Company Post


Here's what our Plastikman Visual Contest winner and veteran TouchDesigner user Achim Kern has to say on numerous topics surrounding his experience with Touch like producing his winning entry (which we can't release yet), multi-touch and projection mapping, pushing limits, real-time 3d authoring, levels of feedback and support- among other things.

Achim also provides valuable insight to newer users, forcasts history repeating itself and tells us why and how TouchDesigner is his tool of choice!

Derivative: Achim, please tell us a bit about your background - what you were doing when you started using Touch?

Achim: I was first introduced to TouchDesigner’s “older brother” Houdini at a demo of Houdini 3 at FMX in Stuttgart many years ago, and the whole procedural concept got me immediately hooked. The following years I spent learning Houdini, which was quite difficult as there weren’t too many tutorials or other users around. After a few years I got a little tired of constantly waiting on renders, so I decided to focus on real-time 3D. At that time TD was still in 017, and while powerful, it didn’t live up to the workflow I was used to from Houdini. So I made a quick detour and learned Ventuz, a node-based tool for creating real-time 3D presentations, shows and television graphics. But that software also didn’t give me the freedom to visualize all my ideas, so I decided to go back and try Derivative’s tools again. Around that time Derivative had a small beta group for the next-gen version of TD, so I decided to buy TouchDesigner and immediately knew that this was the tool to work with.

D: Why did you pick it up... to do what? and why Touch?

AK: It was a natural match, I guess that’s just the way my brain works. I never liked point and click interfaces, and since TouchDesigner’s procedural workflow alongside the intense visual feedback it provides is such a unique combination, I simply can’t imagine working with anything else.

Besides, I already had a strong knowledge of CHOPs and SOPs from my Houdini time, so I was in a comfortable space and able to be productive right away.

I’m currently finishing up a template and state-machine based framework of components (and a matching user interface) to quickly create interactive, movie-like data visualizations, whether for museum type installations, high-end presentations, touchscreen interfaces or as a playback engine for VJ sets with multiple synths. It’s almost like I’m recreating TD inside TD, but at a higher level. You still use TD’s low level operators to create the actual content, but the framework automatically takes care of tasks like flow control, templates, transitions, lightning, object placements …

Another reason for choosing TD is that I’m convinced TD will become a major player in the real-time 3d authoring market and that there will be a high demand for experienced TD users. My forecast is based on how Houdini has evolved into a much more widely used tool over the years, and considering the fact that TD basically brings the same strengths to the real-time market, I have no doubts that history will repeat itself.

D: What's your experience with TouchDesigner and Derivative in general - support, forum, community etc.

AK: I’ve never experienced the same level of support from any other company. Bugs are fixed very quickly, unlike most other companies where you have to wait 2-6 months for a new release. Derivative listens to our feature requests and if demand is high you sometimes can get lucky and the missing feature will be implemented in a matter of days. The community is a very friendly and helpful crowd, and whether they are Derivative staff or other users, everybody tries to quickly help out.

D: Did you learn anything new this time... working on your contest submission?

AK: You always learn something new, even if it’s just a certain combination of nodes that work well together. In the case of my contribution to the Plastikman contest, where I directly visualized the incoming data channels by converting them to geometry (similar to TD’s tile viewers), I found that by using a combo of a trail and shuffle CHOPs I can nicely influence the shape of the “sound objects” on screen.

D: For those who've never used Touch, what would you tell them? Advise...

AK: Try it, you won’t regret it, and if you like it, you’ll never go back to anything else. The learning curve is tough, so I guess for today’s new users the situation might be similar to mine when I first started using Houdini. While we now have video tutorials, which greatly help in getting started, beginners will probably still be overwhelmed by all the nodes and contexts TD offers.

Don’t try to understand them all at once - there are a couple of basic nodes, and if you master these you already can create stunning networks. Then slowly check out the other nodes. Once you have a fairly good understanding of the principles, just check all the tools in the palette and see how they are built.

If you don’t understand one, jump to the next … and when you come back to the problematic one a few weeks later, you’ll probably understand its inner workings right away. If you’re coming from other patching based tools, you should have a pretty easy time adjusting, as TouchDesigner gives you so much feedback while building your networks.

In either case I’d also recommend checking out some of the Houdini tutorials. While both programs have developed in different directions, the underlying concept is still the same, and you can grab loads of valuable information from these. Of course you need to have basic TD knowledge in order to facilitate the knowledge transfer from Houdini to TD, but there even are some theoretical Houdini lessons on 3d buzz which should help people grasp the whole procedural idea.

In addition, TD is very well suited to realizing very contemporary and/or hip stuff like multi-touch and projection mapping. Actually, I can’t think of anything you can’t do with TD - and even if there’s a tool missing in the toolset, TD is so open that you can create that tool with custom code and exchange information via sockets/OSC/MIDI …

D: What else can you do with it? Do you want to do with it - can you imagine doing with it?!

AK: Ever since Marc C. Woehr and I created the augmented painting “the city never sleeps”, I have really been enjoying this unique combination of analogue painting and digital projection - as it, on the one hand, gives me the possibility to create works which remain present (to a degree) even after the projector is turned off and, on the other hand, it brings his paintings to life.

As I come more from a funk, jazz and hip-hop culture background, one of my long-term objectives is to establish good visuals for these musical genres, whether it’s interactive urban art like mentioned above, actual live visuals and stage designs for concerts or something along the lines of the amazing work done by the Graffiti Research Lab. Here's a rough example of an implementation of a GML reader (Graffiti Markup Language) for Touch Designer. More to follow soon.


GML test | Download the gml_reader_demo.toe file

The ultimate goal is creating a “surround” stage design - where the complete stage is a screen and it can interact with the musicians (and vice versa), similar to the digital stage work pioneered in theatre performances. As generative visuals by nature tend to mesh much better with electronic music, the real challenge is to find a way to design and establish something for the funk/jazz/hiphop genres without merely using video sequences. Not that I have anything against the use of video loops - you can do some amazing real-time compositing based sets using keyed footage of writers and breakers glued together with some neat generative graphics - but this tends to raise costs considerably.

Anyway, I also greatly enjoy creating touchscreen interfaces, projection mappings and “all that jazz” and I am always about furthering my knowledge and pushing limits … so I guess you could say I am basically open to any kind of interesting work.