Panasonic HS410 - delay - genlocking - and which projector has the smallest delay?
Sorry if the following is sounding very obvious to you - I'm not very familiar with live switching and I need a little theoretic explanation genlocking cameras.
We have purchased a Panasonic HS410 mixer, controller AW-RP50 and several robotic cameras AW-HE120 and a HPX250 in order to record symphonic concerts. For now we are recording and streaming concerts. But in 6 months we will project selected robotic cameras on a screen in the hall with a 14.000 AnsiLumen projector which will demand the smallest delay possible. I have received diverging info on this particular mixer. "It should excel int its small delay - so you shouldn't need to genlock the cameras since it will only introduce maximum 1 frame of delay". Others tell me I will always benefit from genlocking. When I read the specs it mentions that switching the frame synchronizers to OFF - there is only a delay of 1 line. By using the FS it will delay 1 frame. What is the difference of 1 line and 1 frame? So I need a little advice from you guys. Wouldn't I be able to shorten the delay by genlocking those cameras providing a genlock in? And if positive - what is the best way to implement it? By letting the HS410 generate the signal or by using an external generator - and which generator?
When choosing the right projector i also need an advice. Any of you have experiences with different brands? Again - the projector with the smallest delay has the highest priority for us.
Thanks in advance,
Sounds like a very nice system you are implementing, however you are right to be concerned with delay or as we call it "lip flap".
The difference between genlocking cameras and not is the difference between 1 line (1/525 Frame) and 1 full frame. In either event run all the cameras at the same resolution. I am not experienced with that switcher, but the one frame is the best you will experience. Add a lower third or graphic or an effect betwee two sources and it will increase. We have had good experience with a grass valley synch generator. It provides both bi-level and tri-level synch concurrently. Tri-level is preferred by some HD cameras and some don't care. We use a mixture of Sony robotic and HPX370 cameras. We run Tri-Level on the Pannys, but the Sony HD robotics will only accept Bi-Level synch.
Also, cameras have some built in latentcy and with some it varies according to setup. We have found that our Sony robotics introduce a frame of delay if we invert them. Turns out the inversion calculation is done in embedded software.
As far as projectors, they are notoriously slow if they must recalculate or convert the source. Always feed at native resolution if possible. Projector manufacturers do not publish or disclose the delay numbers. We have found that some manufacturers are faster. There are very few manufacturers of the core projectors. Most are OEM'd and then add their own front ends. The front end interfaces are where the delay comes in. For example, we have found the EIKE projectors that are OEM'd Samsungs are a couple of frames faster. I don't know where to get data without doing your own testing or asking on this forum. There will be variations between models so you canon make a blanket statement between vendors. In any event count on at least 2 frames for the projector.
We made a very simple poor man's delay test setup that you may want to take a look at. The trigger device is a loud switch and a light bulb. Mike the switch so you get the sound of it being thrown and set the light somewhere you can shoot it. Do not use flourescent as they have delay of their own. Shoot and record a setup of the camera wired directly to the record device. Shoot once directly to the bulb, once to the monitor and again to the projected display. Toggle the switch while recording. Take the three shots and import them to your editing software. We use FCP. On the timeline you can see the frame where the light comes on and also the pulse on the audio where you threw the switch. This will give you the delay to within one frame. Compare the bulb vs the projected image and that will give you an indication of the delay introduced by the projector. Do the test again and introduce the switcher to the circuit. Continue by adding devices. Don't forget to try it with each camera you have. With all components introduced you will have a good indication of your total delay.
Total delay is the sum of delays introduced by each component. With a known 1 frame for the switcher and 2 frames minimum for the projector you will start with 3. A good rule of thumb is that anything less than 5 frames of delay will require an expert to detect it, at 5-6 frames a few people might notice, and at 7 frames everyone will see it.
Pardon the long post, but hope this helps. Let us know what you found out.
My $.02 worth
I've used my HS410 for several larger shows, feeding various make/model projectors, with the frame syncs enabled, and doing some up-conversion from SD to HD on a couple inputs. Some of the experts on the most recent crew figured 3-4 frames of delay; by just standing in the middle of the arena and eyeballing the screens. I can't remember if they bothered delaying the audio, but if they did it was minor.
Edit: Oh yea. We use HPX-370 cams and no genlocking. Used some older scan converters on the up-converter inputs for some roll ins.
Mark D. Walleman
Relay House Inc.
Thank you for your comments - very useful!