View Full Version : Clock shot question

01-15-2013, 11:59 AM
Some have expressed interest in a feature that allowed a camera aimed at the game clock to be overlaid on clips for officiating purposes. In that context, I'd love some feedback. (Requirements of various officiating bodies varies only slightly, but tends to be annoyingly vague on precise details).

Specifically, for any who would find this ability valuable:

Should the clock overlay be burned right into the recording?
Or should it simply overlaid when played using a synchronized clock clip recorded at the same time)?
Any position (on the screen) or size requirements for the overlay?
Anything else?

01-15-2013, 02:20 PM
Hi Steve.

We've received similar requests from basketball and hockey officials.

What we've done in the past is one of two things:
1. Use one of our 3Play inputs to record a POV camera with game clock/shot clock, and then play it back synchronized with one of our cameras as 3Play channels A and B (or do a PiP affect using a virtual input in Tricaster).
2. Send the PGM feed back from Tricaster into 3Play for recording as channel 8. The program feed has a score bug (fed by LiveText and Datalink). That way, we can always play back the PGM feed with a clock.

I think it would be great if there was a picture-in-picture playback effect on 3Play, so that you could play back a composite shot of A and B (with DVE and positioner). That could really help out with this, so that the AUX output from 3play could play back A, B or A+B and be sent to the officials.

Alternatively, if there was a way to incorporate DataLink into 3Play, and overlay the clean score/clock feed on a shot (and be able to key it in and out) that would be great.



01-15-2013, 02:41 PM
Thanks for the feedback, interesting techniques.

For the purposes of this discussion, I'd like to try treat this in isolation, as though there were no TriCaster (or other switcher) involved.

Alternatively, if there was a way to incorporate DataLink into 3Play, and overlay the clean score/clock feed on a shot (and be able to key it in and out) that would be great.From what reading I've done, and while some do take that approach, there seems to be some preference in favor of an actual clock shot. (There seem to have been some officiating kerfuffles over miniscule timing discrepancies that, if not eliminated, are largely circumvented by doing so - though some would differ.) As well, of course, connecting a camera is a lot less complicated than supporting diverse data feeds.

01-15-2013, 09:29 PM
Clock cameras are so yesterday. Actually more like so 90's.

DataFeeds are what the majority of sports people we work with are either at or moving to quickly. Steve go over to Jeremy B. ask him for the Datalink code walk back to Mike, tell him Jef said it was ok to put it in. :)

01-15-2013, 09:36 PM
Oh yeah, Answers to your questions...

Time should be high and way above. Most anyone shooting is going to have headroom above the action in the shot. You guys could leverage that headroom to have a good place for the timecode well out sight of a replay of a play in question.

Definitely NOT burned into the recording permanently. I wouldn't want the time displayed if I pulled the clips for Highlights across the network.
Can this be done by MetaData and chosen to overlay or not when needed?

Same kind of feature for the 3Play itself. Show time or not depending on if you need it or not.

01-16-2013, 06:29 AM
I'd particularly like to hear from anyone actually involved in officiating, as their requirements are reportedly rather more specialized than for standard production purposes.

02-01-2013, 01:06 PM
My understanding is basketball officials DO NOT WANT data feeds as reference. As dumb as this sounds, nobody can GUARANTEE the data feed is 100% correct. only the clock camera gives you that 100% verification. in basketball there is a BIG difference between .6seconds and .8seconds on the clock (the actual physics behind it allows a player to receive and let go of a shot in only as little as .8 seconds, thus making it impossible for a shot to happen with .6seconds )

with that being said, if your data feed is off as as little as .1 second, your data feed has now effected the outcome of the game.

Back to the original point. Having a clock cam 'burned' into program would not be good for the reason PIZAZZ states. an Overlay option much like a DSK would work great if it obviously had cropping options. or being able to have such a setup sent out through the AUX output of the 3play would be even the best option because honestly, who is using the AUX output? outside of a seperate iso referee monitor that doesnt touch program.

02-13-2013, 04:22 PM
Jon is correct. No digital clocks are not allowed for review in the NCAA.

02-13-2013, 04:37 PM
Thanks. That was my expectation, and yet - some systems seem to try to fill this need using a direct digital clock feed (much like LiveText does with Daktronics, et al) ... hence the question.

02-13-2013, 11:01 PM
I have argued (okay, just "discussed") this with refs because I don't feel clock camera cutouts or DVEs over a camera are accurate either. In our situation, we're using an analog composite clock camera, it goes into an Analog to SDI converter (maybe 1 frame of delay), Up Converter to get it to HD (another frame of delay), and then cropped and positioned in a DVE (another frame or so). So by the time the clock cutout is actually shown on the screen, it has 3+ frames of delay or ~.1 second. So the most reliable "official review" source will come from our camera operators who include the clock in their shots under one minute of play in each half.


02-14-2013, 08:44 AM
If you televise a basketball game at the NCAA level and your announcers are at 'floor level', officials have the right to ask for a review from the 'truck'. We provide a second monitor at the announcers station and an intercom to the replay operator for their use. in the analog days, we could assure them that what they will see is what really happened. once we went digital i had several discussions about all the latencies that now come into effect. Brian0105 is absolutely correct. But let me add that even the OCR system will add frames of delay. You have to be careful in what you represent to the officials. 1 frame of delay is all it takes to affect the outcome.

Recently, a nationally televised game when under review for a last second shot. The home audience was shown a digital clock embed on program as the replay. Clearly, the shot should have been waived off. But the officials reviewing this at courtside for several minutes called it good. How could they? It's simple. The review at court side was an an analog clock emedded over either live video or SDI video...therefore, it lead the digital representation by at least a frame, if not more. The digital clock, created either thru OCR or a serial link will always have latency.

We shoot 3 games a week for either ESPN3 or an RSN and 2-3 others for local TV...all using newtek equipment. We instruct our cam ops of the back side of a last second shot to include the red light on the bank board. This will never lie if replayed for the officials.

My suggestion is to never represent your video replay with a PIP as accurate in today's digital production world.

05-05-2013, 04:53 PM
great advice. we also instruct our basket camera ops to include the backboards in their shots under 1minute. I agree latency will unfortunately always exist. although this is something the 3play needs to be able to do as a function.