CasparCG wishlist

Client: Option to change the image/video transition default settings. For example, upon loading an image/video, transition is set to PUSH - 30 frames - EaseInQuad - LEFT - Seek 0 - Loop = true, etc.

Or Better yet, have the option to assign separate default transitions for images and video.

I dont know if this is a good idea or if its even do-able, take in webRTC-stream in client like the tool ā€œDecklink inputā€. No idea how it could be practically organized.

Another: possible that in the client define out-transition. If your using Caspar for screenproductions it would be usefull to have better ways to take out layers than transparent PNG:s especially for videos, images and Decklink inputs

Only way to take in WebRTC would be with a HTML template. That also means you donā€™t have embedded sound (until CEF audio works within Caspar). Other option would be to find software that could translate webrtc to another protocol.

Regarding transparency - you can use any image or video as the key layer for another layer with MIXER.

as a wishlist feature, we could have a native webrtc producer!

Regarding tranparency i use the transparent PNG as the following example:
Output on a screen in studio

  • on channel 1-10 a videoloop
  • on channel 1-11 Decklink input with a person joying through Skype
  • on channel 1-11 tranformation size 75%
  • on channel 1-11 opacity 85%
    in client when I take in the Skype-picture I add a Mix 25 frm
    when I take it out I play ā€œtransparent PNGā€ with Mix 25 frm
    then I stop everything in channel 1-11 with a group and run the program logo on channel 1-11 (PNG with transparency).
    This is still simple and managable but if you have a more complicated setup and have things on 4-5 different layers its easy to forget some layer.

My most important wish (not for me but for every new user) INtergrated the Frontend GUI to Server release.

1 Like

I would prefer a web interface as a frontend, which would comply with running the server headless.

4 Likes

I understand that if your running the Server through the frontend you miss out on some nice features like multiply configurations, So I like how it works in 2.0.7 where you can start frontend if you want or just the console if you prefer that. A webinterface would be perfect but still you cant(I think) get a channelgrid or diagnoce window though a webinterface. I have for problems installed TightVNC on my servers which means I can run them headless if I want to.

The frontend supporting multiple config files could be something to add to the wishlist if you want that.
There is a issue on github to replace the diag window with an external tool anyway https://github.com/CasparCG/server/issues/899, which would play well with a webinterface frontend.

1 Like

+1 for rendering to file. This would be a huge feature for us, it is available in other renderers such as Viz RT

1 Like

Automated rundowns :slight_smile:

1 Like

What do you mean @TheYouth?

See the actual topic

Magewell video card support would be great. Iā€™m not a fan of Black Magic Decklink.

we talked to them for a bit at IBC, but they donā€™t have an output card yet. thatā€™s in the works though, and if itā€™s any goodā€¦ who knows.

Interesting to know thanks, Iā€™ll watch this space!

I just saw an interesting product featuring Magewell cards, Iā€™d be great to be able to use it with Caspar but I guess the integration of this cards is not in great demand around hereā€¦

Expose Channel/Layer as HTML video/audio input
I like the idea of having a live camera feed inside a html template, like a classic webcam for browser, and could be nice to have a channel/layer as input, rather than a direct decklink. Maybe also available for more than one template (ex. preview). Could be interesting even for AR in conjunction with threejs.

wild wish science fiction like
Having a server running on a Mac arm, either under linux or natively, with gpu support. Well, because mac usually have a lot of thunderbolt ports and are well build and powerful computers.

Hide media(folders) from media-scanner, like:
user_avatars_hidden/user12312.png
or
user12312_hidden.png

This way when working with large media sets, they donā€™t mess up the media lists.

Also a AMCP command like:
/tls/foldername
Only returns media from that media subfolder.