24/32-bit fixture control

John Roper

Member
I realize that 16-bit control for fixture parameters has been available for a while now, but are there any fixtures/softwares that support 24 or 32bit control?
 
16 bit precision in pan and tilt, assuming a 360 degree angle the light moves over, gives you about an eighth inch resolution of where the light beam ends up with a throw distance of 100 feet. More than likely any moving light would be limited by mechanical and mounting constraints well before the sixteen bits becomes a factor.

In terms of intensity, sixteen stops of control is plenty for most any realistic situation, I should think. Certainly for static looks that much contrast between the brightest and the least bright instruments is plenty, as light spill in all but the very best controlled situations is likely to be greater. Very slow fades to and from blackout could conceivably benefit, particularly if the intensity to control value mapping is strictly linear. (It would need to be the equivalent of something like a 20 minute fade from full to blackout, or a segment of such a fade, for the extra resolution to be of any use at all beyond sixteen bits; any faster, and with sixteen bits you must skip values in the DMX data stream.)
 
If I recall correctly, the Series 200 Vari-lites used 20 bit resolution for pan and tilt(not sure why not a multiple of 8), and computed timing internally, for smooth slow moves. But otherwise, the return on investment from 16 bit to 32 bit is so miniscule, that it is not worth the effort. Some of the Series 300 stuff (VL-5's 6's and 7's-bleh) may have used 32 bit internally, (but less certain on that count)
 
@Ford from you point of view in the product development department, do you see much call for 24 or 32 bit control? Is that something you are eager to have widely available so you can design fixtures to take advantage of or does 16 bit already cover everything you want and then some?
 
GrandMA technically supports 24 bit "ultra fine" control, but I'm not aware of any practical applications for it. Maybe in media server world, but i feel like you're going to run into the refresh rate of DMX being an issue before you need more than 65000 steps of resolution.
 
And to think we used to do some pretty good looking work with dimmers calibrated 1 to 10 and having maybe 40 steps on the plates!

I remember when "moving lights" was something the electricians did, not lights themselves.
 
I was thinking it would be useful for dimming (intensity or color) or pan/tilt.

Moving to higher bit depths seems like diminishing returns. Higher grade LED fixtures already have complex dimming engines to ensure smooth fades so you don't need to move to a higher bit depth for the sake of intensity and there's no reason you would need more than 0.0015% resolution you already have with 16-bit in terms of specifying your "resting value" you're fading into. The smoothness of dimming between values is a function of the engine and not a dependency of DMX resolution -- I think you'll find in most circumstances where there's any reason to have more smoothness of any operation, it's almost a necessity to build that into the fixture natively as part of the engine rather than try to bridge that resolution of control directly between a DMX source and the output motors/dimmers/etc of a fixture.

Color-wise, I could maybe see a justification if you were running in HSI or HSIC mode on a fixture with 6-7 colors -- though I think if you're fine tuning your colors that closely you're already messing around with direct mode anyway.

Maybe it's my lack of imagination, but I have a hard time envisioning where 32-bit control would be practically useful. If you're trying to achieve that resolution of control through a single parameter, you have a fundamental user interface problem where you are trying to achieve a technical goal through a paradigm that a human being cannot comprehend the resolution of.

As for the subject of media servers, if you're trying to achieve that finite level of control, for that to be comprehensible and navigable in any meaningful way you need to be doing that control in the video processor itself. Deadmau5 probably has some of the most advanced video of any modern tour and they're working in a completely different control paradigm than a DMX-controller media server. They can take DMX in as a control source but their dynamic content is largely being controlled/triggered/influenced by Ableton through OSC. Trying to mimic their level of control through any kind of DMX-interface, controlled by a DMX console is simply impossible.

Little bit tangential to the topic at hand, but if you're not familiar with how Deadmau5 is using TouchDesigner to drive their video, I highly recommend going down the rabbit hole. Very illuminating. He's probably 10 years ahead of where the broader industry is in terms of media content dynamic rendering and control.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.


To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.


To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.


To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.
 
Just pick some unused channels and then you can overflow your 16 bits into 24, or 32, or ... whatever you want! :)

Moving those low-order-bits faders is going to do diddly-squat. So you might as well just pretend you have the 32-bits of control.
 

Users who are viewing this thread

Back