The Scanning Project-8: Scanner Terms and Important Specs

 There is no sense buying a scanner or making scans on a scanner that does not have the right tools for the work you intend to do. In this post I will cover the main specs to check and what I consider important benchmarks for a variety of projects you may have in mind. As an introduction to the terms, following is a brief description of each and why they are key elements in determining image quality. These will be covered more extensively within the context of the scanner workflow discussions in following posts.

Resolution: In essence, resolution is a prime determinant of how large you can create a print from the scanned image. Much like the megapixels in a digital camera, the larger the file produced the larger the potential enlarge-ability. Of course, other matters come into play, including the quality of the scanner lens and sensor, but given all other things are equal (which is rare) the higher the resolution you choose the bigger the file produced, thus the greater degree of enlargement.

Scanner resolution is expressed as pixels (or dots) per inch. The important thing to watch for is that the resolution is expressed as Optical Resolution. Some scanners are misleadingly labeled as having high resolution when actually what is being stated is the potential interpolated, or resampled resolution capability. This means the image is crunched, if you will, to extrapolate a higher resolution than the actual optical resolution.

This is an image size dialog box from a scanner 35mm slide that was cropped slightly in scanning. The scan was made at 3200 dpi. The dialog box on top shows the scan when it has been converted to printing resolution, resulting in an 11x16 image size. Note that this can be edged up a bit using resampling at this stage, a topic covered in the workflow sections in later posts. 

Simply put, the higher the optical resolution the larger the image file. When selecting a scanner make sure it has a minimum of 4800 dpi; 6400 dpi scanners are becoming more affordable and are a good choice if you work mostly from 35mm film. But keep in mind that when scanning at 6400 dpi every sharpness and other flaw in the original will become quite apparent, like looking at yourself in a magnifying mirror in the morning. In addition, scanning at 6400 dpi can result in some very large image files, perhaps beyond your needs. It also can point out flaws in the lens that you used to make the photo that might not be apparent even when viewing the film through an 8X loupe.

Here's a scan from a 6x7 medium format negative. Note the larger image print size that can be achieved. This scan was initially made at 3200 dpi.

Bit depth: Again, as with digital cameras, you can choose the bit depth of the image, which means how many bits of information are captured within the three color channels of the sensor. You often have a choice of bit depths, such as 8-, 24-, 36- and in some scanners 48-bit.

Think of the difference between 16-bit RAW file and 8-bit JPEG images made in your digital camera. The 16-bit file simply has more image information available, which means you get a lot more potential out of the scan. But just as with resolution, the higher the bit depth the larger the image file, so consider the tradeoffs. For the most part, 36-bit will do the job, and for some end uses 24-bit and even 8-bit (for black and white images for web) may do just fine.

Dynamic Range: This is one of the main scanner specs to check. Dynamic range is a term you might be familiar with in the context of your digital camera. At the camera’s lowest ISO settings that range might be 10-11 f-stops, meaning that tonal values within that span can be recorded; the dynamic range of any camera inevitably falls as you raise ISO. In scanners, the measure is not f-stops but in something called Optical Density and is within a scale of 0.0 (white) to 4.0 (black).

Watch for how the scanner’s specifications (read capabilities) are expressed in Optical Density terms. A scanner that’s Optical Density falls between 2.0 and 3.0 may work for some prints and “quick” tasks, but generally there will be a loss of highlight and deeper shadow information—in other words, it is quite flat. Scanners in the 3.2 to 4.0 range are more suitable for the kind of work you might be interested in, with the higher number offering a better dynamic range capability. In other words, the higher Optical Density rating will bring out the most tonal (thus color richness as well) qualities of the image.

Flatbed or Dedicated Film Scanner?

There are two main types of scanners for photographers and those working with photographic images. One is a flatbed that can handle most film formats (35mm up to 8x10) and prints up to the platen size, the most common being 8.5 x 11”. Flatbeds can also be used to scan tintypes, daguerreotypes and glass plate negatives. A flatbed can be versatile, but if you have only 35mmn prints and negatives a better choice would be a dedicated film format scanner, commonly for 35mm but also available for medium format (120) film. 120 film scanners can cover all the 120 formats, including 6x6, 6x9 and even 6x17 negatives and slides. (Larger format dedicated scanners are strictly for professionals and will not be covered in this project.) Note that flatbeds can usually do a good job with 120 format and a reasonable job with film as well. How well depends on the scanner’s specs and build.

Thus, the scanner you choose is highly dependent on the type of images you want to scan. If you are a family photographer with lots of old snapshots and prints and only a few boxes of 35mm film, go for the flatbed. If your collection is primarily composed of 35mm film then go with a dedicated film scanner.

Quick Guide: Scanner Specs for Film and Prints

Here is a quick rundown of the scanner specs I believe are best for your work. Of course, budget and the volume of work you plan to scan should guide you, but scanning via a low-spec scanner is generally a waste of your time.

Optical Resolution: 4800 dpi; better: 6400 dpi

Color Bit Depth: maximum, 48-bits per pixel, 36-bit is usually fine

Grayscale Bit Depth: 16-bits per pixel

Optical Density: 3.4 minimum; best: 4.0

Light Source: LED

Platen Size (for prints): 8x10”; better: 8.5x11”

Film Holders and Masks: Dependent upon your formats, 35mm to 4x5” film

Bundled Software: Covered in the next posting

Next posting: Scanner Software

The Scanning Project-5: Editing Color Negative Film

There were literally hundreds of types of color negative film, with various speeds, brands and emulsion stocks on offer. When these were mass printed in photofinishing labs, the equipment processing and printing the film identified those various stocks by their edge coding and made automatic color balance decisions accordingly. Sometimes this worked well and sometimes it didn't. 

But knowing the brand and type of color negative film you are scanning can be quite useful. You might be able to read the brand of film on the perforated strip. (See Posting 4 for a link to a site that ID’s numerous Kodak films by their codes. You can also Google a particular film or type. For example, googling Kodacolor--Kodak color negative film--will yield the years of manufacture and film codes, which can be helpful.) This information is used when applying film profiles to the scan and different profiles can be profoundly different in terms of color balance and contrast. However, knowing exactly which “vintage” (emulsion stock number) of film you are scanning is near impossible and there were variations (though they can be slight) depending on which emulsion stock number the film might have been.

This color negative film, scanned as a positive to show the orange mask, is not an easy "read" in terms of color, since both color and density are reversed and sit within an amber layer.

Here's an image of what the negative would look like when the amber layer is removed. By studying the positive (below) you can see how "reversed" colors (actually the complement to the positive colors) look. This takes some practice, although the easiest way to figure this out is to simply scan the negative, as a scanner will remove the orange mask effect in the process.

Here's the positive made from the scan. This exposure was made on Kodak VR 200, a film sold starting in 1984 and discontinued about 1986. The perf code on the film is CL, which when cross-referenced on Wikipedia led me to the film brand and type, which I used as the profile when scanning. 

While batching the types for scanning sets by film type in your initial edits is a good beginning, it is likely that you will have to fine tune the color balance yourself during scanning, which is not a difficult task, though it can be time consuming.

One problem of course is being able to see through the color mask that was incorporated in virtually all color negative film. This orange/amber coating was incorporated to enhance color reproduction, but also makes “reading” color negatives tougher than slides or black and white. Being a negative, the colors are “reversed", if you will, to their complements when the image is printed (color slide film is processed so the reversal takes place during processing via chemical or light exposure). All in all, the only way to see what colors you have (unless you have a very trained eye) is to scan the negative itself, which will eliminate the mask and make the colors positive.

When high speed films came out in the mid eighties they gave a newfound freedom to low light shooting. They did have their problems, however, as most were daylight balanced, many were quite grainy and contrasty, and I have found that the years have not been kind to their stability. This photo inside a Chicago blues club was shot in 1985 using Kodacolor VR 1000.

This is a scan from the print I got back from the lab at the time. Being daylight balanced, the film recorded an overall amber cast. I am unsure if the lab attempted to rebalance the light but I accepted it as the final output of the image at the time.  It was enough to be able to shoot in such low light without flash. 

When I scanned the negative I had to rebalance the color myself as there was no matching profile provided in the software for VR 1000. Knowing that the image was heavily influenced by artificial light I added blue and cyan and played a bit with the contrast. My aim was to produce an image that looked like it was shot with tungsten balanced film. 

However, color reversal and the orange mask are not the only issues with color negative film. In my experience, color negative films of older vintage suffer from density loss, color shifts (some are worse than others) and emulsion deterioration. There are three main color layers (CMY—cyan, magenta, yellow) within the emulsion itself, with some layers usually suffering more than others. Frankly, manufacturers did not do a good job insuring that color negatives would be “memories that live forever”, as the advertising claimed.

Of course, how you stored those negatives is also an issue, and if you stored them in shoeboxes you will probably have dirt and scratch issues to contend with as well. Not to worry, however, as color negative films can be put through dirt and scratch removal software algorithms, which will help. You can also manually clean the film by rewashing and immersing the film in PhotoFlo, or passing the film very carefully through a chamois cloth.

One option is to remove the worry about color by converting problem or poor color negatives to black and white. While this should be a last resort, it can help maintain the image if not the original color itself. Doing so you can deal with contrast and density issues and not worry about color balance.

Most software will offer a color profile of many types of commonly used color negative film. While these should be considered ballpark algorithms, they are a good place to start; in my experience you will certainly spend more time balancing color negatives film than you will when working with color slides.

So, batch the negatives as best you can according to film type and brand, check each frame with an 8X loupe and be prepared to hunt and peck to find the right color balance. We’ll cover profiling in the coming post on color negative step-by-step scanning.

My guess is that the vast majority of color negative films were shot with affordable cameras and in many cases the negatives are lost or simply tossed as being thought unnecessary. These photos are often the storehouse of precious family memories and should be scanned to protect and preserve them for future generations. I am always amazed at finding family photos in the form of black and white prints from over a hundred years ago that are still in very good shape, but mass produced color prints from only thirty to forty years ago that quite clearly are not. Scanning color prints (those with or without negatives) using a flatbed scanner is an easy matter. This print is from the 1970s and had already shifted color towards a magenta cast, alleviated somewhat by changing the color bias when scanning. 

Unfortunately, many people who shot color negative film of family and travel memories usually used other than stellar cameras and lenses. The plastic or hybrid lenses used in these cameras were notoriously soft and more than not automatic exposure modes, or guesswork, was used when shooting.

When scanning such images, expectations should be kept in check, although getting good scans for web sharing sites and even photo books—if image size is kept “reasonable”—allows these memories to be preserved. Frankly, in some cases scanning from the original snapshot print, if it’s not too far gone, may be a better strategy for such uses.

The Scanning Project-4: Editing Black & White

With color slides you can make a fairly quick assessment of quality and edit out hopeless causes (or dupes, etc.). This is not always the case with black and white. True, you do not have issues of color to deal with, and this in some ways makes it easier to separate the good from the bad and the plain ugly. You do, however, have to understand and work with density, the "thickness" or "thinness" of the negative. Those who have done darkroom work will have the advantage here in recognizing a "good" negative, but even if you don't you will quickly gain an understanding of this issue.

Here's an example of a "thin" negative. There is very little shadow detail and the contrast required to bring detail up will be excessive. If you definitely need to scan such a negative do not attempt to fix the contrast in the scan, but do so later in an image editing program. 

You also should consider quality issues such as the possible deterioration of the film base onto which the emulsion is laid (cracking, splitting, severe curvature due to roll-up storage). Other possible destructive forces can be poor processing (failure to clear hypo properly, thus staining and even loss of original converted silver) and the use of destructive storage materials or containers. 

This is a "thick" or overly dense negative that resulted from overexposure and/or overdevelopment. My suggestion is to batch these and work on them as a group as you will be using certain tools to hopefully cut through that density, although larger and unpleasing grain and contrast problems will be present.  Not all dense negatives are lost causes, but they will be challenging.

 As one of my mentors, Bob Schwalberg, used to say, “density defends density,” which means that old negatives which have been processed and exposed properly and have a rich range of tones will always do better in the long run. Well, there isn’t much you can do about that at this point, but it is self-evident that rich negatives will always yield a better scan than “thin” negatives, since the thin get thinner as time goes on. There is a another caveat: if the material was grossly overexposed and/or over-developed (thus making for a “thick” negative), many scanners may have trouble penetrating the silver (density) and harsh and possibly useless scans may result. There are ways around this (multi-scanning is one), which will be covered in the step-by-step scanning posts, but opaque negatives will always be bothersome.

Here's an example of an excellent negative that has been properly exposed and developed and shows no deterioration. This will yield a very good full tonal print and poses little challenge to get right. Shot on Tri-X 400, the scan was made using that film's "profile", a topic that will be covered in the step-by-step scanning sections in later postings. Photo copyright George Schaub.

If you have made satisfying prints from some of your negatives in the past you might think that scanning the “positive” print will deliver a better result. The problem is that darkroom printing always and necessarily compresses the full tonal values available in the negative. If the negative is truly poor or if you have only prints remaining this is the obvious way to go, but in my experience a decent negative scan will usually allow you much more creative leeway (and yield a better result) than a scanned print from that negative.

Of course, if only the print has survived and you don't have access to a negative you will be doing flatbed scanning to make a digital copy. Post scan processing can often do wonders with original prints, but in all cases you will get better results should you have access to the original negative, given that it is in good shape. This photo of "doughboys" from World War I was found in an old postcard album purchased in a flea market.

Depending on the film used or the way it was processed, you may run into grain issues. This is something you have to live with, although there are certainly ways to reduce unwanted grain when scanning and later. However, keep in mind that all grain reduction methods (called “noise” these days) entail a certain amount of image softening. Check your negatives with an 8X loupe to see if that grain will be ruinous. (It can of course be “charming” as well.) Keep in mind that scanning etches grain unlike anything you might have seen in your silver printing days, especially if you printed via a cold light head. Scanning, like a magnifying mirror, can be cruel.

Another issue to consider when editing is dirt and scratches embedded or etched into the emulsion surface. Unfortunately, many dirt reduction procedures and software programs do not work on black and white (although they do work on color film, except for Kodachromes). Retouching this stuff can be tedious, but it can be done, and in some cases you might be able to remove the objectionable material from the film itself by rewashing or using a kind of fluid as a kind of diffusion mask on the frame when scanning, a technique we’ll cover later. You might consider using a chamois cloth to clean surface dirt off the negative, and while this can help you should be very careful so as not to scratch the film in the process.

Every film type has a certain characteristic curve, and identifying the type of film you have is very helpful when making scans. Using this "profile" when scanning can eliminate a number of steps in the process, particularly with contrast selection. This is a Kodak Tech Pan negative of ice flows on the Long Island Sound. The film yielded very low grain and high sharpness. However, it tended to develop out with higher contrast than other panchromatic films, so setting its profile when scanning can eliminate guesswork later. Photo copyright George Schaub.

 Like all film, each type of black and white film has a “characteristic curve” and gamma, which means a certain contrast, density and relationship between tonal values. For example, Panatomic-X and Ilford HP5 have a distinctly different character. When editing, it is wise to batch all similar film types together. Even though there may be variations in different stocks (dates of manufacture) this allows you to set up various responses to maximize the film’s character when scanning and certainly helps in the time spent in setting up each frame. Many times a film type is not shown on the frame edge. There are web sites that track these numbers (http://www.taphilo.com/Photo/kodakfilmnumxref.shtml is one of them) although you may still have to make an educated guess in some instances. I’ll go over the importance of ID’ing film types in the step-by-step posts later.

There are other types of black and white film you may have in your collection, including black and white positive film, chromogenic film, IR film, high-speed (at the time) surveillance film, Polaroid “Instant” B&W, etc. While some of these can be challenging they can be managed, so do not disregard them in your selection process. However, they don’t usually respond in the way a “regular” panchromatic film does to scanning, and require some special setups, so batch these as well in your edits.

There are many tools you can use when scanning black and white, as well as those you can apply when processing the image in image-editing software later, so do not be discouraged if you have some marginal negatives or prints. I encourage you to work with those negatives you think are lost causes and see just how far you can take them. However, understand that they may require considerably more effort and work. Start your work with "rich" negatives and then move on to those that are more challenging.

Next posting: Color Negative Film Edits