Chasing Focus...

[Vern Dias]

Did anyone have input on why 35PA will seem completely in focus, but then you run a print and it looks like you have a slight adjustment to make still?
I suppose we should make some text heavy scope loops and use those to get a more reliable focus ahead of screenings.

Where there any prudent tips for what format order to go in if we are trying to get our lens collars set better for focus across all formats?

Thanks all on the procedural tips for astigmatism adjustment. 35PA Loops hang next to both projectors, and are used before every film if there are format changes (except double features). The procedure in the link sounds the most intuitive to me, and we have MPAA loops and other loops with text available to check our work, as well as a house print if we want to be doubly sure before running anything else. Plenty of trailers to chop up if we need more scope specific loops. I also have a laster disto and can just shoot the screen and set it from the measurement if I wanted to try the lazy route first.

Our house print is not scope but I suppose if we ran the last reel of it with full crawl it would afford tons of time to feel confident in our astigmatism adjustment without putting more mileage on a distribution reel, it would just be wider than usual.

If your curious our anamorphics according to my inventory notes are:
IscoOptic Ultra MC 5.91in/155mm with Ultra-star Cinemascope Attachment * HD
IscoOptic Ultra MC 5.91in/155mm with Ultra Anamorphic MC 2x

While not identical they are close enough that I could probably just set the ring based on the No. 2 projector's ring position to get back to where we were. But checking if where we were is the best we can do seems prudent.

I like the idea of kapton tape, I have some thin stuff at home but need to pick up some wider stuff. Locking the mating seam between the attachment and the anamorphic, as well as a backstop against pulling out the locking ring seems useful once set.

How we went south this time was, although we moved the ring a little (falsely thinking it was an alternative way to focus) and had good horizontal and vertical focus on 35PA, during the print test reel it seemed we needed more focus adjustment, and I went down to the end of the balcony and tried to call it with my counterpart on radio, but our mistake was we continued to move the astigmatism ring here rather than switching back to the knob. So we landed with the ring not where it was during 35PA, and focus not on screen yet. Once rolling the print we were unaware we had put the ring into a position where focus wasn't going to be possible, and the interplay of the two mechanisms was very difficult to get right on less than ideal focus footage of a very very dark film (There Will Be Blood), especially without the knowledge of exactly what the lens ring was actually for (vertical astigmatism/focus, rather than overall focus).

Venue is film-dark for two days right now, my first chance to correct will be Wednesday ahead of two flat prints, but not absolutely needed until the 19th for next scope print. I'm not worried now that I know what everything does and how to fix it.

It's worth noting that if you have a steep projection angle (based on your balcony comment) you are likely going to have a difficult time achieving perfectly uniform top to bottom focus and also have issues with keystone. In older theatres with steep projection angles, the screen was often tilted backwards slightly to minimize these issues.

[Ryan Gallagher]

It's worth noting that if you have a steep projection angle (based on your balcony comment) you are likely going to have a difficult time achieving perfectly uniform top to bottom focus and also have issues with keystone. In older theatres with steep projection angles, the screen was often tilted backwards slightly to minimize these issues.

Yeah for sure, we know perfect focus across the whole field will not be possible. Our screen is not tilted because it is flown out for live events. Lots of compromises... but just trying to make sure the ones we can genuinely improve are addressed, and it sounds like our anamorphic alignment is one of them.

[Ryan Gallagher]

Wait until the end credit roll, after a movie, when everybody's already left the theater. Focus your anamorphic on that. First the horizontal, then the vertical.

It won't be perfect but it'll be close enough to work until you can get some time to put in a piece of test film and do it proper-like.

An amusing aside, because of the steepness of our balcony, and average age of our film going regulars, our theatre does not usually completely empty until credits conclude and full house lights come up, always a few groups waiting to have better light to navigate the balcony stairs. Historic venue quirks.

[Brad Miller]

Do not use the distance markings on the lens. They are NEVER optimal if you want to resolve as much of the image resolution as possible. Throw the astigmatism WAY out of focus, focus the horizontal lines with the projector's focus knob, then adjust the astigmatism adjustment to the point of sharpest focus. Done, period, end of story.

[Gordon McLeod]

Did anyone have input on why 35PA will seem completely in focus, but then you run a print and it looks like you have a slight adjustment to make still?
I suppose we should make some text heavy scope loops and use those to get a more reliable focus ahead of screenings.

Where there any prudent tips for what format order to go in if we are trying to get our lens collars set better for focus across all formats?

Thanks all on the procedural tips for astigmatism adjustment. 35PA Loops hang next to both projectors, and are used before every film if there are format changes (except double features). The procedure in the link sounds the most intuitive to me, and we have MPAA loops and other loops with text available to check our work, as well as a house print if we want to be doubly sure before running anything else. Plenty of trailers to chop up if we need more scope specific loops. I also have a laster disto and can just shoot the screen and set it from the measurement if I wanted to try the lazy route first.

Our house print is not scope but I suppose if we ran the last reel of it with full crawl it would afford tons of time to feel confident in our astigmatism adjustment without putting more mileage on a distribution reel, it would just be wider than usual.

If your curious our anamorphics according to my inventory notes are:
IscoOptic Ultra MC 5.91in/155mm with Ultra-star Cinemascope Attachment * HD
IscoOptic Ultra MC 5.91in/155mm with Ultra Anamorphic MC 2x

While not identical they are close enough that I could probably just set the ring based on the No. 2 projector's ring position to get back to where we were. But checking if where we were is the best we can do seems prudent.

I like the idea of kapton tape, I have some thin stuff at home but need to pick up some wider stuff. Locking the mating seam between the attachment and the anamorphic, as well as a backstop against pulling out the locking ring seems useful once set.

How we went south this time was, although we moved the ring a little (falsely thinking it was an alternative way to focus) and had good horizontal and vertical focus on 35PA, during the print test reel it seemed we needed more focus adjustment, and I went down to the end of the balcony and tried to call it with my counterpart on radio, but our mistake was we continued to move the astigmatism ring here rather than switching back to the knob. So we landed with the ring not where it was during 35PA, and focus not on screen yet. Once rolling the print we were unaware we had put the ring into a position where focus wasn't going to be possible, and the interplay of the two mechanisms was very difficult to get right on less than ideal focus footage of a very very dark film (There Will Be Blood), especially without the knowledge of exactly what the lens ring was actually for (vertical astigmatism/focus, rather than overall focus).

Venue is film-dark for two days right now, my first chance to correct will be Wednesday ahead of two flat prints, but not absolutely needed until the 19th for next scope print. I'm not worried now that I know what everything does and how to fix it.

35PA is a actual camera original not a print as such the image is on the side facing the lens and not on the side facing the lamp. as such it is in a different plane. There was a red image version that was a print for the purpose of adjusting focus

[Steve Guttag]

Ummm...some right and some wrong.

Both positive print film and 35PA are emulsion towards the lamp. With 16mm, particularly when reversal prints were made, you could get some prints with the emulsion one way and other prints the other...depending on the generation.

35PA, as Gordon points out, is camera original and make on negative stock with the short pitch (1866) BH perforations. Release prints are on standard pitch KS-1870. How the perforations are made are even different. negative stock, particularly the one that 35PA uses are perforated one frame at a time. Release print stock is perforated in a continuous fashion. Aside from printing registration from the duplication process, 35PA can have stability issues when used in projectors that are not the projector's fault. The projector does not pin-register. It uses a combination of edge guidance for lateral stability and perforations for vertical registration. The camera (and telecine) use pin registration. As a result, you can have 35PA with perforation skew (happens more than you would think) and this will show up in a projector as movement. The old trick of pulling the aperture plate to see if it is the film or the projector causing the motion does not work with negative film...if the perforations are skewed, they will be moving but they are moving relative to the edge of the film (what the projector is using). On a release print, due ot how the perforations are made, the pulling of the aperture plate to see if it is the print or the projector is more valid but not conclusive.

So, why the differences in focus (and framing)? Well, on framing, you are using a different perforation pitch on 35PA. The thickness of the film stock can be different...the density of the image on the film (35PA is near 50% opaque so your lens is going to respond to that as will the film to the lamp's heat). Your test film is likely running in a loop and heating with each cycle...causing more deflection in the gate.

Your gate (on most curve gate projectors, particularly Simplex, Century, Christie, Norelco DP70, all use the BASE side as their film plane reference. So, you have film thickness playing into it as well. Kinoton (FP20, FP30, FP75E, notably, uses the emulsion side as their reference though since both the skate and runners are "rigid," it depends on how they wear on how they reference but normally, they wear together).

The bottom line is, 35PA is a MEASUREMENT film...it is NEVER for setting focus/framing as it does not simulate actual prints, with respect to film projectors.

[Ryan Gallagher]

Thanks Steve. I'm aware this topic has been hashed before, even some prior responses from you, this in a 2002 thread.

SMPTE also has 35-IQ which is equal lengths of film (100 feet) in Black and White as well as color (only on the magenta layer). The magenta film is on the top layer so focus can be established but it won't soak up the radient energy as much as the B&W version. The color version, I believe, is currently on KS-1870 perforations and is a better representation of feature film stock though the B&W version is certainly easier to see and to set shutter timing to.

It seems worth restating cause the new person would assume all those measurements that 35PA is intended for translate reasonably well to a print too, which is just not true. Ultimately it is for testing the projector mechanism and optical alignments etc. If i'm learning correctly, it is not for pre-setting a state that is ready to run a print of a given format.

The idea that 35PA is not for focus kinda runs counter to the info sheets SMPTE puts out about the test film though. I suppose referencing itself, you can achieve all the measurements the test film is for, it just doesn't directly set you up for the same on an actual print. If you had 35IQ it would perhaps be more reliable in the color section, we only have 35PA loops in our booth, and accurately or not, I have used 35PA overlaid on a film frame still to get a reference size for setting masking and then filing custom aperture plates before the print is in hand, I think sizing is one area where it does translate reasonably well, assuming of course prints use the SMPTE framing sizes, which they often don't these days.

Due to vertical registration I can see how it may not be the ideal starting point for pre-determining framing though. We are always willing to bump tilt/framing with the print even after setup ahead with 35PA or other format matching loops available, and frequently have to. For us the biggest advantage of pre-screening 35PA is to have the reference lines for our tilt, pan, and masking changes between formats, but with the right section of footage those could be accomplished with the dedicated format loops too, and we would get a focus opportunity that is more reliable.

We have most of those loops, but taking the time to thread another loop is sometimes skipped, especially if we are running a couple reels in test anyway (when permitted by the archive/distributor).

[Steve Guttag]

As I said, it is a measurement too. So, you CAN measure the resolving power of the system (in lines/mm). What does not do is let you set focus with it, then have that translate to another piece of film. However, if you can resolve 80 lines/mm on 35PA, you can on release prints (it doesn't mean that the print has that detail but if it is there, your projection system can do what 35PA says it can do). Likewise, on framing. First off, release prints, since the '80s, and possibly earlier, would use rock-and-roll printing (run the reel head-to-tail on the first pass, Tail-to-heat on the second pass (eliminates the rewind step)...well, that practice will guarantee that there is a vertical shift between the directions by the amount of slop between the sprocket tooth and the perforation. 35PA doesn't have that. Center is center since it is camera original. Also, however, the printer is set up could affect vertical centering relative to perforations. We're not talking about big differences but everything is heavily magnified.

JackRoe had a printed test film that did an okay job allowing people to set up frame/focus as it was printed film. Then again, checking focus needs to be done when the film starts and then again after the lens heats up (and even if the gate parts heat up and expand too). But 35PA is a measurement film, not for setting focus/framing controls.

[Frank Angel]

If people referred to this ring using it's proper name (Astigmatism adjustment) then there would be much less confusion on the subject.

Thank you sooo much, Vern! I can't tell you how often coworkers try to mess with that ring and when I explain it is NOT a focus ring, but a way to align internal elements in the adaptor to correct for astigmatism, I am looked at like me like I am from another planet. "The why is it called a FOCUSING RING...HUH?" they ask. So I tell them to show me anyplace on the anamorphic attachment where it says "Focus Ring."

The adjustment process for tweaking the anamorphot is simple, but tedious, but once done, it will make you beam with pride and satisfaction. Then make it immutable...just lock it down and as Randy says quite rightly, so no one will muck with it going forward. It is like I used to say about the cinema processor adjustment controls, once everything in the B chain has been aligned, I really wish there were that carbonite material that the poured over Hans Solo that hardened him into a solid block of it. There needs to be a way to keep fat, clueless fingers off the controls!

Long-live CinemaScope prints and the anamorphic adaptors to show them!

[Ryan Gallagher]

Thanks again all for the anamorphic technology lessons. We were able to spend a 10 min with PA35 and our problem lens and get the astigmatism alignment re-calibrated, or at least calibrated enough to not worry about it before switching to flat for two films that day. It was easy as described!

Had I known the ring only affected vertical focus/alignment I might have even been able to pull off an approximate correction while that last film was running, if I had a suitable scene and was willing to assualt folks with a larger adjustment than we were attempting. Or could have just used our throw distance approximation if I knew it was operating independently from the focus knob! Certainly would have been more pleasant than scratching our heads while we made small adjustments and not seeing the results we expected.

I'll make a point to pair adjust it again with someone in the house and check the other one ahead of our next scope film on the 19th. We'll put reference marks on the astigmatism rings with where we are at now before attempting any improvements! And then lock it all down with extra tape method after that.

I'm tentatively hoping to pull an afternoon and file new scope plates too before then, our historic scope plates are our most overfied format with quite a bit visible on masking. Will probably 180 the lens that has the factory reference marks away from the operator at that time too, and set up our mounting depth/focus for the center of the focus travel, when I get around to adjusting mounting/focus starting point with the other formats scope will become our reference.

We were also less gun shy this time in flat about making the needed focus adjustments away from what we roughed in with 35PA. We ran two suitable loops to get closer, one had a rating card at least, which landed us much closer to correct on two test reels, where we got the final focus and both screenings looked great. Film grain pattern visible through binos etc.

[Randy Stankey]

I'm glad you got it working!
It is an easy adjustment to make, although not an entirely intuitive process. If one doesn't know what that adjustment does, it's hard to understand what to do.

Most people just use the distance markings on the lens. Most of those people will just count the number of ceiling tiles between the booth and the projector and multiply by two. (Or four, depending on which way the tiles are oriented.) Some people, like you, might know the exact distance to the screen but, even then, the adjustment is only approximate. The right adjustment can't be made unless test film of some ilk is used.

Fortunately, adjusting your anamorphic less is usually a "set it and forget it" proposition. Once done and locked in (with Kapton tape) it doesn't need to be done again. I used to check it every once in a blue moon but, as long as the picture looked good, I didn't worry about it, so much.

For the aperture plates, I think you are right to recut them. If it's been a long time since anybody took care of them, you might have a slightly wider opening than you need.

When I first started as a projectionist, the scope aspect ratio was 2.35:1. When we first opened Tinseltown, that's what all the plates were cut for. A couple of years later, the aperture size changed to 2.39...slightly smaller. As I was told, it was done because labs/film makers started using polyester-base film. Because of that, they had to use heat splicers which left a wider mark on the film than the old, cement splices. (Cement doesn't work on polyester film) Aspect ratios were changed in order to hide the splice marks that would appear on the screen.

If your aperture plates haven't had any attention since about 1995, they are probably the wrong size. You probably should get new plates and cut them to fit your projectors with the new aspect ratio.

You could use silver solder to fix up your old ones if you need to economize. Lay a bead of solder along the edge of the opening then smooth it down with a file and some emery cloth. Once the solder is cool and hardened, you can re-file it to the correct opening. If you do this, use SILVER SOLDER! The melting point of silver solder is about 100º C higher than lead solder. The heat of the projector lamp might be enough to soften lead or SAC (lead-free) solder. Even though it is unlikely, the consequences would be bad if it ever did happen. It's easy enough to just use solder with a higher melting temperature so that you'll never have to worry.

IF it was my call, I'd get new plates, cut them to the right size then keep the old ones in storage. They will still work for older films that were made in the old aspect ratios. Just mark them, appropriately, and put them some place where you won't accidentally mix them up. Your screen masking might change, slightly, but if you have four-sided masking (I think you said so, didn't you?) the adjustment will be easy.

[Ryan Gallagher]

We have unfiled plates to start from, and having done that twice now for oddball under printings, the task does not scare me (anymore), it is just a matter of making the opportunity in our schedule.

[Vern Dias]

Thank you sooo much, Vern! I can't tell you how often coworkers try to mess with that ring and when I explain it is NOT a focus ring, but a way to align internal elements in the adaptor to correct for astigmatism, I am looked at like me like I am from another planet. "The why is it called a FOCUSING RING...HUH?" they ask. So I tell them to show me anyplace on the anamorphic attachment where it says "Focus Ring."

The adjustment process for tweaking the anamorphot is simple, but tedious, but once done, it will make you beam with pride and satisfaction. Then make it immutable...just lock it down and as Randy says quite rightly, so no one will muck with it going forward. It is like I used to say about the cinema processor adjustment controls, once everything in the B chain has been aligned, I really wish there were that carbonite material that the poured over Hans Solo that hardened him into a solid block of it. There needs to be a way to keep fat, clueless fingers off the controls!

Long-live CinemaScope prints and the anamorphic adaptors to show them!

Yep, back in my 35mm days I had to deal with anamorphic lenses. Today in the video, not much has changed, at least here in my HT.

In my HT. I use an ISCO Cinema DLP 1.5X anamorphic lens with my JVC RS4500 projector, using a 54" x 144" x 10" deep curved 2.76:1 1.0 gain screen.

With a screen width of almost 12', a throw distance of 14', and a preferred viewing distance of roughly 9', optimum focus, astigmatism, (and the PJ's convergence) settings are absolutely critical.

The horizontal non linearity and pincushion you see in the photo below which was taken with my pixel phone show the results of a test pattern generated by Cosalient on the PC (which is present in the lens to a degree, but appears exaggerated in the photo) is corrected during media playback by using a pixel shader in the media player.

To deliver a constant height image, I have AR presets for 1.77:1, 1.85:1, 2.00:1, 2.20:1, 2.40:1, 2.55:1, & 2.76:1 which are also defined in the media player. They are automatically applied on playback start based on the actual title's AR.

Here are some preset examples using 4K (3840x2160) projector resolution (x offset, y offset, x size, y size):

SetPlacement(8,16,3830,2140) //1.77
SetPlacement(-76,-26,3996,2224) //1.85
SetPlacement(-228,-100,4300,2370) //2.00
SetPlacement(-452,-264,4740,2694) //2.20
SetPlacement(-678,-368,5200,2906) //2.40
SetPlacement(-860,-486,5560,3150) //2.55
SetPlacement(-1080,-586,6000,3340) //2.76

The photo really doesn't do justice to the sharpness of the in person experience, but here it is anyhow:


PXL_20250531_194140642.jpg