10 February 2019

Processing Film Negatives on Your Computer

Few people use film for photography today, but it's not uncommon to have boxes of negatives around. Digitizing these images from negatives is challenging and many people settle for scanning prints. There's a much better option, though: Either scanning the negatives or using a digital camera to photograph the negatives. Fortunately, the publishers of photo processing software are making it easier to deal with scanned or photographed negatives.

 Click any small image for a full-size view. To dismiss the larger image, press ESC or tap outside the image.

Some scanners have software that converts scanned negatives to positives, but scanners are slow. The fastest way to obtain an image from a negative is with a digital camera and a mechanism that's designed to hold the film. If you're handy with woodworking tools, you can build a device to do this. Not being so handy, I bought one.

The Novoflex device can be used to copy slides as shown here or negatives.

In either case, you'll need a lens that's capable of filling the viewfinder with a negative. Although many camera manufacturers make macro-focusing zoom lenses, these are usually not sufficient for this task. Instead, you'll need a true macro lens. Canon makes 50mm and 100mm macro lenses, Nikon makes 60mm and 105mm macro lenses, and third-party manufacturers make macro lenses in similar lengths. My preference is for lenses in the 100mm range. Regardless of which you choose, these lenses are not inexpensive.

Photographing slides is easier because they are positive images, meaning that the colors are correct, bright is bright, and dark is dark. Negatives flip that. Bright is dark, dark is bright, colors are reversed and color negatives have an orange cast.

Applications such as Adobe Lightroom Classic (at the right) and Alien Skin Exposure X4 (below) have added controls to work with negatives. The process is fairly basic in that it simply involves inverting the tone curve so that light parts of the negative are shown dark and dark areas are shown light. We'll get to the orange cast in a bit. A normal tone curve for a positive image runs from the lower left (dark source, dark output) to the upper right (light source, light output), but this must be reversed for a negative image.

Two types of black and white negatives exist: Those that use silver and those that use dyes similar to what is found in color negatives. Silver-based negatives are neutral, but dye-based negatives often display a color cast. (Most of the images in the top two rows of the image are from dye-based film stock.)

These images were taken during a tour sponsored by the Travel and Tourism Bureau of the Ohio Department of Economic and Community Development for travel writers. This series is from the Marietta area.

An image's tone curve represents the relationship between the image file and the output. I've reversed the tone curve for this exmple image by selecting negative from the options.

The image was also underexposed, so there's minimal detail in the darker areas and it's flat overall.

Adjusting the contrast, shadows, mid-tones, and highlights produces a much better image that captures the natural mood of the scene more accurately.

That's all that needs to be done for black and white images from silver-based film. A color cast will be present in most dye-based negatives and that may be something you'll want to fix or, if you prefer the sepia-like coloring, leave alone.

Color negatives are more complicated, primarily because they have a strong orange cast.

Unlike black-and-white negatives, color negatives contain no silver and in addition to having reversed colors and densities, there's that orange cast that has to be dealt with. The orange hue was added to film to provide masking that corrects flaws in the overall color rendering.

The orange hue, when colors and density are reversed, becomes a blue cast. Getting rid of the blue is probably easier than you might expect. Virtually all photo processing applications have an eyedropper tool that's used to select an area that should be neutral gray. This can be a gray card (if you included one) or something else that's neutral. You can't use pure white or pure black, though. If the photo is a close-up of a person, the whites of the eyes will usually work well. Cement blocks are often acceptable and pavement also works well.

To get a starting point, I used the surface of a highway. The image was taken from an airplane on a somewhat hazy day, but the overall colors are accurate.

Having first applied approximate color correction to all of the selected photographs, I was then able to fine-tune the color correction on individual images.

One point to keep in mind, when working with scanned negatives is that presets won't work the way they do with scanned slides or direct digital images. Presets, whether in Adobe Lightroom Classic or Alien Skin Exposure X4 expect a positive image as a starting point and will convert the image back to a negative,

Also, some of the controls may appear to be reversed. This varies from one application to another and the only way to determine how the controls work is to use them. Once you understand how the controls work, using them will be easy. Then you can rescue all those old negatives and make some memorable prints from the good old days.

Why Not Just Scan the Prints?

It would be a lot easier to just scan the prints, so why not do that? That's a good question and scanning prints is an option. If the negatives no longer exist, that may be the only way to create digital copies of old pictures, but it's not ideal.

Prints exist in three general categories:

• Glossy finish: Scanning these kinds of prints will yield the best results. Glossy prints pick up fingerprints, though, and photofinishers sometimes selected other options.
• Matte finish: Scans of matte prints won't be as crisp or clear as those from glossy prints. The matte finish was popular because it didn't show fingerprints as much as glossy prints.
• Textured finish: This is the worst possible choice for scanning. Textured prints didn't show fingerprints, but the texture pattern is nearly impossible to eliminate when the prints are scanned.

Prints should always be the last resort though, and for several reasons:

• Prints are cropped. Commercial photofinishers always cropped in a bit from the edges of the negative to ensure that the automated process didn't have any blank areas in the prints. In most cases, this isn't a problem, but it's better to start with the entire image.
• Color may be incorrect. The automated color balance might have produced prints with bad colors. Going back to the negative provides a second chance to get the color right.
• Detail will be better from negatives. Prints were made by shining light through the negative and focusing it on the photographic print paper. Each step of the process has the potential to degrade the resulting image slightly. The results will be even worse with matte finish or textured paper.
• Contrast suffers in print scans. Every photographic medium has a contrast range (refferred to as D_max) that describes the range of brightnesses the medium can reproduce. Color slide film has a relatively high D_max, color negative film a slightly lower D_max, and photographic paper and even lower D_max. What this means in practice is that prints will have considerably less detail in bright areas and in dark areas than what was present in the film.

Scanning slides or negatives will always yield the best possible quality and using a digital camera makes the process much easier and faster than using a film scanner. If you like the idea of resurrecting your old images but don't want to do the work yourself, there are companies that will do the scanning for you. Unfortunately, most of them provide only jpeg files instead of tiff files and that limits your ability to work with them in a processing application such as Adobe Lightroom, Thumbs Plus, Alien Skin Exposure X4, or other such applications. This is a good do-it-yourself project and ScanYourEntireLife has a series of step-by-step lessons that can help you learn how to do the job right the first time. Membership may seem a bit steep ($200 per year or $20 per month), but you'll be able to work through all of the lessons in less than a year — so that's something to consider.

Short Circuits

The Scammers Are Becoming Smarter

Even though scammers are smart enough to create phony messages that look almost right, it's still easy for someone who's paying attention to spot a phony in less than 10 seconds.

When I received a message from "Apple" this week, it looked legitimate, but it immediately raised several questions.

• I knew that I hadn't bought anything from Apple in the past week, certainly nothing for nearly $70.
• Apple always lists the name of the device used to make a purchase. The purchase location showed "iPhone" and I do not own an Apple phone.
• I didn't recognize the name of the item I had supposedly purchased.

But there's quite a bit more that identifies this as a scam.

1. The message should have come from apple.com, not apple-support.com.
2. The link should have gone to apple.com, not apple-authenticate.com.
3. The apple-authenticate site was created at the end of January.
4. The apple-support domain is registered to someone in St Petersburg, Russia.

And if that's not sufficient, there's even more. Here's the top section of a real receipt from Apple.

1. Note the presence of an Apple ID. This was not present on the phony receipt.
2. Note the presence of a billed to section that includes an identifier for the type of credit card used, the final 4 digits of the credit card number, my name, and a complete address. The phony receipt had none of this.

So the initial examination, sufficient to raise serious questions, takes only a few seconds and confirming the presence or absence of key bits of information takes no more than half a minute. Researching the WhoIs information takes longer, but it isn't necessary. I perform this step only for my own amusement.

It's also unnecessary to examine the routing header of the email, but this can also be amusing. This message started with an unknown sender with the IP address of 23.82.128.153. The IP address is not assigned to Apple, but to a hosting operation in Arizona. Either the operator of the hosting service has some severe security lapses or (more likely) one of the domains hosted there has a mis-configured form that gives scammers access to the site's SMTP server.

Clicking the link would take the victim to a form that looks like Apple's log-in page. The mark would then enter a user name and password. Depending on how the crooks set it up, their form would either die with an error or pass the victim on to Apple's site after extracting the user name and password. Either way, the crooks have the victim's user name and password for Apple and from there they would be able to do a lot of harm.

Windows 10 Finally Takes the Lead for 2019

After 3 years, Windows 10 now has more users than Windows 7. Many enterprises stayed with Windows 7 because Windows 8 and 8.1 didn't have features that IT managers needed. Microsoft scrapped plans for Windows 9 and released Windows 10 in July 2015. Early in 2018, Windows 7 had about 42% of the desktop market, Windows 10 had about 35%, and both Windows 8.1 and the MacOS had around 6% according to NetMarketShare.

NetMarketShare provides monthly reports based on data collected from websites about the operating system on users' computers. In January, Windows 10 was slightly under 41%, Windows 7 was at 37%, Windows 8.1 had dropped to a bit over 4%, and the MacOS was under 3%.

Microsoft claims that Windows 10 is running on more than 700 million devices, a number that includes PCs, tablets, phones, and Xbox consoles. That's considerably lower than the company's initial estimates when Windows 10 was released that the new operating system would be on one billion devices by mid 2018. Windows phones failed to catch on and more users than Microsoft expected decided to stay with Windows 7.

StatCounter, another company that provides similar analysis (and is used by TechByter Worldwide) shows Windows 10 with 53% of the market in January, Windows 7 at 35%, and Windows 8.1 at about 7% for users of Microsoft operating systems. When other operating systems are included, Microsoft has 75% of the market for all versions of Windows, the MacOS has about 12%, Linux and Chrome each have less than 2%, and about 10% were unrecognized.

The differences between StatCounter's number and NetMarketShare's numbers are the result of differing analysis techniques. NetMarketShare measures daily unique users and StatCounter measures total traffic, so multiple visits can skew the results somewhat.