most innovative interactive and generative works being presented today - I am thinking here about TeamLab, Universal Everything and Adrian M & Claire B. For example a plugin that leverages algorithms for fluid dynamics such as Navier-Stokes would help Isadora users to match their expectations for these standards. This is an ambition of mine (and perhaps many other Isadora users based on the popularity of the post on this forum about e-motion software with over 20k views).
Is the glsl shader provision in Isadora the panacea for these particle effects? It is a fantastic addition but with only 6 (or is it 8?) Isadora configurable inputs it does not match the generative and mutabile variation of a particle array system. Besides, it is a rare animal that can write the kind of shader that simulates a particle system (we have to really appreciate the coders who make such amazing shaders available through CC licences for non-profit projects). The glsl shader integration in Isadora is arguably a game changer, however it requires serious coding skills to author particle simulation effects.
I would argue that there is a gap in Isadora for more particle system based actors and it is this gap that has driven me to want to invest the time and resources putting together a team to develop new plugins.
dbini last edited by
I agree with Bonemap on this. I'm dedicated to Isadora, and am constantly introducing people to the possibilities of using it in their workflow. When I teach people how to use it, I really don't like to admit that its not the most powerful option for generating particle systems. I explain that the 3D particles actor is wonderful, and can produce some incredible results, but particles that are generated are passive - there are plenty of options to manipulate the environment that these particles exist in, but no way to effect the behaviour of the individual particles themselves once they are generated (even the video texturing has a blanket effect)
If it would be possible to attach algorithms to particles so that they have their own dynamic properties, or are aware of each other, or have a continuous relationship to an external variable, then that would be a serious game-changer for Isadora.
bonemap last edited by
I notice Derivative have recently released a MacOS version of Touch Designer ... perhaps we can use the particle generators there and Syphon the output over to Isadora? Still it is not the same as having the particle interface contained in one program.
tomthebom last edited by tomthebom
@dbini That was exactly my question about, if there could be parent/ children particle systems...! Imagine, if one particle system would emit particles, where a second or even third emitter could be applied... ;o)
without giving anything away...
I want to add that some new and very interesting capabilities are in the works.
No need to jump into additional tools ☺
@mark you are being noticeably absent from comments on this thread. It would be great to get a response to the question about the sdk. As a guess, the silence on this suggests we are close to something going on that is related to current development
If it is related to fluid dynamics and particle system embellishment I would say it is definitely needed for Isadora to remain competative with other node and wire based applications.
mark last edited by
Dear @bonemap and All,
Thank you for your patience while awaiting my response to this thread. I am always juggling priorities, trying to ensure a good balance between solving urgent bugs, making sure things like the High Sierra release doesn't cause serious problems for the entire user base, and adding new features you all want. In that long list of priorities, the I'm afraid the SDK is nearer to the bottom of the list because – historically – it hasn't been utilized by so many. Of course that's not optimum if the SDK is really important to you, but necessary as I attempt to serve the greatest number of users.
The thing that held me back up is Isadora's reliance on QuickTime. We have an internal build of Isadora for Windows that does not use QuickTime at all. But this version is missing some features to replace those lost when divorcing it from QuickTime. I was putting off the SDK hoping that I could release it when this new version was ready, but it's going to take a bit more time before that milestone is reached. (For those of you who don't know, fifteen years ago I used QuickTime as a way to make a cross platform version of Isadora. That may have been clever at the time, but seems less so now, because removing it now is requires widespread, deep changes to the entire program.)
That said, @bonemap's comments here and a recent message sent via the ticket system has prompted me to get something out there. I am going to get the SDK to him first so he can use it for his project and offer any bug reports/feedback. By the middle of October, we'll get it up on github so that everyone can access it.