Film-Tech Cinema Systems
Film-Tech Forum


Post New Topic  Post A Reply
my profile | my password | register | search | faq & rules | forum home
  next oldest topic   next newest topic
» Film-Tech Forum   » Operations   » Digital Cinema Forum   » S2 MCGD calibration through 3d system ... TI bug?

   
Author Topic: S2 MCGD calibration through 3d system ... TI bug?
Daniel Morez
Film Handler

Posts: 16
From: Hollywood, CA, USA
Registered: Jun 2010


 - posted 07-21-2011 01:08 PM      Profile for Daniel Morez   Email Daniel Morez   Send New Private Message       Edit/Delete Post 
Has anyone had any experience performing an MCGD calibration though a 3D system with a Series 2 machine? If so how did it go?

Reason I ask is that two different brand units I have tried this on so far both lose 3D sync output when the MCGD native uncorrected patterns are activated. The loss of sync out to the 3D system causes the calibration results to be inaccurate due to the 3d systems not actively shuttering. Its essentially like trying to calibrate through the 3D systems with them powered off. Then going to test the results of the calibration with the 3D system turned back on.

Both 3D systems used (Xpand and RealD) have LED indicators that show the presence of the 3D sync signal from the projector. Its clear the signal turns off and on with the activation and deactivation of the MCGD patterns. Results are the same with both manufacturer software and T.I.'s S2 DCC.

Manufactures have both shrugged their shoulders and said it is a TI issue. Just wondering if this is something others have run into.

The only unofficial work around Ive found is using the Xpand's internal test sync signal mode while reading native patterns. Its gets me a ballpark cal result but never as repeatable or precise as a Mono cal or 3D cal with a S1 projector.

 |  IP: Logged

Pete Naples
Phenomenal Film Handler

Posts: 1543
From: Dunfermline, Scotland
Registered: Feb 2001


 - posted 07-21-2011 04:12 PM      Profile for Pete Naples   Email Pete Naples   Send New Private Message       Edit/Delete Post 
First off, unless you know different, the only 3D system that requiers colour calibration is Dolby. I've been told by both Xpand and MasterImage to use the projectors native MCGD file.

However, if colour calibrate it in 3D you must, then do it the Dolby way, run the test patterns from the server.

 |  IP: Logged

Daniel Morez
Film Handler

Posts: 16
From: Hollywood, CA, USA
Registered: Jun 2010


 - posted 07-21-2011 05:57 PM      Profile for Daniel Morez   Email Daniel Morez   Send New Private Message       Edit/Delete Post 
I have very limited experience with Dolby and only anecdotal knowledge of its special calibration you are speaking of. Its something like because the 3D optics portion of the system is before the DMDs the calibration patterns are generated via the server and customized per each eye, which I guess corresponds with different sections of the polarization disk. That is different from what I am doing.

The method I am following is exactly the same as a mono calibration but with the 3D system optics setup in line with the meter being used, in mycase either Xpand Glasses or ZScreen/Passive glasses. there seems to be the lack of internal protocol that allows the 3D sync signal to remain on when the projectors uncorrected patterns are activated.

This is something that's tried and true with S1 DC projectors but has become an issue with the couple S2 projectors I have tested.

 |  IP: Logged

Mark Gulbrandsen
Resident Trollmaster

Posts: 16112
From: Bountiful, Utah
Registered: Jun 99


 - posted 07-21-2011 06:30 PM      Profile for Mark Gulbrandsen   Email Mark Gulbrandsen   Send New Private Message       Edit/Delete Post 
And... the Dolby calibration is done in the server... not in the projector.

Mark

 |  IP: Logged

Daniel Morez
Film Handler

Posts: 16
From: Hollywood, CA, USA
Registered: Jun 2010


 - posted 07-21-2011 06:47 PM      Profile for Daniel Morez   Email Daniel Morez   Send New Private Message       Edit/Delete Post 
Yes Mark, correct. Im am attempting the in projector MCGD cal with Xpand and RealD systems. The unwanted behavior is the same with both the manufacturer communicators and the T.I. DCC for S2 which is really making me lean towards this being a bug. From what Ive found from the projectors tech resources Im using current on all levels of firmware.

 |  IP: Logged

Steve Guttag
We forgot the crackers Gromit!!!

Posts: 12084
From: Annapolis, MD
Registered: Dec 1999


 - posted 07-22-2011 12:06 AM      Profile for Steve Guttag   Email Steve Guttag   Send New Private Message       Edit/Delete Post 
I too have found that the 3D sync is not there when running internal test patterns...it would be a TI issue...I never looked to far into it as I generally like to use test patterns off the server since that takes into account everything when doing a 3D.

I have indeed done separate 3D measurements for Real-D and MasterImage...the offsets were NOT the same as for 2D.

I don't get too crazy about it though...I question just how color neutral the cheap lenses are in the glasses.

If doing it from the player, use the TI ICP program and start with MCGD of "Nominal" and a Target of "DCI_Xenon". After running the three colors and white, save them to your new MCGD and then make that active. Use "Color Verification" as the target and and then remeasure to verify that the color correction worked properly.

-Steve

 |  IP: Logged

Daniel Morez
Film Handler

Posts: 16
From: Hollywood, CA, USA
Registered: Jun 2010


 - posted 07-22-2011 12:52 PM      Profile for Daniel Morez   Email Daniel Morez   Send New Private Message       Edit/Delete Post 
Thanks Steve. Im in a post environment where we are doing final color correction for features, mono and stereo. I must follow greater than average color calibration protocols for both. Appreciate the suggestion but I would worry a work around that bypassed the standard practice of using the chips native primaries could have adverse affects to our color work flow. I could be wrong and its a completely valid method, however Im going to follow up on some feedback I just got from TI.

They basically acknowledge the issue is in the current hardware/firmware release. They included a suggested work around from one of their software engineers. It was mentioned they just proved it in their R&D hardware yesterday [Roll Eyes] . Ill give it a go and see how it works in our NEC and demo Barco.

Breezing over it, it is definitely not as convenient as the S1 single button pattern activation.

Ill update with my findings.

 |  IP: Logged

Steve Guttag
We forgot the crackers Gromit!!!

Posts: 12084
From: Annapolis, MD
Registered: Dec 1999


 - posted 07-22-2011 04:27 PM      Profile for Steve Guttag   Email Steve Guttag   Send New Private Message       Edit/Delete Post 
Watch out on the NEC...when one sets the white level...I'm finding that the NEC will use UNCORRECTED white rather than apply the MCGD. I almost always do that from the TI ICP program to avoid unexpected projector quirks.

So, since you are a post house...for 3D, what type of screen do you have (material/gain)?

Steve

 |  IP: Logged

Daniel Morez
Film Handler

Posts: 16
From: Hollywood, CA, USA
Registered: Jun 2010


 - posted 07-22-2011 06:29 PM      Profile for Daniel Morez   Email Daniel Morez   Send New Private Message       Edit/Delete Post 
Honestly Ive never liked the NEC gui. Its been years since I last used it for anything other than the NEC only functions, tile edits, lamp control, etc. The issue you mentioned reaffirms my dislike. Ill certainly keep that in mind though just in case.

Stewart SnoMatte 100 (1.0gain) for mono and Xpand. For RealD Stewart silvers (latest processing spec, 3.0gain i believe). All seamless, no perf. Had used MDI silver but moved to Stewart because of preferred overall characteristics.

 |  IP: Logged

Steve Guttag
We forgot the crackers Gromit!!!

Posts: 12084
From: Annapolis, MD
Registered: Dec 1999


 - posted 07-23-2011 07:35 AM      Profile for Steve Guttag   Email Steve Guttag   Send New Private Message       Edit/Delete Post 
Oh, I definitely use Stewart as my "go-to" screen for the best looking image...and the Snomatte 100 has a great dispersion characteristic. I'm not as familiar with the Stewart Silvers (have not worked with them much nor measured their characteristics).

You have me thinking on this calibration thing...

Tell me, if, in 2D mode, you measure your DMDs native colors...and then use the procedure I mentioned above to measure those same colors but via a player...if they measure identical, would that not completely validate them as suitable to measure for 3D (all 3D types)? Next time I get a minute, I plan to do just that (I've always taken them at face value and they "look" identical to the projector generated colors...but if my eyes were THAT good, I'd be doing color timing and such! I will also investigate further about getting S2 projectors to put out the sync pulse while in test mode (by all means, tell us what you have found from TI too).

Okay...now back to 3D evaluation. For Xpand you use the Stewart Snomatte100...with that screen you are going to get light readings in the corners that are going to be in excess of 95% of your center reading (depending on how well your optical path is set up...it won't be the screen that is darkening the image). If you set up your center brightness to be say 5.0fL and evaluate the image with the corner brightness around 4.75 with sides/top/bottom closer to 5.0 too, you then go to test your Real-D on the SAME image (same player, same movie, same files...etc) and use the silver gain screen...if the center brightness is again set to 5.0fL...the corner brightness is bound to be about 1.6fL (or in that ballpark).

How can these two images be considered equivalent? How can they both be right? I claim that the method of setting image brightness is very wrong. It is a reason that silver-screen 3Ds tend to look dark. Because the average brightness of them is LOWER than the lower gain screens.

If the movie is "timed" for 5.0fL then the black levels for that movie shown with XpanD or Dolby should be fine (detail is not lost, blacks don't become grays...etc). However, in a MasterImage or Real-D...the image should only look right in the MIDDLE (if you are sitting in the middle too)...however, everywhere else, the black details will be lost due to insufficient brightness.

I am currently doing experiments/measurements that show that if the center brightess on a silver screen system is increased to about 7.7fL, the image will have an equivalent image (in terms of OVERALL BRIGHTNESS) to one of the white screen systems on a low-gain screen. That is, if you raise the center brightness to 7.7fL, the overall brightness (average) will be on the order of 4.8fL which is about the same average brightness you will find that XpanD/Dolby will yield on a matte-white screen running with a 5.0fL center brightness. It also means the the average brightness will still be in the acceptable range that the movie was timed for. Yes, the center will be bright and one will have gray instead of black but rarely is the important action of the screen in dead-center...peoples faces are towards the top (about 2/3 to 3/4)...etc and yes it is a compromise (which a silver screen starts with) but it is a better compromise since most of the screen will fall within acceptable light in terms of how the movie was timed. It also allows for a better experience for the bulk of the people that are not sitting dead center in the theatre. The corners will still be dark and below 3.5fL but most everywhere else will be much better.

Now as we ebb the center brightness levels up a bit to say 5.5fL...for Silvers, that would mean going up to about 8.6fL to keep up.

Note, if you curve your screen (not the silly 1:20 curve, but using ray-tracing), you can lower the requirement for silver screens quite a bit. A curve with a radius of about 5/6 of the throw, can yield a result such that a center brightness of 6.2fL has the same average brightness of a matte-whites screen of 5.0fL.

-Steve

 |  IP: Logged

Daniel Morez
Film Handler

Posts: 16
From: Hollywood, CA, USA
Registered: Jun 2010


 - posted 07-24-2011 01:41 PM      Profile for Daniel Morez   Email Daniel Morez   Send New Private Message       Edit/Delete Post 
You're right in that could test would be telling. From what Ive been taught the difference in the native and externally generated primaries is that an ext generated pattern is setting the DMD to a maximum video code level, while the native pattern is the mirrors being driven to their maximum "on" and "off" values. In theory this should make them different on some level (even if just minute) since the projector should have head room beyond defined video color spaces. But, yeah, ive never done a practical test to see. Im curious now and will give this a shot the next time I have some time in a room.

Ill have to follow up on the rest another time.

 |  IP: Logged

Daniel Morez
Film Handler

Posts: 16
From: Hollywood, CA, USA
Registered: Jun 2010


 - posted 08-08-2011 04:14 PM      Profile for Daniel Morez   Email Daniel Morez   Send New Private Message       Edit/Delete Post 
TI S2 3D Calibration - Loss of 3D Sync Output Manual Fix:

1)Activate MCGD pattern.(this action disables 3D sync output, if not you don't have this issue)
2)Go to Test pattern tab of the TI S2 control program and change pattern freq to 48hz.
3)Go to 3D Control menu and select desired stereo flash multiplication. (6:2 for triple flash)
4)On the same menu set Input Frame Reference to "none provided"
5) Optionally, one can also set dark time settings to match the 3D system specific settings.

The pattern should be on screen now with an active 3D sync outputting to the 3D system. Taking a reading though the 3D system optics.

When switching to the next pattern repeat the steps listed above. Apply all readings to file and proceed with routine 3D title/preset creation.

In Series1 systems this was done automatically in the software commands with each press of the MCGD pattern buttons. Seems it was missed in current S2 TI, Barco and NEC S2 programs. Barco has acknowledged they are working this into their next S2 communicator release. I am unsure if this is an issue with any other brands.

TI is sending a notice to all vendors that this should be incorporated into there control software (if not already). They did not indicate if this will ever be built into their control program though I'm guessing it will be. So hopefully this wont be necessary for much longer.

This applies to MCGD calibration procedure for RealD and Xpand stereo systems attached to S2 Barco and NEC DLP projectors (when using internal native patterns). Software and firmware releases this was tested with are current as of the date is this post.

 |  IP: Logged



All times are Central (GMT -6:00)  
Post New Topic  Post A Reply Close Topic    Move Topic    Delete Topic    next oldest topic   next newest topic
 - Printer-friendly view of this topic
Hop To:



Powered by Infopop Corporation
UBB.classicTM 6.3.1.2

The Film-Tech Forums are designed for various members related to the cinema industry to express their opinions, viewpoints and testimonials on various products, services and events based upon speculation, personal knowledge and factual information through use, therefore all views represented here allow no liability upon the publishers of this web site and the owners of said views assume no liability for any ill will resulting from these postings. The posts made here are for educational as well as entertainment purposes and as such anyone viewing this portion of the website must accept these views as statements of the author of that opinion and agrees to release the authors from any and all liability.

© 1999-2018 Film-Tech Cinema Systems, LLC. All rights reserved.