Getting back to silver gelatin on glass

A first attempt at making an anamorphic pinhole image inside a marmalade jar, and it won’t be the last!

At the back of a cupboard I came across a bottle of SE1 Emulsion. It was the remains of the emulsion I had used when I last coated a batch of glass plates, way back in July 2016 and even then it had been several months since the bottle was first opened. From my notes on the box, the emulsion was diluted with 16% Photo-Flo and amounted to about 50ml. I wondered if it would still work.

With Worldwide Pinhole Day 2018 just six weeks away, I have also been trying to think of something new (at least to me) to do this year. How about preparing and coating the inside of a glass marmalade jar with a pinhole in the lid, to create an anamorphic image around the inside of the jar that could be viewed from the outside? It’s not such a crazy idea but subject matter would need to be carefully chosen and exposure might be tricky.

emulsion and marmalade jar

I made up a small quantity of gelatin (1g in 200ml water) with chrome alum hardener (4ml of 2% solution in water) to sub the inside of the cleaned glass jar which I did by pouring in the hardened gelatin solution and flowing it over the entire area by gently rolling and rotating the jar before pouring away the excess. The jar was left aside to air dry, ready for coating with emulsion.

To protect the emulsion from light, I prepared a two-part light-seal from black card and Duck Tape. The jar would fit into one piece and a second piece with a hole in the top for access to the pinhole, would slide down over the first and the hole sealed to the lid of the jar with black PVC tape. A pinhole was punched/drilled into the centre of the lid and a piece of black PVC tape used for a shutter. From a high resolution scan of the pinhole I measured its diameter as 0.37mm.

light seal for the marmalade jar

Now sometime over the winter I had tidied out some of my early glass plate attempts, cleaning off the images by immersing them in very hot water and bleach, scrubbing them clean and then re-subbing them ready for re-use. The 50ml of emulsion that I had would cover quite a bit more than the marmalade jar so I used the excess to coat a few glass plates too: four each of 5×4 (for my Intrepid plate holders) and of 5×3 (for my pincam constructed for WPPD2014). Coating the marmalade jar was achieved in the same fashion as I have already described for subbing it. The coated jar and plates were left in a cool, dark place for a few days to dry completely.

With everything prepared, I was now ready to test out whether my well out of date, diluted emulsion would still work. I planned to set the jar on the floor of my old garden shed, thus exposing an image of the underside of the shed roof on the base of the jar and the shed interior around the sides. By way of a check I would also expose a plate of the outside of the shed, in the WPPD2014 pincam.

the marmalade jar pincam in position in the shed

From past experience I rated the emulsion at ISO 5. For the shed interior I metered EV(100) 8 and thus a 60 minute exposure, and for the outside of the shed EV(100) 13 which gave a 2 minute exposure.

For developer I used Ilford Multigrade diluted 1+19 at approximately 20ºC. The marmalade jar image was developed, as for subbing and pouring, by pouring developer, stop bath and fixer in turn into the jar and swilling it around to cover the surface. Fixer was poured back into the jar and left to stand until the image cleared. The plate from the WPPD2014 pincam was tray developed as normal. Both the jar and the plate were then rinsed in fresh water for a couple of hours before being left to dry.

Unfortunately the marmalade jar image was completely overexposed. The fact that the emulsion is black at least tells me it was still ‘active’ and it is just possible to make out some faint detail in the dried image.

grossly overexposed, it’s not what I’d been hoping for!

On the other hand the plate from the WPPD2014 pincam was perfectly exposed!

a perfectly exposed plate, 2 minutes on ISO 5 rated emulsion
the plate image from the WPPD2014 pincam scanned and inverted

I’m glad to have decided to expose a plate as a check. It tells me the emulsion is still good and that I need to be more careful with my metering for the marmalade jar. For the marmalade jar exposure I think I was fooled by the amount of light entering via the windows and door (which was left open during the exposure), having metered for the dark recesses of the shed. Unfortunately I only prepared one jar so it will be a wee while before I can have another go … but there’s still plenty time before Pinhole Day!

Negative scanning experiments

I spent a day scanning and rescanning at different settings, the same negative to discover what works for me.

My ongoing frustration with viewing scanned images on my Light Moments blog and Flickr with my MacBook Air led me to spend a day experimenting with various methods of scanning and processing 35mm negatives.

Whether or not the method I use to scan and process my negatives has any bearing on their being viewable on any particular device is unlikely to be determined by these experiments but it might at least give me an idea of what output quality is achievable from my scanning setup.

My scanner is a flatbed Epson Perfection 4990 Photo, capable of scanning negatives up to 10″x8″ at up to a claimed optical resolution of 4800 dpi and a Dmax of 4.0. My scanning software is Epson Scan as supplied with the scanner and my editing software is Serif Labs’ Affinity Photo.

Poor weather during the week left me effectively snowed in at work for three days with little work to do. Fortunately I’d taken my camera and a couple of rolls of film with me and was able to spend some time taking pictures of the snowy scene in which I found myself. I’ve chosen a single frame from the processed negatives to illustrate the results of my experimentation.

The film is Kentmere 400 developed in Ilfosol 3 at the standard dilution of 1+9 and at 20°C, Ilfostop and Ilford Rapid Fixer. Scanned frames were output to TIFF files for processing. The files uploaded here were all resized to 1200dpi wide JPEGs at 85% compression quality.

For my scanning experiment I started with a 16-bit greyscale scan at 1200dpi to a TIFF file.  Exposure and image adjustment settings were the standard auto settings provided by Epson. The resulting TIFF was so rough that I didn’t bother attempting any further processing. Scans 2 and 3 were made with the same settings but at 2400dpi and 4800dpi respectively. At full size they show some improvement in resolution which is just discernible here, but not in image quality.

Scan 1: 16-bit greyscale @ 1200dpi. Epson Auto settings. TIFF file size 3.4MB.
Scan 2: 16-bit greyscale @ 2400dpi. Epson Auto settings. TIFF file size 13.5MB.
Scan 3: 16-bit greyscale @ 4800dpi. Epson Auto settings. TIFF file size 53.9MB.

Based on what I could learn from the first three scans I decided to use 2400dpi for the next three.

Scan 4 was also a 16-bit greyscale scan but with manual over-ride of Epson’s auto adjustment of the histogram. I set the black and white points to just left and right respectively of the ends of the histogram, the grey point value to 1.00 and set the output to stretch the histogram from 0 to 255. I also unchecked the unsharp mask setting. The output gave me full histogram values to work with using Levels in Affinity Photo and to my eye produced a much more acceptable result.

Scan 4:: 16-bit greyscale @ 2400dpi. Histogram manually set in Epson Scan, adjusted in Affinity Photo. B&W conversion in Affinity Photo. TIFF file size 12.8MB.

Scan 5 was made just as Scan 4 except as a 48-bit colour file. Not only does this give me the option to make adjustments to levels but also allows control over the conversion to Black & White and the opportunity to emulate the use of filters on the camera.

Scan 5: 48-bit colour @ 2400dpi. Histogram manually set in Epson Scan, adjusted in Affinity Photo. B&W conversion in Affinity Photo. TIFF file size 40.5MB.

Pleased with the progress I seemed to be making, Scan 6 was also 48-bit colour but this time with all Epson settings turned off: No auto exposure or colour management, no auto histogram, no unsharp mask and no auto setting of the scan marquee. I was able to manually select for scanning, a little more of the negative than had been automatically selected by the Epson software, hence the slightly larger file size. I was also sufficiently pleased with this one that I took the time to spot and straighten the file once in Affinity Photo.

Scan 6: 48-bit colour @ 2400dpi. All Epson Scan auto settings turned off. B&W conversion and all adjustments made in Affinity Photo TIFF file size 46.3MB.

I was really very pleased with this. So much so that I repeated it at 48oodpi just to compare the resolution. The 4800dpi at full size is just noticeably better. I also made a slight change to the brightness – not sure I made the right call on that but like everything else that’s a subjective judgement!

Scan 7: 48-bit colour @ 4800dpi. All Epson Scan auto settings turned off. B&W conversion and all adjustments made in Affinity Photo TIFF file size 185.1MB.

As I said at the top, none of this is likely to have any bearing on my MacBook Air problems (I’m beginning to see the problem being something to do with it, either hardware or operating system) but it’s been an interesting day making these comparisons. For the extra effort and disk space, scanning at higher resolution in 48-bit colour and with no Epson software intervention  makes a huge difference to what is achievable. And the beauty of making edits in Affinity Photo (and I guess any Photoshop-like software) is the flexibility and ability to go back to make adjustments.