Hi there,
I have discovered an issue with NDIWatcher on Isadora 3.0.8 on some Mac operating systems. Seen on both Mojave and Sierra on intel macs: NDI watcher limits maximum resolution dimension to 640 px and scales the other dimension proportionally. So e.g. if you send 1920x1080 you only get 640x360. If you send 1280x1280, you get 640x640, etc. Also, if you send resolutions larger than 640 in either dimension, NDI playback becomes very stuttery / low frame rate.
On Big Sur (with otherwise same computer specs), these issues do not appear. On Mojave and Sierra, using NDI Watcher with isadora 3.0.7 also does not have these issues. It seems to be a combination of 3.0.8 with the older OS, but is enough to prevent using 3.0.8 for a project with NDI input on these computers.
Dear All,
As I complete a short but much-needed holiday, I've been reading up on NDI 5 which the upcoming release of Isadora will support.
As it says on the NDI Tools page:
Securely share NDI sources between remote sites, anywhere in the world, using a simple and secure network setup.
- Maintain native end-to-end NDI workflows, reducing complexity, cost and latency across any distance.
- NDI Bridge delivers local convenience on a global scale as it handles all the complexity of remote sources and still works with alpha channel, KVM, PTZs, tally, metadata and much more!
*NDI 5 Bridge BETA offers NDI Bridge Local Mode at present. We are working hard to bring Bridge and Host Mode to you, to expand your productions even further.
So while they don't seem to have the "host mode" working yet, the promise of an easy method to ship video and audio from anywhere to anywhere is tantalizing for anyone creating about remote performance work.
Our beta testers have received a beta build of the upcoming release which is still NDI 4. But the next release they receive will have NDI 5. We'll dig in to do some testing with NDI Bridge to see what it means and report to all of you here once we know more.
Best Wishes,
Mark
I am using 3.0.8f12 on Windows 2019 (EC2 instance) and when I create an output with NDI, the app just crashes. Works fine on other machines. I am running on a T3.large instance with an Elastic GPU. Other OpenGL apps work fine. I'm not seeing any crash reports or anything.
Hello internet,
My apologies if this has been addressed somewhere else, but the search function was being very buggy for me. Unusable on Firefox and barely usable on Chrome.
I am using Isadora to output to more stages than I have ever before. The project calls for sending (40) 640x480 NDI streams out over a gigabit network to (40) Raspberry Pis, each connected to an individual CRT TV built into a huge wall. I will be using Izzy to map a single video image over the wall of TVs. So far I have tested the setup with (27) Pis. Each NDI output is a separate stage defined in Izzy. With this many outputs Izzy is lagging. All I'm outputting is a simple test pattern I created consisting of a background color cycling through hues, a text draw and a few shapes. Load hovers around 100% in Izzy, and framerate doesn't quite get to the 30fps I want, BUT the playback computer has plenty of overhead. CPU, GPU and RAM loads are only about 25%. The receiving Pis use about 75% CPU to decode on their end. So, why is Izzy so loaded down when the PC has more power to give? Any ideas on where the bottleneck is and how I could address the issue?
Playback PC:
Izzy 3.0.7
Windows 10 Pro
AMD Ryzen 3 3100
16GB RAM
NVIDIA GeForce GTX 1660
I've been using OBS.Ninja and Syphoner to get remote video into Isadora, but have just found a nice alternative. Vingester will take a web URL and stream the video of the page via NDI, which can then be grabbed into Isadora with the NDI 4 Watcher actor. (is this still BETA?) if you have enough CPU you can add more streams in Vingester. Its faster, cleaner and more stable than using a browser.
Hi everyone,
when I change scene, the NDI watcher actor in the scene to come takes a second or two to get the stream to the projector,
I tried to preload the scene in the one before but as I thought, preload works for medias and no video real time stream,
of course I could put NDI watcher in an activated scene and use matrix send value to fade or switch the projector,
but is there any other way to preload the stream to get it right when I activate the scene?
all the best !
Hello,
I would like to have the possibility to choose the NDI source of an NDI watcher plugin from a control without acting on the actor himself.
I'm creating a patch to control PTZ cameras and I would like to give the users (that don't have access to actors) the possibility to choose the video source relative to the PTZ camera they are controlling.
Thank you
Stefano
Thanks for your powerfull work Here is a small video of my work in progress...
https://www.youtube.com/watch?v=qpWqI-MVMfk
More I test Isadora, more I like it!!
Hi!
Have now used Isadora for a couple of remote performances, and completely loving the experience!
Just wanted to look into one issue I had when using NDI output (which has been needed a few times) -- still frames often have a tiny flicker. I wasn't sure if it was a frame rate thing or something, but decreaseing frame rate didn't seem to affect it.
It's quite subtle, and I'm sure it's not noticeable for an audience... but was hoping to get to the bottom of it, because if it's something I can fix on my end I'd love to know!
Here's a GIF zoomed in on some Izzy subtitles as an example (sorry for reduced quality, seems like the easiest way to demo):
I haven't had the issue in Spout, so I'm guessing its something about the switch to NDI.
Many thanks and all best,
Django
Greetings! I have been a long time user of Isadora (since 2008) and I am currently working on a virtual production during the times of covid.
We are using keyed NDI feeds piped into Isadora and I am saving resources by doing the chromakey and crop before it reaches Isadora. Although I can see the alpha channel premultiplied in the stream out of the vid-gpu port, I am still having to apply the alpha tool to get the correct feed. Although I do nothing with the alpha tool, no setting changes on this actor at all.
The issue I'm worried about is I am taking 7 (seven) ndi feeds at once, and by passing through the alpha tool there is a considerable load appearing on the machine (in the 85%+ range). I would love for there to be a way for the alpha coming out of the NDI v4 feed to be already be usable without the alpha tool.
I can build the show and the system seems to be able to handle the load, but this feels like a bug perhaps.
Also, to maintain connection to the feeds, I have placed them in their own (Active at all times) scene, and I am broadcasting them to listener actors in appropriate scenes as necessary. I haven't seen this be a problem performance wise, but I am open to suggestion if this is terrible idea.
If needed I can post a patch, but just wanted to get some insight from this wonderful hivemind. I appreciate any assistance!
Thank you!