thanks for your replies!

Your were right, i've used non-usb version.

Thanks Mark.

The no-loop problem is using 2.4.5b10, not the new 2.5 beta you sent me. I can’t use that at the studio… the patch works great without the if-then calculators. The movie player remains active when it freezes onscreen in the sense that if I mouse-over the cable to the projector, the movie is playing, but on external monitor it is frozen. Moving the speed input to the player makes the image begin to move again. I tried a "end" trigger into a value=1 into the speed control, but that didn't work either. I tried resetting the players, and that worked for about 2 minutes or so, then froze again. The whole point was to have 3 movies playing, not 6. But things work fine with all 6 playing as I cross-fade, so for now I can use this. But it is weird.

John

@Michel Perfect! Now I understand how this works! And what's the system behind. Thank you so much. R

have to try that, sounds pretty cool 

@DusX Thank you for the precision.

I knew it was mutable, but somehow thought it would still show up...

I guess that's the kind of mistake you do only once! :)

I think its ok.
When I do a white stage with 2 edges blended as you would need. I get something like:

Notice the corner is double faded (both gradient alphas are applied)
When I duplicate this layer and flip it about so that 4 copies are laid out like your example, I seem to get them adding together smoothly.

@DusX interesting point about the cycles, I had that problem once and solved it with the float to integer actor and comparator set to equal.

Instead of using a single grid to cover the entirety of each column, you may have some success splitting the input into multiple grids or slices first. This would allow you so adjust for the increasing distortion as you move up the column, as the grid on the bottom/ middle might not need as many sections as the grid on the top. 

@Maximortal

Yes the difference sure is the time zone. The JavaScript code takes your system time and compares it with the future time, so the result will always be the date and time of your current destination. To compensate you have to add or subtract the offset hours to the country or timezone you want to have.

Even if you compare different online counters for that election date they often show different values. Some code takes the time of the server location, other take your system time and other have fixed the time for that certain timezone they want the countdown for.

Best Michel

Sequential trigger did it!  Thanks so much folks!

@Michel Hi!

Ohhhhh great thank you!!

Have a nice day

Thanks. Now I got it to work : )

I just wonder how to blend Video and Audio at the same time... 

Is this somehow the way to go? 

Thanks for all your insight... Best . Erik

Hi David,

Can you give me a bit more information about your 3D model - It is hard to tell how complex it is? Is it the same model and Isadora patch in both videos?

The presentation setting in the second video looks great!  

I make an assumption that there are limitations to applying texture with the '3D Player' actor. The more complex the model the more issues arise with rendering surfaces and applying textures.

Regards,
bonemap

Thank you for your replies! 

Here’s an example I tried to solve my problem (making a realtime depth map in Isadora 3D space):

Image 1.jpeg, a 3D object with lighting on

Image 2.jpeg, the same 3D object with a different tilt

Image 3.jpeg, these both 3D objects combined

Image 4.jpeg, both 3D objects combined, lighting off

Image 5.jpeg, with 3D Particles actor constructed a series of vertical rectangles traversing in z-direction, each rectangle black with 4% intensity. This creates a kind of solid cube with “black fog” in front of the 3D objects, and when depth test is enabled, it creates a “depth map” of the objects in 3D space.

However, this is consuming a lot of qpu power and is not “stable” – if changes are made, for instance, to the tilting of the 3D objects, the transparent black rectangles are not delivered evenly => the “depth map” looks uneven and banded. See the following image 6.jpeg

You can see lots of "banding" also in the darker areas, if you look carefully.

So, I was wondering if there would be a better way (and less gpu consuming) to do this? Any ideas?

Thanks!

What I use now is Artnet Proxy on OSX , OSC to Art Net with mixed experiences.  Before that with the Kineme QC plugin, but that is unfortunately outdated. Worked also for years with Lanbox but their software is so out of tune these days, but if you get it working it is a safe workaround and a secondhand LCX cost you a hundred. 

Hi,

Both solutions from Mark work well. I used the first one by changing the range of the Projector Zoom. I learned a lot !

 Thanks to all of you.

Beyond using a GLSL Shader, there's a few other ways you could achieve a "sepia" look.

You can use a second movie player or picture player displaying a vintage filter overlay -- a search on CCsearch should turn up lots of free-to-use filters which you can combine with your livestream using an additive projector, alpha channels, or even a video mixer.

You might then use HSL adjust, contrast adjust, or even colorizer to get the desired tint.

@hatfab

Yes - you will need to download the FFGL installer to access it. Look for the FreeFrame installer link on the download page called GET IT.

https://troikatronix.com/get-i...

@eratatat

Can't be sure without having the hardware. I know that often you must connect these cameras to independant USB hosts (many machines only have 1).
If you can connect to separate USB hosts that may help. (on my laptop, I have 1 usb host that connects all my USB ports, except for 1.. I have 1 that is a USB/Esata port and it has a different controller.)

As Mark has mentioned, scaling before measuring is your best choice. 
I have found that scaling by an even factor seems to work best... so divide the width and height by, 2, 4, 8, or 16 etc... until its small enough to allow a good framerate.
You could also, try adding a 'multi blocker' before scaling, and limiting the frames thru.
A 'time' setting of 0.08 for the MultiBlocker will give a framerate of 12.5 fps thru  [ (1 / (25 / 2)) where 25 is your isadora target framerate ]
Half as many frames to process. This will help, but adds some latency.