PDA

View Full Version : Genlock Card - Anyone using to gain live switching frames back?



tarkken
03-15-2004, 02:20 PM
Hi. I primarily use the VT for live-IMAG. Currently, it appears as though my latency from the cameras to the projected image is about 4 frames. I notice it, but no one has complained, and there hasn't been a 'lip syncing' factor yet (more noticable on drums, etc). So far it's acceptable, but I wouldn't mind tweaking if I could.

Is anyone using the Genlock card - and if so, do you know if it is indeed giving you a frame or two back on your camera inputs, or will the VT always have an internal video buffer (must be around 2 frames or so?). I know my scaler (Analog way) has a 1 frame delay - but it's a necessary evil at this point, and probably helps me from any projector buffers, as everything is scaled up to 1024, and run five wire RGBHV.

I plan on Genlocking everything that I can (all Cam/CCU's, etc) on the way in. If the VT Genlock card will let me get a frame or two back (and even help in switching stability, or allow quick integration to other mixers (Grass Valley, etc), it would be awesome. However, I need to also be assured that it's bulletproof, and not another piece to cause troubleshooting dilemmas in a live scenario.

Good for inputs, or designed for synching the toaster to other machinery (or both)?

Thanks again!
Shawn

PIZAZZ
03-15-2004, 02:29 PM
The genlock card would definitely help you get synced up with other gear. That is it's primary design.

As far as getting your latency down, make sure and genlock your cameras. Use the analog out into the SX8. Component prefered for best quality. The Genlock card would not necessarily help you get the latency down though. DV inputs should never be used for live switching in my opinion there is way too much latency.

What cameras are you using BTW??

Also what are your system specs.

tarkken
03-15-2004, 02:35 PM
JVC DV-550 cameras with the RM-P210u CCU's. I've been very happy, all things considered.

I have a dual Xeon 2.8, with the RS-8/SX-8 - constructed by the good folks at Video Hardware (Blaine Holm, Chris, etc). I heartily endorse them.

There's a picture of the rig on my website now: www.renaissancevideo.com. Click on the 3rd photo on the main page ('Hearts Conference"), and you'll see the live rig, and some live shots.

I'm attempting to Genlock everything with a Horita BSG-50 Blackburst signal (CCU's, decks, and perhaps the VT3). If I can gain a frame or two back by adding the card, it'd be great.

For some reason at the last show, I was getting sketchy composite outs (which I send to the preview monitors) from the CCU's when the Horita was on. I'd turn it off, and everything seemed fine. Not sure what's up - more tweaking this week.

Any other Genlock card users?
Thanks.

PIZAZZ
03-15-2004, 03:10 PM
We use a similar setup but with JVC DV5000 cameras and Sony D30s for larger shoots.

We also have used a Horita Blackburst with success. You can also loop the VT's composite output into the genlock in on the CCUs to accomplish pretty much the same thing as the blackburst if you are in a bind.

That is one of the reasons we use Dual Coax with our DV5000 system. One for video out and the other for video return.

Gordon
03-15-2004, 05:58 PM
Hi. I primarily use the VT for live-IMAG. Currently, it appears as though my latency from the cameras to the projected image is about 4 frames.
Don't forget that today's projectors are adding at least 2 frames to the delay as well. (Digital keystone correction and rescaling to the projector's native output are the main reasons.) So if you are already down to 4 frames with IMAG then that is as good as it gets.

PIZAZZ
03-15-2004, 06:21 PM
Gordon!!!! You are ALIVE!!!

Hadn't seen you around in ages.... Hope you have been busy making tons of money!!!

Give me an email to chat about some script ideas.

jef at pizazzme dot com

Later!
and welcome back!!!!

podro
03-15-2004, 06:22 PM
I also have the 4 frame delay with IMAG and the vt3. And very little delay (read I never noticed after 3 years of watching the screens) when i was switching with a grass valley 110.

I can even notice the delay on my program monitor right above my quad monitors holding the 4 cameras (DSR-390's through CCU-D50's)

I have a genlock card coming in soon, I'll let you know the results.

PIZAZZ
03-15-2004, 06:28 PM
I would estimate our setups to be 2-3 frames depending on cameras used.

This is definitely useable in most all applications.

Frame delay is something that is going to happen with computer based switching. 2 frames is the least that the VT system will ever get. Probably 1 more frame at any projector. and 1 more frame at a scalar. Sucks but that is life on the edge.

podro
03-15-2004, 07:30 PM
I've had several people tell me in our church service that "it may just be me but the screens seem a little behind..."

you can really tell when the pastor is moving his arms a bunch. I'm hoping 2 frames will minimize that difference...

ted
03-15-2004, 09:02 PM
There are Modules that can adjust the Audio to match the video delay.
I don't have a need for one, but for those of you who use a Digital Switcher, (of any kind), and go through a projector system, I'd look into one.

PIZAZZ
03-16-2004, 10:03 AM
Rane Makes a good audio/video delay unit.

Only problem in a church type setting is then the audio would be delayed from the pulpit or preacher to the congregation. It would match the video for the congregation but the person speaking would hear a delay.

This could be a small issue.

Gordon
03-16-2004, 11:26 AM
And delaying the audio is not necessary. The speed of audio in a church at room temp and RH of about 70% is about 38 feet per 1/30 of a second. In a large church, which is where you need IMAG, the front people are watching the stage not the big screen whereas the people in the back of the church are watching the big screen. In our church it is about 150 feet from the back of the stage where the big screen is located to the back of the main auditorium. It takes the audio 4/30 of a second (4 frames) to reach the people at the back (not counting the balcony(ies). Because light travels much faster than audio, the people at the back of the church see the video in perfect sync with the audio. People in the balcony see the video first before they hear the audio and people at the front of the church hear the audio before they see the video. However, the people at the front of the church tend to watch the people on stage anyhow so lip sync for them on the big screen is not an issue.

The only people who notice the delay in video are those whose line of sight allow them to see both the people cleary AND the screen behind them.

If you worry too much about reducing the video delay then the people at the back of the church and in the balconies will be the ones noticing that the lip sync is out because the video is AHEAD of the audio! These are the people that need the IMAG the most so it only makes sense to leave the delay to about 4 frames so that the audio and video reach their eyes and ears at the same time. OTOH, in most larger churches, 3 frames delay in the video would be preferred to 4 frames.

PIZAZZ
03-16-2004, 11:40 AM
Well there you have it.

I couldn't have said it better, Gordon.

3 cheers for Gordon, Yeah Yeah Yeah.

BTW Gordon, got your email. Thanks a ton! See you at NAB right???

Gordon
03-16-2004, 01:52 PM
Thanks.

BTW, I first had to do the math when we did a live switch for a 3000 seat public auditorium. This auditorium was a bit more rectangular than our fan shaped auditorium and here the A/V was in sync at the back of the main floor. The first balcony was still okay but the distance to the back of the second balcony put the audio about 3 frames behind the video. The Toaster with a four frame delay on the video was perfect for this venue!

I won't make it to NAB. We have a large conference here at that time and will need to do a good job on the closed circuit television for the 1500 people that won't fit into the main auditorium. Another great opportunity to show what VT[3] can do.

tarkken
03-16-2004, 02:01 PM
Gordon,

Thanks for your posts and your math. In my last event, it was a large auditorium, and there were about 4000 people in the Saturday session. I jumped out front to where the FOH audio was to check on things, and ask them if they were delaying the audio at all (because the screen looked like perfect sync with the audio as we received it a couple-hundred feet away). They said there was no FOH delay added, and they only noticed the talent latency when I showed it to them (sharp hand gestures, etc). Good to know, considering they were in front of it the better part of a full day.

Your math is very beneficial for the future though - we can use it in the event that conditions/venues change.

After all of this tweaking (and a bit of fretting), it's nice to have a system up and working, and a great forum to check on things.

Thanks again everyone for your valuable information, support, and community.

Best regards,

Shawn

ted
03-16-2004, 09:34 PM
Gordon, you could write PR for the Digital Switcher Industry!:p
"We intentionally added a 4 frame video delay in consideration for all those in the back of the Church".
I like it!:p

tonsofpcs
03-25-2004, 02:23 PM
If I remember the specs right, the VTNT cards have a 400 ns delay inherent, due to the PCI bus being DC based with a timer on the motherboard [long story], this is one of the reasons I like my Amiga toaster better for live switching; since the Amiga was designed as a gaming system, it's internal clocking speed is 59.94 Hz [video speed], thus enabling the toaster to integrate and switch live [it genlocks the whole system to the first input, if you watch closely when you plug in a signal to input one, you will see your Amiga's monitor jump a bit]