Oscillating Bounce (Tips Wanted)

[Ryan Gallagher]

Suggestions for routes to diagnose wanted. Take up clutch condition or tension? Drive belt condition? Bent reel arm spindle?

Our No. 2 Projector (Century JJ) tonight decided to exhibit this new behavior on R3, 5, and 7. Applying more gate tension reduced it but did not eliminate it. Turning the knob to 3'o-clock was about all I was comfortable with. But more might have mitigated it. Intermittent shoe tension adjustment also had little impact. Our impression was more tension reduced it, but that might have been in our heads.

I would describe it as an oscillating bounce or pull that was not timed with reel rotations. It did it with multiple different house reels so I don't think reel condition is to blame. Print could have certainly been a factor but it was behaving much better on No. 1. It did not seem to be any worse or better (or timed any differently) at the start versus end of the reel.

Here is a video, crappy blurry zoomed in phone shot, but it does display the bounce if you watch near the masking for vertical movement. This was the result AFTER cranking up the gate tension to calm it down as much as I was comfortable doing.

Our next 35mm is Wed the 13th, and I'll certainly throw a house print reel on No 2 ASAP to see if the behavior was specific to the print or not, but my hunch is I'll still see the bounce. I'll also try loops to see if it goes away when only the gate and intermittent are involved.

It was a reprint on polyester of All The President's Men from WB Library... hard matte 1.85. The matte lines seemed to move a touch between scenes and reels too. I was in the booth but not the operator, but jumped in to help tame the beast slightly.

My best guess after the screening is something take up related? But even in my head that doesn't jive, as the takeup is isolated form the gate by a loop.

[Ryan Gallagher]

For fun I asked Chat GPT to see what it has learned about this ancient craft. Decent response... and a clue. IMHO the operator tonight tends to error on the side of a large lower loop... not so big to contact anything but still large. I'll have to run a loop size test with my house reel. It might be as simple as that... and explain why it did not do it on reel 1 which he might have threaded differently. I did point out loop size as a possible cause after R3 did it, he went a little smaller, but still probably on the large side.

I need to show everyone in this booth the trick of threading from picture start to sound start to convey the idealized minimum lower loop. I think people will be shocked by how much larger they are making it.

Here is the query, AI hallucinations and all:

Prompt: What is the most likely cause of a 35mm polyester film print to exhibit an oscillating vertical bounce in the projector gate?

Response:

An oscillating vertical bounce of a 35mm polyester film print in the projector gate is most commonly caused by: ➤ Excessive film tension or poor film loop setting, especially due to incorrect loop sizes (top or bottom loop) or poor threading.


However, specific to polyester film, the most likely cause is:

---

? Slippery, stiff base causing intermittent movement instability


Polyester film (as opposed to acetate) has less flexibility and greater tensile strength, which can cause:

• Inconsistent engagement with the intermittent movement (sprockets/claw)
• Springiness in the film path, which causes it to oscillate if loop sizes are even slightly off
• Improper loop cushioning, especially if loop buffers or rollers are not absorbing the tension correctly

---

Other possible contributing causes:

• Worn or misadjusted intermittent movement or claw: Not grabbing the film consistently every frame.
• Gate pressure plate issues: Too loose or tight pressure causing the image to "swim" vertically.
• Bad splices: Particularly with thicker tape splices on polyester that don’t flex well.
• Projector not optimized for polyester prints: Many older projectors were designed with acetate film in mind.

---

What to check or correct:

1. Re-thread the projector and check loop sizes — follow exact manufacturer recommendations.
2. Check gate tension and ensure pressure plate is functioning properly.
3. Inspect intermittent movement for wear or misalignment.
4. Use proper roller alignment and film path maintenance to reduce stiffness-induced oscillation.
5. Avoid overly tight tension from take-up reels or rollers.

---

Let me know what projector model you're working with or if the bounce is rhythmic or random — that can help narrow it down further.

Prompt: Rhythmic bounce and Century JJ Projector.

Thanks — that’s very helpful. Given you're using a Century JJ projector and experiencing a rhythmic vertical bounce with a polyester 35mm print, here's the most likely cause and solution path:

---

? Most Likely Cause


Improper loop size — especially the upper loop being too small or bottom loop too tight — in combination with the inherent stiffness of polyester film.

On the Century JJ, a too-small upper loop doesn’t give enough slack for the film to feed smoothly into the intermittent movement, and because polyester has less natural give, it causes a repeating “bounce” in sync with the claw pull-down (hence the rhythmic nature).

---

?️ What You Should Do

✅ 1. Re-thread and slightly enlarge the upper loop

• On the Century JJ, aim for a generous upper loop (usually around 8 perforations above the trap).
• Be extra careful to follow the original Century threading chart exactly, but bias the loop slightly larger if using polyester film.
• Ensure the loop isn't too tight or “slapped” against the sprockets or rollers.

✅ 2. Check intermittent movement & sprocket timing

• If the intermittent is even slightly out of phase, polyester won’t “forgive” like acetate might — check for:
  • Wear in the claw cam
  • Loose or misaligned shutter/intermittent shaft
  • Worn movement bushings or pivots

✅ 3. Adjust film tension

• Make sure take-up tension isn’t too strong. Polyester doesn’t stretch, so overtension gets transferred back through the film path.
• Use a proper loop setter if available — or manual "hand roll" the film through the gate and watch loop stability before starting motor.

✅ 4. Clean rollers and gate path

• Even minor dirt or misalignment in film guides/rollers can cause polyester to "walk" slightly in the gate.

---

? Why Polyester Makes This Worse

• It’s stiffer than acetate and doesn't absorb small threading errors.
• It's slippery, meaning it doesn’t grip the film path rollers or the intermittent claw as easily.
• Loop oscillation reflects through the whole film path like a taught spring when even slightly off.

---

If you've checked threading and loops and the bounce persists, the next step would be:

? Watch the film running in the gate with the trap open (carefully!) at a slow speed. If the film is visibly "tugged" or "jerking" once per frame, it’s almost certainly loop or intermittent-related.

Let me know if you want help walking through a proper threading diagram or if you've also noticed this issue with acetate prints — that would confirm whether it's polyester-specific or a mechanical wear issue.

[Jim Cassedy]

About 7 years ago, I had a rhythmic 'bounce" like that on a new print. I spent a lot of time
trying to figure out what it was, but since it was only one reel, - and it happened no matter
which projector I ran it on, it turned out to be some sort of 'slippage' issue in the printer
which made that reel. They (Deluxe) acknowledged that there was a problem with several
prints made on a particular printer on a particular day, but that they thought they had gotten
all the 'bad reels' before they were shipped out. Apparently they missed at least one!
They did overnight a replacement reel though.

[Brad Miller]

Flat prints will show this more than scope prints.

Prints made at Deluxe in Canada will be every bit this bad or worse.

That's not your projector. That's printed-in.

[Ryan Gallagher]

Thanks Jim and Brad, that makes me a bit less worried I'm gonna uncover something mechanical. Will still run a poly house reel at first chance for confidence reasons, perhaps tomorrow. I've got booth tours scheduled before the DCP, maybe the season subscribers get to see some film running between the yacking.

If we were smart and not ready to go home we should have played one of the reels exhibiting the problem on the alternate projector after the show. That would have definitely pointed the finger accurately. But alas in the cans and out the door already.

[Marcel Birgelen]

The cans are out the door, so what I propose is no more an option... but in order to see if the issue is in the print itself, you could throw a piece of film that's affected under a flatbed scanner, scan it in the highest resolution available and check if you can detect the oscillation between the frames in there in e.g. Photoshop. Flatbet scanners nowadays are light and cheap enough that this doesn't require any extraordinary gymnastics anymore.

[Mark Gulbrandsen]

That looks like printer bounce to me. And as Brad mentioned it is 1:85, so the bounce is magnified. I had a friend that worked for B&H and flew to labs all over the country maintaining their printers... It may be a worn part, or just something out of alignment. I used to even see this on dailies, mainly from Technicolor in New York. DP's always blamed the projectors first. So I always carried a dead steady roll Of Dye Transfer Technicolor, and running that shut them up really fast. And then they had to call the lab and complain.

[Ryan Gallagher]

The only part that made me worry it was the interaction of this print with the no 2 projector, is that it was consistent-ish on that projector only, for 3 reels out of 4, and not the other.

Another clue is that the optical sound head tension also bounced a little while it was exhibiting this behavior. We also had a slight buzz from what sounded like the lower drive sprocket.

Could it also have been an issue with the way the perfs were cut into the stock?

Had it been both projectors I would have pretty easily concluded it was something to do with the print itself.

Loops and a house print will either confirm or allay my concerns.

[Ryan Gallagher]

Tested this yesterday, all evidence points to it indeed having been "printed in". None of our loops or my personal poly flat print did it even in the slightest on that projector. I even tried to make it angry by using excessively large loops. Nothing.

[Steve Levy]

If you pull out the aperture plate and are able to see the perforations on screen and they are steady, the projector is steady.
I have seen cases where the perforations are rock-steady and the image is bouncing all over the screen. That is printer slip as described in previous posts.

[Ryan Gallagher]

If you pull out the aperture plate and are able to see the perforations on screen and they are steady, the projector is steady.
I have seen cases where the perforations are rock-steady and the image is bouncing all over the screen. That is printer slip as described in previous posts.

I keep forgetting about that method. Of course on this screening it did not do it for the two reels ran in test, only appeared on other reels once we had an audience... so not an option unless we had stayed after to run one of the problem reels again in the other projector. Which we should have done in hindsite. But only showing films once easily puts people into "it's over now, let's go home" mode.

Our side masking is manual from stage, so a bit hard to make perfs highly visible... but some wise old projectionists from a bygone era built a clever lever mechanism into our pedestal risers, it's very easy to pan our projectors, and we have reference marks.... so could easily pan the perfs into view.

[Steve Guttag]

If you pull out the aperture plate and are able to see the perforations on screen and they are steady, the projector is steady.
I have seen cases where the perforations are rock-steady and the image is bouncing all over the screen. That is printer slip as described in previous posts.

That is a necessary but not sufficient method. I have found the opposite...the perforations was not well registered so you'd see moving perforations and it wasn't the projector's fault. Since the image was pin-registered to the faulty perforations, it made it look like the projector couldn't hold the film steady...switch films, rock-steady. I later learned that some films are perforated one frame at a time (negative film in particular) where as print film is, normally, perforated continuously. Since the projector holds the film vertically via the perforations but laterally via the edges of the film...how the film is perforated can also play into the final result.

Of the various test films, I have found that Schneider's target film was more steady that 35PA though NT-Target (uses the RP40 pattern with some higher resolution line pairs in the middle) always had good results too.

The moral of the story, never use one source as a determinant of the system performance.