Welcome to our community

Be a part of something great, join today!

Register Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

AJA Gen10/RED showing genlock but it's not real, what gives?

Phil Bates

Phil Bates

Well-known member
Joined
Apr 4, 2007
Messages
685
Reaction score
0
Points
0
Website
www.artbeats.com
We've used the AJA Gen10 for months with out a problem, genlock spot on everytime. However, in the last month we've noticed the odd clip with shutter out of phase. We chalked it up to us not paying attention to the "padlocks" and must of missed something. However, on our most recent shoot, we lost a third of the shots to shutters being out of phase. I KNOW we checked the padlocks this time. We rarely powered down, so the genlock is changing between shots while everything is powered up.

So, I can't trust the padlocks now?

Anybody else having trouble with this? Any idea what I can do?

I'm dead in the water until I can figure this out.

Here are my settings:

RED ONE MX
Shooting 3K, 60fps
Shutter speed 120th/sec
AJA Gen10 jumpers set to ON: 1, 2, 3
How I know the genlock is out of phase: It's obvious in the clapper motion blur and in the subject which was moving at high speed.

Thanks for your help,
Phil
 
Phil,

In all the shooting and testing (included those I had Pedro do for JuliaX) we never got good results at speeds that are not multiples of 24. I'd try shooting at 48, 72 or 96 and see what happens. Also, I'm not looking at one right now, but I am pretty sure you want those Gen10 dip switches set to 1,2,7
 
Thanks for responding Jason. Question for you, in all your testing did you ever come across a scenario where the padlocks were showing a lock, but the resulting footage didn't?

Thanks,
Phil
 
The padlocks always show, the sync is always a little off, and Red has acknowledged that. JG, have you tested Epics yet? Red claims epics will be perfect.
 
3D Sync

3D Sync

And the same can be true of RED ONE - how are you judging that the sync between cameras is off?

And which build are you on ?
 
Stuart English said:
And the same can be true of RED ONE - how are you judging that the sync between cameras is off?

And which build are you on ?
Click to expand...

Hi Stewart,

I can tell by looking at the slate motion blur, then at the motion of the subject, which in this case is pyrotechnics. Some shots are absolutely dead on. Others are off, nothing is just barely off. The hits and misses are pretty clear. I could send you images if you want.

The build is 30.5.0.

Thanks for helping,

Phil
 
3D Sync

3D Sync

So are you judging this "live" by comparing monitor feed to live action, or by comparing the .R3D's recorded by the two cameras ?
 
Comparing the R3Ds. I bring the R3Ds into AE then do snapshot A-B comparisons. Sometimes I'll match geometry first, if needed.

Phil
 
60fps on RED One will not sync for 3D AFAIK. It may show the padlock, I don't recall. The very popular thread Pedro posted about genlock is actually a report on my research, despite his presentation of it as his own. Re-read his posts to see exactly what we did and the results we got. One thing you could try if you need 60fps is to set your camera's timebase to 30 or maybe 29.97 and make sure your Gen10 is set for that timebase as well. Not sure if this would work, but that is where you should start.

I was under the impression that EPIC was like Fight Club in that we are not supposed to talk about EPIC, but if Mark is, then I will say this. While I cannot post the photos, I did conduct the same tests we did for JuliaX when we started The Amazing Spider-man back in November. Of course this time I was using EPICs. Sync signals were provided by Aja Gen10 as usual. Tested at 1/1000th of a second 23.976fps, genlock and shutter sync are dead nuts on. Not even a slight offset (as with the RED ONE). So there it is, they fixed it. Exactly what speeds the camera can genlock at remains a moving target and probably a RED secret for the time being so that's all I'll say.

It's been brought up a million times but it is worth mentioning again. The rumor that RED ONE's cannot be genlocked, or have genlock problems is unfair and false. We have shot quite a bit of 3D RED footage at speeds up to 120fps with 180 degree shutter and have NEVER found the shutter phase issue to be noticeable in real world footage. In fact, some of our 120fps 3D footage is among the most sought after material in the 3D stock footage market. Anyone who has seen Light The Wick has seen Sammy Carlson, John Spriggs, and Byron Wells breaking freestyle skiing world records in beautiful slow motion with no visible artifacts whatsoever. Same with the Monster vs Mini Cooper commercial (although the really, really high speed stuff in that was shot with Phantom), same with 3D Bikini Car Wash.

The way to test genlock sync and accurate shutter sync is to photograph rapidly changing LEDs like those that make up the numbers on a smart slate. Some stereographers have also created LED devices specifically for testing sync, those are good too. Point is LEDs refresh really, really fast. Set your shutter speed to 1/1000th and set whatever frame rate you want to shoot at. If your slate numbers look identical (no partial numbers in one eye and whole numbers in the other) you are in sync.

Do not assume that one frame rate being in sync means that all will. I encourage you to test the exact setup you intend to use and not waver from that, only way to guarantee your results.
 
BTW, a good way to compare left eye/right eye when checking sync is to lay them on top of one another in your AE timeline, put your slate at zero parallax and then ride the opacity control from 0% to 100% and back. AE will do the opacity change in real time and you can very easily evaluate any changes in the numbers and or motion blur of the clapper.
 
Using a Cinetal or a transvideo, I find that even at 24 the sync between the cameras is only correct 1/3 of the time, even thought the padlock comes on. We can resync repeatedly and randomly the sync will snap in. Using an SBT with a T. 60fps syncs up fine for us when the genlock is at 30 and leaving the camera base at 24.00. By using various genlock speeds of 24, 25, and 30, I can get sync on both cameras at 24, 25, 30, 48, 50, 60, 75, 90, 100FPS with only 72fps conspicuously unable to sync. Not all monitors or stereo display processors will fail when given out of sync feeds, and you actually need to have at least one monitor that is sensitive to the out of sync and keep jamming the cameras until there is good image on that monitor. We have convergence carts with 3ality 3play Pro boxes feeding transvideo 245W monitors over hdmi and the sync always looks good, until the cinetal or transvideo wires in and the sync can truly be seen.
 
Again we need to differentiate between sensor sync and HD-SDI sync. If there is a "failure" to sync, it will be at the HD-SDI output level not at the sensor level. So recordings are correct, it's a monitoring issue.
 
Not entirely. During prep, when i noticed the problem with the sync, I checked the files and they were indeed out of sync despite the lock being displayed. Having now done 4 weeks keeping hdsdi sync correct, I have not seen any further problems with the sync between the cameras. I think it would be more correct to say that if the HDSDI are out of sync there might be some chance that the files are indeed correctly synced, but with the HDSDI in sync the files are more certainly in sync.

If there is a known issue with the HDSDI sync, why is there no bulletin or warning or manual addendum explaining this?

Are the end users of this camera system really expected to wade through this 1000 monkey typing at 1000 typewriters festival of a user forum to find such important information?
 
I'd first need to know which build you are using Jasper.

w.r.t HD-SDI sync, there are two independent issues with monitoring 720p / 60 when you are imaging 24 fps.

One is the actual frame sync, and one is 3:2 pulldown phase. Even if the frame sync is accurate, if the 3:2 pulldown phase is different, some monitors will indicate the HD-SDI outputs are "out of sync" even when they are not.
 
Stuart many times I will have to toggle the genlock to get my HD-SDI sync to bite, but I always get there with the toggle.

OP: I agree with Jason, everything we shoot that is a multiple of 24 has worked swimmingly.
 
Stuart English said:
Again we need to differentiate between sensor sync and HD-SDI sync. If there is a "failure" to sync, it will be at the HD-SDI output level not at the sensor level. So recordings are correct, it's a monitoring issue.
Click to expand...

Came here from another thread.... How do we then acheive output level HD-SDI sync? Especially when trying to send an L and R HD-SDI signal to a stereo monitor that requires a genlocked signal like the Panasonic 2550?

Build 30.7
25 Frame Timebase (To be able to shoot 100 Fps, 16X9, Redcode 36)
2K 16:9
Ambient 203 for sync and timecode
Padlock and Chain icons are present

Should I switch timebase to 24 fps? Perhaps a 3:2 pulldown issue?
 
I decided to revisit this thread because I have just learned of the sync box that everyone who wants to shoot 3D on RED should be using. I have enjoyed the AJA Gen10 but will be putting 3 on ebay now that I am using the Denecke SB-T. This wonderful product is similar in size to the AJA but also provides timecode. THIS IS THE FIRST PRODUCT I'VE SEEN THAT KEEPS THE TIMECODE OF 2 RED ONE'S IN SYNC!!!! Works very well, cannot recommend it highly enough. Ambient Clockit, not for me at all.
 
You must log in or register to reply here.
Back
Top