Question about media players/upstream key limitation on ATEM

Questions about ATEM Switchers, Camera Converter and everything live!
  • Author
  • Message
Offline

NicWassell

  • Posts: 67
  • Joined: Thu Apr 18, 2013 8:57 pm

Question about media players/upstream key limitation on ATEM

PostThu Nov 25, 2021 10:47 am

Hi all

I've been doing more livestreaming jobs, have an ATEM Mini Pro which I think is an amazing piece of kit for the money and have been enjoying it very much. I just had a question that I wondered if anyone could answer about why there are are such restrictions on upstream and downstream keys?

I do understand that higher end switchers have more upstream and downstream keys but I really don't understand why there is such a limitation on displaying graphics or split screen etc at all in any of the models, especially just one of each for the Mini Pro. It seems the sort of thing that is just a holdover from older, less powerful machines where a image of a couple of megabytes might take up a significant amount of memory. The inability to play video clips is also extremely surprising. I know there are other boxes like the Hyperdecks that provide video playback etc but is the only reason the switchers can't do it to protect the other product lines? I could probably understand the limitation on upstream keys perhaps as I'm sure real time processing of HD streams is still fairly intensive, but in 2021 is it really so hard to show two or three pngs on screen at the same time? I can get round it of course using OBS but then it just nullifies the benefit of being able to stream directly from the deck (and uses up the USB port than I could otherwise record to USB disk over).

I still love the deck and I know that compared with other systems the Extreme is still very reasonably priced which would open up more of these abilities, but it's still one of those things that just seems very odd compared with the other technological achievements we're used to seeing. I'm sure it's down to my just not knowing the technical challenges of this kind of operation but I'd still be very interested if anyone knows?
Offline
User avatar

Roman Pytkin Pekarek

  • Posts: 2312
  • Joined: Tue Oct 30, 2012 1:11 pm
  • Location: SVK

Re: Question about media players/upstream key limitation on

PostFri Nov 26, 2021 12:03 am

all mini series models are cheap models, but with big possibilities .. If U need more, u need another switcher .. Look on bigger switchers .. For example hiend CONSTELATION .. It have not streaming, recording .. etc .. U have it on AMP .. every switcher is for another user ..
http://tally.pytkin.sk
http://chucktv.eu
http://www.stonepp.tv
http://www.media-planet.sk
Offline

NicWassell

  • Posts: 67
  • Joined: Thu Apr 18, 2013 8:57 pm

Re: Question about media players/upstream key limitation on

PostSat Nov 27, 2021 11:35 am

As I said, I do understand that bigger switchers have more capacity so I’m not sure that “if you need more buy something bigger” is that helpful. It was a question of why it seems to take so much processing power that the mini/mini pros can only manage to show one png at a time.
Offline

Dave Del Vecchio

  • Posts: 1603
  • Joined: Mon Nov 25, 2013 10:25 am

Re: Question about media players/upstream key limitation on

PostSun Nov 28, 2021 11:14 pm

Video processing tends to be very resource intensive. This is true across various types of video transforms whether it's compositing (keying), scaling, color correction/grading, or something other effect. Obviously some video transform operations are a bit more complex than others, but they all get more processing intensive as the video resolution and frame rate increase.

Consider something like full HD 1080p60 video which has a frame size of 1920 x 1080 or just over 2 million pixels per frame at 60 frames per second for a total of 124.4 million pixels per second. And for color video such as RGB, each pixel is made up of 3 different values for red, green, and blue (there are other color representations such as YUV may be used depending on the situation, but the main point is that there are multiple values per pixel). So in order to do a single video processing transformation that outputs RGB video, you would need to do calculate roughly 373 million values each second.

So even for relatively simple video transforms like compositing with an alpha channel (which mostly involves some multiplication and addition to calculate each pixel) there is a fair amount of data being processed. And these calculations are per transformation, so the more video operations are done, the more processing is needed.

In the case of the ATEM Mini Pro, compositing is just one of the video transforms it is doing. That unit supports scaling/frame rate conversion on each input, picture-in-picture (which involves scaling and compositing), multiview output (more scaling for each input and the program output), USB-C webcam output (which uses MJPEG compression), and live streaming output (H.264 compression). So all of this video processing adds up.

And pretty much anything that does video processing will run into some kind of processing limitations eventually. For example, if you are using video editing software on a computer, and you apply enough effects or transformations on top of each other, then there often comes a point where you can no longer do full resolution, full frame rate playback to preview the resulting video. And where you reach this limit will depend on the number and type of video transforms being done, the resolution and frame rate of the video being manipulated, and the processing speed of the underlying hardware. So if you want to be able to handle more video processing, while maintaining the same resolution and frame rate, usually you would buy a computer with a faster processor and faster graphics chipset.

Return to Live Production

Who is online

Users browsing this forum: No registered users and 32 guests