Wolverine-Hawkeye Telecine

[Stan Jelavic]

V12 update. Still working on cam sync. 3 more components to add circled in blue. Then test.

[Bruce Davis]

Hi everyone, finally received the 2mm ID x 3mm OD brass tubing from China (has taken 9 weeks, no metric tubing this size was available in Australia). Wanted to experiment by not only extending the shaft length slightly for more support to the capstan printed part, but also a slight increase in diameter would also be helpful to allow truer running (larger bearing surface to reduce wobble).

Just so happens that the brass tube is a snug press fit over the 2mm dia motor shaft, used a small vice to press a 9.5mm long piece of tube onto the shaft.

Made the necessary changes to the printed parts and assembled the capstan system in the new center location (thanks to Kamel for the updated motor mount).

I am happy with this improvement, much less fiddly adjustment with the grub (set) screws. The extra bearing surface available to the bore of the Back_Toothed_Pulley_Capstan in both length and diameter makes for much better support and therefore true running (as true as any printed part can be made).

This may all be a waste of time, as I understand, Stan may no longer be implementing the capstan in the future (probably not needed), with all the difficulty of having parts printed accurately at a reasonable price.

Would like to thank everyone for the continuing developments .

Regards - Bruce
LAST Printed USED Capstan Parts_3mm Bore_08-05-2020_1.zip Toothed_Pulley_S8_Capstan_Mine_15.90dia_Latest Kamil Teeth_3.zip Toothed_Pulley_S8_Capstan_Mine_15.90dia_Original Kamil Teeth_3_USED.zip

[Stan Jelavic]

Nice work Bruce. I stopped working on capstan for now. Just too busy getting the boards finished (in addition to other projects). But once I am done with the board kits I will re-evaluate capstan at some point.

[David Brown]

Thanks Stan for that image composite of the board. That will help and so will your suggestion to use a secondary mount. Best to keep things as simple as possible.

Bruce, I remember this idea, glad to hear that it helps. I just ordered that tubing. Sorry to say it will arrive Wednesday! 9 weeks is enough time to completely forget about your order. Like trying to buy toilet paper. I ordered a mask in late March. It still has not come from China. None of those around here.

Tubing on Amazon: https://www.amazon.com/gp/product/B0...0?ie=UTF8&th=1

[Stan Jelavic]

Hope you can fix it just by opening up the holes a bit David. I finished the cam sync on the first board and tested it with a 100 ft clip. It needs some tuning at 2 FPS . Now that the trigger is at the point where the claw is going down, I think I am a bit too early and the camera catches the top of the image while the claw is still engaged. The top is affected only since the cam uses the rolling shutter. Will experiment with timing next.

[Bruce Davis]

David thanks for the link, did not even check Amazon for this brass tubing (should have) would have taken much less time.
A section view with changes made to to the capstan printed parts.

[Stan Jelavic]

Extending the trigger delay to 4000 fixes the issue at 2FPS. Reran 100 ft without a glitch. Will have to retest at slower speeds as well. The new takeup cct still has some jitter. Will look into it tomorrow.

[Chuck Tsocanos]

Ordered V12 boards qty 15.
Updated the qty sheet.
https://docs.google.com/spreadsheets...it?usp=sharing

Hi Stan, Im a recent joiner/lurker seeking to scan family Super 8 inventory, discovered the Wolverine, and then this thread. You, Kamel and a few others are doing amazing work here so Im sharing an understatement of some appreciation of what you guys are designing here. This will save me thousands of dollars having my dad's old Super 8 reels shipped and scanned in HD so I don't mind the $400 unit plus whatever the upgrades cost. Ive read through this and the older forum over multiple hours and want to get the best I can out of a Hawkeye modified wolverine pro - even if it ends up costing closer to $800. I did sign up on your list and looks like Im #43 on the list. Im curious if you are still tracking the list and feel you are able to fulfill all the orders e.g. if its just a matter of time e.g. wait for a few more months? Again, not putting any pressure, just curious, as I know you aren't making much money here and I don't want to scan too much film I will want to repeat if I get the Hawkeye installed. Lastly, while I probably can do the assembly, is it an option that I can I ship you the whole unit, the new camera, etc. and pay to have you put this thing together so I can focus on scanning. If not, I'll make a run at it, but wanted to ask. Im more than happy to compensate you for the time and effort. Thanks! Chuck

[Stan Jelavic]

Thank you for Chuck for the words of encouragement and support. Pretty happy with the final video results and that was the main objective of the project. The qty list is still active. I am completing 14 units for the people that put their names on the list a long time ago. I still have extra parts and could do the board for you but ran out of V12 boards and would have to order some more. Not a big deal but just letting you know.
The video resolution is 1280x1024, not HD+ but it is very good quality. HD may also be possible - see previous posts from David Brown.
In any case I will plan to do the kit for you and will let you know when it is done and then decide if you want to send me the unit or you do the mod yourself. Regards, Stan.

[Stan Jelavic]

Looked into the jitter issue today. I slowed the motor down to a slow crawl and took a video of the claw, the takeup and the camera capture. Perhaps hard to see but if you look at the claw movement you will notice that the takeup gets engaged at the beginning of the claw cycle at the point where the claw start pulling the film to the right. The capture happens when the claw is all the way to the right and going down. The two events are decoupled and in theory there should be no jitter caused by the takeup. But in reality, the 2FPS is so fast that there is some "bouncing" in the takeup that will cause some jitter if the film is not routed through the rollers. The first three rollers did the trick in my tests but more testing will be needed.

https://photos.app.goo.gl/fzBXXMAZn74tZkS58

The frame smear can happen at 2FPS. Again, here, the 2FPS timing requirements are pretty tight. If the trigger is moved a bit earlier, the top of the image gets smeared for dark scenes and if it is retarded, the bottom gets smeared. I set it in the golden center in the firmware but still dark scenes could cause issues at 2FPS. If that happens, increasing the gain will fix it (gives more grain that can be reduced during post) or switch to 1FPS. Again this is for very dark scenes (long exposure).

[Bruce Davis]

Hi Stan, thanks for the information, if 2fps is problematic what if this high speed were reduced slightly (range between 1.5 to ~). Did I understand correctly that with the latest changes you made on the v12 board, the take up motor timing and duration can be altered in firmware so that the claw and take up are not necessarily working in sync.

My concern is that if the take up is tugging / pulling the film at the same time, as the take up real changes diameter (say a 400 ft 7 inch reel) so does the tension or pulling power on the film change, this will cause the film frame location to change position in the gate over time.
https://1drv.ms/v/s!AtIUdJCJJJGigROD...p7CRa?e=n7pfYi

Capturing the film frame image while the claw is still engaged (on its way down) I would have thought might not be ideal as this could cause the film to vibrate or deflect slightly at the same time. In my Wolverine / Hawkeye the claw applies a reasonable amount of pressure against the leading edge of the film perforation, so much so that I can hear it scrape against the film as the claw is moving down (the intensity of this does vary).

Regards - Bruce

P.S. I am no expert but I think movie projectors do not allow the take up reel to have any effect on the film being fed through the film gate.
Also perhaps a faster lens (from 2.8 to 2 or even lower) and or a faster camera might be an option, instead of only 5fps perhaps double that, just a thought.

Limited lens information available on the current lens being used :-
TBL 12-2 C 5MP Lens
Specifications
Format 1/2"
Resolution 5MP
Mount M12x0.5
Focal Length (f) 12 mm
Iris F2.8
IR Cut Filter Yes
Working Distance (mm) ?
Distortion (%) ?
FOV ?

Lenses are a very complex subject for me, will a faster lens have any advantages/disadvantages?

[Stan Jelavic]

Thanks for the input and suggestions Bruce. Yes, the speed can be reduced and the takeup timing can be changed.
The video that I posted shows that the camera trigger and the takeup are out of sync. The takeup is at the beginning of the cycle when the claw starts pulling the film and the camera trigger is with the claw down going back to the left. That is actually the optimal setting. The trigger has to be while the claw is down and to the right. That will give the longest delay before the claw engages the film again. I am testing at the beginning and at the end of the reel. If I reduce the force, the takeup does not catch up with the film advance at the beginning of the reel.
I actually forgot to set the FPS to the max in IC Capture, my bad. Will try that today. This may fix the smearing issue.

Regarding the lens, yes the brighter lens requires shorter exposure, which does help with this issue but on the other hand the DOF suffers.

The FPS setting did help. Still some very dark scenes show some smear at the bottom.

Bumping up the gain fixes it.

Going 1FPS fixes it also.
Going with a faster camera would solve this issue but is it really worth the extra money? I am putting this issue on the backburner for now and getting back to the boards. Almost done with the trigger cct on all 14 boards.

[Bruce Davis]

Stan thanks for the reply, today put a 7 inch reel through the Hawkeye, first one after the capstan relocation, it is defiantly performing better, there is less sideways movement of the film in the gate and less up/down shifting of the film frames (jitter), the jitter remaining is small and at 0.5 FPS almost none.

Regarding a faster camera, which make/model would you recommend and how much more expensive is it.

Regards - Bruce

[Stan Jelavic]

Glad to hear the new capstan worked for you Bruce. The camera that we use
DFM 72BUC02-ML
is spec'd at 6 fps at full resolution (2592x1944

For higher fps we would have to go with the USB3 camera and > 25 FPS because I tested our cam at 17 FPS and it still has some smear at low gain.
Here is the list from imagingsource
https://www.theimagingsource.com/pro...usb-3.1-color/

Look at the shutter and pick one with the rolling shutter since it is lower cost. I picked this one:
https://www.oemcameras.com/dfm37ux178ml.htm

60 plus FPS (plus at lower rez) but quite a but more money. There are other suppliers with possibly cheaper models but the mounting etc may be different.

[David Brown]

....mounting will be different and a new lens will be needed for the dfb37ux178-ml, it has a larger sensor. It seems easy at first, but picking a new camera consumes a lot of time!

I'm content to run captures at the slowest speed. It seems to have the most stable capture. Takes a long time, and if I feel I need to monitor, it will take place over days! But that's just me, I know there will be more demand for speed.

I was excited for the dfm 37ux226-ml at 4kX3k pixels. OEM is the only place I know to find prices and they don't list it yet. I could then use crop and center on the sensor and not need to build a new lens mount. It's likely expensive, and I would probably need a new lens too.

Best thing that happened, I discovered Eagle Software is AutoDesk and it integrates with Fusion360. I have over drilled all the mounting holes on my V6 board including the camera mounting holes. I found the .brd file for V12 and 3D printed it! It fits! That should help me get started.