Sensor-Pan system

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

ray_parkhurst
Posts: 3413
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Sensor-Pan system

Post by ray_parkhurst »

For well over a year I've been contemplating building a Sensor-Pan system for use with the Rayfact 3.5x (RF3p5x) lens. Integral with this system concept is a FF camera to minimize the number of tiles required to shoot a Lincoln Cent. The diameter of a US Cent is 19mm, and at 3.5x this projects a 66.5mm diameter image onto the sensor plane. My measurements of the RF3p5x show it has a high quality image circle of at least 82mm, so is perfect for this application. The FF camera would let me continue doing the 2-wide x 3-tall panoramas I've been doing with APS-C and 95PN at 2x for the last few years, just on a larger scale and with higher resolution.

Unfortunately after testing several FF cameras for this application I have still not settled on one for various reasons, and have continued to use the APS-C / 95PN Subject-Pan system for my high-res coin photos. The system has worked fine, and produced some very nice images, but I ran into a snag on an important coin. The problem is that even though the subject + lighting are moved together, the lighting still changes as the camera + lens are panned across the panorama. For coins that are less brilliant, this has not been a significant problem, and the stitching program has been able to handle the transition between the tiles well enough. But a more brilliant coin has more contrast and variation of reflected light intensity from tile to tile, posing a bigger tiling problem. Unfortunately, the Subject-Pan method failed on this coin, leaving me in a quandary.

After reviewing options, I decided to move forward with the Sensor-Pan method, but with APS-C / 95PN. This allows me to use existing equipment to validate the Sensor-Pan method for coins, so was far more accessible and less risky than creating a whole new system, buying a new camera, etc.

It took me 3 days to put the system together and get it functioning, and I'd like to share with you all my first Sensor-Pan image of a Lincoln Cent:

https://easyzoom.com/image/239899/album/0/4?mode=manage

This system also gives me an interesting capability: wide-range image circle evaluation. I configured the XY panning system to allow panning across up to a 100mm image circle, so I can now more effectively test lens coverage on new or old lenses which may be candidates for the Sensor-Pan technique.

dickb
Posts: 342
Joined: Mon Jul 05, 2010 1:54 am

Re: Sensor-Pan system

Post by dickb »

Very interesting. Do you have a description of the Sensor Pan system you used for this? Some sort of automated XY system to move the camera through the large image circle of the lens, I presume? Do you connect the lens and camera with a bellows or another system that allows the necessary movement?

ray_parkhurst
Posts: 3413
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Re: Sensor-Pan system

Post by ray_parkhurst »

dickb wrote:
Sat Jan 09, 2021 6:20 am
Very interesting. Do you have a description of the Sensor Pan system you used for this? Some sort of automated XY system to move the camera through the large image circle of the lens, I presume? Do you connect the lens and camera with a bellows or another system that allows the necessary movement?
My Subject-Pan system config is:

XY base (2xKR15) with subject stage and lights
Z column (KR20 + idler) with camera and lens on bellows

I modified the system to Sensor-Pan as follows:

XY base (200mm square) with subject stage, lights, and lens on bellows lens standard
Z column with camera on bellows camera standard

This means I'm only moving the camera for stacking (where I was moving both camera and lens in the Subject-Pan system), and keeping it fixed for panning.

I've built the system so I can easily convert from one system type to the other.

dickb
Posts: 342
Joined: Mon Jul 05, 2010 1:54 am

Re: Sensor-Pan system

Post by dickb »

Thanks. An interesting approach. So the focus is done by varying the extension between lens and camera. Is telecentricity a limiting factor for lens choice here or are your subjects so flat that that doesn't become an issue?

ray_parkhurst
Posts: 3413
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Re: Sensor-Pan system

Post by ray_parkhurst »

dickb wrote:
Sat Jan 09, 2021 8:55 am
Thanks. An interesting approach. So the focus is done by varying the extension between lens and camera. Is telecentricity a limiting factor for lens choice here or are your subjects so flat that that doesn't become an issue?
Telecentricity is always an issue with stacking, but the 95PN is for sure not telecentric, so not a big difference either way. My total stacking distance is typically 800um, with 50um or smaller steps.

Interestingly though, it's the lack of telecentricity that forced me to change the configuration, but not due to stacking issues. If the 95PN were telecentric, I'd have fewer issues with lighting as I pan across the subject, and I could continue using that method. With Subject Pan, I am shooting 6 separate images, each with its own lighting angles. I move the lights with the coin to minimize this, but the angles still change. With Sensor Pan, I am shooting 6 tiles of a single image, so lighting is perfectly consistent from tile to tile.

Lou Jost
Posts: 5943
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Re: Sensor-Pan system

Post by Lou Jost »

Great to hear you have this working. I have been doing this as well, and there are some very easy ways to do this for those of us with more limited ambitions. Limited shift adapters exist for MFT, APS, and FF; of course one needs to use a lens with a larger image circle than the sensor size. These are off-the-shelf techniques which work really well for small shifts. Infinity objectives can be used for this as well, using larger-format tube lenses.

If larger image circles are available, you can use a view camera with a FF shifting back. I've had great results with this.

Ray's subjects, being round, are a little easier than subjects that go out to the corners of the frame. Digital sensors can produce color casts and artifacts when the angle of the incoming light is not perpendicular to the sensor. This is partly due to the micro-lenses on the sensor which are mis-oriented when light comes from far off axis. It's also partly due to the light having to go through the thick sensor filter pack at an angle. So there are some limits to this method. It works best with long lenses, and works very badly with wide-angle lenses.

ray_parkhurst
Posts: 3413
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Re: Sensor-Pan system

Post by ray_parkhurst »

Thanks Lou, great input.

I decided to test the usefulness of the system for lens image circle measurement. My first couple of tests of some microscope objectives showed disappointing results, so I mounted the 85mm Macro-Varon to see if there was an issue with the system. I was able to go out to a 90mm IC before the bellows started complaining from the strain. Here are shots taken each 10mm on subject out to 30mm from center ("8" on the scale). Magnification was 1.5x, so each 10mm on subject is 15mm at the sensor plane. Thus "5" on the scale is 45mm from the center of the sensor, or 90mm IC. My lighting leaves much to be desired, as it is set up for a 19mm dia coin, but this at least shows the system capability. Looks to me like the 85MV is softening a bit at 60mm, and a bit more at 90mm:

Center of IC:
IMG_2765_1.JPG
30mm IC:
IMG_2766_1.JPG
60mm IC:
IMG_2767_1.JPG
90mm IC:
IMG_2769_1.JPG

Post Reply Previous topicNext topic