Correct use and interpretation of 70mm Testfilm SMPTE RP-91

[Jerry G. Axelsson]

Hello fellow technicians,

We did an alignment and adjustment of a pair of 70mm projectors recently.

We used 70mm testfilm SMPTE RP-91 for this purpose.

After the fact I got some thoughts regarding how to interpret the testfilm and the patterns correctly?
What is acceptable for min/max image height and width and stay within spec. for 70mm 5-perf 2.20:1?

Any clear advice would be appreciated.

Thank you. SMPTE RP91.jpg

[Phillip Grace]

This looks like it might be the projected image from the right hand projector. SMPTE Standards (I think) give a figure for maximum image crop at 5%. Keystone is something similar. If you run your set-up through the Schneider TD-Pro software you will get a very accurate result. From memory, the text turns Red when a parameter is out of tolerance.

[Steve Guttag]

I don't have the relevant SMPTE RP (projectable image area) in front of me but, in general, the is no tolerance for over-cutting the aperture plate. The maximum area would be the outline in the 70PA target.

I can say, while on the P3 committee, I never got a satisfactory answer to the question:

"If one sets their masking exactly on the dimension lines defined by a format (in this case 0%), and then cuts the aperture plate such that no shadow from the aperture intrudes into the picture, is that considered correct? Generally, the answer was yes.

I then asked, if the masking was moved out to reveal the extent the aperture plate is showing on the masking is beyond the maximum projectable area, is that not over cut since we're depending on the density of the masking to hide ALL things that can happen outside of the maximum area (splice flash, wavering optical soundtrack doping, clear frame lines, clear guard bands, particularly on the inboard side, DTS time code, in later years).

To that, I got no satisfactory answers.

Since nothing is defined beyond the maximum projectable area, and the masking system employed in the theatre is NOT part of the standard (if it was, one would need to specify its density and ability to absorb light and quantify that), I conclude that a properly cut aperture plate never shows more than the maximum projectable area. This also, necessarily, means, in a masked theatre, you will have some amount of cropping.

It has generally been considered good practice to limit cropping to 5% or less in all dimensions with 10% being an absolute maximum. Traditionally 10% cropping has been considered the "safe action area" with 20% considered the safe title area. in the digital world, they do not respect the safe title area...thinking every pixel is always visible.

[Ryan Gallagher]

There are some pretty good guides on how to use PA35/RP40, the same concepts apply to all the scientific markings (such as measuring image stability etc).

For size, care to share what is making all the edges in your photo? Pure aperture plate, or combo of plate/screen extents/masking?

[Jerry G. Axelsson]

Thank you everybody for your insights and help. The image you see is a screenshot of the image with adjusted screenmasking and a re-used aperture from another cinema. It is apparent that the aperture is over-filed in height and that the maskning should be slightly tightened in terms of height.

[Jim Cassedy]

I know this is going to cause some of you to literally gag when you read it, but
when I first started working at my current venue about 6 years ago, one of the
former projectionists told me that they often ran 70mm without ANY aperture
plates. < pausing here for gag reflex>

While I was somewhat skeptical at first, I tried it with a some 70mm trailers
and shorts that I bought in from home and it did work. At that time the only
70mm aperture plates I had at the theater were, in my opinion, really badly
cut. I even questioned if the person who made them was sober at the time.

Since then, I've either bought or acquired through other means, over a dozen
70mm aperture plates, of various aspect ratios, and a few that were brand new.
With all the oddball aspect ratios coming from distributors lately, I can usually
find a pair of plates that works for what I need. However, I still occasionally run
70mm without any plates. I did this with our recent "Project Hail Mary" shows.

I can get away with this on most newer titles that may have digitally mastered
or blown up from 35mm, but on most older titles, or on a new title that's been
photographed and edited totally photo-chemically, the lab splices in the
negative become visible on screen if run plate-less. "Licorice Pizza" was
one newer titles I had to use plates for when we ran it in Y2k21. But I was
able to run a restored 70mm print of VERTIGO last year without plates and
without the negative splices being visible.

So, every time I we get a 70mm print, I look closely during inspection to
see if there are negative splices visible in the framelines between scenes,
and I'll also usually run at least one reel on each projector "in my spare time"
to double check how it looks on screen.

[Ryan Gallagher]

Thank you everybody for your insights and help. The image you see is a screenshot of the image with adjusted screenmasking and a re-used aperture from another cinema. It is apparent that the aperture is over-filed in height and that the maskning should be slightly tightened in terms of height.

Depending on the model projector.... FilmTech can arrange for better plates to start from if yours are over-filed as you say.

Masking can hide a lot of blemishes, but in a perfect world the plate is there to do that too. I've only had a chance to remake a few of the 70mm plates here, and produce a few new ones we didn't have. But our 35mm stock are almost all overfiled and in need of a re-do as well. Scope is especially bad. Ohh the things I'd do if projection was my ONLY responsibility here.