Image Sensor / Optical Components


#81

Eager to learn about your progress. How far did you go with the project.
Thank you.


#82

Hi,
I have already poster several queries on the forum with no formal introduction. Sorry about that and here the intro.

I am from Sri Lanka. We have many films that need digital restoration. Since a professional restoration was prohibitively expensive we needed a solution to do it ourselves. I came to know about the Kinograph at it’s early stages, while researching for a DIY solution. And Mathew was kind enough to provide me with design files of his rollers. I went with a DSLR solution (like the early Kinograph) and found a Nikon J1 (with electronic shutter) as a camera to start with. However due to several reasons I had to shelve the idea for some time.

But now there is a reason to get back to it. Our country’s most respected film maker Dr. Lester James Peiris is now 95 years of age. His first film Rekhawa (Line of Destiny) https://en.wikipedia.org/wiki/Rekava is 50 years in December. At least I could restore a part of that film (If not the whole film) to e shown at its 50th anniversary it will be a great way to honor the Master.

So, I am back with the project.
I still have the camera and the rollers. I need to start from there.I have seen a sample scan posted by the AACA Library. It seems that they have succeeded to some extent. Hope to hear from them also.

Thanks again for your assistance and advise.


#83

Hey all,

There’s a new camera I’m wanting to buy and thought I’d get your reaction. What do you think of this little guy: http://z-cam.com/

It has a couple stops less dynamic range than the BlackMagic or Red but for the price I’m thinking this is a really great option for us. Go ahead - tear into the specs and tell me what you think!

Also, looks like there’s a way to get single frame capture on a trigger with the Black Magic camera. Want to do a test on the BMCC 4K soon to verify.


#84

Hey so we have finally gotten the machine to a place where we like it. Here is a playlist of the results.

We’ve been using the Panasonic Lumix GX8 which is fully electronically controlled sensor/shutter, but has the cmos shutter. After getting the images I use a combination of adobe premiere and after effects to stabilize the images/video. We decided to run a film projector and extract the audio through the audio out and match them up in post. It got us much better results than from the aeo sound capture. Much quicker as well.

The rolling shutter is still a hindrance but for the sake of getting the machine up and running we completed what we wanted. Now we would like to upgrade the camera, but the complications come with the triggering mechanism. @matthewepler does that Z-Cam have an easy to way to trigger the camera similar to the original setup of a shutter release cable?


#85

Hi Matt, the Z-Cam does look pretty good, and for a good price. Although according to their Kickstarter page, the camera uses a rolling shutter: https://www.kickstarter.com/projects/2081787381/e1-camera-4k-uhd-interchangeable-lens-camera

Not sure if that’s a dealbreaker, if the film is traveling by at a fast rate.


#86

This looks so much better. Interesting idea on the sound capture. I need to take apart a projector and see exactly what the sound hardware consists of. Anyone have knowledge in that area?

M


#87

I’ve taken it a step further since then. I implemented a step motor into the sprocket/switch roller. Changed the design of that roller and added 2 more sprockets to make it an even 16 so it jives with the 4 steps in the motor. This moved the tension points around so a lot has been implemented to make do for that. I’ll post a more comprehensive list and such a little bit more down the road. We added some more changes to the gate and such. The result is a bit slower to capture but saves a bunch of time on editing. Here is a playlist of the films I’ve scanned so far.


#88

Looking forward to seeing how it all works!! Post a video if you have time.


#89

Here is a presentation we did at a recent meeting for our club, the AACA. At slide 31 is where a lot of the new additions are added, but feel free to look at the whole presentation. Not a whole lot of background info is built into the slides, but the images are there and some of the slides have videos of the new parts working. Some of the clips look like images on google’s presentation mode, so try clicking the images to see if they are videos.


Los Angeles.... anyone have a functional kinograph built?
#90

INCREDIBLE!! I"m emailing you now so we can schedule a phone call. I have questions. Love the additions I’m seeing. Would you guys also be open to doing a voice-over version of your slide show so we can hear the full explanation of what we’re looking at? I’d love to get your take on how you tackled the problems, and what parts of Kinograph weren’t working for you.

This is great!!!


#91

Yeah, I’ve been planning to do a mini documentary style explanation of the machine. Just haven’t had the time to do it yet. With all this excitement, maybe I’ll push it out soon.


#92

Your machine and your videos look awesome! I am looking forward to see your documentary.

It would be great if the original plans of the kinograph could be updated, too.


#93

Here is a Machine from an university, used for scanning in 4K


#94

What do you guys think of this camera?


#95

Good find @robinojones. This was at the top of my list about 2 years ago and when I called Point Grey they suggested that I would need additional hardware to handle the data at the speeds I wanted (24fps). The additional hardware and the camera came to $10K.

Perhaps they have upgraded the guts of this camera since then to help handle the offloading/writing of data and buffering so that additional hardware is not needed?

Would you be up to contacting them and finding out if you would need additional hardware (I believe it’s called a Frame Grabber) if you ran this camera at full resolution at 24fps?


#96

I talk to them and they directed me to a knowledge article (10827)

Based on the “Getting Started” guide (https://www.ptgrey.com/support/downloads/10211) the only extra hardware is a USB interface card, I would think this would be integrated in a PC motherboard already - maybe dedicated USB 3 cards are faster?

Speed wise I only need 3-4fps for my project (35mm 4 perfs scanner built from optical printer projector with pin registered movement) so It looks like a fast raid setup might be able to handle it - I’m not sure about 24fps though, maybe at 24fps you need GigE or something or maybe Raid or SSDs might help with USB3. I’m no expert in data rate but will know a lot more very soon as I build my scanner.

Based on my research this seems to be the only “affordable” camera option at 4K with good dynamic range (71db in mode 7, or 66.15db in mode 0, see page 101 of Technical guide https://www.ptgrey.com/support/downloads/10125 ) since the other cameras (Imperx etc) are 10K +

An issue with the grasshopper for my project is that mode 7 is only allowing 4096x2160 instead of the full 3000 vertical which is not good for 35mm 4 perfs.

In mode 0 with full resolution it’s great for 4 Perfs - What do you guys think about this camera dynamic range for full resolution @ 66.15db?


#97

@robinojones thanks for looking into this. I’m excited to see that a frame grabber is not essential for a 4K capable camera.

I think the dynamic range + global shutter + USB3 + software SDK makes this camera a strong choice for a custom solution. If you don’t mind setting all of it up, then it is a very good camera. It’s definitely at the top of my list for a 4K solution.

Matthew


#98

A question: Could one use an imaging sensor that is approx 25mm x 20mm (or larger) and have the film run in front of it, without any optics, just collaminated light?


#99

Interesting thought, Martin! I would assume you’d have to control aperture somehow and also figure out how to make sure the sensors didn’t get dirty. I think focus would also be an issue.

I did a quick search and found these links to projects where people are attempting to do what you’re imagining:


http://www.hitachi.com/New/cnews/month/2016/11/161115.html


#100

Had a talk with some colleagues, and basically there are various problem.

The first one is that, to achieve focus, the film has to be as close to the sensor as possible, preferably direct (i.e. physical) contact. But the sensor is very fragile, and would get damaged.

Thus, one would need a thin, optically neutral layer over the sensor, and be resistant to scratches (such as diamond); something that I have not found and thus the cost would be prohibitively expensive to make.

Even if one could get such a sensor, the film would then have to move along that surface, which would introduce scratches.

Bottom line: for now it has to be optics.