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 :)

Red One genlock tests for 3D....

Out of interest has the sync issue discussed on this thread been fully resolved now Build 30.6.0 has been released?

I'd be interested to hear or see results of tests with the new build. (a test similar to the TC-clapper/fan example which Pedro posted earlier on in the thread)

Cheers

JohnF
 
I'm shooting a 3D feature for the next 2 months (not red) so unfortunately I can't replicate my tests until after then since I'm on location out of state.

I'm sure it has been improved on if not entirely addressed. But still someone should test it out.

if not when I get back I setup up my test again....
 
Cheers Pedro,

All the best with the shoot!

JohnF
 
Right now I'm working on a 3D feature in Seoul. We've got two Red MX's running build 30.6.0, genlocked with an Ambient ACL203.

What I've seen so far:

If I boot the cameras with genlock applied, we can shoot up to 120fps and the sensor sync is solid.

Video reference from the Lockit is 23.98/1080p.

We can shoot varispeed at 48, 72, and 120fps. Strangely, the camera absolutely will not indicate good genlock at 96fps varispeed, despite it being a nice even 4x multiple of the base rate.

With 25fps clock reference, we can shoot 100fps, but I would prefer to avoid the downstream hassles caused by changing timebase.

I tested these frame rates at 1/2000 shutter angle with smart slate and spinning fan blades. The sensor sync between the two cameras looks good.

However....

The Transvideo does not like the HD-SDI signal coming from the Reds. About 70% of the time, the overlay/3D mode on the Transvideo is out of sync when we switch it to 3D view. Or if I do something like put color bars up on one of the cameras, then turn them off, the Transvideo will lose sync. Or changing the frame rate from 24 to 48/72/120, the Transvideo will lose sync (despite the Reds always indicating good genlock).

I came up with one fix where if I set the Transvideo to Input 1, then toggle the PVW Refresh on the Red to 50Hz, enter, then back to 59.94, enter, then toggle the Transvideo back to 3D view, sometimes that will fix it on the first try and it will sync. Other times I do that procedure 3 or 4 times and then it syncs. Other times, particularly at varispeed frame rates, the Transvideo will not lock at all.

Meanwhile...the 3ality box feeding the director's Hyundai monitor always stays in sync, at base frame rate and 48/72/120.

I recall a thread elsewhere on this forum where someone documented that the signal quality of the HD-SDI output on the Red is subpar, so I'm thinking about getting a couple of AJA DA's to reclock the HD-SDI output of the Red's, and see if that improves the situation with the Transvideo.

-Bryan
 
Genlock at 96 fps

Genlock at 96 fps

We can shoot varispeed at 48, 72, and 120fps.
Strangely, the camera absolutely will not indicate good genlock at 96fps varispeed, despite it being a nice even 4x multiple of the base rate.
Click to expand...

Yes, that one puzzled us for a while, but we have recently found and fixed that bug..

The fix will be in the next update to the Beta build 31 firmware.
 
Thanx...
 
We regularly experienced the issue you describe with Transvideo when using AJA Gen10 for genolock. What I do, and I do this many many times on every production, is to cycle the power on the Gen10. Usually 2 - 3 restarts will get the monitor to respond properly, sometimes it takes more restarts but each one takes only seconds. Maybe this is easier than changing around the PVW refresh, maybe not. Good to hear that 96fps will finally work!
 
I simply remove the genlock cable from the lockit, then re-attach. It causes the Reds to re-sync and the HD-SDI output to re-align. With the transvideo in 3D mode, re-do the process until the picture is completely frozen, then cycle through the inputs on the transvideo. When you get back to 3D mode, it should be stable. Sometimes cycling inputs on the transvideo will work when the 3D mode is more than 50% frozen, but if the image is 100% frozen in 3D mode then when you cycle through it usually will be perfect.

We also have the AJA FS1s in service, and on the few times we have run the transvideos straight out of them there has been no issue. 3D mode works out of the box. Of course the FS1s are genlocked to each other, which makes a difference. Most of the time we use the transvideos as a local camera alignment helper directly attached to the cameras, so we need to go though the above procedure to be able to use them.

We dont use the Gen10 on purpose, BTW. Its never made sense to me to have a band (timecode) without a drummer (genlock) in sync.
 
Mike Gross said:
At 1/2000th, your exposure time is 1/2 a millisecond or a ~4.6 degree shutter. Position on screen with a spinning or focal plane film shutter or a rolling shutter would cause the objects to shift a bit. That's just par for the course and I don't believe it's related to sync.
Click to expand...

Well, that might be but I just tested our Phantom HD Gold sync'd in a beamsplitter on the 3D feature I'm working. We were setting the package up and I thought lets throw a slate in there and try it out....

So I'm not saying your wrong just saying this is the results I got. The Phantom locks it's shutters to each other via a cable. No external sync generator was used.

The phantom is also a rolling shutter like the RED. I did the test at 144fps...and 24fps with 4 degree shutter.

The test setup was very make shift inside on one of our camera trucks..

IMG_0488.jpg


You can see a tiny difference but alot less than I seen before
(this was taken with my phone, sorry for the quality)
phantomsync.jpg


anyways......thought people would be interested....
 
Jason Goodman said:
The difference seen here could be an exposure difference. Shutter phase looks dead on.
Click to expand...

Yeah absolutely dead on.

Also,

We just finished our aerial portion of the feature and I have a couple of Kelslow's RED MX bodies with the latest 30.6 build.

During setup of the cameras I threw a slate in got this. Seams like it's still off a bit.

ONCE AGAIN......this is minuscule difference the one can rarely if ever see. Just did this to continue the info with a new build.....Once again at 24fps everything was just fine...

here was the setup in the helicopter,
IMG_0367.jpg


Results...the same. (sorry for misalignment, 2 cameras on the camera cart during setup)

synctest1.jpg


synctest2.jpg


Anyways unless you plan on shooting with fast shutter speeds this is pretty much a NON-Issue at this point.
 
So a very good friend and CAS sound man shared with me that there was a recent big meeting at Sony for the sound guys and the overall view was very negative on "tri-level sync" but rather straight sync or black was perhaps preferred. Evidently tri-level can create more problems than it solves.

I have allways used tri-level and the Red Opps manual calls for tri-level. I assume that works best...but:

Has anyone tested the different types of sync?

Thanks,
Rich
 
First of all, this is one of the best and the most useful thread I ever read in Reduser. Cheers for everyone :)

I have upcoming 3D shoot with two red MXs next week and have schedule to test them later this week. I have used RED before but haven't setup for 3D yet. This is my first time to setup with Reds in 3D. Here are my questions,

1. Checklist: To sync two RED MXs, what I need are
a) AJA10,
b) 3D field Monitor,
c) Beam-splitter Rig
d) some bnc cables, right?


2. I am thinking to shoot with shutter angles of 45 degrees (or 1/192) in 48fps or 60fps in some shots. Hope it doesn't generate out of sync problem

And I direct a magic show (I didnot have much budget to get color grading so some of color is bit off but over all I was pretty fun project! )

http://www.youtube.com/watch?v=e7TKtr6hW9Y

Nix Hong
3DPlan
 
I finally tested RED MXs yesterday
We were using Aja Gen10 to sync two Cameras. It seems the one of camera's timecode started late after a couple of times recording even with 23.94fps. We couldn't even shoot 120fps because the slaved camera were locked, We were able to shoot 100fps but the timecord were un-synced during the recording.
The setting we were used...

Tested Frame rate: 23.94/48/60/100 fps
Gen10 Jumper switches: 1-2-7
Gen10 Outputs 5 and 6
RED MX Camera Build: 30.6.10
Visual sync method: Ipad Timecode Slate
Want to know what did I do wrong??
100fps.jpg
gen10_setup.jpg
 
Hi Nix,

First off (no offense) get a real slate.

One of a the easiest methods to check syn is the "clapper" of the slate. Usually by looking at motion blur of the clapper you can tell.

Forget about TC matching on Red MX. The only way I have ever seen it is when someone re-clocked and built in a 2 frame delay then the 2 TC matched.

In any case I wouldn't worry about TC too much at this point.

Also it's the LED style of the "smart slate" the gives us the more detailed look at the sync. Since at higher frame rates you can see the refresh of the LED "roll" thru the display giving you a good look at each camera.


Yesterday, I did the same sync tests(with deneke slate) on a ATOM rig with 2 Epics.....when I have time this weekend I will take a look at the footage and see the results....

Picture4-1.png
 
You must log in or register to reply here.
Back
Top