Differences in 3D between 2K NEC and 4K NEC Laser Projectors

[Chris Haller]

Today I spent my morning setting up a pair of RealD XL 3D systems with our service provider for a pair of NEC Laser projectors ahead of Avatar 3's release. One is a 2K NP-NC2403ML and the other is a 4K NC2443ML. The 2K is in a 30% smaller house than the 4K, and I know in general that the contrast is a little more flat with 4K projection across the board, but we both noted that the 3D was a little deeper and more intense in the house with the 2K projector.

The screens have the same gain, but the 4K is on a much larger screen for sure. And we couldn't quite get the size adjustment to 100% align on the bottom image in the 4K house, but both of us felt it was within acceptable limits. We also couldn't quite match the brightness in both houses - the 4K output 13.2fl in scope at 100% power while the 2K output 14fl at 90%. Both output 14fl in Flat according to his colorimeter.

Could these differences really make that big of an impact on 3D depth? Either way I'm pretty pleased with the light output and general setup in both houses.

[Marco Giustini]

RealD works with silver screens which are high gain.

I suspect that a smaller screen might have a better light uniformity so the perceived brightness is higher on the smaller screen than the bigger screen. Check the brightness at the sides and corners of both screens to get an idea.

Alternatively, measure the brightness over 9 or 15 points on screen and compare.

I'm a bit behind: did you get 14fL in 3D, through the glasses?

[Chris Haller]

I'm a bit behind: did you get 14fL in 3D, through the glasses?

Our service tech used light measurements through his colorimeter, and made sure to not measure the hotspot, but multiple points across the screen.

[Steve Guttag]

The tech cannot give you a single number if s/he measured in multiple spots around the screen. This is a "typical" distribution of a flat silver screen:

image.png

What number would you assign to it? 13.8? 13.4? Add up all 45 boxes and divide by 45 to arrive at a number? If so, and you have 13.2fL, what would the middle be? In the above grid, the average is 8.3fL. Even if I normalized it to 14fL in the middle, the average only jumps to 8.7fL. So, I'd say, anything other than than placing your meter in the center of the seating area, pointing it at exactly the middle of the screen and providing a single number is nonsensical. I'm not saying that the single number reading is representative of the screen...as the above real-world example shows, distribution of the light is part of what gives an image its appearance. The is why the standards have a minimum (for cinemas) of 75% of center, on the sides...something that is and any screen over 1.3, that is flat, cannot meet (all of the sliver screens I've encountered have gains above 2 (sometimes well above). With the gain so goes the intensity in the center (hot spotting).

As to the differences in appearance...there are multiple factors. The 2K projector is using a conventional .98" chip, like was used on the NC2000C. The 4K projector is using the TRP chip (Tilt/Roll/Pivot) that, unfortunately, has show that it tends to get stuck pixels. Barco has already switched to the new SST chip (typical DMD construction in a .98" package) and Christie never adopted the TRP. Hopefully, Sharp has a plan there, In any event, you have a different optical element there that could be giving a different look, particularly how the projector/lens interact. Which brings me to that you likely have different lenses in the two machines owing to the different screen sizes. If so, you need to account for that in your observations too. Without a doubt, I've found 1.2-1.8 lenses to have a combination of poorer light throughput and lower contrast than 1.4-2.05 lenses (both Minolta). The 2.4-3.9 has a good look but also seems to eat more light. Generally, the closer you are to the middle the lens line up (not on the very long or very short) tends to get you more efficient lenses. Additionally, where you are in the zoom range of a lens will affect your light output.

There is also your meter. They are not all created equal. While on an SMPTE committee that was working on how to measure light in a cinema, it was notable the wide variances of how meters read light, particularly how they rejected light outside of the intended measurement area. Some would factor in the surrounding area into their reading much more than others. A way to combat that is to have test patterns that only have circles of light so the meter is just reading the intended spot.

image.png
So, depending on the meter and the method used to make the measurement, plus the particulars of each auditorium, you could get different results.

Really, all you have in common between the two rooms are the light source. The DMD, lens, screen and the physical layout of the theatres are all different. It is hard to draw conclusions from that.

[Chris Haller]

This is a reasonable approach that I suppose I was not considering in factoring in the means of measurement - I was not with him when he took his measurements, so I cannot verify what measurements he took, or what tool he used. I had to dip out after we calibrated the RealD equipment to the built in alignment slide on the NECs, and the rebuilding of the screen and lens files for 3D.

Thank you for your insight, and your serious response. Quite humbling to come up against such thorough knowledge. I've been managing our projectors for over a year, and I still know so little.

[Jim Cassedy]

Only the largest auditorium here had 3D, and we junked all the 3D equipment
AND the silver screen back in January of this year. Mostly, I"m glad we got rid
of that stupid screen. I hated it. The big auditorium it was in was quite wide, &
trying to get, decent, even light distribution on screen for normal shows was
almost impossible, especially when running film. It seemed foolish, in my
opinion, to have 90% of our shows not look as good as they should, just for
the sake of the occasional 3D release. And trust me, if the corporate bean
counters here thought 3D was actually making us any money, they would
have kept it, and put 3D systems in some of the other auditoriums too.
I'm glad it's gone!

[Chris Haller]

Only the largest auditorium here had 3D, and we junked all the 3D equipment
AND the silver screen back in January of this year. Mostly, I"m glad we got rid
of that stupid screen. I hated it. The big auditorium it was in was quite wide, &
trying to get, decent, even light distribution on screen for normal shows was
almost impossible, especially when running film. It seemed foolish, in my
opinion, to have 90% of our shows not look as good as they should, just for
the sake of the occasional 3D release. And trust me, if the corporate bean
counters here thought 3D was actually making us any money, they would
have kept it, and put 3D systems in some of the other auditoriums too.
I'm glad it's gone!

Our two largest auditoriums at my theater are a Screen X installation, which is frankly a disaster for image quality for the most part, and the above mentioned 4K laser projection auditorium with a silver screen, which is fine for the most part. We'll never run film here, so I'm never going to have to be concerned with anything but the 4K laser. It was just really interesting to me how similar they were in terms of perceivable light through glasses to me yet how different the depth and intensity of the 3D was.

[Leo Enticknap]

Our service tech used light measurements through his colorimeter, and made sure to not measure the hotspot, but multiple points across the screen.

Holding (or taping) the 3-D glasses in front of the colorimeter's lens?

I'm currently tech checking for an Avatar screening at a residence theater (killing time while the 700 GB DCP ingests, hence reading F-T), and am only getting 10.2 through the glasses (XpanD) with the lasers at 95%, using a Barco SP4K-27BHC on a Stewart Snowmatte 100 (microperf) screen that is 21.9ft x 11.8 in flat.

[Marco Giustini]

What Steve said

But the bottom line is that regardless of the number, you might just have more overall/subjective light being reflected by one of the two screens. Maybe the smaller screen has a "larger hotspot" than the bigger one?

What do you see if you drop the brightness of your 2K screen by, say, 20%?

[Steve Guttag]

Since you are saying that the 3D depth seems so different...I'm curious, have you tried loading a 3D eye chart up (one that lets you verify that there is no/low crosstalk)? Make sure just Left eye is getting to left eye and likewise on right eye. Make sure that your Gamma is at 2.6 on the inferior one (it should be 2.6 on the other one too). Since it is Real-D, make sure that ghost busting is set on your server.

Not all screens are the same and if the screen in the bigger theatre is not retaining the polarization, then you'll get soft 3D with some double-imaging as the two eyes are being seen by both eyes.

Note, Real-D, with MDI (and possibly others) have a white screen that does retain polarization (MDI: HighWhite). I believe they have it in 1.4 and 2.0 flavors. Real-D calls their technology "Precision White."

[Jim Cassedy]

Our two largest auditoriums at my theater are a Screen X installation, which is frankly a disaster for image
quality for the most part, and the above mentioned 4K laser projection auditorium with a silver screen,
which is fine for the most part. We'll never run film here, so I'm never going to have to be concerned
with anything but the 4K laser. It was just really interesting to me how similar they were in terms of
perceivable light through glasses to me yet how different the depth and intensity of the 3D was.

Our 3D system used dual Sony SRX-515 projectors, so brightness wasn't really an issue. If a venue has
the resources, dual 3d projection is the way to go. Yes, you've got check the alignment occasionally, but
IMO the image is much better, and brighter, since you're not 'multiplexing' the left & right frames. There
are only two or three theaters in the SF area that could do dual digital 3D. (The Grand Lake, in Oakland
is one of them) One of the best 3D presentations I've ever seen was a dual projector 35mm Halloween
showing of "HOUSE OF WAX", about a decade ago at The Castro Theater here in SF.

(Just last week, during a major theater cleanup, we dumped several hundred pairs of 3D glasses that
were still sitting in storage back stage, since 3D is officially 'dead' here)

[Marco Giustini]

alignment in Dual Projection 3D - that is, one projector is Left eye and the other Right eye - is really not critical. I did some tests myself, moving the lens of ONE projector while watching 3D content and I started noticing something when I was more than 1 metre off

I dislike 3D in general but if I have to watch it, I cannot stand the "flashing". Dual projector 3D at 14fL is the minimum I can accept as a watchable image!

that "triple-flashed" 3D on a silver screen with 3fL on the hotspot was steaming garbage!

[Steve Guttag]

The problem with non-aligned dual 3D projectors are the torture you are putting a person's eyes through as they are being asked to move independently. Additionally, when 2D content is up on the screen, it should look "normal."

[Marco Giustini]

Why would the eyes be asked to move independently? An offset between L and R is the base of 3D to create depth. Introducing a "fixed" offset by misaligning the projectors will introduce a non noticeable "permanent 3D effect". Of course I'm not suggesting that you should leave the images 1m away but a few pixels won't matter.

won't even matter in 2D-through-3D as each eye will see a perfectly sharp image, you won't be able to distinguish a minor offset.

Now, that is mainly horizontal offset. Vertical can be a bit more tricky but our brain knows how to deal with that, our heads can rotate and our eyes don't get perfectly aligned images in our day to day life!

[Gordon McLeod]

Why would the eyes be asked to move independently? An offset between L and R is the base of 3D to create depth. Introducing a "fixed" offset by misaligning the projectors will introduce a non noticeable "permanent 3D effect". Of course I'm not suggesting that you should leave the images 1m away but a few pixels won't matter.

won't even matter in 2D-through-3D as each eye will see a perfectly sharp image, you won't be able to distinguish a minor offset.

Now, that is mainly horizontal offset. Vertical can be a bit more tricky but our brain knows how to deal with that, our heads can rotate and our eyes don't get perfectly aligned images in our day to day life!

Years ago there was a gimmick lens that delivered (haha) fake 3d from normal prints. It was a wedge divice that misaligned the same image slightly