![]() First draft is here (just one universe so far): We can hopefully offer a richer interface integrated into the lighting console too with RDM (remote management) to allow patching and config from within the existing tools people are using. The main benefit of the OLA support should be a richer range of protocols supported (about 7 other IP ones) as well as about 10 hardware dongles that support DMX input. I haven't done anything as clever as what you're talking about yet, although we've hit similar sync issues when driving strings of SPI pixels directly from a Pi, when you pass a certain number, they need multiple universes and you have to decide when to flush out. I'm one of the OLA dev's and I've gone about this from the other end, adding support for the Fadecandy's USB protocol into OLA. I didn't realise you'd got so far on with this, as I didn't see anything in the issue: Vdmx5 specs update#(MadMapper won't update fixtures that aren't changing, so we aren't guaranteed the packet rate is a multiple of the frame rate.) Then to detect phase, I'd look for the largest inter-frame gap within each of these periods. The clock rate would be detected by looking at the change rate of the most frequently updated OPC/Art-Net channel. If you're curious, I want to solve this by having fcserver retime incoming frames to a consistent clock. So, the naive solution you currently get means that there's some subtle tearing and the interpolation won't be effective because Fc thinks the keyframe rate is some multiple of the actual keyframe rate. With Art-Net, now we have this problem at only 170 LEDs instead of 10,000. If you updated half your LEDs with channel 0 and half with channel 1, the interpolation algorithm would act like your frame rate is twice the actual rate, giving you choppy edges. If you used multiple OPC channels, though, fcserver would treat each as a "frame". Most people do this, and so fcserver's 1:1 mapping between "set pixel colors" packets and "frames" turns out to be fine. If you have under 10,000 LEDs or so, you can fit the entire thing in channel 0 and this is fine. With OPC we've managed this by encouraging people to use a single channel. They might belong in the same USB packet, and Fc's interpolation algorithm requires a single consistent frame clock. It likes to do this globally, because the arbitrary mappings can mean that it's impractical for different parts of your display to have a different frame clock. The issue here is that fcserver really likes to know when the end of a video "frame" is, so that it can flush out data to all of the attached devices in synchrony. Lots of people won't notice, but it bugs me and I have a solution planned. So uh, why isn't this merged to master yet? The timing isn't quite right, so your video will look *slightly* juddery. This basic mapping gives you two FC channels per Art-Net universe. They map each channel from 1 through 4 to 128 consecutive LEDs (two output channels) on a Fadecandy board. You can also use this mapping over Art-Net if you want, but doing so would produce flicker because it splats the output across all 512 pixels. So, this does two things: Maps channel 0 one-to-one, giving us access to all 512 LEDs just like usual over Open Pixel Control or WebSockets. Learning the VDMX InterfaceĬontinue reading this manual, starting with the first section of this covering the VDMX Inspectors.The fields in the "map" lines are. See the Templates section for more details and a list of examples. The fastest way to get started with using VDMX is by using some of the pre-made interfaces that can be loaded from the Templates menu. Vdmx5 specs movie#
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |