30 March 2000 (Updated 14 July 2001)
Let us begin with one test of a camera's resolving capability, lines per mm. A test target would have alternating black and white lines of equal widths. There are other tests, such as MTF, which when coupled with resolution tests can provide a more accurate overview of the performance of a lens system, but they are more complex and harder to compare results.
Average point and shoot cameras (lens ratings of very good, not excellent or outstanding as rated by Popular Photography magazine) manufactured today will be able to achieve around 50 line pairs / mm of resolution, that is, in each millimeter on a 35mm film frame, it can clearly resolve 50 black lines between 50 white lines distinctly. Professional 35mm cameras coupled with the best 35mm fixed length lenses and film (such as the superb Fuji Provia 100F slide or Kodak Tech Pan B&W film), can easily record 100+ lp/mm of resolution in optimal studio conditions.
Knowing that a 35mm film frame is 24x36mm in dimension, we can calculate what
this equates approximately as a digital image.
24x36mm * 50 lp/mm * 2 pairs per line = 2400x3600 pixel digital image equivalent, that's a 8MP (megapixels = million pixels) image.
From this alone, we can say on theoretics alone that a digital camera must be capable of taking true 8MP images in order to match the resolving capability of film. Lower resolution digital cameras will not be able to resolve fine hairlines, eyelashes, and such details in as crisp detail, if at all. Better 35mm cameras and film, or even medium and large format cameras, can and will record much sharper images than average 35mm consumer cameras.
Of course, because film is analog and a continous recording medium, unlike digital sensors which have discreet sensing cells, film can record in-between pixel information whereas digital cameras cannot. However, film possesses grain, and unless you are using one of the best films available (Provia 100F or Tech Pan), the larger grain particles will overwhelm the recording of super-fine details. Digital cameras can have better sensitivity and dynamic range than films, and can record much more information (not necessarily resolution, but color information) than film can.
When printed, quality and sharpness of prints depends heavily upon the operator of the print labs, and average supermarket and store prints will produce poorer prints due to lack of proper focus and adjustments. Naturally, professional printing produces the best film prints. Digital prints can have similar problems if done through an online photo print lab (although their quality control tends to be higher). At home, prints made to a photo printer will be as good as the printer, and can usually be depended upon to produce the same quality (whether great or not) all the time.
For professionals requiring exacting detail and quality, contact the makers of professional 5000x5000 pixel 25MP+ digital camera makers such as Betterlight.com and PhaseOne.com for further help in selecting a suitable digital camera. These digital cameras, albeit costing $5,000+ USD, will easily match and exceed the capabilties of current 35mm, medium and large format film cameras, and have replaced MF/LF cameras in numerous product, model, and professional photo studios.
In general, for consumers, 2+MP digital camera images can match and exceed the quality of 35mm film prints made from consumer level 35mm cameras such as point and shoots on 4x6" prints, and 3+MP digital images for 8x10" prints. (Naturally, larger print sizes will require a larger digital images to maintain print quality that has minimal pixelization and artifacts.) This may or may not be true for professional and other photographers who have taken additional steps to ensure maximum film image quality and have seen the differences quality equipment and processing allows.
Why is this the case despite the theoretical calculations above suggesting 8MP+ is needed?
There are many factors in the capture, development and printing of 35mm film which heavily affect print quality. The most important factors include the selection of a sharp lens (or lensed camera), film, the use of a tripod and quality film processor.
Most consumers today buy and use point and shoot cameras, or buy SLRs with long zoom lenses. These cameras and lenses tend not to be the very best available to buyers (such as the Leica/Contax systems costing thousands of dollars, or even the $400 Ricoh GR1s point and shoot), and often, they are not used at their optimal settings (in fact, for most point and shoot cameras, there's no way to specify the aperture to be used! In most lenses, stopping down to about f/8 or f/11 otimizes the optical capabilities and resolution.)
Most P&S cameras used by consumers will simply test out to a 'very good' rating or about 50lp/mm of resolution at best, and often perform worse; better lenses will test out to 'excellent' and 'outstanding' and easily exceed 90lp/mm of resolution.
Most consumers use negative films of faster speeds than 100 ISO, and even at 100 ISO, grain becomes noticable quickly as print sizes increase. They do not use the most grain-free films available to professionals, such as FujiFilm Provia 100F or the Kodak Tech Pan, and as such the large grain interferes with the recording of super-fine detail. In fact, even most professionals don't use the finest grian films available, instead, picking films for the features desired.
Another quite important factor that is often overlooked is the use of a tripod. Many times, putting the camera, digital or film, on a tripod will result in significantly sharper images. Even at high shutter speeds, tripod mounting of a camera will aid greatly in the production of clear, crisp images that are free of shake and motion, especially at long zoom lengths. Most consumers do not use a tripod, and often, have never been taught the basics of even making sure they're holding the camera very steady and to depress the shutter button smoothly to aid in taking steady shots. Many simply 'point and shoot' and hope the fast films they're using will capture an acceptable shot.
Despite the use of a tripod and good film, many consumer level point and shoot cameras depend heavily upon automated focusing mechanisms, many of which focus in steps rather than in a continous fashion. As a result, focus is oftentimes not exactly made on the centered target, but slightly closer or farther depending on which focusing step is closest to the target.
Printing of film involves many steps and the operator of the machines is the most critical in the production of sharp prints. Many consumers leave the prints to automated print labs where little is done to ensure each and every print is accurately focused, or for that matter color corrected. The majority of yucky prints consumers see today can be partially attributed to poor quality control during the print process. A point and shoot camera like the Ricoh R1 or Olympus Stylus Epic can easily make sharp prints on 100 speed negative films when printed at 8x10" sizes.
Some commercial online image printers, like Ofoto, use laser scanning to create prints of digital images. Such equipment are often highly sensitive to minor errors but produce excellent prints. They are adjusted and controlled by skilled operators to a better degree than most local film developers available to the consumer. Quality online can and will vary, so don't expect quality to stay the same over time. Home printers tend to produce the same quality, whether good or bad, regardless of time since they usually are set once at their highest quality settings, fed the same inks and papers, and consumers usually do not make many adjustments after they're setup and working acceptably.
Aside from all of these factors, the consumer themselves often rate lesser prints better than expected due to their lack of experience in carefully examining and comparing print samples. Many simply accept the usual bright, super-saturated prints as great prints. Most consumers have never seen a crisply printed test target image that has been sharply focused and properly color corrected, and most usually see their photos on 4x6" paper. As a result, anything that looks good to their eyes for the brief 10-20 seconds most enjoy prints before putting them away for decades will appear good enough for them. Even with properly made prints, few view their prints under controlled lighting conditions to ensure the best image quality, nor do consumers usually examine prints at 6" or closer or with a loupe.
So what do most average consumers see?
While individual results will vary from my findings below, you can generally expect similar performance that won't be grossly out of line with the findings. They compare the image performance of an average 2.4MP digital camera to an average 35mm point and shoot camera and disposable APS camera.
Digital vs. APS disposable
In a direct comparison of shots made by a FujiFilm Quicksnap APS disposable camera and the FujiFilm FinePix 40i at the beach, the 40i produced more detailed, higher resolution images than the Quicksnap. A ballpark figure would put the 40i images at least 10-30% better in quality than the Quicksnap film prints; subjectively, you can 'easily' see the quality in the 40i 2.4MP digital images vs. the APS prints and fine text and other small details were more clearly rendered in the 40i images. Naturally, APS uses a smaller film size than 35mm film, so quality and resolution of APS prints are lower, in general, than 35mm film prints.
Digital vs. 8x10" 35mm point&shoot film print
#1 Film Image (3.8MB)
#1 Digital Image (328KB)
#2 Film Image (1.6MB)
#2 Digital Image (298KB)
These images were taken handheld on film using a Ricoh R1 point and shoot camera loaded with Kodak 100 film and the Fuji Finepix 40i digital camera at 2.4MP at the highest resolution (while the 40i produces a 4.3MP image at the highest resolution setting, the actual sensor is a 2.4MP sensor and upsampling is done inside the camera to fake 4.3MP; we will use the actual hardware maximum rating here). Images were taken from approximately the same locations, and minor differences in perspective result from the 30mm Ricoh R1 lens vs. the 36mm 40i lens. The film images were printed professionally at 8x10", then scanned in on an Epson 1200S scanner at 600dpi. The lens on the Ricoh R1 has been rated as 'very good' by Popular Photography, and generally reflect the same 'good' to 'very good' lens performance of most zoom point and shoot cameras consumers buy today. The quality of the FujiFilm 40i images are very good, but not top of the class among digital cameras. These choices would reflect the general purchases of most consumers, who do not find and buy the very sharpest lensed, highest performing digital and film cameras available, but rather, usually buy very good choices instead.
Note: When comparing these images please ignore differences in color and contrast. Do not consider the brightness or color of an object to mean film or digital is superior to the other! Either, when printed, can be adjusted to produce the best prints possible.
To examine these images, download them and open them up side-by-side in a good photo editing program such as Jasc.com's Paint Shop Pro or Adobe.com's Photoshop, or any image viewer which supports viewing multiple images at once.
Now let's examine them.
If we zoom out and look at these images together at 25% or 50%, we see two excellent images with very good detail and resolution. In fact, if you saw 4x6" prints of both, this is what you'd see - crisp, colorful, detailed prints.
Now, zoom in to 100% or preferably 200%. Scrolling about to the same locations on either image, you can now see what the differences are between a 2.4MP digital image and 35mm film image. In particular, notice the lower detail in textured areas such as the grass lawn in the digital image. This is due to the lower resolving capability of a 2.4MP digital camera and to the JPEG image compression commonly used by many digital cameras. In fact, in some areas, additional artifacts are particular visible in the digital image caused by the use of the JPEG compression method. Thus, for the best image quality on digital cameras, select and use a non-lossy image format such as TIFF or RAW, if you intend to print a 8x10" or larger -- you can continue using JPEG for 4x6" prints because at that size, the compression artifacts are not generally detectable.
In the pair of #1 images, also examine the curved, stepped wall area to the right of the center maze. Film again records more resolution, and here we can see that the steps are rendered more clearly than the digital image, which suffers from both JPEG compression and limited resolution.
In the pair of the #2 images, examine the banners especially those at farthest away. Notice that the text on the banners are more legible on the film print than on the digital image.
Overall, the conclusion can be made that the JPEG artifacts and lower resolution of the digital image, especially those containing fine text or complex textures, makes less able to record certain areas as well as film can. However, notice how in general, the digital images seem crisper than the film images. This can partially be attributed to the use of a low-resolution 600dpi flatbed scanner on the film prints (rather than professionally drum-scanning the negatives, which produces much sharper images), as well as the fact that on a digital camera, the edge transitions are recorded more sharply because the sensors in digital cameras behave digitally, recording items as either on or off whereas continuous film grain allows for more gradual, smoother edge transitions.
On a 8x10" print of both pairs images from quality labs, the digital image falls behind due to the visible JPEG compression artifacts as well as the lower fine detail resolution vs. the 35mmm film print. The film print, despite 'softer' edges, is very good and sharp, and most consumers can happily frame and mount it without much worry. Due to the sharper edge transitions, the digital print looks, at first, better, but upon closer inspection, the artifacts and lower resolution lead to a lower quality print on closer viewing distances. At greater viewing distances of 3+ feet where the human eye is too far away to see the artifacts, they, naturally, look identical in quality. This is a good reason to buy at least a 3MP, if not more, digital camera if large 8x10" prints are desired. Thus, depending on the subject and viewing distance, the digital image has the potential to look identical in quality to a film print to most average consumers.
Digital vs. 4x6" 35mm point & shoot film print
Representative image used for comparison (25KB)
#1 Film Print vs. Digital Image & Printed Image (106KB)
#2 Film Print vs. Digital Image & Printed Image (102KB)
However, the majority of consumers will not print at 8x10" and instead pick 4x6" prints instead. At this size, the comparison of both film and digital image prints yields the interesting result where both will appear identical for the most part. Yes, you can still use a loupe or examine them very carefully at 6" or closer under good lighting and spot minor differences, yet even these differences are at the very edge of detectability by one foot distant (and not only that, you'll have to be examining these prints under controlled, bright lighting for dozens of seconds in the most ideal of cases to see any differences at one foot distant).
Under normal viewing distances of 1 1/2 feet or greater, one simply cannot see any differences at all - the differences are of such small size that no human, even with perfect vision, can see them at that distance. Fine text in the far away banners and other such details are simply too small on a 4x6" print (less than a millimeter in height!) and any defects are for the most part 'invisible' to the average consumer.
This applies whether they're using a quality, online photo printer, or one of the better 6/7-color photo inkjet printers from Canon or Epson, or a dye-sub printer like the Olympus P-200/P-400. Naturally, the use of a poorer quality printer or online print lab will result in poorer quality prints.
(ie. Do not buy any HP printers made today - any HP printer with RET III technology and running at 2400x1200dpi or older!! These printers use only 3 color inks in photo print mode vs. 6-7 colors that the Canon and Epson photo printers use. As a result, HP prints create very grainy skin and face tones. You will only =waste= your money. Please see my other article in Dotty Spotty to compare actual print samples for yourself and to see how poorly the HP printers perform.)
As an example, two sets of comparison images are provided. These reflect what most typical consumers would see today from 4x6" film prints vs. home inkjet photo printer prints from 2-3MP digital cameras.
Prints of the same image taken at nearly the same time from the same location by the Ricoh R1 35mm point & shoot camera and the 2.4MP FujiFilm FinePix 40i handheld were made to ~4-4.5x6" paper. The Ricoh R1 used FujiFilm SuperHQ 100 speed film. The FinePix 40i was set at the highest quality JPEG compression at 2400x1800 image size for the best image quality. The 35mm print was made by Mystic Color Lab and appears as sharp as an average consumer could expect to receive from any good print lab. The digital print was printed on the Epson Photo 1200, 6-color photo inkjet printer at 1440x720dpi onto Epson Photo Paper.
The prints were then scanned in at 600dpi on an Epson 1200S scanner with unsharp mask turned on and automatic color adjustment. All three images, the film print, the digital print, and the original digital image, are displayed at 1:1 ratio. Subsections of the actual image have been selected for comparison - the first comparison uses an area close to the cameras, the second comparison uses an area farther away towards infinity. A metric ruler was also scanned at the same dpi and scale and added to each image for size comparison - the section of the ruler displays a length 1 centimeter long with ten 1 millimeter divisions.
The scanned images were compared with the actual prints under a loupe for accuracy, and to me, they are good representations of what one would see in actuality under ~8x magnification.
As you can see, the images are mostly identical in the level of detail and resolution. You can see that the film print has slightly more resolution but only in certain areas (eg. the #4 is better defined and legible on the lifeguard tower in the film print, however this difference concerns an area is smaller than 1mm sq.!). The digital image has a broader exposure range, so shadowed areas possess more detail vs. film, and has a better edge response, resulting in the appearence of a crisper image.
Overall, the differences are very small and one has to look at differences smaller than 1mm in length and width on the actual 4x6" prints to see them. For the typical consumer, this difference is only visible under scrutiny at 6" and closer under bright, controlled lighting; invisible at greater distances and under typical viewing conditions.
Because there are so many factors involved with comparing film and digital images, it is not always clear which is the better format.
So what about the 8MP+ size noted above? That still applies, but only if you are to use the highest-quality film and lens, have prints properly focused and made, perhaps use a tripod as well, etc., etc.
For professional photographers, the best way to proceed is to take similar identical shots on film and digital equipment to compare the quality for yourself. If you can see and dislike the difference lower resolution digital cameras provide, then you must seek out higher quality digicams. On the other hand, if the prints look identical, then you can rest assured that you can move into the digital imaging realm without sacrificing expected image quality. Also, talk to BetterLight.com and PhaseOne.com to see how 25+MP digital cameras have replaced film in the professional workflow today.
For the mass consumers today, most can buy a good 2+MP digital camera today and with the use of a quality online photo lab or a top-of-the-line home photo printer, produce 4x6" prints which will look as good as film prints. At 8x10", quality of <8MP digicams can still fall short of normal film prints, but generally, using 3-4MP images will be acceptable to most consumers.
Finally, keep in mind that most computer and digital imaging equipment prices typically fall 50%+ per year based on the past 20+ years of computing. Expect to see digital cameras to become cheaper, better, and more affordable at a very rapid rate, quickly replacing film cameras in the near future.