@dusx Great thanks Ryan !

@dillthekraut

Thank you, that all makes sense.
Making progress :-)

@Fred @Juriaan Thanks for your suggestions. I'm finding a way through...

Will post results and workings when I've got the final thing going.

Much appreciated everyone.

Hi @woland

Any Update on this topic? If not, is there a good working workaround without big latency? 

I bought a BirdDog Play yesterday an like to do a small show on Saturday with it - i don't like to invest money into mitti ;-)

regards, Tom

@kazapp

Hi,

Eight years is a long time - so some incompatibility would be expected. I assume that some of the video codecs of your older files are no longer supported - this article might assist.

Fading between scenes can be problematic with some video codecs and with 3D Actors.

Best Wishes,

Russell

@jingyi3027

there are many ways to do this. a simple way to adapt your current patch would be to use Trigger Value actors instead of your Toggle/Envelope Generator.

Connect the True output of your Comparator to a Trigger Value of 100 - connected to the Intensity of the Projector. Add a Trigger Value of zero connected in the same way to the False output.

Remember to Intialize the Intensity at zero, so that when you enter the scene, the visual effect is not shown until the Comparator recognises a value of 1 or higher.

I think the problem you are experiencing with the Envelope being triggered at the wrong time is because you are using the Toggle actor in an unconventional way.

@bonemap thats great thanks Russell, will do that. Much appreciated! 🙏

@goodiebags

This seemed like an interesting project to tackle... so I made a demo file using a different approach.
I created a feedback loop into a Matte++ actor using a freeze actor. I think this approach is very flexible. Currently it has a random shape from a shapes actor as the 'brush' but you can use any image (alpha around will help :)
I might expand this a bit and make it a new User Actor offering on the Add-Ons page, what do you think?
Hope you like it!  

Brush Drawing

@mark_m

It is high on our list of things to work on next; Isadora 4.x released, and IzzyCast allowing us to spent time on other plans.

Audio is really high on the list.

@juriaan 

Thanks. I'd not discovered the Comparitor actor, I'll have a play with that, I can see this being useful for me moving forward. I've managed to get the right effect by limiting min and max values in the Control Watcher value parameters (0-8) then putting it out through a value changed actor with a min change of 5 and through toggle to switch effect bypass on and off when knob passes 8. Seems to work consistently. 

@dillthekraut Thanks for that.

I worked up the math based on the measurements of the screen and then used the vertical projection resolution of 1200p to determine the width in pixels. After that, building a new blend master worked *very* well.

However, I had to tweak the projector "actors" so that the images & movies that weren't at that aspect (roughly 11:4) would look good. I used the zoom property to bring them to the correct width, then decreased the height factor to fit the screen.

After a bit of wrestling with the control grid to align the two projections, I was able to do a fairly good job with the blend.

Thanks for all the good ideas!

@reload2024

Actualy, this isn't really that simple to answer, as there are many factors, which would influence the possibilities.

USB 3.1 Gen1 does 5Gb/s data throughput theoretically. But in reality it is a bit less, because of the overhead. Anyway, this should be enough in almost all cases for a small to mid complex show. But still, there are more things to consider. Most USB drive doesn't offer the possible speeds which could be possible with the USB mode they propagate to support. Often it is much less then that. Even if they offer faster speeds for USB 3.2 (10Gb/s), they might be much slower in USB 3.1 mode. You need to look for tests, to see the real speeds of the devices. Then, some drives get hot quick on heavy load and then throttle their speed after a short while. Again something which could be taught by a test.
But I mean, in general with most USB drives you can copy your content much faster, then playing it back in real time, right? There shouldn't be no big difference between those data transitions. But copying/ streaming multiple files in parallel,

depending on the controller quality of the drive, the actual speed could be much less, as the management of multiple parallel data streams reduces the actual throughput alot and sometimes more then 50%.

Last but not least, the codec and therefor the resulting bitrate of the videos define, what drive speeds you'd need.
You can kind of calculate what minimum of data speed on read mode you'd need, by multiplying the videos bitrate with the number of videos you going to play in parallel. It might be good to add at least 50% overhead.

A ProRes 422 would go with around 150 Mbit/s on FullHD. On real world USB 3.1 Gen 1 speeds this would mean you theoretically could play 30 streams of it in parallel.

A ProPres 4K 4444 does around 750Mbit/s, which would still be up to 6 streams, just looking at the possible USB 3.1 speeds.

The fastest USB 3.1 drive I could find on a quick look, does 400 MB/s (3200 Mbit/s) sequential read (which means 1 long continuous reading like in video streaming instead of a lot of little file chunks which adds overhead). Which would still be enough for 3 to 4 HQ 4K streams.

But one more thing is a nono for me: USB often isn't very reliable connection wise. There always is the risk of an unconnect in mid show. I would not risk it, if not really needed.

@demetri79 said:

could you please share if you are able to track it down.

 here it is: ISF PLUGIN | TroikaTronix Forum

Trying to get eyes ++ to connect to live drawing so bounding box tracking will make a drawing when following the movement from a camera.

Here's screenshots of what I have tried to do to connect the two but have had no luck any help is useful thank you

@jfg

I tried it but the EQ didn't take effect, it seems that works with the audio input of the system, in my case a Midas M32 mixer, but I need to EQ the audio output of the source, video and/or audio.

Also I tried using the Loopback App, that allows to route the audio from your apps and sources to different outputs, but It didn't work.

Thanks for you help.

@reload2024

I don't think you're going to be able to achieve what you want, not just in Isadora but in any 'affordable' system, if the 'difference' method isn't working for you.

Any method for isolating one thing from another relies of a way to do that isolation. Greenscreen is one way, IR masking is another, using a Depth Camera is a third.  So if you can't use any of these, how else can you isolate the dancers from the background?

@eattreebark never thoughts of trying AI to convert 32 to 64! Worth a shot!

@eattreebark said:

a way to convert the old Pete warden free frames into shader code

 unfortunately no, most of these plugins use CPU based processing, and simply won't convert to GLSL.

@demetri79 don't forget the UVs need reworking though. Blender can do this. Mesaage me if you need help with it. 

@jfg. Ah thank you so much, I will try that out.

I ended up giving up and just using all the sound for the level watcher.  I was using Izzy to created a haunted house room, where the ambient sound played all the time but when triggered, the jump-scare scream goes off and light from projector hits the scary skeleton...  reflecting in plexiglass so it forms a hologram, like a pepper's ghost.  The sound level made the light flicker in time with the scream.

(edit: what I was missing was the AU sound level watcher, didn't know about that)

I'm sure there are hardware ways to do this especially if it was a permanent exhibit, more reliably over years, but for a one-night party Isadora was my instant choice.

vimeo demo of scary room