My film, my way, my processing:

[arthurking83]

Hi folks, back again with some more film stuff.

No, I haven't converted to film from digital too!
Never will even tho I have purchased an old film body just for some fun, and am still looking for an old film Nikon(most likely F100, maybe F5), but this new play thing is not so much about using more film, it's more of a fun project thing at the hardware level.... something to do when I feel a need to do something!(if that makes sense).

The film body I recently got, is a 60's Topcon Auto 100, with a 53/2 lens which was apparently broken.
Ebayer said that aperture wasn't working properly, and I felt a need to pull something apart just for the hell of it .. and another reversing lens to have access too, so the $9.99 bid was placed and it eventually found it's way to my house
Turned out that the camera and lens were fine, aperture works fine, shutter sounds a bit slow, but viewfinder is disgustingly filthy with the dust and debris of the insides of the camera's sealing strips and other foam bits.
Eventually got it apart and cleaned most of it out, and now you can see through it, but with the amount of sealing dust and debris, I doubt that it will provide a good light seal, so there's no point in trying to pass a roll of film through it (yet).
The main purpose for this bit of gear was the lens. It comes apart easily and I want to mod it to make my own lenses and stuff .. eventually.... down the track.

So, back to the film topic.
CaptureNX2. seemingly a woeful bit of software for converting negatives into usable prints, until you ask the wrong questions about other software, to the right people.

Posted a question on another forum(as well as here) about any info on how to convert a negative image into a print image using either CaptureNX2, or Lightroom.
The only other two software I have to do this have the easy option of converting an image with the click of a button .. the good ol 'invert colors/image' option.
Both Paint.net and FSViewer have this, but their raw file handling is not very appealing compared to CNX2 as my priority, or LR as my next choice.

Firstly there's nothing wrong with using the colour inversion tools in either FSViewer or Paint.NET, but the process workflow was slow tedious and frustrating.
Because the process of capturing the negs onto the D300 was what I wanted to do, I wanted to dodge, burn and re process the negs to achieve more dynamic range in as many of the images as could be recovered, compared to the scanned jpg images provided by the processing shop.
This is where the bellows with film attachment idea came from, rather than the scanner option I originally wanted to go with two years ago.

Anyhow, I got a reply on how to process inverted colours using LR3, but it didn't work very well for me, so I tried a few similar tricks in CNX2 instead and bingo! .. eventual success.

So this is in part a tutorial for CNX2 and hopefully some info here to help others find a solution in whatever software they prefer to use.

First I capture the neg via the D300 and get this weirdo looking raw file(the raw file is the important part here, and working directly with this raw file made all the other captures of negs a lot easier, and with some careful use of grads and ice cream sticks I could recover a lot better than I could if I only had a tiff or jpg file to work with.
The only step in the process where the jpg file comes into it is final image type for upload to here.

CaptureNX2 with initial negative file captured:

1:

Hard to judge exposure when the negative is in this state, but after a couple of images, it becomes second nature.
But working in reverse is still confusing!

CaptureNX2 colour inversion process:

2:


a badly made exposure of a ho hum negative exposure:

Without the grad, I could only expose the neg either for the sky or the land, and could really only recover the sky in limited amounts but with detail deficiencies in the clouds .. I wanted a better balance

3:

The sky is either blown out with the ground perfectly exposed, or the ground is unrecoverable if the sky is exposed well from the neg.

So I used an old scratched grad to blend the light accordingly from behind the neg.
Darken the ground and brighten the sky huh!

4:


I used a 50W halogen desk lamp to provide the back lighting for the negative, and the grad had to be placed dark side on the dark sections(reverse the exposure remember).
So filtering the dark sections to brighten them up naturally, which then leaves minimal processing work later for the inverted raw file.
If I use +1.5Ev compensation adjustment, the sky comes up ok, but the ground now loses colour from this contrast anomaly.

5:

In these image, I placed the grad(hand held) across in an inverted manner to how you would normally use it.
That is, upside down dark section over the land, clear section over the sky, but I kept getting this washed out low contrast look on the grass no matter how I placed the grad, and eventually the penny dropped and I realised that I was over doing it, and that it needed just a bit less grading.

So it's starting to make a bit of sense now.
I need slightly less graduation, but I think even a very light grad will still not work properly either.

So for the next version, I only placed the clear part of the grad over the ground and had to be quite still in holding the filter with the edge of the filter right along the horizon, so in this next instance of the image, the sky is not filtered in any way, and is receiving the full extent of the backlighing. From memory, I think a clean grad has about 0.3 stops of transmission loss through the clear part of a 3stop grad.
Holding the grad steadily across the middle of the image, with an exposure time of 5 or 6 sec is very difficult, but in many cases near enough was good enough, and eventually I ended up with my version of the event.
For this actual shot, I use a 3 stop grad to get as best exposure as I could at the time(same with the D300 equivalent image at the time of capture), but the lens is so small the grad is pretty useless.
For the actual capture of the neg onto the D300 I coudl have used any piece of glass, but I had the grad readily at hand, and the graduation did come in handy for other images . just not this particular one.

Final result:
6:


Note the steps required for processing the neg into a properly viewable image.

Step 1. the curve had to be inverted. This was the info I was told for doing it in LR3, but for some reason, it didn't work for me, and no amount of tweaking was getting anywhere.
In CNX2, there was immediate but not quite 100% success in inverting the curve graph. you could see the image beginning to form, but other tweaks were needed. And that was simply a matter of experimentation.

Step 2. finally found the last piece of the puzzle in the -100% contrast slider edit.
Have no idea on how or why, but those two tweaks did it perfectly.

Step 3. exposure compensation. Great more so for getting the best possible exposure from the neg at the bellows stage, rather than using it for tweaking the image itself.
And not that it works in reverse(even tho the neg has been reversed now.. exposure compensation still works in reverse.. +Ev to darken, or recover highlights, -Ev to brighten or recover shadow detail.
So I used the exposure compensation to get a feel for what the next exposure should be.
if I needed +1Ev to darken the image, then I'd slow down shutter speed speed by 1 stop to brighten the neg. and so on and on and then once you get the hang of it, it's easier to manage this inverted workflow.

The basic editing steps were saved as a batch file, and then used for all other images loaded into CNX2 to effect the colour reversal process.
One problem was that each image required careful handling exposing and editing, so they couldn't just be loaded in CNX2 and processed with the batch job. The batch had to be applied selectively and with moderation due to the different nature of each slide.

It must be noted tho, exposure is critical for the shadows more so than the highlights. Recovering the shadows always introduced nasty low contrast anomalies, whereas the highlights were easily more recoverable, at over 2 Ev or more. So I had to expose the negatives more for the shadows than the highlights. And work from there.
Never ever played with colour film, and this is my first attempt at any like this, did do media studies in upper high school, but spent little time in the dark room due to the ill effects of the chemicals.

This has been some thing I've been trying to do for over two years now. I could have just scanned the images, and got some high quality 2400dpi scanned tiff files in return, but I had tiff files on another occasion, and don't particularly think highly of them for processing, compared to raw files, and then there was the issue of capturing the negatives with a bit of tweaking.
But in the end, and why I preferred the bellows over a scanner, was to adjust brightness levels accordingly directly from the negative, rather than work on the raw file on the PC.

As for this Ektar 100 film ... well I'm not overly impressed by it.
Produces a red cast in many situations, or alternatively a blue cast in bright sunlight, but I can see it as an ok film for low contrast situations.

Next roll will be something else, but not for a while yet.

My only regret in all of this is that I wished there was a way to store the negative film in a manner that doesn't scratch it easily... even without use!
Yeah film is OK, it used to be a relevant media type, but unfortunately film fan folks .. it's way past it's use by date when compared to digital mediums.

[fabian628]

interesting use of grads to get the most out of the exposure. might be more difficult if there isnt a distinct line between highlight and shadow.
I have found ektar 100 working very nice, nice warm colours, possibly your light source (or expired film?) causing colour cast. I use a scanner to get mine digital so I am not too sure what is applied in that software either.
an ektar 100 shot from me, i love the warmness of this film for landscape + i am a sucker for slightly warm images

[fillum]

possibly your light source (or expired film?) causing colour cast.

Been a while since I looked at a colour neg, but doesn't the film base have an orangey cast after development? For the neg "digitized" above, a straight inversion will result in a blue image. There is probably a way to filter out the cast when "digitizing" but probably easier to do in PP. I don't think a white balance setting in camera would have enough latitude to remove it (although I've never tried).


Cheers.

[arthurking83]

Thanks for your responses guys.

I found that overall, compared to the same images taken with my D300 and Tammy 28-75, that the Ektar exposures were much more red.

Of course an small compact Rollei 35 with it's dinky 40mm lens from a million years ago is always going to render it's images differently to the D300+Tammy combo, but just a point I noted for now.

I've been looking for a cheap SLR to suit a couple of the lenses I have and do more comparisons at some point in the future.

Currently got a bid on a F100, and it's looking pretty at $109!! and a few hours to go, but I don't like my chances (yet again! )

For WB, I can't actually set any in camera .. there is no setting warm or cold to suit, so this is why I had trouble in using the raw process and then opening the tiff image in Paint.net.

It was taking forever to try to get a decent WB value as my process in CNX2 was to find a reasonable grey point.

I found it in the D300 image, and simply used that same point, on the captured neg image.
ie. the same point in the image that results in a good WB setting with the D300 image has the Ektar image looking more red. It's only very slight, but still there.

WB was set with a grey/white point dropper, depending on image.
it was so much more fluent doing it in with a raw file in my raw editor, and I quickly found a good practise.
WB alterations from one shot to the other with the film wasn't as dramatic as it is say with the digital medium.

For what it;s worth, the eventual values for WB turned out to be 0.23 for the red and between 2.80-3.05 for the blue channel, when using the dropper tool.
This is a vlaue of variation from the standard values of 1 for each.
it's pretty wild, when you compare those kinds of adjustments to the average change of about 0.05 either side of 1.
Any more than those kinds of values of change in wb values from the standard of 1, when using the dropper tool ..... and the image takes on an 'alternative' look!

My problem for a long time was getting the gear I wanted to make this as easy as possible.
Not 100% there yet, and I may need to look at other lens options again if I can't find a filter step up ring adapter for the lil Nikon lens.
At the moment, I'm wrapping the gap between lens and negative with variations of 240GSM photo paper, cut to suit and folded, and an old cloth nappy(cleaned of course )

I know I'm losing a bit of contrast in the air gap between lens and film, as there isn't a perfect seal.(I think).
if I can locate this odd ball adapter ring(34.5mm to any size that allows me to get to 52mm), then I can use the bellows between lens and film to give it light sealed environment.

I've 'scanned' all the images so far, and not 100% pleased with the results, but with a couple of the images I've managed to get a good correlation between the Ektar and the D300.

Also, I have to note that the images captured both on the Ektar and the D300 reference images are pure snaps. The Rollei doesn't really allow for accurate framing, and the film was brand spanking new, but in a sense free(as a pseudo gift) so all I wanted was the reference point to work on doing it right later on down the track.