in general, resizing on the gpu isn't nearly the resource hog that cpu resizing was. This provides a lot more freedom when working gpu video actors.

Still some processes are heavier than others so it's good to keep an eye on your Load. Specifically things like Eyes, or Measure Color will still use a lot of cpu and can benifit from scaled down video as an input. Using scaler before these types of actors is my approach.

I also tend to use 'get video size' rather often so that I can easily see the size of video in different areas as I build a patch. Once it's working as I want, I generally remove them.

I will often use 'Get stage size' combined 'scaler' to force a video to the size of my output before running any processing. But in the end you really have many options and a lot of flexibility.

Quoting @Mark from another thread:

If you double click the "eye" icon on the movie player, and enable 'audio device', 'snd out' and 'audio trks', you'll see these options. 

Right now, these channel related inputs currently only function on MacOS if you set the mode to 'interactive' -- which means you are using QuickTime to play the movies.

If you set the number of 'audio trks' to three, the Movie Player looks like this.

The audio device input allows you to send the sound to a specific output device. The 'trk' inputs specify the volume, panning, and routing for each track. 

@cam3ld

Try hap in either a AVI or MOV wrapper.

also worth noting that all CI and QC video actors are Mac Only, as well as the all of the AU Core Audio actors.

[EDIT by Mark for clarity; I've updated my answer above to make this important point]

@juriaan thank you ...good idea !!

Best,

@michel said:

This seems not so easy, the only thing coming to my mind is to trigger only up to a certain size. In theory: use a blob decoder there you will have a "obj width" and "obj height" output. Play around with the values and run them through a inside range actor that will only trigger if the object is in-between the size you have defined.

 This seems like the best solution

Make sure you have the right settings going on your Ion (pg 40 has a picture in my readme file)

just to add another option here.

I recently made a user actor / javascript script that provides another type of value decay to help a user address a need. 

You can check it out on the plugins page: https://troikatronix.com/plugi...

@dillthekraut thank you very much !!

Best,

@eva said:

when I close Isadora and starts again, or change to another window and come back to Isadora, the device turns off

I've not experienced this myself. For sure, when you quit Isadora it's going to shut off. But you're saying simply click on the window of another application will and clicking back to the Isadora window will disable the device?

Also, when posting about issues to the forum, it's super important to tell us 1) What platform you're on (MacOS or Windows), 2) What version of that operating system you're using (MacOS High Sierra, or Windows 10, etc.) and 3) the version of Isadora you're using. If you want to, you can add that information into your signature line here on the forum (click your icon at the top right and choose "Edit Profile" to do this), so we don't need to ask.

Best Wishes,
Mark

I found it now. Thank you

I think it may be specifically about the vertical dimension, and that you may be able to get away with horizontal resolutions that are not multiples of four - at least, that is what I have encountered in some cases, like QLab.

However, there are certainly implementations of HAP that do not have this limitation (HAP files played by disguise on Windows for example do not have any multiple-of-4 limitations).  Ideal of course would be that any resolution could be used.

another thing to consider is that the need for power injection is less if you are not turning all leds white at the same time.

For example if you wanted to move a Red block of leds (say 1m wide) the other 3m of leds would be off and you wouldn't need power injection. 

As Juriaan said, connecting the power to both ends is probably the easiest way to extend the normal (single sided) setup.

I have done this a lot using a single video feed into Izzy but connect the six cameras to a processor like those used in video surveillance systems.  They come in 2, 4, 8, or 16 inputs as the standards.  Here's an example of a 4-input for $50.

https://www.amazon.com/Evertech-Security-Splitter-Processor-Controll/dp/B004TN18UU/ref=asc_df_B004TN18UU/?tag=hyprod-20&linkCode=df0&hvadid=242012522334&hvpos=1o1&hvnetw=g&hvrand=12483842741987355182&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9019794&hvtargid=pla-583456088599&psc=1

The advantage is only 1 input stream into Izzy (you zone the input trigger areas in Eyes) and there are fewer computer processing issues with only 1 input stream.  The big disadvantage is the video quality is greatly reduced and thus does not work well if you are trying to send actual video out of Izzy as a part of the work.

@eva said:

You guys are awesome:) Thank you! 

 I love it when people say that. 😉

@dbini : Thank you. So obvious once I saw it.

This forum is SO great. Thanks, Time of Day in periodic mode is just what I was looking for! Sorry I didn't see it in the manual...

Thanks you both for taking the time to respond.   - John

oh. whoops. that should have been obvious. thanks.

Thank you Michel for the fast answer.

This seems to work for what I need.

@fred hi,

Thank you so much! I have an Apple Airport Extreme that’s usually pretty fast.

My machine will be sending and receiving osc messages to the musician’s Ableton Live.

But I’ll experiment point network to compare.

The video will be 1080, my machines can’t handle 4K real-time animations 😄 yet.

No I don’t have any notch Hardware... I’m still learning.