Lens Resets Itself

[Harry Smolin]

I tested Joker again in our largest auditorium today. I initially had it set to the flat version in the 2.20 masking and lens based on testing it yesterday, but found that the lens had reset itself and was hyper-zoomed in, which caused it to overshoot. I had to switch it to the scope version and hyper-zoom that version in the 2.20 lens to get it to look presentable. We had some maintenance done on the projectors this morning, so that may have something to do with it, but is it a common occurrence for projector lenses to reset themselves like this? The flat version in 2.20 masking and lens still looked fine in our two smaller auditoriums. All three theaters have Christie projectors, the one for the larger auditorium being much bigger, so I'm wondering if this might be something common to that brand.

[Frank Cox]

The lens shouldn't change its setting unless you tell it to do so using the local control panel or send it an appropriate cue.

I have a library of cues on the server that start the lamp and set the appropriate screen file. So for example I will put Black MOS 10seconds as the first clip in the playlist, and put "2D Flat on Screen" at the start. (Or 2D Scope or 3D Flat or whatever is needed for that particular movie.)

Then every time you start the movie the screen ratio is automatically to what's needed for that particular movie.

[Marco Giustini]

It's important to "reset the lens" at service time, that will take the lens back to a known position and reset the software position. If you don't do that what can happen is that someone adjusts the lens manually (which the software cannot know about) and THEN do the lens file. That lens file will work UNTIL someone resets the lens.

I'd imagine the technician might have reset the lens in the morning (or, to be honest, it might be that the lens has reset itself, it wouldn't be the first time I see that). On a properly maintained system, where the lens is reset periodically and nobody touches the ILS, that should not cause any issues - or some minor ones.

If you did your lens file on a out-of-sync ILS (that is, the software thought the lens was in position X but it was in position Y), then the first time someone resets the ILS, what you describe happens.

To be clear: RESET the lens doesn't mean that the ILS files are reset. It means the mechanical and software positions are sync'd. Oh, on a Christie you'll know when a lens reset as it makes a HORRIBLE noise like the ILS is falling apart. Well, that's normal.

[Ioannis Syrogiannis]

From what I read and what others are saying, it seems to me that the server and projector do not have a preset and cue for the settings you go for during the specific screening.
The simplest solution would be to create those (yourself or having someone to do so, best case scenario, they will create a lamp preset for that as well, instead of using Flat or Scope).
Then, the proper settings will be part of regular operation the same way that Flat or Scope are.

[Ryan Gallagher]

I tested Joker again in our largest auditorium today. I initially had it set to the flat version in the 2.20 masking and lens based on testing it yesterday, but found that the lens had reset itself and was hyper-zoomed in, which caused it to overshoot. I had to switch it to the scope version and hyper-zoom that version in the 2.20 lens to get it to look presentable. We had some maintenance done on the projectors this morning, so that may have something to do with it, but is it a common occurrence for projector lenses to reset themselves like this? The flat version in 2.20 masking and lens still looked fine in our two smaller auditoriums. All three theaters have Christie projectors, the one for the larger auditorium being much bigger, so I'm wondering if this might be something common to that brand.

It's hard to advise without knowing what the Tech's were onsite to do. Were they there specifically to work on the presentation of that film?

It's easy to get into the weeds with these non-standard aspects, especially on films that use more than one aspect.

The first things I would check:

Do other films using other projector channels also look wrong?

Does the in-projector framing chart look correct for your flat and scope setups?

Did the film provide a framing chart DCP, or do you have one to use? They definitely SHOULD have for this one if they want anyone to get it right.

From what I read about Joker, and your decision to show the FLAT version. I would think you actually want to be in the FLAT 1.85 channel/masking, not 2.20 as you stated?

The reason depends on how they handled the aspect changes. Did they size their 1.33 and 1.90 sections to fit within the box that is otherwise contained by the 2.20 image? Or did they use the full container instead and show the 1.33 section at the full height of a normal flat container, and the 1.90 section with very slight letter boxing relative to the normal flat container? If they used the full container you need to show it in 1.85 Flat masking even though the bulk of the film will be 2.20 letterboxed. I expect that is the way they built it because most cinemas cannot properly mask to a 2.20 image these days for the other approach.

But if they did build it with the aspect changes landing within the 2.20 image, then you were on the right path to show it with fairly custom 2.20 channel/masking, and your zoom issue might be found elsewhere

[Ryan Gallagher]

Circling back to the projection note in the feature offsets thread:

"When in 2.20, the film will intentionally have small black bars on the top and bottom in the flat version and on the sides in the scope version."

Assuming that came from the studio and was not you paraphrasing. This to me reads that they intend you to use the FLAT or SCOPE settings depending on which version is appropriate... and not show it setup for 2.20 sizing/masking as you stated you were trying. That may not solve your other issue but it's worth noting.

[Marcel Birgelen]

If you show the scope version, you zoom to "SCOPE" and mask your screen to 2.20 if you have such a preset. Do not zoom to F-2.20 when presenting this movie in scope, it will cause it to become window-boxed.

Since lens zoom and focus stops drift a little over time, it's normal to recalibrate those presets during maintenance using proper framing charts. Maybe a wrong framing chart was used during the last maintenance while re-calibrating for F-2.20. S-2.20, on the other hand, does not need any special lens macro.

To test this theory you can try a framing chart yourself and put the projector into F-2.20 and see if it overshoots.

[Harry Smolin]

That actually was paraphrased. Sorry if I threw anyone off; I should have made that clearer. A good portion of the projectionist letter is diagrams and that was my attempt to translate the diagrams into text as I wasn’t sure I could post them here. Still figuring out how embedding pictures in these posts works. Attached is the full projectionist letter with the diagrams; hope it worked here.

The studio’s actual words in case anyone can't see them were:

“The feature has been provided in both FLAT and SCOPE DCP containers.

NOTE FOR ASPECT RATIOS: Please note this feature has multiple aspect ratios. These shifts are intentional and the flat and scope DCPs have been optimized to provide the best viewing experience for the size of the screen it is being displayed on.

Do NOT apply any additional zoom/scaling as this will result in active image being cropped.

The FLAT DCP opens in a 4:3 ratio, followed by a 1:90 ratio for a short period then transitions to the 2.20:1 ratio for the remainder of the film.

The SCOPE DCP opens in a 4:3 ratio, before transitioning to the 2.20:1 ratio for the remainder of the feature.”

I ended up switching to the scope version on the largest screen and setting that to run in a zoomed-in 2.20 lens and with 2.20 masking. The image on that version appears to be contained within a 2.20 frame (it at least never exceeds that), and the 1.33 section is pillarboxed in that version, so nothing is being cut off, and with the masking and the aforementioned zoom-in the image fits perfectly. It looks nice in standard scope too, the only discernible difference being the small black bars on the sides. I may switch the two smaller auditoriums from flat to scope as well as that version looks much better. The flat version still looks presentable with the 2.20 settings and I haven’t noticed anything being cut off, but the scope is easier to keep in one place with fewer aspect changes. The 1.33 and 1.90 sections do appear the way a standard film in those ratios would in that version.

The techs were just doing standard maintenance and pulling logs, nothing involving a specific film. I’ll likely never know if it was that or something else that caused the lens issue. If something like the lens reset mentioned earlier were to randomly happen again though, I might just set the scope version to project in standard scope as I know that lens will reliably stay where it should.

[Marin Zorica]

Besides diacussion what and how Joker screen should be fit, about lens preset problem. Does it "remember" saved position when you call it or that is issue? As other advised from time to time lens needs to be calibrated, or that can happend during time. But if you can save position and call it back, and it works that everything is ok. Othevise, i have sometimes had issue with encoder on lens that had bad output and you could not or had bad position when call it.

[Marco Giustini]

besides the lens issue, the studio went the extra mile to make FLAT and SCOPE versions of the DCP so you shouldn't need to make any 2:20 preset, the studios did that for you as shown in the nice paperwork delivered with the DCP.

A different story is when the studio only releases one format - say Scope - and you have a Flat Screen. Then it's a good idea to make a 2.20:1 preset.

Still, lens should not reset

[Gordon McLeod]

We set up all our Christie machine to reset lens on powerup as they are notorious at just drifting. The automation will cue the correct lens memory and the lens tends to be more repeatable

[Harry Smolin]

To answer your question, Marin, usually it remembers the position you last saved the lens in and runs that. It sometimes slips out of position a little, such as slightly up or down, but nothing too extreme. I'll assume this hyper-zooming example was just a fluke.