Wolverine-Hawkeye Telecine

[Stan Jelavic]

Tested with the USB2 port Bruce. 8 bit is around 4 FPS and 12 bit around 2 FPS at higher rez.
At the lowest rez it is around 16 FPS.

Lens:
Possibly blue series from Edmund
https://www.edmundoptics.com/f/blue-...-lenses/13096/

Or something similar. Would have to check if the FOV is OK. No IR filter either.

The 25mm lens would not work without a mechanical mod to the Wolverine front panel. Will double check my calculations. I ran some tests and they did not match what I calculated. Back to the drawing board as they say.

Missed David's post. Still looking into 25mm. Good input on different adapters David. Could come in useful because the cable that I use is heavy and puts strain on the camera.

[Bruce Davis]

Posted this before seeing Stan's post above

Hi David, thanks for this information, have downloaded the Edmund optics catalog as pdf, it has a wealth of information.

Quote "There is plenty of light for f5.6 if using HDR" and without HDR is it borderline, is the frame capture speed too slow to use the 2fps without HDR.

For daylight capture to take a 1/20th of a second at f5.6 seems very slow (I may be wrong), is there an ISO value, its not a sophisticated camera so maybe not.

Just thinking allowed :- The sunny 16 rule, at f16 if the ISO is 100, then the shutter speed will be 1/100 second, with f5.6 there is a 3 stops difference, 16mm test lens may not be comparable to a normal digital camera lens.

Stan was concerned about DOF with an aperture of f2.8, David mentioned it was more difficult to focus his 16mm f5.6 lens than the 12mm lens due to more movement/rotation between out of focus to in focus.

Kamel has the f2.8 16mm lens, is he having problems with focusing, DOF, how to do a comparison?

Thanks again - Bruce

Stan thanks for testing the frame rate for USB2, its a considerable reduction.

[Stan Jelavic]

Lots of good questions and inputs. There are two FPS specs. One is the scan speed that we set with the switch and the other is the camera FPS. The camera FPS is the delay after the image acquisition. The sequence is
trigger
image acquisition
FPS delay.
In our case the delay should be as short as possible and is calculated by the USB driver.
My experience with the current camera is that at 2 FPS scan and rez at 1280x1024 (obviously 8 bit)
you need close to 20 FPS camera setting. Otherwise will get smearing for very dark scenes.
The alternative is go slower scan or increase gain (effectively ISO to answer Bruce's question).

The test camera has to be verified at 2 FPS scan and higher rez.

Did some tests with the test camera and 16mm lens. The magnification depends very much on the lens performance and how close you can get to the film with the lens and still get good resolution. The lens I got from David is lower rez and I can go as far as 2500 x 1900
The distance from the lens to the film at this magnification is 41mm. For full magnification the 16mm lens would have to be 28mm.
The 25mm lens for full magnification even closer at 24mm.
16mm seems like a good compromise but will have to check previous messages to see if David pushed it to the limit.

[Kamel Ikhlef]

Hello evrybody,

I have a lot, a lot of focus problems (very little blurry/sharp margin) when screwing up the optics.
Your different tests are very interesting and will be very useful to buy another lens.
I have a very shallow depth of field too limited !

very difficult to make a precise focus, however I think that we can get at least the 80l on the SMPTE clip,
I have not yet succeeded (maybe another better M12 lens).

My actuals M12 optics :
IS OPT_TBN 16C 1/2" F1.8 5MP low distortion
IS OPT_TBL 12C 1/2.5" F1.6 3MP
(both IRCut)

[Stan Jelavic]

I think you need at least F2.8 for 12mm and F5.6 for 16mm Kamel. Can you get sharp focus in the center?
Note on new camera rez:
David Wrote:

There is no way to improve on the sensor, that I can find. You can't have a 4k final image from a 4k sensor. Lenses jump from 16mm to 25mm and few have aperture choices. In the end my image will be a 2k at best. If I put this on DVD it will drop down to 720P (1.3k).

Question to David: Have you tried different spacers with the new camera and your good quality 16mm lens? How far can you go before the corners go soft or you cannot get the focus at all.

Here is the theoretical magnification curve for 16mm:

[David Brown]

Kamel, you should try less magnification. With less lens to sensor distance, you could get better sharpness. I tried different spacers and was able to fill the sensor with full width of the 8mm film. That is the magnification I had with the 72BUOC02. But with the new sensor it was a soft image. Besides that I would have to physically center the lens over the film. This new sensor 37UX226 is larger that the last one.


I don't remember seeing any M12 lenses for a 1" sensor, so 2/3" is the largest. I lost optical magnification by changing the 72BUC02 (1/2.5") for the 37UX226 (1/1.8"). I did increase the image size with more pixels.

After I get the new USB cables, I will try to find some dark footage and test the increased dynamic range. I don't know if that will be handled by the HDR in IC or if I need to process the tiffs another way.

Let me refine my complaint that the f5.6 is hard to focus.

Yes, with the lens stopped down, the depth of field makes it tough to easily get precise focus.
But there's more.

1. It is difficult to grab the lens. The end of the 16mm lens does not extend much below the opening.

2. I was using a lens mount where the threads were very tight. That's good and bad. The lens will not move once in focus, but is difficult to rotate into focus.
I'm now using better thread fit, easy to rotate without a wrench. It can be difficult to hold that focus when tightening the locking ring.

My wrench.

When I open Hawkeye for the last time tomorrow I will measure the lens, sensor, & film distances.

I was keeping notes earlier, but stopped before the end.

[Bruce Davis]

Stan, that's a nice informative chart - thanks, on my Hawkeye with the 12mm TIS lens :-
Film to front of lens = 31.9mm (compared to your 41mm with 16mm lens)
Front of lens to underside of camera board = 26.3mm, that is a total of 58.2mm film to board.

Stan Quote "For full magnification the 16mm lens would have to be 28mm" from the film.
That seems a reasonable distance, only about 4mm closer than the 12mm lens.
Do not know how focusing would work out though.

Not sure what you mean by full magnification, does that mean utilising all 4K X 3K pixels or just increasing the number of pixels from your current 2500 x 1900? Some room is needed on the sensor for "Partial Scan" and "Region of Interest" settings (you know all this anyway).

Kamel thanks for your lens and focusing details.

David thanks for clarifying the focusing problems you had, totally different to my original impression.

Regards - Bruce

[Kamel Ikhlef]

Wow! Congratulations David,..

I am far from getting the quality you get on the SMPTE target!

It would be very interesting to share your final configuration because you get a very good quality.

Yes Stan, I agree.
I would like to buy other lenses but I don't know which ones,

according to your tests and experiments, that you advise me as optics ?
Thank you all for all your suggestions

[Stan Jelavic]

Hi Kamel,
The problem with the lenses from TIS is lack of info. David send us this chart

The issue is that the 12mm lens that I use (TBL 12-2 C 5MP) has this spec:
Specifications Format 1/2"
Resolution 5MP
Mount M12x0.5
Focal Length (f) 12 mm
Iris F2.8
IR Cut Filter Yes

And that is it. Pretty pathetic. Most of the scans I posted are done with this lens. It was a bit of luck there that the lens worked out since the specs are so skimpy.

So for now all that I can recommend is 12mm TBL 12-2 C 5MP lens. The issue is that the optical magnification is 0.5.
David's 16mm lens looks like a good candidate for better optical magnification and higher resolution but based on his tests I do not see what the final number is.
David, correct me if I am wrong but the latest number that you published was

2048x1536 is my final crop size.

This is with the IMX226CQJ sensor.
The sensor spec is 1.85um for pixel density.
2048x1536 = 2560 pixels diagonally
2560 x 1.85um = 4.7mm diagonal image

optical magnification = diagonal image / super 8 diagonal = 4.7 / 7/04 = 0.64
Not that much better than the 12mm lens at 0.5.
Can you push the lens any further David?

I am still looking at the 25mm lens because the chart I published earlier is for ideal optics and single element lens. Will have to check if I can get some more optical specs on the 25mm. Without that I cannot do good estimates.

Bruce - I did some measurement tests with the 12mm and 16mm lenses that I have and the chart that I presented is not very good because it is for ideal single element lens. Will see if we can get some more specs.

[Kamel Ikhlef]

Thanks Stan,
TIS recommended 25mm optics but was not sure of the minimum focusing distance.
It would be interesting to know your settings and configuration (Camera / lens : 12 and 16mm).

To have a good working base for the best possible result.
see the drawing below.

[David Brown]

WOW, that's a huge framerate increase with the USB 3.1, 10 Gbps !

Here is a link a cnAICO 25mm lens. There are more specs than others give.

https://www.aico-lens.com/product/25...ns-ach2525mac/

Yes, 2048 x 1536 is MY final crop size for Super 8, Regular 8 will be smaller.



To capture only Super 8, the image can get a bit larger with the 16mm lens. I didn't make notes and was limited to the fixed stand off spacers and extension tube multiples. It should be easy to test outside the Hawkeye. I used a small microscope mount with a rack focus. Keep adding extension tubes and refocus.

I don't want more magnification without "building" a Y axis stage. Edmund Optics sell those. They are the size of our sensor! More $$$ than I want to pay! Even if I bought one, mounting it would be a challenge.

My Dimensions



I did try a full width (8mm) capture. It lowered the lens to a comfortable focusing position. At that magnification the best image was softer than I wanted. Not the corners, the whole image. There is also a problem with the lens colliding with the panel cutout if it is too low.




I found a film with dark interior & exterior scenes. There might not be any detail in the dark shadows on the film. See what happens at 12 bits (aka 64bit in IC). That should be 36bit, 48bit with a 4th channel.

[Stan Jelavic]

Very good input David. Thank you for for hard work putting this together. I did the measurements on my 12mm TBL 12-2 C 5MP lens setup with the BUC02 camera. Here they are:

film to top of hawkeye board	1.873	inch	47.5742	mm
hawkeye board top to camera board	0.375	inch	9.525	mm
Film to sensor	2.248	inch	57.0992	mm
top holder to top lens	0.407	inch	66.6242	mm
holder	0.56	inch	123.7234	mm

This gives the 1280x1024 resolution and overall very good image quality including corners.

Dave provided the 16mm dimensions.


My opinion is that David's 16 mm is the best choice if you are going for highest pixel resolution.
The 25mm is a problem. Here is why:

Using the lens formula 1/F = 1/a + 1/b , sensor to film distance formula d = a + b and magnification d = a/b and if you plot the d vs magnification,
you get a curve that has a minimum when a = b = 2F
So, the total distance d = a + b = 2F + 2F = 4F
In case of 25mm, the total distance from the sensor to the film is at the minimum at 4x25mm = 100 mm and that is very close to to top of the front panel cutout. Also would require extra long standoffs which makes it less stable. Another issues is the minimum distance to object causing lens distortions. And before I forget, the F stop will have to be around F8 increasing the exposure time requirements.

So anyways that is my angle at it. I am now running a side by side scan that compares the old and new cameras. Will be interesting to see if there is any noticeable difference.

Forgot to mention another advantage of the 16mm lens. This is still to be tested but the 16mm may have less less flare and better contrast. The reason I believe is the lower angle of the light coming in (lens sits higher).

[Bruce Davis]

Hi Kamel, I can only give you dimensions for the TIS 12mm FOV lens setup on my Wolv/Hawk :-


12mm TBL 12-2 C 5MP lens & DFM 72BUC02-ML camera board, 1256 x 942 (4:3) resolution.

Regards - Bruce

[Kamel Ikhlef]

Thank you very much everyone for all your contributions and your work.
All this will allow us to progress well.
your settings will help me a lot

David, Please can tell me the reference and where can I buy your 16mm lens.
I would like to order it. (mine does not give me as good results as you (F1.8 vs F5.6, 5MP vs 10MP)

[Bruce Davis]

Hi Stan, is this lens https://www.edmundoptics.com/p/175mm...ng-lens/27047/ as good as the 16mm that David has from cnAICO.
Min focus 150mm for the Blue Series and 100mm for the cnAICO, does this matter as David's WD is considerably less.

Would it be true to say that (David mentioned above) as the Working Distance is reduced, the contrast of the lens is also reduced and causes a softer image (sharp focus suffers) or is it that it just can not focus at that lower WD.

Is the Blue Series a 10Mbit 200lp lense or something less, it does not really say, that I can find, in the specs.

Regards - Bruce

P.S. it is interesting that your dimensions with the "12mm TBL 12-2 C 5MP lens & DFM 72BUC02-ML" camera board are different to mine.