Separate pages:General Photo Tips
The Future of Silver Based film
Tom Philo Work Experience
Scanning B&W Film
|DPI LPI and PPI relationships
Optical Density of Film
Hue, Color, Saturation
LCD vs. CRT
Initially I had all my photos transferred to Photo CD from the original Kodachrome slides, Ektar 25, Kodak Gold or VPS negative films after the film had been processed. This quickly became cost prohibitive at sometimes $2.00 a frame for a quality scan. If you digitize the film at the same time you turn in the film for processing you can get them scanned for around 30 cents a frame. So I went out and bought film scanners.
The high per scan expense makes it practical to go out and dedicated film scanners such as my old HP 5100A scanner and the newer flatbed HP G4050. I still use film and get CDs created at processing time from my local lab; however, I re-scan the negatives and slides when necessary at 9600 DPI for any quality work.
If I want to send out samples of work via email or put them online, like the images that I put onto this site, I scan them downsample the images to 72 DPI - which is the normal monitor resolution (96 if you have a MAC).
I use several software programs to read the CD images: Photoshop or Paint Shop Pro. Most labs store the images at 300 DPI on the CD. Old versions of Kodak had 5 resolutions. Most labs save them as JPG uncompressed format - some will save in TIFF.
I usually use Paintshop Pro to add in the © symbol, my name and web site information onto the images. I do minor sharpness and contrast corrections since they are normally destined for the web site. The effort in using Photoshop to get quality images are wasted when most people do not calibrate their monitors. Thus extra correction and time do that for routine images using Photoshop is wasted.
I also will sometimes use the watermark system to embed the copyright into the image. However, the distortion is noticeable at anything above the lowest level so I avoid watermarks usually.
The HP 5100A scanner scans at up to 2400 DPI optical and it is very very good. However, it will not run on a Windows 7 sytem so it is now on my backup system.
The HP G4050 scanner has better D-Max than the 5100A and will go up to 9600 optical and 120,000 via software. With it you can scan 20 slides at a time - though the software is tricky to use and is prone to crashing. But it does work on Windows 7.
Here are some of the things I have learned doing photography and then posting the pictures onto the Web.
I shoot for the highest quality lighting will allow when I take them. I take all my photos and have now scan them as noted above. I have multiple digital cameras and with them I just use the highest quality setting and set the camera to save both JPG and RAW.
Purchasing a high end scanner like the HP G4050 ($259 on sale in 2010), easily pays for itself. I spent $1,600 putting all my photos from Europe in 1994 (46 rolls) onto Photo CDs after I got back. Sending to a lab will still cost around $1 a frame at discount - so time or money is the key here. If you have the time scan them yourselves, else let someone else do it.
Nikon stopped making film scanners. There are other firms still making film and slide scanners which have bulk feeds so if you have more than 1,000 images to scan spend a thousand and get a really high end scanner.
The operator of the Photo CD transfer equipment can take great slides/negatives and transfer them really badly. I got so-so results with Kodak. 70% of the time the transfers came back way too dark and I've had to use lots of editing tools to lighten them up. The original slides were exposed perfectly and look great via a slide projector. My advice is to find a local firm with a machine and use them if you can afford it. If you get a bad operator you can complain and show them the originals and the transfers and get them to redo it. One time I sent my slides back twice before they did them correctly. Now of course I can only blame myself for a bad scan using the HPs.
For those images still on a CD (if good) I place it into my PC, bring up my CD Software that reads the Photo CD format Paintshop Pro, Photoshop and then save them in JPG format set at 20% compression.
To get good thumbnails is tricky. Sometimes you can screen snapshot the browse of the images in Paintshop Pro, Windows (view files a thumbnails). Photoshop 6 creates the thumbnails for you. But they are rather large in size — 8 to 12k for a 100 pixel wide image. PSP and Photoshop both allow scripts to be run that will take a folder full of images and make thumbnails out of them. DO IT!
Once a thumbnail image is available I then tag the photo into a page [<img src="normandy_sm.jpg">] and then it shows up on the browser. Then I will add in a hyperlink to a larger image. This is done by tagging the images to be a hot link by using <a href="normandy.jpg"><img src="normandy_sm.jpg">Normandy Coastline</a> tag format.
Scanning a print at 300 DPI is OK - as long as the print is 3x5 in size, 2400 and higher for a negative orslide. There is always loss of color and sharpness due to the scanning process. This is only noticeable if you enlarge way above the image size that you scanned. If you post a 2x2 photo scanned at 300 DPI at a 2x2 size you will never see any artifacts from the scan. Also, if you scan a 5x7 photo and then reduce it to a 2x3 you will also never see any scan errors. If you take a slide, scan it at 1200 DPI, and then make it 4 x larger you will see some color and sharpness loss unless.Any 35mm negative or slide scan should be done at 2400 at minimum.
JPG uses lossy image compression and that causes a noticable degrading of the photo if compressed more than 15%. PNP format is out there but I have not seen much use of it. PNG is supposed to have higher compression ability than JPG for the same quaility. I have not seen that.
Note: This whole section came from HP's Smartbits newsletter.
SMARTBITS: Types of Scanners
Scanners can be thought of as a bridge between traditional and digital photography, enabling us to convert film, slides, and printed images into digital files. Like much of digital technology, scanners started out as large, expensive devices out of reach for consumer use. However, as the technology has improved and the prices come down, scanners have become a common presence in the photographer's home. Just like your digital camera, most consumer scanners rely on charge-coupling devices (CCDs) to capture images from your film, slides, or prints. We'll introduce the basic types of scanners and how they differ.
* Film: The best digital conversions come from film, as they have a higher dynamic range than prints. Film scanners are designed specifically to scan film; they pass light through the image for the CCD to record. These scanners must have very high resolutions because film and slides are so small, the images must be enlarged substantially. Film scanners are used to create photo CDs.
* Flatbed: Flatbed scanners are similar to copy machines in that you lay the print face down on a sheet of glass and close the cover. The scanner illuminates the print from under the glass so the CCD can move past the print and capture the image line by line. The CCD reads the light reflected off the image.
Flatbed scanners can scan film using optional transparency adaptors, and optical character recognition (OCR) software can even convert scanned text into an editable digital form.
* Print/Sheet fed: Comparable to flatbed scanners, sheet fed scanners were originally designed for snapshots, though some can also scan slides and film as well. You feed the snapshot into a slot, and it is drawn past a stationary CCD. The CCD reads the image line by line as the image passes by it. Although sheet fed scanners take up less space than flatbeds, a disadvantage is that you can't scan images from items such as books and magazines.
* Drum: Drum scanners are the top of the line and are rarely found outside of professional studios or service bureaus. Instead of a CCD, the drum scanner uses a photo multiplier tube to capture images. Prints or film are placed inside a rotating glass drum. A bright light is reflected off prints (or shone through slides), and the photo multiplier tube records the light.
Drum scanners capture much greater detail in shadows and highlights than flatbed, print, or film scanners, and they offer the highest resolution and best color. They can also capture handle larger prints. Drum scanners can cost tens of thousands of dollars and are very labor-intensive, so service bureaus mostly charge by the scan for use.
* Scans almost always need to be corrected, as unwanted noise can be generated by electrical interference or device instability. The most common solution is to sharpen the image with photo-editing software. Sharpening cleans up the boundaries between colors to increase an image's definition, and is also helpful when greatly enlarging an image.
* Resolution: Resolution describes how well a scanner can capture detail. If you divide the scanning surface into square inches, resolution is determined by how many pixels per square inch the scanner reads (or samples per inch SPI) in each direction. A typical scanner samples 300 pixels vertically and 300 pixels horizontally for a resolution of 300x300 ppi (pixels per inch). Thus, it would produce 90,000 readings per square inch. Higher resolutions yield higher readings per square inch, capturing greater detail. Off the shelf scanners can reach up to 4,600 pixels per inch optically.
You may have heard of two types of resolution: optical and interpolated. Optical resolution actually refers to the amount of sensors the scanning head has per inch (a 300x300 scanner would have 300 sensors per inch). Interpolated resolution refers to the amount of pixels a scanner adds to the image, by guessing at the proper light reading from the surrounding pixels. Optical resolution is the true measure of a scanner's quality, as you can accomplish interpolated resolution using photo editing software.
* Bit-Depth: While resolution measures how many pixels per inch a scanner captures, bit-depth refers to how much information the scanner records for each pixel - - that is, the number of bits used to represent each pixel. Greater bit-depth allows more shades of gray and colors. A one-bit scanner would allow only images in either black or white, while an 8-bit scanner would allow black, white, and hundreds of shades of gray (256 tones in all). A 24-bit color scanner assigns eight bits each of red, green and blue to describe each pixel, tremendously increasing the detail in a scanned image. 24 Bit depth means the scanner can differentiate 16.8 million different colors.
Most scanners on the market are 24-bit, but more and more are appearing that have greater bit-depth. Though not many systems can represent all the color possibilities such scanners provide, the greater bit-depth does allow the scanner pick up more detail in dark areas of an image, and it can help reduce noise in the final image.
* Dynamic Range: Dynamic range refers to the tonal range of the scanner, its extremes of light and dark. A scanner's dynamic range measures how well the scanner captures detail in shadows and bright areas, and how well it records an image's transitions between brightness levels.
The range is expressed numerically on a scale from 0.0 (absolute white) to 5.0 (absolute black); this number is the difference between the extreme values of light and dark the scanner can handle. The higher the value, the greater the scanner's ability to distinguish information in shadows and highlights. A dynamic range of 2.0 is on the low end, while top-of-the line drum scanners can reach up to 4.6. Most flatbed color scanners have a dynamic range between 2.4 and 3.2. The quality of a scanner's optics, in combination with its bit-depth, contribute to the dynamic range.
Different combinations of these scanner attributes are suited to different scanning purposes, so bear that in mind when comparing scanners; you might not always need the very highest offerings in resolution, bit-depth, and dynamic range. However, understanding these basic specifications will help you choose the best scanner for your needs.
Have you ever had trouble getting clean scans from printed materials such as books or magazines? Printed images are already composed of tiny dots, and scanning them often creates cross-hatch patterns between those dots and the pixels. These interface patterns are known as moiré (more-ay) patterns. To reduce the problem try applying a Gaussian blur filter in your photo-editing program. Just don't go breaking any copyright laws.
Light and pigment: The difference between the two color models boils down to the difference between light, which is projected to form color, and pigment, which reflects color. Monitors project light to form colors, while printers mix pigment (inks) to form colors.
RGB: Red Green Blue. This is the method that a monitor uses. Additive color method. The cones in our eyes are sensitive to the three primary colors of light--red, green, and blue--which combine to produce approximately 16 million colors. Computer monitors and TV screens duplicate the way our eyes interpret color, though the range of colors a monitor can display is much smaller than what our eyes can interpret.
Add it up: The RGB model is an additive model in which the three colors (red, green, blue) combine on screen to make white. For example, when working with web-safe images, you can adjust each color in a range from zero to 255. If all three are set to 255, the color is white. If all are set to zero, the color is black. Different combinations in between account for all other colors.
CMYK: Cyan Magenta Yellow Black. If you mix CYM ink on white paper, you'll get a muddy brown instead of white. That's because now you're dealing with inks (pigment), which reflect light, not a monitor that projects light. The colors in the CMYK model-- combine to produce black. They do not have the same intensity as RGB due to the reflective nature of the viewing method.
Take it away: The CMYK model is known as subtractive, as the inks subtract other colors in the spectrum, leaving their own hue to be seen. When you mix all the subtractive colors, you take away all color to leave black. However, inks are imperfect, so they don't quite make true black. Black (K) ink is included in the system to adjust for these imperfections.
Color-matching: When you print an image from your computer, the colors must be converted from RGB (what you see on-screen) to CMYK (the system most color printers rely on). The problem is that RGB monitors can display more colors than what's possible in print. Some CMYK colors also cannot be matched on-screen. The trick of digital imaging and printing is to match up the colors in the RGB and CMYK color models as closely as possible.
Conversion: Fortunately, most photo-editing programs work with your printer to make the necessary conversions automatically. In programs like PhotoShop, you can convert images back and forth between RGB and CMYK to see the differences and account for them as necessary. Though your imaging programs and color printer often convert the images with no surprises, sometimes things don't turn out as you expected. In these cases, you will have to experiment with your images until you find the ideal print. Keeping notes on what you did to correct the images and prints will help you prevent similar problems in the future. You can also adjust your monitor to reflect more accurately how prints will appear.
ICC Color Profiles were created to match monitor, file, and printers for a WYSIWYG with computers.
Basic rule for your monitor is to be calibrated and the Gamma set to 2.2 and the white balance is set to 6500 degrees Kelvin for the color temperature. If prints are darker than your monitor raise gamma. Lighter lower gamma, before you change luminance of the monitor after calibration.
It is best to scan B&W negatives using color settings. You will get more information by using the four channels that is used when scanning color settings than using a scanner's B&W setting.
After the scan you convert the scan to a B&W image using your photo editing software.
To convert back and forth remember that if you double the LPI you get the equivalent DPI: Thus 150 LPI = 300 DPI.
Note DPI is NOT PPI: Pixels per inch. Pixel per inch is the resolution of a monitor. most are 72 PPI while MACs tend to be 96 PPI. Scanning at 2400 DPI then showing it on a monitor that has a resolution of 72 PPI compared to a scan of 72 DPI results in the same quality at the same image size.
Film has a typical OD of 3.4 to 3.6. So you should always get a scanner that can read at that optical density or higher. Some of the newer scanners can read up to an OD of 4.2.
What this means that the scanner can read from pure whites to details in the blacks in the film. If the light source of the scanner cannot pass through the film it cannot discern the colors, tonal quality and all the other subtleties that you are trying to scan.
Histograms give you a visual representation of the color spectrum of an image.
This is a fairly well balanced exposure with colors throughout the spectrum with some bright areas. Thus very little can be done when exposing this image to make it better.
The basic rule is:
To move data to the left in the histogram subtract light when exposing
To Move data to the right add light to exposure.
Right is white (lighter) left is black (darker) on a histogram.
Reviewing the histogram of an image right after taking it with a digital camera is very important. It can tell you at a glance if you captured all the details that you wanted in that image.
The better you capture it the less post-processing work needs to be done in the graphics program on the computer.
To compensate for the blurring effect of a camera's low-pass filter, and any image after it has been scanned (slide or negative) you must set the sharpening filer with a low radius value: 0.3 to 0.6.
All pictures will need to be sharpened: the Unsharp Mask filter settings will vary based on the intended output. In general, the coarser the printing, the higher the radius. For instance, when sharpening an average-sized photo for newsprint reproduction, the radius will be at 1.0 to 1.5. To print the same photo on a premium quality inkjet photo printer, the radius need be no more than about 0.6 to 0.8. As print size increases, so too does the Radius value. The Threshold that triggers the program to sharpen should stay low — under 10 — while the radius amount will vary picture to picture based on image size and output size and is highly dependent on the final output process.
Hue determines the basic color-whether it appears green, red, blue, orange, and so forth.
A color's hue has a strong effect on how it is perceived. A color from the center of the red part of the spectrum will be perceived as more threatening, or perhaps more passionate, than one that is closer to orange; while red with a pinkish hue is a calmer color.
Saturation refers to the amount of gray in a color, and determines how vivid it is.
As a general rule, women tend to prefer less saturated colors, and men more saturated ones.
Luminosity refers to the whiteness or intensity of a color; luminosity for any one color will range from pure white though to black.
"The contrast on LCD screens runs from 150:1 to 450:1 while CRT screens range from 350:1 to 700:1. The color range still is not comparable on LCD screens." From Tom's Hardware web site. This means that when doing photo editing on a laptop - what you see is NOT what most other people will see via a CRT. It also means that printing will be vastly different than what is seen.
By far the best way is to purchase a slide scanner that has an auto-feed function so you can do an unattended scan up to 40 slides at a time. Most modern scanners also accept uncut rolls, click start, go eat dinner, come back and all scanned.
Speaking from experience a bulk loading slide scanner will pay for itself VERY fast if you have even 400 slides to scan.
Going with a flatbed you will be sitting there for many long hours moving slides around and scanning them.
Going with a flatbed you still load one slide at a time - you will spend around at least 5 seconds a slide - if you are fast - even when putting them into a flatbed. Then you have the scan time of 40 to 120 seconds a slide. You do not want to go with any scanner that does only 1 at a time. Most flatbeds scan 4 negatives/slides at a time and some will do 12 slides at a time.
If you send them out you will spend over well over a few thousand with no guarantee of any quality results (cheapest around is like 79 cents a slide, most charge $2.00+ per slide as of July 2005) or any way to control that they keep them in sequence.
You can use a slide projector to project onto a wall only if the wall is absolutely smooth and no color casts exists on the wall (no eggshell color). If the wall is spackled your camera will pick that up and be impossible to remove.
What another person stated to do was:
Glass is the answer I use a pane of app 2' x 3' the projected image does not have to take up the full screen. The surface is painted (I made it 2 coats) of vivid white super flat paint this then eliminates glare reflection (pretty cheap but works). Projecting onto this you may need to use a filter depending on your projector, if you use a fairly current machine you should be ok, if you use an older model the you may get a red tinge on the result, hence maybe the filter. I place the camera above as close to the lens as practical and have the projection unit app 3' to 4' from the glass screen. Remember here that the larger the projected image the more imperfections you will pick up with the camera. The results that I have achieved are very much superior to a couple that I had made by a professional. (Which incidentally cost me an arm and leg almost) Good Luck with your venture.
There are multiple firms out there that will take your movie film and transfer them into digital format. The process is just like scanning a single frame of film and it requires special gear to do it right. Movies are recorded at 24 frames a second so that is a LOT of scanning. As of 2012 prices can vary from 10 to 50 cents per FOOT of film.
You can transfer movies yourself if you have a video recorder, a movie projector, and a good screen. You just show your movie and then record with your video camera. This works, but you better have a good video camera that has selectable frame rates and an audio input jack for movies with sound.
You do not want to record the movie so that people notice the interframe blanks!
So this mean you should take it at 1/30 of a second or faster so that the multiple images are caught and if the interframe is recorded it is of such a short duration people will not notice it. Recording it slower than 1/24 of a means that multiple frames are caught and the blanks between frames is also recorded (but nothing is there so it does not impact the previous frame captured) but it will then blur the movie.
A Portland area firm that does movie film transfers is www.keepsakefamilytreevideo.com - 503-253-7027. I saw their work at the Sportsman show at the Expo center and it was well done. Their film transfer cost as of January 2012 is 15 cents per foot for 8, Super 8 and 16 mm film transfers. They also digitize VHS tapes that you may have. They also offer audio transfers. They output onto DVDs a full production so you have narration, still images, movies etc.
[If you are in the field taking photos using modern digital phones / cameras it is a good idea to use online storage to prevent loss of a REALLY important image file by uploading an online storage immediately.]