This is topic Fix ghosting on MasterImage or RealD in forum Digital Cinema Forum at Film-Tech Forum ARCHIVE.
To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=16;t=002630
Posted by Justin Hamaker (Member # 2165) on 03-24-2016, 03:11 PM:
It appears one of my theatres is having a ghosting image with Batman vs Superman on at least one screen with RealD and a NEC2000C projector. This site has 3 RealD screens, our other location has 2 MasterImage screens.
Is there anything that can be done to adjust for ghosting with RealD or MasterImage systems? This isn't my home theatre, so I can' say for sure, but to the best of my knowledge there have not been any ghosting issues with other 3D movies. This includes Zootopia which has had 3D showings in the same auditorium(s) using the same formats.
Posted by Monte L Fullmer (Member # 2797) on 03-24-2016, 04:29 PM:
Using "Z" Screens, or XL?
If "Z" screen, make sure that the bubble level gauge mounted on top of the "Z" frame shows being level, for if that gets knocked off alignment, that can cause issues.
"XL" requires a tech to align.
Good luck
Posted by Marco Giustini (Member # 4544) on 03-24-2016, 05:22 PM:
I've heard of the same issue from other cinemas with Batman Vs Superman - any chance of an issue with the 3D mastering?
Posted by Dave Macaulay (Member # 813) on 03-24-2016, 05:27 PM:
There are a few things that one can call ghosting.
- 3D sync issue: this means adjusting settings in the 3D control data. Usually the default data provided by Real-D or MasterImage is ideal though. If the 3D was good before and nobody with itchy fingers has been "optimizing" the 3D settings this is not a likely problem.
- Bad glasses? They both work with circular polarization and all "OEM" glasses should work equally well. I think one even works properly with the other (ie L/R polarization directions are the same). I have seen 3D glasses that were bought online from somewhere other than Master Image to save a few bucks... they were not at all good.
- Trouble with polarizer: burned or otherwise damaged Z-Screen or MI wheel/panel. I don't think it's possible to burn an XL.
- Reflections: if your port window is vertical, a Z-Screen or XL, Wave3D, even an MI wheel, can cause ghost images from light reflected off the port glass, back to the 3D unit, and back out to the screen. Most visible in credits where a slightly larger fairly dim ghost image follows the crawl. The port glass should be tilted to show the ghost images off screen.
- Connection: The Z-Screen "DIN plug" is notorious for bad connections, it can look perfect but be just a bit out: this can cause weird on/off ghosting (severe though!). Reseat both ends and make sure they are seated all the way in.
"Ghostbusting" - RealD has ghostbusting done in the server, you need a license to activate it though - free to RealD customers. Call RealD support if you don't have this active, it does make a difference but I don't know what it does or how. I know Doremi will do it, don't know about other servers but they should.
- XL misalignment looks sorta ghosty but it doesn't change: you get double images that look the same in both eyes. Easy to check by showing a 2D image or a test pattern through the XL: the best possible setup will lose focus compared to no XL in the light path, bad will have double images at one or both sides - or all over.
Posted by Monte L Fullmer (Member # 2797) on 03-24-2016, 06:10 PM:
I'll go with that DIN S-video cable issue for the Z screen.
Had to replace one of mine, for the display box was displaying available 3D content and the screen was new.
We have a motorized screen mover and was odd that a cable would go out like that.
Posted by Justin Hamaker (Member # 2165) on 03-24-2016, 08:11 PM:
I just watched a stretch of the movie for myself (MasterImage system), and what I noticed was the foreground images were almost always steady and clean, but there were a lot of ghosting and other issues with background images. I found it to be somewhat distracting because it would occasionally create odd patterns.
As an example: there is one scene where the focus is on a foreground character who is standing in an office. Over his shoulder you see a computer monitor, except you see 3 monitors: 1 is bright like a monitor should be, and then off to either side you see a a portion of a monitor with a 'half bright' screen. Another example is you see the side of a bridge where there is a brick pattern. If you look at the pattern you see odd breaks and overlaps in the pattern. But in both instances the foreground characters are nice and clear with no hint of ghosting.
Posted by Dave Macaulay (Member # 813) on 03-24-2016, 09:25 PM:
That sounds like the timing is off or the polarizers are tired. The foreground objects probably look OK because they have no 3D offset - you can see this if you take the glasses off. Objects at the screen plane will look fairly normal, in front or behind will have two images beside each other.
Does the wheel seem to when viewed from the lens sidestand still with 3D image showing? If it seems to twitch around a lot the motor drive could be failing, it should lock quite firmly to the sync signal. If you can manually get it up and running with 2D playing, starting a 3D clip in 3D - 3D selected in the server and a 3D preset in the projector - you will hear the wheel speed up to sync speed and see the wheel segments lock into sync.
If the sync looks stable, MI wheel adapters come with a spare wheel: I would install that and see if it's better.
But... some crosstalk/ghosting is unavoidable. It's worst with high contrast bright objects behind or in front of the screen: it looks like what you describe, but if the system is working right with a good wheel the ghosting shouldn't be bright and annoying.
Posted by Justin Hamaker (Member # 2165) on 03-24-2016, 10:15 PM:
Thanks for the info Dave, I'll take a closer look. However, we are getting the same thing on systems with RealD and MasterImage.
Posted by Steve Guttag (Member # 268) on 03-25-2016, 01:07 AM:
I'd start with a 3D test image from your server using the same frame rate and using the same preset in your projector. A good test pattern will give you a feel for any ghosting. Note, the pattern need not be MasterImage specific. I tend to use Dolby's test pattern and my own since I know them best. Check both eyes to see that ghosting is equal and minimal.
With polarized 3D...anything in the light path can degrade the polarization, including the port glass and the screen as well as dust. But start with just a test pattern and start your experiments from there. Compare this theatre to the others without the problem and note any variances in the 3D settings for the "Title".
Posted by Marcel Birgelen (Member # 6801) on 03-25-2016, 01:19 AM:
I guess it's the overall dark tone of the movie, combined with the occasional very bright spots that's causing the ghosting issues on polarized systems.
What servers are you using at both locations? Maybe the ghostbusting algorithm in use just fails to do its job properly for this kind of content.
Posted by Marco Giustini (Member # 4544) on 03-25-2016, 05:10 AM:
Marcel
That's what I thought, lots of high contrast content which the ghost busting simply cannot address.
And being a big title, staff bothered to have a look in the auditorium on this occasion
Posted by Mark Gulbrandsen (Member # 72) on 03-25-2016, 11:16 AM:
I watched about 15 min of it yesterday and didn't see any ghosting at all. NC-2000 with Real-D XL, system set at about 6 foot lamberts. GDC server. Yaaaaawwwwnnnnnnnn.... What a POS movie!
Mark
Posted by Justin Hamaker (Member # 2165) on 03-25-2016, 03:24 PM:
The server is GDC SX-2001 Non-IMB.
Posted by Monte L Fullmer (Member # 2797) on 03-25-2016, 06:04 PM:
Ya, kinda gave Chris Nolan, exec prod., a black eye with this one.
Posted by Jorgen Larsen (Member # 5302) on 03-26-2016, 04:09 AM:
Hi
We have the same problems with Batman in our cinemas here in Denmark.
It seems to be a general problem. The ghosting is on RealD Volfoni and Timewaing systems Dolby DSS200 Kinoton/barco projectors.
Justin... is it possible you can find the timekode where the scenes you are talking about are, so I can use them as a reference?
Thanks
JL
Posted by Marco Giustini (Member # 4544) on 03-26-2016, 04:46 AM:
Can you disable one eye and check on the affected scenes?
Posted by Justin Hamaker (Member # 2165) on 03-26-2016, 01:31 PM:
Jorge, I have noticed it a a number of different points in the show. I did another cleaning of the port window and filter disk just to be safe, and saw now improvement.
quote: Dave Macaulay
Does the wheel seem to when viewed from the lens sidestand still with 3D image showing? If it seems to twitch around a lot the motor drive could be failing, it should lock quite firmly to the sync signal.
The image looks nice and stable on the filter disk when viewed from the projector side.
Given that I'm hearing similar complaints from other theatres, I'm resigned to the fact it's in the movie, not our equipment.
Posted by Marcel Birgelen (Member # 6801) on 03-26-2016, 03:35 PM:
quote: Justin Hamaker
The server is GDC SX-2001 Non-IMB.
I'm not sure if those servers enable ghostbusting in combination with MasterImage, but your RealD screens should do ghostbusting in any case.
I've seen about 20 minutes using a RealD XL setup, a DP4K-23B and a Dolby DSS220 and there was quite a lot of ghosting. This ghosting wasn't due to bad calibration, but due to limitations in the polarizer technology. The contrast was just to great to entirely avoid it. It was somewhat reduced when I tried my Polaroid glasses instead of the stock RealD glasses.
I guess they would've done people a favor and release it as a 2D-only feature, this content just isn't a fit for most 3D screens out there...
Posted by Carsten Kurz (Member # 5396) on 03-28-2016, 10:58 AM:
Apparently same in Germany, and here also most prominent with Real Systems. Some people mention some double contours even when playing the 3D DCP in 2D. Looks like a serious flaw with the mastering or 3D conversion?!
Warner germany has admitted the issue, but only solution so far is to play the movie in 2D only on affected systems.
Many movies show some ghosting on RealD systems, especially with a bad combination of screen and portglas. But this one really triggers a lot of loud complaints from the audience it seems.
First sightings of the IMAX 3D version seem to suggest excellent 3D...weird...
Maybe a strange interference with RealDs server-side ghostbusting?
- Carsten
Posted by Marcel Birgelen (Member # 6801) on 03-28-2016, 05:07 PM:
I've seen it in full tonight in Dolby 3D and the 3D is definitely defective.
I haven't seen any form of "ghosting", but in several scenes, your brain doesn't seem to be able to correctly fuse the images in the background together. It's like the parallax of some of the objects is wrong or totally out of whack between the left and right image. Maybe this, combined with rampant ghosting issues in polarized systems explain the many complains.
Some people gave this movie positive reviews regarding the 3D qualities, I must have missed something.
Posted by Marco Giustini (Member # 4544) on 03-28-2016, 05:29 PM:
sounds like the 3D "clashed", where objects which belong to the foreground are put in the background by the parallax. Your brain cannot understand - because that cannot exist - and the picture kind of breaks down.
Like showing 2D subs on 3D picture, text cannot be behind objects so your brain basically farts.
It makes sense to me that the average customer/projectionist may identify that as 'ghosting'.
Posted by Carsten Kurz (Member # 5396) on 03-29-2016, 04:49 AM:
A colleague quoted a letter from RealD acknowledging the issue, asking for more reports, and saying that they are working with Warner to solve the issue.
- Carsten
Posted by Carsten Kurz (Member # 5396) on 03-31-2016, 06:38 AM:
Someone suggested it could actually be L/R frames out of sync.
At least that's an explanation how it could have happened somewhere in the mastering process. It doesn't explain why it seems to be more visible on Real D screens. But maybe that is because with out of sync 3D material, the Real D Ghostbusting will cause very weird effects.
Enough people have commented on bad 3D with Dolby 3D and IMAX 3D as well, so this clearly is not a Real D only issue (I would have a hard time to imagine any reason for any Real D only issues - other than their lower stereo separation and ghostbusting).
- Carsten
Posted by Marco Giustini (Member # 4544) on 04-01-2016, 02:31 PM:
What puzzles me is that WB is not saying anything. What are they waiting for?
Posted by Carsten Kurz (Member # 5396) on 04-01-2016, 03:48 PM:
Maybe it's sheer embarrassment...
- Carsten
Posted by Marcel Birgelen (Member # 6801) on 04-04-2016, 04:38 PM:
quote: Carsten Kurz
Someone suggested it could actually be L/R frames out of sync.
Out of sync as in reversed? Which could explain the weird parallax issues I've seen now and then.
quote: Marco Giustini
What puzzles me is that WB is not saying anything. What are they waiting for?
As long as it doesn't hit a major news outlet, they will probably say nothing at all.
Powered by Infopop Corporation
UBB.classicTM
6.3.1.2