digitising film

[CS]

Not sure if I have any of my current setup, it's very basic, D800 mounted conventionaly onto my PB4 Bellows, atop my heavy surveyers tripod, AIS micro Nikkor 55mm f/2.8 with the slide copy attachment in it''s normal configuration.  This allows me to get 1:1 or better, so I can actually crop if I want to but I seldom do.  The illumination is a 10 Watt LED floodlight about 2 feet away, screwed to a plank of wood on top of a filing cabinet.  I tape the diffusing plastic screen to the plank about 6 inches away from the lamp.  To turn the light down I add pieces of toilet tissue draped over the LED floodlamp, two pieces usually get the job done, the LED floodlamp is very cool so there is no risk of fire.  It's all very basic, I used to be an avid fan of Blue Peter, a kids teatime TV show where they did projects with bits of paper and glue, they always had the next stage of the project ready under the desk and would pull it out announcy this was "one they made earlier"!.  It became a catch phrase at school.LOL

I have also focus stacked with a particularly difficult slide which was curled, I took a couple of shots focused differently and the result was an improvement, getting both areas of the slide in focus. I have tried taking the slide out of the cardboard holder and mounting it free in the negative carrier, but once buckled or curled they are hard to get straight. Maybe a glass carrier?

I will dig out some images made with this setup so you may see the results.

Thanks, Robert. That pretty well explains your setup, minus the color temp pf the LED lamp. 

[CS]

Double post error.

[Seapy]

Thanks, Robert. That pretty well explains your setup, minus the color temp pf the LED lamp. 

Good point Carl, sorry, I had intended to mention that...

It's just an 'off the shelf' lamp I don't know the colour temperature but I did try using flash, which I have in the past, bouncing it off a white board but I found it too harsh despite trying the trick with the toilet paper! LOL

I compared the images when I started the copying and decided to go with the LED, it's warmer, there may well be gaps in the spectrum but frankly the results look good enough for me. Anyone who has seen them on my computer has been amazed what I have managed to extract from what were effectively rejected slides.  In fact I have used one of the images my father took back in 1965 and was in the reject box as my 'wallpaper' for my iPhone.  It's comforting to carry that with me and to help remember my times with him and his extreme scepticism about computers.

[Roland Vink]

No Roland, you misunderstand, I think...

Thanks, that makes good sense.

I think the explanation is that the latitude for the correct exposure of a slide is very narrow, you have to get it correct when you press the shutter. You can't adjust the brightness in post (unless the slide is scanned). With negatives it is possible to adjust exposure errors when printing, and with digital images it is easy to adjust in post processing.

However the dynamic range of slides is probably greater than most give credit for. Slides generally have high contrast (which is what makes them so vibrant) so mid-tones quickly become dense or washed out. However the response into highlights and shadows is not linear. Deep shadows which appear near black can still contain a lot of detail. And although care must be taken not to over-expose slides, it takes a lot for them to be washed out completely. I have beautiful sunset slides where the sun is a distinct disk against the sky, while a similar image with a digital camera would render the sky around the sun as completely washed out. Digital sensors, with their more linear response are good at separating tones in the mid range, but once they max out, or are near black, there is little you can do, except to use multiple exposures and HDR techniques.

[Seapy]

Thanks for explaining that Roland, it definitely increases my understanding of the issue,  I sort of stumbled on the HDR technique by accident while hunting for an optimum exposure.


OK, I finally found the energy.  I had shingles a few weeks ago, the symptoms have gone but my energy is still sorely missing.

I have gone through some of the images I made from the slides as described above.

This is a good example set of five exposures which I ran through Lightroom HDR process last Christmas Eve.  In the median exposure the flower parts are blown to nothing and the shadow area is devoid of any detail.

The first image is the median, two darker, two lighter, all at f/2.8, the exposure varied by shutter speed, ISO200, the median was 1/13th sec. using D800.

The final image is of course the HDR compo of the first five.

Starting with the median exposure.











This is the HDR version:



Finally the image I have as my iPhone background...  a 5 exposure HDR, not the sharpest but I like it and considering it was taken in 1965 as a chrome slide I think it marvellous.



Thanks for looking.

[Roland Vink]

Nice examples.

I assume the 5 exposures are all a stop apart: -2, -1, 0 +1, +2
I wonder if you could achieve the same with two exposures: -2, +2
Wouldn't there be enough overlap between the two to fill in the mid-tones, or does the HDR software need images close together? It might save you some effort.

Also curious why you shot them at f/2.8, wouldn't f/8 or f/11 be better? Wide open does not give you much DoF if the film is not perfectly flat, and you'll get some mechanical vignetting. Sharpness is usually better stopped down too.

I have a lot of slides I would like to digitise some day, so am interested to see what works...

[Seapy]

Thank you Roland.

1/5 Sec, 1/8 Sec, 1/13 Sec, 1/20 Sec and 1/30 Sec.

I think that is 1/3 stop intervals?  I did try different increments, I found mostly five exposures gave best results, occasionally seven, seldom three

I feel my way to the correct exposure for each slide, checking the histogram and looking at the image to see when there is no advantage in going further.  For really dark slides I move the lamp closer, to help recover brights I attenuate the lamp with tissue.

I feel the exposures are long enough, I don't really want to move from base ISO to give the sensor its best chance of capturing the maximum range.  I suppose I could go to ISO 400 and close the iris one stop but the Nikkor micro performs very well at f/2.8 I don't notice vignetting, in a sense my main aim is to recover the detail lost in the darks and lights, the well exposed parts look after themselves.  I am sure my technique can be improved upon but most of these slides were rejected anyway, some in the box are either suffering other defects or simply of no interest.  There are 1019 images in the folder, they produced 206 final images, a box full, more or less.

The effort required to take three, five or seven exposures is little different, set the camera to rapid shutter repeat and hold the remote release down, Lightroom takes only a few minutes to create the HDR, in Lightroom I found a way to batch process, setting up maybe 5 sets of images, went to make a cup of tea, returning to find them all done.  The latest Lightroom has made this even easier now.  The only downside is storage of the large files. Once processed, the originals could be dumped.  I doubt I will revisit the processing, it's as good as it's going to get.  The images are not so important to justify more time but as I mentioned, I evolved a technique which will be useful when I get to the better slides, yes I will try closing the lens a little and increasing the ISO, maybe also a more powerful lamp with a better spectrum.  I may even try to use the flash again, in a different way, I will experiment with various aspects when I revisit the project later this year.

[PeterN]

I can recommend using the Negative Lab plugin for Lightroom when you scan with a camera. It worked very well for me, especially with a good film (and good light). IMHO it works better than just using PS's feature or mirroring the tone curve. The developer recommends shooting with ETTR and setting the white balance on the film strip outside the frame. 

[Seapy]

Thank you Peter, I will look, at that.