Skip to content
forked from hrtlacek/rtv

Comparison of programmable real time video environments

License

Notifications You must be signed in to change notification settings

scomicstrip/rtv

 
 

Repository files navigation

Real Time Visual Content Creation - Overview and Comparison

This Repository tries to compile information about programmable real time video processing environments.

Not included are environments that are so broad that one could do just about everything (like C++) or complete tools without the need or possibility to program anything(or very little, such as resolume).

Discliamer: The author of this document is basically a TouchDesigner/Max/python/GLSL guy doing live performances and installations in an art context. Any bias towards these tools/needs are tried to be avoided but probably inevitable.

Examples

The examples directory is supposed to hold a highly simplistic example for the specific language. This example should play back an audio file or use live audio input and use this data to somehow (RMS, FFT) modulate the scale of a simplistic 3d object (Sphere, cube) that is rendered.

Environments

Short List

Environments listed here are headed towards creation and/or manipulation of real-time video content as a big part of their whole purpose. Environments excluded here are tools that arguably have a different purpose as a whole. Also excluded are unreleased, unstable or outdated environments.

Unreleased/Beta/coming Soon

Outdated

Scientific Data Viz and preprocessing

tools for scientific visualization, neural network video stuff and data parsing

Other

Realted tools and tools I don't know much about.

pure GLSL/HLSL coding environments

There are several tools that make (live-)coding in GLSL/HLSL easier/more interesting/convenient or make coding of GLSL relevant in a live show context. Trying to list all of them is not the idea here. But it's probably important to mention their existence/relevance as a whole.

Evaluation

This repo also tries to evaluate the different environments numerically. Please take these numbers with a grain of salt. What does it mean that tool x has a documentation with a value of 0.7568? and not 0.879? Well, it is an attempt to structure the wood of environments but it is of course a bit hard to evaluate some parameters. Others, such as 'does environment x run on Linux' are of course objective and trustworthy.

The dataset and an ipython notebook in this repo is used to plot/explore it. You can look at it/play around with it via jupyter notebook or via collab online: https://colab.research.google.com/github/hrtlacek/rtv/blob/master/notebook/compareEnvs.ipynb

What do these Vague Categories Mean?

These are categories that are meant to characterize the purpose a specific tool is used best for. It is of course a bit weird to say environment x has a science score of 0.6. That is why there are two versions of the really vague Categories(SCience, Experimental, Creativity,Show Production) :

  • Opinionated
  • Calculated

The opinionated version still is useful since the calculated version fails to capture some stuff (too little data). For example vvvv and Max/Jitter have a pretty low rating for Show Production. This is because Max/MSP for example does not support DMX out of the box and it has not time-line editor out of the box. Both of these Environments are used a lot in a live show context (which is a feature that is hard to enumerate). Below you can find how both the calculated version and the opinionated version came up.

Opinionated

Science

An environments get a high score for this category if the following is true:

  • It is often or sometimes used in a science context
  • It is advertised as being a good tool for research
  • it is open source
  • It has some text programming capabilities
  • Its files are not stored as binary (but as JSON for example) and general openness
  • It's highly extensible
  • support for computer-vision
  • linux support

Creativity

An environments get a high score for this category if the following is true:

  • It is often or sometimes used in an art context
  • It is advertised as being a good tool for design
  • It's beginner and artist friendly and does not need a lot of know-how of inner workings
  • It's design-centered
  • Ease of use is an obvious priority of the environment (technical things are not in the way)
  • Reinventing the wheel is nearly not necessary
  • It supports some kind of graphical programming environment(not only text programming)
  • has built-in out-of-the-box 3d viewports (such as unity,unreal,smode, notch, touchDesigner)
  • Offers 'fancy' output out-of-the-box, without tweaking a lot (anti-aliasing etc)
  • Offers good procedural modelling and 2d post-processing features

Show Production

An environments get a high score for this category if the following is true:

  • It is often or sometimes used in a professional Live-Show/Set-design context
  • It is advertised as being a good tool for professional large-scale live shows
  • It offers great out-of-the-box integration of current technology(Kinect, NDI, Dante, posiStage)
  • Extremely stable
  • out-of-the-box 2D/3D Mapping capabilities
  • Good interfaces to other professional software that is widely used (Substance Designer, Photoshop, various 3d Modelling Tools etc)
  • DMX/Artnet support out-of-the-box
  • It has some kind of Linear/Keyframing (time line) functionality

Experimental

An environments get a high score for this category if the following is true:

  • It is often or sometimes used in an experimental/art context
  • It is advertised as being a good tool for art based research
  • It's open source
  • It supports current technologies such as HIDs and VR
  • It's highly extensible
  • It's not very specialized or very specialized on a not so common technique (such as live coding)
  • It's not extremely expensive
  • linux support
  • built-in support for computer-vision

3D plot

Calculated

In an attempt to support the numbers there is a calculated rating of the above vague categories. These are just averages of specific features as explained below. Some of the features that go in the calculation are still subjective(such as 'easiness' or 'out-of-the-box fancy output'. One could for example calculate the 'fancy output' thing by averaging a number of by default enabled features such as anti-aliasing, SSAO of DOF. But this really goes a bit far already.

Creativity:

  • out-of-the-box 'Fancy' Output
  • out-of-the-box 3D Viewports
  • visual programming
  • 2D Post-pro
  • Procedural Geometry
  • Easiness

Science:

  • Text Programming
  • Linux support
  • Open source
  • Computer Vision
  • Audio Analysis

Show Production:

  • out-of-the-box time-line
  • out-of-the-box 3D view-ports
  • DMX/Artnet support
  • out-of-the-box 'Fancy' output

Experimental:

  • Text Programming
  • Visual Programming
  • Computer Vision
  • Open Source
  • Linux Support

3D plot

About

Comparison of programmable real time video environments

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 100.0%