A: There are definite advantages to film, particularly color negative film. We believe that because it has been somewhere between difficult and impossible to get good conversions of color negative film to digital images, the decline of film in recent years has been unduly accelerated. ColorNeg and ColorPerfect level the playing field so that valid comparisons of digital and film can be made. Having provided the tools we leave most of the comparisons to others - image quality, resolution, subtlety or richness of color. However, one key advantage color negative film has over current digital is exposure latitude. Blown highlights, that trademark of digital camera photos, is not a problem, nor is it necessary to bracket exposures (when possible) to be sure you can reconstruct what you want. Meanwhile, there are billions of already existing color negatives - most people have a few and many photographers have large archives. No one can go back and take those images again using a digital camera.

As to whether and when digital will win out over film, consider that for the past 50 plus years color negative and color positive films have coexisted. They both appear to do the same job but positive film was just better for some things than negative film and vice versa. It is very likely that as long as film continues to do some important things better than digital the two will coexist.

A: The short answer is because it doesn't work properly with film negatives - what it does is completely wrong for inverting a color negative. The technical answer involves some simple arithmetic. Photoshop inverts pixel values by subtracting the scanned negative's pixel values from a constant value 255 ("white") to produce pixels for the positive image. The way film actually works, to produce pixels for the positive image it is necessary to divide a constant exposure value by the negative pixel values. Imagine the trouble it would cause if you subtracted instead of dividing somewhere on your tax return! Of course the full story is a little more complicated, involving photographic gamma and a few other factors, but that is its essence. See our CFS-244 for the mathematics and physics involved or see the question on why 16-Bits/Channel for a simplified version. This isn't a problem just with color. We have users that apply ColorNeg or ColorPerfect mainly to B&W negatives, where they have found it gives the accurate tonal range and values they have not been able to get otherwise.

A: No. Some do, but some others are based on a method called "profiling." You probably are using another form of profiling to get your display colors to display correctly and get your printer output to match your display. Technically, profiling looks like it could work with color negatives and it can, sort of, under exactly the right conditions. The problem is that profiling was specifically designed to take an image that correctly displays in one form, as on a CRT, and by going through a common intermediate form, make it match as well as possible on a printer, an LCD display, etc. Unlike with positive color slides, with color negatives there is no correctly displayed image to start with. There are additional technical problems such as the difficulty of using available profile forms to approximate negative film characteristics, and the extended dynamic range of color negative film. In inverting color negatives, the profile tool is being used to do a job for which it was never intended, rather like trying to drive a nail using a measuring tape - it can work if you are not overly worried about the quality of the result. The profiling method is normally completely automated, but what it does is very similar to the what we describe on our web page on Color Negatives which was written before we produced the original NegPos. As to whether there are color negative inverters besides ColorNeg and ColorPerfect that operate correctly, there may be, but in the five years that our earlier plug-ins have been available on our site we have never been informed of any or found any others mentioned.

A: Our usual answer to this is a terse - and correct - "No!", the orange mask is no problem at all if you are going about inverting the negative correctly. It just automatically disappears using ColorNeg or ColorPerfect whether the mask is orange, red, bluish, or whatever. The technical reason for this is explained in CFS-244 (page 9-11). Basically the mask is a color cast exactly equivalent to placing a colored filter over the image and is automatically removed in the same way and with the same tools that any other color cast in the image is removed. The orange mask is a scapegoat incorrectly used to explain away the poor results from using one of the incorrect methods for inverting a color negative. When an incorrect method is used to invert the negative it can distort the mask so that it no longer acts as a simple filter. So, for incorrect inversion methods the mask can appear to be a serious problem, but the real problem is that the inversion is being done incorrectly.

A: As nearly as we can tell, the main culprit is the scan - see the answer to "trickiest part" just below. Next in line is that the film is mis-characterized - the film may be different than the standard due to processing, storage or other reasons. Although ColorNeg and ColorPerfect provide several tools to help characterize problem and unknown films, it takes skill and experience to learn how to use them properly. Then of course all negatives are not created equal. It is never easy to get a good result from a bad negative.

A: In our experience the trickiest part is in getting a good, clean 16-bit/channel linear scan as a starting point. If you want really good results from color negatives you have to spend the time required learning how to get the proper scan from your scanner system. That is very easy with some scanners and can be quite trying with others. We have found that scanners and scanner software very often make adjustments to scans without telling you they are doing so. These adjustments are always aimed at trying to make the scans of positive images look better than they really are. While the changes may be harmless and barely noticeable in positive images, they can do serious damage when the scan is used for a color negative. For instance, a small scanner "adjustment" to the deep shadows may be barely noticeable in a positive image, but in a negative those shadows become highlights and cause serious distortion of color throughout the image. This is not just a problem with inexpensive scanners. Some of the most expensive scanners fudge the data in an effort to impress you. Fortunately, figuring out how to turn off all the tweaks and color management to get good scans needs to be done only once. We have a page on linear scanning as well as an older scanner page which shows how ColorNeg users did it for some different scanners and software. Even if your scanner is not listed this will give you an idea of what to look for.

A: Color negatives are normally designed to have a low photographic gamma. In common terms, they would be called "low contrast." The dynamic range of color negative film (often called DMAX) normally also is fairly low, meaning the blacks in the negative are not extremely black and the whites are, well, orange. However, the low gamma (contrast) effectively compresses the dynamic range of the original photographic scene so that it fits within the narrower dynamic range of the negative. In fact, the compression is normally more than is required so that the negative actually has a dynamic range greater than can be expressed in any normal positive display image created from it (and a greater dynamic range than a digital camera image would have had). Color print paper to be used with color negatives is designed to have a corresponding high gamma - "high contrast" - that effectively re-expands the dynamic range back to what it should be. Those who have worked with color negatives in wet darkrooms are aware that the color negative normally has considerably more highlight and shadow detail than a straight print will show. Dodging and burning in - both of which amount to simple exposure controls - can bring out substantially more detail. This ability is lost once a color negative is digitally converted to a positive. Part of digital color negative inversion is selecting an exposure (lightness) for overall interpretation of a negative but a good conversion must also not lose the highlight and shadow details that are present and would not be present in a digital camera image of the same subject. See our page on A 16-Bits/Channel Highlight Detail Secret for more on this topic.

A: The short, non-technical answer is no. The purpose of color management is to make color consistent for an image whether viewed on CRTs or LCDs or printed on paper. For input devices, color management tries to insure that a displayed target image will look as much as possible like the original target when the scan or digital camera image is taken under prescribed conditions. Color management is not intended to produce good color under arbitrary conditions, only to help retain good color once it has been produced, and then only under carefully controlled conditions. If an original image has poor color, color management will duplicate that same poor color.

Technically, color management systems hold the image (literally or virtually) in a special intermediate form and use profiles to convert the image into and out of that intermediate form. Each device, CRT, scanner, printer, has its own profile to achieve this. The profile assumes the device has constant, very specific characteristics. You probably know that as a CRT display drifts with time it is necessary to periodically generate new profiles for it. If you even use a different type of paper in a printer it will require a different profile. In addition, profiled devices generally have a well-defined and known set of primary colors, or in the case of printers, ink or dye colors.

Color negatives do not fit well into that pattern of use. When you use a color negative you first take an image of a scene. The film uses at least three silver halide layers sensitive to red, blue, and green. (Nowadays there are usually more than three layers.) Some manufacturers provide spectral sensitivity data for these layers and some do not. After the negative has been developed, the results of the color separation into this set of additive primaries red, blue, and green are expressed as three layers of subtractive primary dye images, cyan, magenta, and yellow. If there are more than three layers there may be more than three different dyes involved. The spectral behavior of these individual dyes is rarely given by the manufacturer of the film. This negative is then scanned, where it is yet again separated into red, blue, and green primary colors using sensors or filtered sensors in the scanner - but the red, green, and blue primaries will be different than those of the film. The spectral sensitivities of a scanner are at least difficult and often impossible to find. Added to this is the fact that film manufacturers make it clear that the spectral sensitivity and other data that they do provide are typical rather than being exactly applicable to any specific roll of film and that the data will vary depending upon both manufacture and processing. In principle one might be able to calculate or more likely experimentally measure a profile for a color negative film when using a specific scanner. In practice it is very difficult.

Technical history comes to our aid in this. Although red, green, blue primary colors do differ, most of the technology has had very similar goals and targets in this regard. Thus if we are reasonably careful, the RGB image we get from this mixed system will fit with the RGB systems in common computer usage. In Photoshop assign whatever profile works best for you to preserve the good color as you continue. The uncertainties involved make it so that you will not be any more or any less accurate whichever of the standard working profiles you use. It becomes a matter of preference. To get the really good, accurate color in these circumstances we believe that the best approach is to use a calibration technique like one of the calibration methods built into ColorNeg. These calibration techniques take into account the general physical behavior of the film and other components in the process and result in a form which accounts for and is forgiving of variations in the process. However, for most cases calibration is not really required to get quite acceptable results as long as the film is known or otherwise characterized.

Some readers may have noticed that most of the above appears to apply to color positives (slides) as well. There is a major difference in that with the color slide, a color managed scan can produce an image file so that the display will appear the same as the original slide - that cannot be a useful goal with a color negative. Note that although the color managed images will match the color slide, the slide itself may not be a particularly good image of the original scene.

A: Blackpoint is an adjustment of the very darkest areas in an image, normally used to drive the darkest areas in a image to pure black. This adjustment also has a pronounced effect on apparent image contrast. For an image with color integrity it is frequently the best way to adjust image contrast. Be aware, however, that Photoshop does not do a blackpoint adjustment correctly. The standard Photoshop blackpoint adjustment is the Levels tool, using the lower, shadows slider. It does not properly correct for system gamma in making this adjustment, resulting in considerable color distortion well up through midtones see CFS-243 (Pages 12-15). Our plug-ins do a correct blackpoint adjustment, as does the Photoshop RAW input for digital cameras. Technically, the requirement for a blackpoint adjustment primarily results from the non-ideal behavior of image sensors and viewing situations, whether film or electronic. Particularly in working with film there is an interplay between blackpoint and gamma. In general and for film in particular, gamma is a measured characteristic of the image sensor and stays the same for a particular type of film. Blackpoint, like exposure, may vary according to the individual image. Since both the blackpoint adjustment and the gamma adjustment will change apparent image contrast, it is best to leave gamma fixed for any particular type of film and adjust blackpoint, taking care to leave proper shadow detail.

Photoshop uses an adjustment in going between color management profiles which Adobe has chosen to also call blackpoint (as if there were not already enough confusion on the matter). They have used this adjustment for quite a few years now to "correct" what they see as a flaw in color management profiling. Finally this year Adobe has issued a document describing this correction. (Look for AdobeBPC Blackpoint.pdf on the Adobe website.) We are not yet prepared to comment on this document.

A: Gamma is number that describes how responsive a film or image sensor is to lighter spots and darker spots in a scene. As the lightness increases, sensors with lower gammas do not change their response as much as sensors with higher gammas. When properly used, gamma is a property of a film or an image sensor and so remains the same for different images from the same source. The Photoshop Levels tool "middle-gray" adjustment is the same form as gamma and thus tempts (even instructs) the user to treat gamma as an adjustment tool to be set differently for individual images. This is wrong.

The three layers of a color negative normally each have a different gamma, with each gamma being less than one, meaning that the density range of the color negative is compressed. Color negative films are "characterized" by determining these gammas for the film, either from manufacturer's characteristic curve data, from calibration, or from estimation. A gamma adjustment can then be made to expand the compressed density range of the positive image back to normal. The same gamma adjustment should be made to all normal negatives of that type of film and gamma should not be further changed for different images.

System gamma (called gamma c in the plug-ins) is an old idea that has lingered into the present. It was invented to solve a hardware problem in early television. CRT display tubes do not respond conveniently to an electronic driving signal - they slowly brighten with increased signal level and then rapidly become brighter only as the signal approaches its highest level. It was found that this could be changed to an evenly increasing brightness response if a gamma equation was used to adjust the signal voltage before it was applied to the CRT. When computers first started storing images to be displayed on CRTs, for faster operation the pixel values in the images were stored so they were already "gamma adjusted" to create the correct response of the CRT. By sheer coincidence the gamma adjustment warped the pixel values in just the right way so that 8-bit pixels were slightly more than sufficient to insure the eye would see continuous tones from such image whereas if the 8-bit pixels represented light intensities that were not gamma adjusted, the image would be uneven and banding would result. Essentially all digital images are still stored and normally edited in gamma-adjusted form. Although necessary if 8-bit pixels are the standard, gamma-adjusted images have caused no end of problems, largely because image editor programs such as Photoshop often do not take into account that the image has been gamma adjusted where they really should.

A: An image with color integrity will either have no color cast in the shadows, mid-tones, or highlights, or will have the same color cast in the shadows, mid-tones, and highlights so that the color cast can be removed using a simple colored filter. This definition applies to an image that has a consistent illumination source or within an image, to areas which have a consistent illumination source. An image with color integrity is easily adjusted with regard to color and lightness. Many standard operations in Photoshop can destroy color integrity. The practice of "color balancing" an image by using the Levels control to do a "white balance" followed by a "middle-gray balance" is once of the worst, and can destroy color integrity right from the beginning. A primary goal of our plug-ins is achieving color integrity and maintaining it once it has been achieved. See Color Integrity in Digital Photography for a detailed explanation and examples.

A: The short answer is that image adjustment of any sort eats away at the ability of an 8-bits/channel image to resolve color, Really extreme image manipulation like inversion destroys this ability. But basically this is a numerical question that requires a numerical answer. Keeping it as simple as possible, the correct way to produce pixels for the positive image is to divide a constant exposure value by the negative pixel values. Let's say we are using 8-Bit pixels. 8-bit pixels take on values 1 - 255 (0 is excluded). For simplicity say the exposure value is 255, which is very convenient because it uses the whole range of 8-bit values possible on the negative. When the negative pixel is full white (255) the positive pixel is full black, 255/255 = 1 and when the negative is full black (1), the positive is full white, 255/1 = 255. But look what happens when we take the next whitest negative pixel value, 254. The positive pixel value is 255/254, which is 1.004. Since pixels can take only integer values, that is still 1. You have to take the negative pixel value all the way down to 170 before you get 255/170 = 1.5, which could be rounded up to a positive pixel value of 2. You don't get a positive pixel value of 3 until the negative pixel is down to 102. In fact, by the time you get down to a negative pixel value of 10 you still are only getting a positive pixel value of 255/10 = 25.5, or 26, which still is deep shadow on the positive. So, pixels ranging from 255 clear down to 10 in the negative result in positive image pixels of only 1 to 26. There remain only negative pixels values of 1 through 9 to produce all the tones in the image from fairly deep shadows (27) up through all the mid-tones and all the highlights (255). Obviously that is not nearly enough. To get a really good negative inversion - in fact, to get a negative inversion that is correct and usable - you need the negative pixel values that fall in between the 8-Bit values of 1-255, and that is why a 16-Bit/Channel scan is necessary.

In the real case there are mitigating factors, including gamma and a better understanding of the exposure value that make the situation not quite as bad as the simple example above. But the example does illustrate what happens and does not greatly exaggerate the real situation. Our tip A 16-Bits/Channel Highlight Detail Secret has a neat demonstration (Colorkid1) of the power of 16-Bit/Channel images.

A: In most cases the gamma-blackpoint equation fits quite well through the region of greatest interest. In any event, film characteristic curves are a measured quantity, subject to variation due to manufacturing, storage, and processing. It has been our experience in such situations that an exact fit to one measured curve is often less reliable than a fit to a simple form that that is known generally to describe well data of that type. It provides meaningful comparison with other measurements of a similar nature and has physical meaning as well. For example, the relationship between the gammas of the three layers normally responds in a known way to variations in processing. Additionally, the gamma portion of the equation normally is characteristic of the film type while the blackpoint portion can vary from image to image. This fundamenat separation might not be possible with a more complicated form.

A: There are ColorNeg and ColorPerfect users doing this with reasonable success. However, setting up a system that works well is definitely not easy. The primary problem again is getting a good 16-Bit/Channel linear image to work with. Many digital cameras will deliver 16-Bit/Channel RAW images, but RAW image converters are normally designed to work with positive images and may apply tweaks to make their results more impressive. These tweaks may be relatively harmless in positive images but very harmful to color negatives.

There is a section Using a Digital Camera as a Scanner in our scanners web page with information and suggestions. If you wish to try using a digital camera as a color negative scanner we strongly suggest reviewing that information first. In particular, we explain why the Photoshop RAW converter will not work well for color negatives but we do detail a method whereby it can be used for tests.

A: Every now and then someone will ask us about this. This seems like a clever idea for producing conventional color prints from precisely controlled digital images.

Unfortunately, it is very difficult - perhaps impossible - to get satisfactory results using this technique with available printers. When this idea comes up, our suggestion is always to forget about the negative. Just print the positive digital image as a positive color transparency and then use a conventional wet process to print that positive transparency on positive (reversal) paper. It is possible to achieve as much control via this route and, given equivalent care in preparation, the final result will be better than could be achieved by going color negative route. But that also translates to mean if the positive route does not work well with your equipment, be assured that the color negative route would be even worse.

We are aware that this seems to run counter to conventional wisdom. If you have a color film transparency to print and want the best possible result, the recommended traditional method is to make an internegative and then print this on color negative print paper. But here we are starting with a positive digital image rather than a positive film transparency. It is possible - although tricky- to produce a color negative digital image using ColorNeg. The problem comes about when printing that image to a transparency. Making the negative requires the same sort of division- rather than subtraction-based inversion that is required when making a positive image from a scan of a color negative, which we describe elsewhere. Just as it requires a 16-bits/channel scan of the color negative to make a good resulting positive digital image, it is also true that if we want to print a digital color negative that has been produced from a digital positive image we will need to have a printer that will accurately reproduce 16-bits/channel. In general digital printers are not even close to producing 8-bits/channel accuracy on a fine scale. They require that you step back a bit to get something that appears to approach the effect of 8-bits/channel accuracy in color. So, while a printer might possibly produce a positive transparency that could be satisfactorily printed, the same printer will produce a color negative transparency that when printed will produce poor highlights or shadows or both.

There are other problems as well. Many (most) color printers use pigments rather than dyes and the performance of pigments in color transparencies is usually quite poor since the pigments are by definition opaque. Another major problem is that printers routinely make pronounced enhancements when printing color positive images. These "enhancements" are intended for positive images and will do serious damage to a printed color negative. Unlike the similar problem with scanning color negatives, in the case of a printer there is rarely any provision for turning the off the enhancements.

So, while there may be some hope of printing a color positive transparency that can be directly printed in a wet-process darkroom, producing a good, printable color negative is very difficult at best - approaching the impossible when using standard digital equipment. If we at C F Systems were interested in taking digital positive images to the wet process darkroom with high quality results, we would explore the possibility of making a single large digital print, probably on paper, possibly as a large transparency, and photographing that on a larger format color negative film under carefully controlled conditions.