View Full Version : New to VT4, Questions About sync & lag (I searched already)

09-17-2007, 05:28 PM

I have searched the forums for 2 hours now, and I've reached the exhaustion point (aka the post a new thread point). Most of the solutions that people come up with say "remove downstream sources of lag"

I have just installed and set up my system, and I'm running a direct signal from my Sony BRC-300 to my VT-4 via my SX-84, and then back out via the SX-84 to an LCD monitor. Tests hooking the camera directly up to the LCD produces almost 0 lag (not even a frame, more like maybe .3 frames). My test is doing a live motion (dropping a pencil onto a table) in front of the lcd screen and recording it. Watching frame by frame, now I time the difference between when the pencil hits the table and then how far behind the LCD screen is. Not incredibly scientific, I know.


Doing this test through my VT4 I'm getting ~7.5 frames of lag. 3 frames is ok. 4 I might be able to live with, but 7-8 is 1/3 of a second and very noticeable.

Of course this is the lag I get before we put in the line driver between our camera and VT4, the switcher and splitter/line driver we put between between our VT4 and our projectors, and of course the projectors them selves. Now we may have to simplify our process later, however I'd like to get it running optimally just out of the VT first.

Possible Cause

Here's one thing I can think of that may be causing this and it very well could be the culprit and would like your input. Right now I'm running only 1 processor. An Intel Xeon 3.20 GHz. The plan was for 2, but one was DOA and had to be shipped back. We were debating on whether 2 was a necessity or just an extra in case things ever got really complicated. So, is the 2nd processor, or lack there of what is likely causing this in your opinion, or are there other things for me to check?

Other System Info

We are running a GForce 7600 GS (which I believe doesn't matter for the video output anyway, please correct me if I'm wrong) and 3 GB of RAM.

All your help and information is of course much appreciated. I'm glad to be part of the community and will most certainly work to give back information to others once I'm more familiar with the system.

09-23-2007, 05:58 PM
ieee1394, Component, or SVideo In & Out?

09-23-2007, 06:05 PM
I have never seen anything suggesting that a single processor would introduce more latency! In fact, I know there are others out there running VT on a single proc for live production and don't have that issue. The areas you'll see a single proc slow you down is in multiple DDRS with playback, record, etc, or in rendering during post.

as the last post suggests, firewire input will definately give you latency.

Your first point - "remove downstream sources of lag" - i differ with that statement that most people on this forum will tell you that exclusively. Yes - that's true... but genlocking your cameras is another major issue, and many people have talked about that. There's some work to do behind that to get it all happy.

Did you work with a vendor or system integrator? If so, they should be helping you. If you didn't, then this forum will remain your best bet. There are alot of discussions on the topic... i know, because I've had many of them, and followed many others.

Another search would be to look for threads by "cineman". Nes seems to be the resident expert around here and has had much discussion about it.

09-23-2007, 06:18 PM
I have seen this with the BRC300 cameras once before. Are you using them inverted or sitting on their base?? Reason I ask is I remember a Sony tech mentioning that if they are inverting the camera the processor could add lag/latency. You should test it both ways.

How are you connecting the cameras to the VT? that is the next question.
Composite, Y/C, Component or Fiber optics. IF it is Fiber then I would eliminate the convertor box to test the latency through VT. It kind of sounds like you have in your post but I just wanted to mention it.

Did you genlock/sync the cameras together by chance? You can just use the composite output of the VT to sync them all.

Please list exact model numbers and signal flow for the rest of your gear after the VT also. More info the better.

I would not worry about the single processor being the issue. If you could list exactly all the parts in the VT PC then do that too.

09-27-2007, 03:32 PM
First off, I'd like to apologize. I jumped to conclusions that my problem wasn't just like everyone else's and started a new thread. Of course, after analyzing my processes, I indeed found out I was, indeed, like every other problem. My problem was the HD TV that I was using as a test monitor. When I plug the camera directly into the TV using component, no latency. When I plugged in the SX-84 directly to the TV, 7 frames of lag. So therefore VT is causing 7 frames of lag, right?

Oh no. For what ever reason the signal from VT has to be processed more by the TV or something (despite both using the same format and compression) because sure enough, as soon as I hooked it to a CRT monitor, there were only 2-3 frames of lag. Anyway, I will try to keep a more open mind. I'm sure that I'll need a little help at some point later, and thank you for all of your responses.

The consensus seems to say that you can cut lag down from 3 to 2 frames with manual genlocking as opposed to letting the VT internal gen.

So I'd like your opinion. I have two BRC-300 that are genlockable. I have one Cannon XL-1s that is not easily genlockable (I haven't looked into external devices yet.) To start out, should I genlock the two BRC-300s, or should I leave all three unlocked to get a more consistent look?

We've been looking at buying a new camera. Any suggestions on a genlockable camera? Right now I'm eyeing the Cannon XH-G1 and XL-H1 (the G1 is definately more in my budget, but the H1 has the ability to swap lenses.)

Last but not least, I will have a line driver between my camera and my SX-84. If I'm sending the composite signal back as my sync (genlock) signal, and I don't put a linedriver on it, and the signal has some ghosting, (but no phasing), will that effect my picture that comes back?

Thank you for all of your help.

09-27-2007, 04:14 PM
Swapping lenses isn't the only strength of the XL-H1, but if that is it's big plus for you compared to the XH-G1, you might ask how much you want to budget for additional lenses, and specifically what other lenses you will use regularly. I've often heard people say they chose the XL-1 or 2 (predecessors to the XL-H1) because of the abilities to change lenses, yet I've never seen anyone use anything but the stock lens (which is a fantastic lens) on them.

09-27-2007, 05:02 PM
I already have 3 cannon lenses. The stock is great, but I have 3 lenses for our Cannon XL-1s. We have one that is all manual, which means no auto focus but the focus and zoom are mechanically connected to the lenses so the lenses move exactly as you adjust them (as opposed to the electronic that sometimes take a second to react and often don't react quite as you thought they would) and a wide angle lens which is just for those special shots such a lens requires.

Seeing as we already have those lenses for the XL-1s, which are compatible with the XL-H1, and the fact that we do use them more than maybe your average person, that feature on the XL-H1 definately interests me. Is it worth the extra 3Gs? We'll have to see, but any other suggestions of cameras with genlock would be great.

09-28-2007, 07:50 AM
Moving into a camera with a CCU is a big jump in price, but it certainly is nice. Apart from having genlock, you have complete control over the camera from the control room. Most Sony and JVC's ENG cameras have a CCU option, but it doesn't come cheap.

We've gone that route, and been happy with the results.