Hawkeye-DSLR

[David Brown]

FIXED!

I'm surprised that just cleaning repaired this. It now runs totally silent, no grinding or claw snapping. I added a piece of paper to the left side of the gate and it is as jittery as any new Wolverine. Aggressive take up stopped by holding the reel.

I'm going to check on the gate lid replacement now.

[Stan Jelavic]

Wow, great job David. The DSLR focusing problem is definitely caused by the camera shutter shake. In video mode the camera is steady but the pixel resolution is lower than in image mode. Here is the side-by-side in video mode.
https://photos.app.goo.gl/ieyj2GFLPMZMGS4XA

Two possible enhancements:
1. Use e-trigger camera for frame-by-frame. Actually this is the only viable option in this mode since you do not want to wear out the mechanical shutter. I know some people planned to use a lower cost dslr with the mechanical shutter and are not worried if it wears out. In that case the vibration from the shutter can be a problem and they should consider using the Hawkeye board with lower scan rate. But that is higher cost.

2. Use video mode with a newer DSLR with higher resolution. The video quality may not be as good because the video mode uses some sort of moving image compression but with the Mark2 I have not seen any significant difference. The only difference between the video and still image mode quality is the pixel resolution as far as I could tell.

[Stan Jelavic]

Read a post today from Barry.
https://8mmforum.film-tech.com/vbb/f...er-new-hopeful

This gave me an idea for the DSLR project.
Essentially, instead of going with the expensive DSLR and alternative could be a web cam or similar cheaper cam. The rig consisting of Wolverine film transport is an open platform that can accept almost any camera. I am thinking of cheaper camera with the C lens mount. The C mount lenses are superior to M12 and the selection is much better. With the camera running at 30 FPS and the rig running at 2 FPS exactly the two systems stay synchronized without any additional components and it is very easy to extract the relevant frames, 1 out of 15.

Any thoughts on what camera we can try?

The issue may be that old pink glow again...

[David Brown]

I was able to get a great white balance of the Winait's led using in camera custom WB. Dead center on the vectorscope!

'DSLR' option is the only way to get 4k, if that's what you want. Using 4k video the image would need to be framed tight. Still capture can easily do 4k, assuming your sensor is 16-24 mb.

I have only 2 cameras that can be used with the DSLR option. A Sony full frame and a microscope. The microscope is a board mounted sensor in a c-mount housing. It has a LONG lens, would need another lens. No aperture and too close to the gate. Video is 1080p, and the still capture 4k. I don't know how fast it is.

So a board camera could work here with a c-mount housing.

There's a lot to be said about IC capture software. Sadly, my microscope can't use it.

Video mode does not need a trigger, right?

I have a small super macro lens for the Sony that collides with the stepper motor. I think it might be 'easy' to reverse mount the stepper. That lens could work with an extra 1/2" clearance.

or

I'm using a 90mm macro with 42mm of extension in crop mode (16mp). Image is still too small for 4k video. Will need to try a longer lens for video. How close can a 135mm focus? Will see.

So any video camera could work if it can focus and stay in tight. I'd want an aperture at this magnification. They still make camcorders??

Phone? Not without a clip on macro lens or ............ probably not an easy choice.

I think I'm rambling on now.

[Stan Jelavic]

Lots of thoughts here David. The video mode does not require any extra hardware and will sync up with any camera running at 30FPS. Probably the cheapest solution. If there is a web cam that would accept the 12mm or 16mm lens (pink glow issue) then that would be a really good low end solution. Probably, the whole rig can be made out of wood cutting the cost down further.

This would not give you 4k but would still be better than stock Wolverine.

Reversing the stepper is doable. It could mount on the metal plate. The wiring would have to switched for reverse operation or a jumper can be added to the PS for that.

[David Brown]

Hi Stan

Don't feel bad about your camera support moving. I'm using a copy stand. It' very convenient, but it's not stable at these levels of magnification. I was surprised when doing a 9 exposure hdr capture. The camera does the work and it's done in about a second. Although I started at 1/250 second, it steps down the shutter. I had to switch to the electronic shutter. No blurred pictures, but the framing moved slightly creating ghosts in the blend.

The longer lenses I tried do not magnify the image more, even with extensions. They can't focus close enough.

I flipped the motor and yes it runs backwards. The motor is temp mounted with a magnet. Hope it doesn't mess with the motor.



I can use the super macro now.




I think there will need to be a connecting brace from the camera to the gate assembly. I captured some S8 short clips running film off the desk and into a box. Strange. I will check the settings and try that again. Problem might just be too much movement. The camera on the copy stand is not rock solid, and the film gate is on a plastic box. There are stable frames, just too many smears.

I was looking for a longer plug for the motor and found this.



Would this work as variable speed control? If so, the take up reel is not a stepper motor. Would less voltage slow it down?

https://www.amazon.com/15-160V-Adjus...YM5VT&pd_rd_r=ac1005b6-9d0a-454d-9ace-76ca0d271e34&pd_rd_wg=krxAU&pd_rd_i=B07HNSVMVH&psc =1


The other option I'm thinking of is the Hawkeye ver 6.0 board I have. I like the video option the Wolverine can do and you say the Hawkeye runs at an out of sync speed.

Add this to the growing list of differences of Winait/Wolverine MMP. The gate window is slightly smaller. I could reprint a smaller mask or just file down the one I'm using with the Hawkeye. Maybe these don't come from the same factory?

[David Brown]

My 3d printer uses 2 stepper motors for the Z axis. Though I have not checked the wiring, I wonder if they are not just wired in parallel?

These motors (like the Wolverine Claw) are ~ $10 on Amazon, what if there were 3 motors, Supply, Take up, & Claw sharing the same connection?

[Stan Jelavic]

Good work David. Most of the steppers including the one used in Wolverine use two phases to generate the rotating magnetic field. That is why the cable has 2 pairs of wires. Connecting the steppers in parallel should work as long as the driver is capable of driving them i.e. it should be able to support the required current.
The easiest solution for you I think would be to add the xtal control to the Hawkeye board. I will check how this could be done. Maybe tacking the xtal to the MSP430 pins or use a small adapter board that plugs into the MSP430 socket.
Another possibility is arduino or RPI controller. I did that a while back in the old forum but it may not be accurate enough. Could check.

Here is the xtal control.


The easiest way would be to pull the XIN and XOUT pins out of the socket and solder the xtal and the two caps directly to them. Alternatively, I can make a small daughter board.
Here is example code:
http://www.msp430launchpad.com/2012/...z-crystal.html

[Stan Jelavic]

Ordered a few 8Mhz and 16Mhz xtals. In theory they can be attached between Xin and Xout pins. If my calculations are right these frequencies should produce 1 and 2 second scans exactly , or very close to that depending on the accuracy of the crystals. Will see how that pans out.

[David Brown]

I've been in a quandary on the final application for this DSLR project.

1st, I was speculating on the simplicity of 3 stepper motors running in sync. I don't see that working now. The supply and take up reel's speed are constantly changing relative to the amount of film wound one each. The claw is geared, so all three motors would run at different speeds, 2 of those constantly changing. Not so simple.

2nd, the stock board runs in sync with video capture and for single frame capture, it can already trigger the camera's shutter.

3rd, 2fps might be too fast for stable film transport. Not sure of that, I've only been capturing at the slowest speed with the Hawkeye. Most of the films so far have extreme exposures so I've been running slow in manual mode. It's easy to stop and adjust with little rewinding.

4th, I need to mount all this hardware! Too tedious to make any test trials with loose parts!

Note: I've not tried this, but I think ffmpeg could extract the unique frames from a video capture and recombine them in a new stream. All this in a single batch file! I might be dreaming! There are 3600 frames (200 seconds) in a 3 inch reel! The video could have 54,000 to sort thru by hand.

I'll get out my V6 Hawkey board. I will need to flash the msp since I was toggling the later versions between 2 of them. It might have Versions 10-12. Maybe that would work anyway? I don't know how to check that, just flash with the highest ver that works with the board.

[Stan Jelavic]

Wow you have several chores at your hands David. The V6 board does not have trigger control so it will default to the Wolverine trigger at the beginning of the claw travel. I believe the latest non HDR version should work with the V6. Let me know if you run into issues. It does not work well in video mode. I will let you know if that works within a week or so.

[David Brown]

Small steps, this is a side project. Next, I need to find a larger angle to support the Wolverine claw, the motor is not installed here.

[Stan Jelavic]

Pretty impressive setup David. How is the image quality and how much sensor coverage can you get with it?

[David Brown]

I've had the geared macro sliders since this all began years ago! They are cheap and sloppy. This way I have some fine tuning. It's been sitting here for weeks waiting for inspiration and time. So I started with things I already have.

I just now looked for a picture, I'll post one later. I was shooting some stills yesterday, but not with this setup above. The image fills most of the sensor and the file size is enormous.

I was shooting stills from a reel I had just finished digitising last week thinking stills would look better from the DSLR than the tiffs. Not so much.

More on that later, I shouldn't read this forum so late!

[David Brown]

Another problem with these cheap macro sliders is after they are in position, turning the set screws will change the position. I need to anticipate the offset before tightening.
Even after tightening, all parts are screwed to a board, there is some wobble. Enough that the mechanical shutter is blurring the image. I can't wait until the motor and claw are moving!

Designing with physical parts is faster up front!

Here's what the camera and this lens see.The front of the lens and the film are .5" apart. There is a lot of flaring, it's not an expensive lens.

https://drive.google.com/file/d/1hes...ew?usp=sharing

The above tiff is much larger than even a DNG. If capturing raw files, the process would be limited to the card size. Some cameras can be tethered to a PC, transferring each file as captured. I don't see that happening at 2fps unless without shooting jpgs. That takes the wind out this process and video is nothing more than streaming jpgs. That could be ok if the exposure and color is corrected before capture





The mask made for the Wolverine will not fit this Winait. The opening is abt. .4 mm smaller. I filed the mask and now it fits, but the opening is not aligned with the film framing. I will need to design another.

I've reached my max post limit. Waiting for the next window.