Blackmagic Forum

Hey all! Quick question for you gurus.

I do a few sports broadcasts and are looking to up our production value by adding live timing & scoring statistics as an overlay. Right now, my overlay could be considered static, non-animated. I have a JSON/XML file that is constantly updated and I need to ingest that file, translate into the text of an overlay and push to ATEM.

What I'd like to prevent is having to layer scenes using key/fill to create the transparency. We've used an automated script and XSplit to do this in the past, however I'm sure there has to be better solutions out there which are made for this exact purpose.

Ideally, I'm looking for a solution to ingest the JSON/XML, it updates the text of the graphic, and then exports to a downstream keyer. Kind of how the ATEM Photoshop plugin works. It's so seamless and quick... looking for something like this.

I know that Fusion is capable of something like this, however feel it's like bringing a bazooka to a fist fight. Looking for something lightweight and intended for this purpose.

Anybody have any ideas / options?

Thanks all!

Anybody ever run into an issue with pushing constant updates to the media pool? I'm wondering if the flash memory on any of the ATEMs would be harmed if static images were pushed every 10 seconds.

You can use CasparCG to output fill and key and write your own Json scripts for html5

Robert Schemitsch wrote:You can use CasparCG to output fill and key and write your own Json scripts for html5

Seconded, Caspar would be a great candidate for this.

Robert Schemitsch wrote:You can use CasparCG to output fill and key and write your own Json scripts for html5

Thanks for the CasperCG tip. I'll dig in more, however it looks like it's outputting to an SDI/HDMI card, which then in turn takes up inputs on the ATEM. This is what I'm trying to avoid. Does this sound correct? Or does the CasperCG server have the ability to output directly to the ATEM media pool? I'm having trouble finding proper documentation for Casper.

Thanks

Yes, I'm afraid you will need a key and a fill unless you fancy using a luma key or something like that which would only take up 1 input. It might be possible to write something that uses Caspar as a frame producer which would then push something to the media pool when a template is updated, but that sounds similar to what you already have.

If you're desperate to keep inputs, you could run the output of the ATEM into Caspar, overlay the graphics and then output Caspar to programme. Not exactly elegant though and would add some extra frames of delay.

Which ATEM are you using, out of curiosity?

live input with caspar add about 5-6 frames of delay ..

but caspar is great compared to images in media player , becasue u can add animations and motion graphics ..

frlawton wrote:Yes, I'm afraid you will need a key and a fill unless you fancy using a luma key or something like that which would only take up 1 input. It might be possible to write something that uses Caspar as a frame producer which would then push something to the media pool when a template is updated, but that sounds similar to what you already have.

If you're desperate to keep inputs, you could run the output of the ATEM into Caspar, overlay the graphics and then output Caspar to programme. Not exactly elegant though and would add some extra frames of delay.

Which ATEM are you using, out of curiosity?

Thanks for this! Sorry, I wasn't notified when a reply was posted. Using the ATEM Production Studio 4k.

I'd really like to stay away from using Photoshop's export function as it seems like that's really taxing the system. I'm definitely going to look into using Caspar as a frame producer... do you have experience with this?

Roman Pytkin Pekarek wrote:live input with caspar add about 5-6 frames of delay ..

but caspar is great compared to images in media player , becasue u can add animations and motion graphics ..

Have any alternatives to suggest?

CasparCG can do live keying with only one frame delay if you use a 4k Decklink card in Internal keying mode (rather than key and fill or as just a consumer with CasparCG mixer functionality).
In this mode, the video passes through the decklink card, from input to output, and CasparCG adds just the fill onto the video.
You would use this post program out from the ATEM.

You bring up some good questions. If you're doing sports broadcasts on a regular basis and looking to up your production value (which should allow you to generate more revenue) I'd look into a solution built for it. Unless you don't have 2 channels on your switcher to give, New Blue Titler Live 4 Sport and a PC running with a decklink or Mac using an Ultrastudio HD mini is my best go to solution.

I've tried Caspar CG and found it to be less than intuitive. YMMV.

Best,

David

nullfoxgiven wrote:Hey all! Quick question for you gurus.

I do a few sports broadcasts and are looking to up our production value by adding live timing & scoring statistics as an overlay. Right now, my overlay could be considered static, non-animated. I have a JSON/XML file that is constantly updated and I need to ingest that file, translate into the text of an overlay and push to ATEM.

What I'd like to prevent is having to layer scenes using key/fill to create the transparency. We've used an automated script and XSplit to do this in the past, however I'm sure there has to be better solutions out there which are made for this exact purpose.

Ideally, I'm looking for a solution to ingest the JSON/XML, it updates the text of the graphic, and then exports to a downstream keyer. Kind of how the ATEM Photoshop plugin works. It's so seamless and quick... looking for something like this.

I know that Fusion is capable of something like this, however feel it's like bringing a bazooka to a fist fight. Looking for something lightweight and intended for this purpose.

Anybody have any ideas / options?

Thanks all!

Thanks for this! Sorry, I wasn't notified when a reply was posted. Using the ATEM Production Studio 4k.

I'd really like to stay away from using Photoshop's export function as it seems like that's really taxing the system. I'm definitely going to look into using Caspar as a frame producer... do you have experience with this?

I guess you probably have enough inputs to spare for a key/fill arrangement then.

We use a very simple HTML scorebug with a javascript timer that Caspar plays. It's worked really well so far, but as others have said - it's not very intuitive, and can require a bit of lateral thinking to get things to work the way you want to. If you're interested, then I'd encourage you to look up posts relating to using Google Web Designer for HTML templates on the CasparCG Forum, I think that's the easiest way to get basic motion graphics with editable text.