MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

LZX Shape Changer Overview
MUFF WIGGLER Forum Index -> Video Synthesis  
Author LZX Shape Changer Overview
samn
Hey, the new LZX Shapechanger has been announced and I did a quick overview with a simple patch example for it! It's a pretty amazing addition to shape processing lineup.

ebbflo
great job, the bouncing shape at the end is phenomenal
nerdware
Oh, yes. A very tasty demo vid! Thanks.

Fine beard, BTW. thumbs up
FetidEye
yeah, thanks for the clear explanation!

I'm wondering ; if I wanted to rotate a rectangle of video input signal (from a camera or other video source), how would you map that onto the keyed image?
ebbflo
FetidEye wrote:


I'm wondering ; if I wanted to rotate a rectangle of video input signal (from a camera or other video source), how would you map that onto the keyed image?


I am sure you can get some cool effects keying a external source into the shape but to spin the external video source I think is beyond the scope of the shapeshifter.
I could be wrong and hopefully will begin experimenting soon
Dofkev
FetidEye wrote:
yeah, thanks for the clear explanation!

I'm wondering ; if I wanted to rotate a rectangle of video input signal (from a camera or other video source), how would you map that onto the keyed image?


I'm going to guess this will sort of work....It seems the video input would need the horizontal and the vertical frequencies separated before going into the shapechanger. Since navigator offers low and high pass filters on X and Y, filtering before the shapechanger might be able to acheive something like rotate....

Amazing module by the way!
nerdware
You could (sort of) do it using vector rescanning and a Navigator, but I don't see how the Shapechanger could rotate a video signal. You're mapping every pixel from one place of the screen to another place. That's a job for a framebuffer. In the LZX system, position is a timing thing. Neither the Navigator nor Shapechanger are changing that - in Sam's demo, they just process ramps generated by the VC. The result is changing video patterns generated entirely by the modules.

Look at it this way: a video mixer, like the V4 or V8, can do mirroring with framebuffer effects, while the LZX modules do it by waveshaping the ramps, e.g. ramp to triangle.

The rotation effect on the Navigator is done using a geometric transformation of the ramps. This is explained elsewhere in this subject. The Navigator is basically most of the vector rescanning patch in one module. VC ramps into the Nav, Nav outputs go to the X/Y inputs on the scope, and then you send the Luma output from the VC to the "Z mod" input on the scope. Finally, point a camera at the scope, ideally without the graticule, and "rescan" the image. Add the Shapechanger between the Nav and the scope for extra fun.

So, yeah, you can do it, but it won't anything like you imagine. In fact, you can do a lot more. Things that can be very hard to do digitally. Best of all, you can do it in realtime. razz

These threads all refer to the old Visionary series of modules, but may still be useful.

vector rescanning for dummies
LZX system & vector re-scanning newbie questions?
Vector rescanning with Tektronix 464

I can't remember where I found these images, so I'm just posting them here. I found them in various old threads.





FetidEye
mmh. When the Navigator was announced, I thought I saw a video of someone rotating video. but I guess I misunderstood
hmmm.....

Anyway, all this rotating and positioning is very cool.
What would happen if the H and V ramps had other slopes?
nerdware
I've only seen demos of the Navigator generating a rotating pattern. E.g. the rotating ramp demo on Lars' bench. That was fun - until that point, I'd only seen simulations. Thanks, Lars!

Look at the 2nd image I posted above. That explains how you rotate ramps. Use a phase quadrature oscillator and a pair of 4 quadrant multipliers, mix the 2 X outputs and the 2 Y outputs to get the rotating ramps? I think that would work.

The Navigator does all this and a bit more. Quite a bit more! It doesn't rotate video images. You can, however, use the rotating ramps to rotate video on a vector display, like an oscilloscope in XY mode. Use a patch to do what the first does, but process the ramps with the Navigator. Then you can rotate arbitrary B/W video signals. Add a Shapechanger for real madness, like in the 3rd image. Maybe use another Shapechanger and a Mapper for MY ASS IS BLEEDING MY ASS IS BLEEDING MY ASS IS BLEEDING MY ASS IS BLEEDING w00t effects?
FetidEye
Thanks for the info!
For now I have Cadet modules only (system 21)
(plus some DIY Sandin Differentiators + Function Generators)

So this moving and rotating would be a huge complementation
nerdware
The difficulty with rotating arbitrary video with the LZX system is that we're not using a framebuffer. We're processing the signals the same way the video is scanned: left to right, top to bottom. So the XY position changes over time, as each frame is "scanned". This is fixed. We can't change it using anything in the LZX system. It all works in the time domain. Instead, we need a frame buffer - like the long-awaited LXZ framebuffer module. Maybe that can do rotations? I've no idea. If the playback scanning can use something other than left to right, top to bottom, maybe it'll be possible. Mirror effects should be easy, as that would be a relatively simple memory address translation - like waveshaping a ramp into a triangle, but done totally digitally. Arbitrary transformations would be harder. So I can't say anything useful about the framebuffer module. I can only talk about framebuffer effects in general. E.g. the mirror effects in some video mixers. The V4/V8 mixers don't do arbitrary rotations, just reflections. This is just basic address arithmetic. Nor can they do scaling. Yet this is available in any decent video editor.

So many of us here do "rotations" using a camera pointed at a monitor. This works great for video feedback, but not so well as a general effect. We get unwanted artifacts like the monitor case creeping in. Of course, this too is great for video feedback, esp for creating "spinning" patterns, but it might be nice if we could get spinning patterns without using a camera - and the Navigator helps us do just that.

We just need something to help turn the ramps into a shape - something just like the Shapechanger. Well, I guess there's a clue in the name. lol
nerdware
Haha, yes. I'd hate to build complex patches like this only using Cadet modules. I love all the recent complex modules, including BSO crossfader. I don't want to think about how many VCAs and mixers that must use internally. Thanks, Nick!
samn
Thanks everyone

@ebbflo glad you liked the box! I wasn't sure if it was a complex enough patch for everyone liking but I thought it was a fun practical one smile

@nerdware thank you for the beard comment!

@fetideye that is beyond the scope of the shapechanger and nerdware is correct about the video rescanning portion. The shapechanger requires H+V ramps to rotate. You could of course key an external video in using the doorway but that will be standard keying.
MUFF WIGGLER Forum Index -> Video Synthesis  
Page 1 of 1
Powered by phpBB © phpBB Group