Some further efficiency fixes and related notes.

For an index of all my Film related articles, click here.

In Part II I looked at batch inversion and color negative imaging. Here I look at tethered capture and focusing considerations.

Tethered capture: Lightroom has long supported tethered capture with many Nikon and Canon cameras; the camera is connected with a cable to the computer and after tethered capture is enabled (File->Tethered Capture->Start Tethered Capture) this panel will appear on the display:

The tethered capture information panel.

Comments on user fora suggest many have issues getting LR to recognize supported cameras – indeed I had that issue – and one fix from Adobe goes as follows. Go to this directory in Finder:

This is the file to erase.

Erase the highlighted file – this is for LR6. It will be recreated when you next start LR. After doing that my tethered D800 was recognized fine.

I wrote about tethered capture here and because Nikon uses the fragile USB2/USB3 Mini USB connector I strongly recommend you concoct a strain relief at the camera end, whether a piece of wire, as I show, or the more elegant Nikon USB Cable Clip. Damaging the camera body will not make your day when that sub-optimal connector is yanked/torqued out accidentally.

Why tether? Because you see the image on a large display seconds after pressing the shutter button. Nothing is recorded on the camera’s CF or SD card. The image is sent straight into LR where sharpness and alignment are easily judged. This speeds work flow and you can bang away at those old negatives with abandon, watching the images pop up on the display. You no longer have to remove the SD/CF card from the camera and insert it in your card reader.

Film strip orientation. There a lot of nonsense written about making sure the matte side of the negative faces the lens. I tried both ways, flipping the image for comparison. This image has microfine detail in the child’s pullover and there is no distinguishable difference between matte side and glossy side to the lens. So I use glossy side to the lens as there’s no need to flip the image in LR. Here are enlarged comparator images:

Glossy and matte sides to the camera lens, respectively.

Here is the processed image:

Glossy side to the camera lens works fine.

Focus point: Once inserted in the film strip holder, the film is slightly bowed in a convex shape on the glossy side, the one which faces the camera lens. I do not use a glass film strip holder to try and remove this bowing, as holders which use plain glass can cause “Newton rings” interference on the image. These manifest themselves as rainbow like patterns. There are “anti-Newton ring” glasses available but they are lightly etched to prevent this phenomenon and reduce definition in the image as a result. Further, four glass surfaces mean four more areas on which dust can accumulate.

I use an aperture of f/9.5 on the Micro Nikkor AF-S lens which is the sweet spot where depth of field and definition meet. You do not want to use the lens at full or minimum aperture where definition suffers. Further, depth of field at full aperture is very slim, making critical manual focus that much harder. Given that the film surface is not plane, does it make sense to focus on an area half way out from the center of the image to optimize depth of field and related sharpness across the frame?

In the comparator images below, the first used spot focus in the center while the second focused midway between the center and the edge of the long side of the negative. I used the directional pad on the D800 to move the autofocus point, having earlier determined that auto focus is superior to manual focus. The latter uses the viewfinder LED indicator to show best focus and I have found that nailing that is so critical that deferring to auto focus is easier and more repeatable.

Center and half left spot focus points.

There is no discernible difference between the center and half-left spot focus point images with the grain sharp at all points. Conclusion – center spot focus is just fine at f/9.5.

How fast is the process, a key goal of this project? Read about that here.

Batch processing and color negatives.

For an index of all my Film related articles, click here.

One of the problems when round tripping digital camera film ‘scans’ to Photoshop for inversion from negative to positive is that PS does not permit batch processing. You have to select one image at a time, hit Command-I and then move on to the next. Not a big deal if you are processing just an image or two but a pain if you are inverting dozens or hundreds, which is my goal.

Some research disclosed that there’s a Lightroom plugin which does the inversion within LR, no round trip necessary and preserves the original RAW file in which the image was rendered. There’s none of the file size bloat you get when PS converts the round-tripped image to TIFF format and LR is the only software required, in addition to the plugin. That plugin is named Negative Film Lab. My download yielded a Mac Pro lock up in LR (a quick reboot fixed that – not something I have to do too often with that machine) whereafter I set to converting a handful of images in batch mode. Highlight the ones for inversion, hit Ctrl-N and each image takes some three seconds to invert on my speedy machine.

The original negative scan is preserved in the LR History of the file (no need to erase or stack your original negative scans) and the inverted RAW image shows no size bloat. Plus, if anything, the rendering of tones and sharpness is even better than that I obtained in my original test. The download comes with 24 free tries so you can decide if the $99 license (good for two computers) is worth it. Not cheap, but the gains in efficiency for my project make the investment worthwhile. The plugin appears to be actively supported, with a forum showing many expert users where much can be learned. I have found that no special camera settings are needed, meaning one thing less to remember.

A small batch of 2 files ready for inversion in Lightroom.

The result, the very first image I made on my Leica M3 in August, 1971. A significant improvement on the decades old original scan:

Girl on Train. Leica M3, 50mm Elmar, TriX, D76 developer.

How about color negative inversions? The added snag here is that most color film includes an orange mask which has to be removed and Negative Film Lab does that well. In LR’s Develop module you use the eyedropper on a portion of blank film on the edge of your scan to set the reference color mask reading for the film batch (they vary in color so resample with every new film roll) then perform the inversion with one click.

Color negative image awaiting inversion in Lightroom.

As my ‘scans’ are slightly larger than the film area I set the ‘Border Buffer’ to 12% which automatically crops the image – one step less in post processing work. Mask removal is excellent and some fine tuning of color is all that remains to be done, either in the plugin or using the Develop module controls in LR. Batch inversions for color work the same was as for monochrome.

LR compatibility? I use ‘historic’ LR 6.4, the stand alone, non-subscription version, and Negative Film Lab works perfectly. Apparently it also works in the current subscription version but there’s no way I will ever be checking that. The plugin is recommended for anyone with lots of inversions to perform, especially if they require color negative mask removal.

In Part III I will look at tethering of the camera to Lightroom and focusing considerations.

For archival quality files.

For an index of all my Film related articles, click here.

I am at that point in life where legacies take on serious meaning. Wills, descendants, and so on. One of the precious legacy aspects of my photo archive is the significant number of images made on film. For the most part I scanned these some 20 years ago using a Nikon Coolscan 2000 film scanner. That machine did a good job but the speed of scans would see you to an early grave. Reckon on 5 minutes for a high quality 2000 dpi scan yielding a 5mp file. The quality is good but I thought I could do better. In addition there were many more images I wanted to scan which had missed this first pass because …. well, I got jolly tired of the waiting, and my new Canon 5D was impatiently waiting to be put to use. You see, I had finally given up on film and those Leicas and Rolleis had all been sold. They could not hold a candle to FF digital.

So the idea of doing a better job stuck in the back of my mind and when a used Nikon D800 was acquired it occurred to me that this would make for a good film ‘scanning’ device, with its excellent 36mp sensor. Extensive research on the best hardware setup concluded that the Nikon ES-1 and ES-2 film holder lens attachments would not cut it. Made from plastic with mediocre mechanicals and no lighting system, the whole thing resembled the chore of old with Rube Goldberg quality added to the mix. Alternatives using a film copying stand ($$$) and some sort of cockamamie lighting setup were quickly trash canned. Poor mechanicals, poor repeatability. The goal here it to be able to make 20″ x 30″ prints from the scanned files, not mementos for your wallet. And the whole thing has to be fast – fast to set up, fast to use.

Then I chanced on this JJC ES-2 Film Digitizing adapter. Mostly metal, it includes film and slide holders and, importantly, an LED light source. Be sure to get the version with the LED light source. Once assembled the whole thing makes for a robust whole, with the film held parallel to the D800’s sensor. It is very well made.

This tool calls for a Nikon Micro Nikkor macro lens and two are stated to fit – the 60mm AF-D of old and the latest 50mm Z optic for mirrorless bodies. I determined that the last DSLR version, the AF-S G Micro Nikkor 60mm f/2.8 was the way to go – an easy resale – and picked one up for all of $265 mint and boxed. Nice.

Assembling the whole thing I had this:

All in one piece.

Before going further it bears pointing out that there is a host of variables in this process and the more these can be standardized and saved as presets the smoother the workflow.

Image size: My preference for the AF-S Nikkor over the earlier AF-D Nikkor ($150 or so) was a mistake. Using the smallest possible extension on the JJC, meaning the stock #1 tube and the #3 62mm filter adapter ring for the lens, I could not get close enough to the film to fill the frame. The #1 tube has a lip which prevents removal, but removal of four Phillips screws and replacement of the #1 tube with one of the pair of #2 tubes supplied reduces the lens-film distance to where the 24x36mm film frame almost fills the frame.

The JJC lens attachment assembly dismantled.

With the #2 tube replacing the #1 you lose the limit stop but that is of no consequence as the device will be used with the tube tromboned all the way in.

#1 and #2 tubes compared.

I reckoned that a small amount of cropping was no big deal given the high pixel count of the Micro-Nikkor, so I went with the modified version above. You can see how much shorter the tube extension is by comparing the image below with the first one above:

#2 tube fitted.

Lighting: Perhaps the major consideration in choosing the JJC device over alternatives is the included LED light source, with adjustable intensity. Use it right and film damage from heat is not an issue. See ‘Heat risk’, below. The ten levels of brightness saw me opting for maximum as I wanted to use the fastest shutter speed in the D800 so as to avoid blur from mirror slap. The LED light runs from a USB power source and a connecting cable is included. It uses one of those truly wretched MicroUSB connectors at the LED end so be careful, These are easily damaged.

The D800 shooting menu: Go to Shooting Menu->Set Picture Control->Vivid and set Sharpening to Max (level 9). Hit OK to save the setting. ‘Vivid’ works better than ‘Monochrome’ in my experience.

Exposure: You want ISO 100 (not Auto ISO) and Aperture Priority exposure automation. Set the lens to f/9.5 which is the optimum mix of depth of field and sharpness. Smaller apertures see diffraction limiting come into play and definition falls. Auto exposure? Yes. It works well. Forget what you read about using Manual everything. Using f/9.5 the D800 sets the shutter at 1/180 most of the time, short enough to obviate mirror slap blur.

Focus: Stick with autofocus. Nikon does a better job of it than you can using the ultra sensitive Manual focus indicator in the finder. Again, forget the nonsense you read about using Manual everything. I use the film strip holder with the glossy side of the film toward the lens. The film is slightly bowed so there is an argument to be made for focusing half way off center to minimize the effect of the bowing on focus. However, using f/9.5 I have found that edge to edge film grain definition is excellent, so I simply use the default center spot focus.

Cleanliness: Dirt in, dirt out. Handle your negatives using fresh cotton gloves – not the ones you just used for the oil change on the car – and blow away/antistatic brush away any dust on the negative before insertion in the provided glassless negative strip holder.

Heat risk: The device comes with a clear warning label regarding heat output of the LED light source:

Warning label.

It’s important to note that there are two film/slide holder slots:

Two film/slide holder slots.

Because I get a more frame filling image, and because it’s further from the heat source/LED, I use the one nearest the lens.

How hot do things get? There’s a big difference between the temperature at the LED surface – very hot – and the one at the film negative in the more distant slot. I measured these using a laser thermometer:

Temperature rise though steady state.

So, simply stated, if you propose keeping the negative illuminated – I am using the max power setting above – for long periods of time, the safe way to do that is to use the slot most distant from the LED light source. 141F is too hot. 95F is fine – and that is reached after 10 minutes of inaction. Kodak’s TriX film survives fine in the tropics, so 95F is no big deal. But 141F is 10 degrees hotter than Death Valley, the warmest place on earth!

Horizontal alignment: The holder is excellent, allowing only a small amount of slop with the end images in a strip of six. I turn on the Grid display in the finder (Custom Settings->d->d6) to ensure proper alignment before taking the picture/scan. There are detents as you slide the holder, making frame alignment easy.

Taking the picture/making the scan: No camera support is necessary as the assembly makes for a robust whole. I just rest the assembly on my desk and click away.

Lightroom and Photoshop: I use non-subscription versions of LR (v. 6.4) and PS (v. 12.0) on a 2010 Mac Pro running OS X High Sierra (10.13). Later versions of LR and PS add nothing, but do draw down your pocket book as greedy Adobe seeks to make you a lifetime subscriber. Heck, they probably even charge your estate after you have croaked. Another legacy consideration. What a dishonest business Adobe has become.

Anyway, insert the D800’s CF or SD card into your card reader and import the negative images to LR. You then need to round trip them (Alt-E) into PS where you hit Command-I on each, which converts them to positives. Don’t bother with obscure solutions in LR which would have you messing with curves. PS knows how to do inversions perfectly. File-Save gets the converted images into LR where upon you can erase the negative originals. Crop and adjust contrast/tint/clarity in LR on one image, save those settings for the session (Alt-C) and apply them to all the other positive images in LR (Alt-V). ‘Boom’, as Steve Jobs would put it.

Comparisons: The image comparator in LR makes for easy tuning of the D800 scan with the Coolscan 2000 original:

Nikon Coolscan vs. D800.

An enlarged comparison shows the TriX film grain is perfectly rendered to the edges and smaller and sharper than with the original Coolscan scan. Then again, the image from the D800 is some ten times the file size of the 5mp Coolscan original and I would guess the Micro-Nikkor’s optics are a step up from those in the old Coolscan.

Greatly enlarged comparisons.

Speed: How long does it take you to press the shutter button, move the CF/SD card to LR and round trip for inversion through PS? A whole lot less than the five minutes a scan in the Nikon scanner. And do not be tempted to buy that now 20 year old scanner. No one has parts to fix it and it will probably break, if you do not give up the ghost first. Plus try finding connecting cables which work on your modern computer.

Cost: The most commonly used scanning tool appears to be the $400 Epson V600 flat bed scanner. You are looking at 30-60 minutes (yes, 30-60) for the highest quality scan and will die of old age before getting the job done. Quality will not be all that great – I know, I tried it after careful dismantling of the scanner for proper lens collimation. Ridiculous. Fughedaboutit. Plus good luck reselling that behemoth when you are done.

Here is what this little lot cost me:

• Nikon D800 body – I use that for picture taking, so it’s not an incremental cost factor.
• Nikon Micro Nikkor 60mm f/2.8 AF-S lens, used, mint, boxed – $265 plus tax and shipping. Easily resold for cost. I would prefer the earlier AF-D version for $100 less, see above.
• JJC film scanning attachment – $95
• Total cost – $360

It’s not nuclear physics to solve that one.

Other bodies: I see no reason why this should not work with any number of bodies. For the FF set there’s a large selection of Nikons at bargain prices as everyone dumps them for mirrorless, which they mostly do not need. A good 12mp D700 can be picked up for $250 and is still an excellent daily user. If you use Canon, MFT or other makes, check the JJC listing linked above to see whether your camera/lens combination will yield full frame ‘scans’. The large number of filter attachment adapters provided, covering 46mm through 67mm should do the trick at the camera’s lens end.

In Part II I look at the additional variables which arise when ‘scanning’ color slides and negatives. I will also look at an approach to batch processing of conversions of multiple scans in one step, further speeding this tedious process.