Hi! Well, a lot of things happened since there as you can imagine.
Get way more experience and worked on a full-blown commercial studio, so I can share the experience.
First of all, let’s get the things straight: the usefulness of TD with scalable and Vioso. Basically, what you have as an out after calibration is a distortion map (like UV map) and blending map applied by multiplying color. You then just feed your texture and it’s predistorted with the calibration file (remap) loaded in the VIOSO or SCALABLE TOP.
There are couple of caveats:
- VIOSO in TD doesn’t have a blackpoint correct. Apprerently, it’s being reworked now, but no news about this since at least 1 year. I can understand it’s not a priority for guys with so much amazing stuff to be implemented, but this is smth you should now.
- don’t know about scalable blackpoint.
About 2 systems: if you have money - go Scalable. It’s way faster, more straightforward and less buggy. But their license system is smth to shoot yourself in the head if you need to travel with. Basically you need to pay for breathing and living on Earth They have dongle system, but it’s not enough, you still need to pay for aux systems. Whatever… It’s got quite intense with their sales representative at one point. But… if you have one install to do and leave it on and have money, you will loose soooooo much less time with Scalable.
Vioso… Well, it’s way cheaper, like twice less expensive than Scalable. It’s not using the same algorithms and principles, but the results are very comparable (I preferred Scalable). It takes way longer to calibrate - for 4 projectors the whole structured light routine and processing takes around 10-15 min instead like 2 min for scalable.
The worst part about Vioso is the buggy software at least at the time. I mean there are bugs and there f****g bugs you can’t begin to explain. The interface is much more modern than Scalable and uses kind of nodes (yohooo) which helps to visualize the system. But in 9 cases out of 10 you can’t go through the calibration procedure because of crashes. Believe me, on 100th time of 10 min calibration you want to shoot someone or yourself. It’s a wizard procedure. I ended up using Manual procedure which is much more stable, but very obscure because not very well documented. Sorry, Vioso, as you know, I tried and even spent 1 full day with the lead DEV with a remote support session.
RIght now, I’m considering other solutions:
-
Fly Eng Immersive Calibration Pro - we have licence and everything but the dev guy is really not forthcoming about how to work with custom meshes.
-
Screenberry Server software - might be overkill for us, but quite a complete solution with TD engin in it.
There is an obscure native solution made in Montréal UQAM for using opencv etc… (opensourced) and I think it was then augmented in Thinkwell studios in MTL (or completely new), but the latter is private and not shared.
I guess that’s it about calibration…