dacs.doc electric

 

A Brief Introduction to Digital Photography
Part I - Capture Your Photographs

By Richard P. Ten Dyke

 

I believe that black and white photography, using traditional chemical processing, will survive as an art form for many years to come. It is a powerful artistic medium and the prints are archival. But for almost every other purpose, digital photography is proving to be the better choice. Many magazines today are totally digital. Even paid advertisements must be submitted in digital form. Advances in resolution and archival processing make most wet processes obsolete.

A discussion of digital photography can be divided into four general topics: capture--how to get the photograph into a digital form; storage--how to archive and retrieve the digital image; edit--how to modify the image for artistic purposes; and publish--how to print or distribute the finished image.

Also, this will be a personal story; I will relate my own experiences in digital photography. It is not intended to be a complete and balanced treatise on the subject, which could grow into a large book. Instead, it is intended as an example and a guide to those who may be new to this rapidly expanding field of interest.

Part 1:
Capture Your Photographs

Today the photographer has three ways to capture a digital image: He can scan the film (either a positive transparency or a negative) into a computer. He could have his films developed and printed, then scan the printed images using a flatbed scanner. Or, or he can use a digital camera that records the image directly to a digital memory.

Let's look at
film scanning first.

In 1994 I purchased a Nikon Coolscan 35mm film scanner (since replaced with a newer model). At the time, film scanning was the best way to get a high quality photograph into digital form. I had acquired a fairly large inventory of 35mm Kodachrome slides, and my first experiences were to scan the best of those into the computer.

The Nikon slide scanner transforms a 35mm Kodachrome slide into a 28 megabyte digital file. For images, we talk about pixels (Picture Elements) rather than bytes. When scanned, a slide image is separated into three images based on the primary colors, red, green, and blue. These are the same colors that are created on a computer monitor, hence the term RGB Monitor. A pixel consists of three color dots of red, green and blue. White is made up of a combination of the three. Yellow is made up of equal parts of red and green. In other words, yellow is white minus blue. Conversely, blue is white minus red and green, and therefore, white minus yellow.

Since we can express 256 levels (from 0 to 255) in an 8-bit byte, the scanner was designed to provide for 256 levels of intensity for each of the three colors. Thus, any pixel being a combination of levels of intensity for red, green, and blue would require one byte for each of the three colors. The 28 megabyte image is the equivalent of 28 divided by 3 or 9.33 megapixels.
What does this mean? How many megapixels is enough? It all depends upon the intended purpose of the image. One way to look at it is to think of creating a print.

Professional printers--those who want to create for sale or publication--like to print an image using 300 dpi, or dots per inch. That is the equivalent of 300 pixels per inch. Therefore, each square inch of printed image requires 300X300 or 90,000 pixels. That yields a very good quality print, and the human eye could probably not tell the difference between a print at 300 dpi and one at 600 dpi. But suppose you want to print a photo of your baby or your horse or your favorite vacation spot. You want to look at it and not be disturbed by the "fuzzy" quality of the image. You probably do not need 300 dpi. You can do a very good job at 180 dpi. With that in mind, here is a table of how many "megapixels" you require to achieve a print of a given size.

Print Size Megapixels
4X6 0.8
5x7 1.1
8x10 2.6
11x14 5.0
16x20 10.4

So it is easy to think about printing 11x14 from a scanned 35mm transparency. It works. The results are excellent.

But suppose you want to make posters or very large enlargements of the pictures. The scan from a 35mm image is not good enough. You could use a large format camera, such as, a 2 1/4 by 2 1/4 Hasselblad. Then you could have an image that has an area that is 3.8 times the area of a 35mm transparency, and when scanned, a file of 35 megapixels. I have never done this, but I have seen the results of someone who has, and he achieves print sizes of 30x40 inches.

If you don't have a scanner or a digital camera and don't want to buy either, what do you do? Knowing that the digital revolution is fast overtaking wet processes, your local one-hour photo store is likely to offer you the option of getting your images scanned to a compact disk. They can scan either transparencies or negatives. I recently had three 36 exposure rolls of negative (also called print film) scanned to a CD for a cost of about $100. Some of these images were to be sent to a magazine where they would be reproduced to a maximum size of about 3 by 4 inches. The lab scanned each exposure to an 18 megabyte (6 megapixel) file, which is not the maximum possible but, in this case, good enough for these purposes. I suspect the cost for scanning will be coming down rapidly as more people want to use their film cameras to produce digital images.

What about
digital cameras?

A few years ago, a popularly priced digital camera could create images that looked good if not enlarged to more than 5x7 inches. No more. Today's popularly priced digital camera produce images up to 5 megapixels, approximately one- half the "quality" of a well scanned 35mm transparency. and can create very good prints up to 11x14. By popularly priced, I mean under $1,500. Professional cameras can run to $30,000.

Which is better to use, a film camera or a digital one? Image quality is not the only question that determines which is better. Quality is important, but so is cost, convenience and ease of use.
The images that I had scanned at the lab were taken at a party where celebrities were present. Certain photos were very important to capture. Also important were the facial expressions of the folks in the pictures. I needed a strong flash unit. With these requirements, I set aside the digital camera and used a film camera instead.

Reason #1: Popularly priced digital cameras come with a built in flash unit that runs off the camera's batteries. It is a weak flash that works if the subject is within 8 feet from the camera. I needed a flash unit that would handle a distance of 20 feet.

Reason #2: When you press the shutter, the camera starts to process the proper settings for that particular picture. One of these is focus. It can take up to a second for all of this processing to complete. Between pressing the shutter and taking the picture, the subject has changed facial expression, blinked the eyes, or something else. What you see is not what you get. I created a lot of ruined images before I learned this lesson.

That was a few months ago, before I acquired the Olympus E-10 camera. Unlike most digital cameras, this particular camera overcomes both of those problems. It allows for manual overrides, an external flash, and shutter delay can be reduced to a tenth of a second. If I had owned this camera at that time, I would have used it. (Olympus has since announced the E-20N.)

So much discussion about digital cameras focuses--excuse the pun--on "how many megapixels does it have?" There are other important considerations in picking the digital camera that is best for you.

  1. How much image resolution do you need? Three megapixels is probably enough for even the serious amateur. More is not always better.
  2. Does the camera use proprietary or expensive batteries? Are they rechargeable? Batteries are an important cost item. How long will they last until they need to be replaced? The number of pictures you can take will depend on how much you use flash.
  3. How many images will the camera hold before you have to download? This will depend on the size of the camera's memory, on how large (in pixels) each image is. This is becoming less of a problem, as flash memory cards can now contain hundreds of images.
  4. What kind of a zoom lens do you need? Many 35mm film cameras come with interchangeable lenses, so you can go from 18 mm to 1000 mm by zooming or changing lenses. Most digital cameras have fixed lenses, so you are limited to (a) the zoom power of the attached lens or (b) using supplementary lenses. Digital cameras have fixed lenses for a reason: to seal the camera and prevent dust from falling on the photosensitive element, probably a charge coupled device, or CCD, which is the chip that substitutes for film.
  5. How long do you have to wait between successive shots? More pixels per image means that the camera will take longer to write to memory, and so will cause a longer delay.

Popular digital cameras are loaded with automatic features. They also include software that will allow you to transfer the images into your computer. Because of the richness of features, learning to use the camera may be daunting. The manual that comes with the camera may become your new best friend.

But if you take it seriously, you will be richly rewarded.

Finally

To keep things simple, I made assumptions. I stated that 8 bits were sufficient to represent a range of a particular color. This is a kind of standard, but it is arbitrary, and some users believe that 8 bits per color is insufficient. Some professionals use 10 or 12 bits per color, which requires more memory.

If you are using only black and white, you may use one or two bytes instead of three to store the pixel, and the memory requirement is reduced.

I have not yet mentioned scanning a printed image. With a photographic print and a flatbed scanner, you will get an image quality depending on the quality and size of the print, and the resolution of your scanner. Consider scanning an 8x10 image at 200dpi. That will give you an image size of 3.2 megapixels--clearly a good amount of data to work with, but not equal to having scanned the original film. The print does not have the same level of detail. Nevertheless, you may be surprised at how good it really is, so do not dismiss the approach. You can achieve an image that is very close to the quality of the original. If there are scratches or bad areas on the print, you can even improve on it by using a photo editor.

If the printed image is from a book or magazine it is not an original photographic print. It has already been created with a screen of pixels. Screening it again will create mioré patterns which will be difficult to eliminate. However, if you are planning to reduce the image size you can get a very acceptable copy.

To summarize, you can capture your digital image using a film camera & scanner, or you can use a digital camera. The quality you need is governed by how you intend to use the result. For amateurs, you can still get the best quality by using film and a scanner, but with improvements in technology, digital cameras now provide all the quality most people really need. And new digital cameras are being developed which will exceed the quality available from equivalent film cameras.

In the next sections, we will discuss how we store and retrieve our images. In following sections, how we edit and improve them with photo-editing software and how we print and display them.


Richard Ten Dyke is a member of Danbury Area Computer Society who has had a long interest in both photography and computers. He started his photogr aphy career with a Leica IIIC in 1952, and his computer career working with an ERA 1103 in 1956. He currently is retired from IBM and resides in Bedford, New York. You can reach him at tendyke@bedfordny.com..
Copyright © 2002 Richard P. Ten Dyke

BackHomeNext