This page is a work in progress. Some aspects of my processes are pretty solid, but others may change as I improve my methods and find better ways to present the images I have. As such, this page may be updated from time to time.
For the most part, scanning is probably the easiest part of putting images on this site. I have an Epson Perfection V600 Flatbed Scanner. It cost about $230, and is probably the best scanner I could buy without having to spend over $500. It also has inserts designed for holding and aligning slides and negatives for scanning, although I ended up needing to purchase a few extra inserts for some obscure film sizes I encountered.
I use the included scanning software, Epson Scan. I have other software available, but I find it more limiting. The settings I use are pretty straighforward:
Documents | Photos | Negatives and Slides | |
---|---|---|---|
Resolution: | 300 - 600 dpi Documents such as a page from a book tend to be text-only, and there generally is no need to scan at a resolution higher than 300 dpi. If the document includes any hand-written information, I recommend scanning it at 600 dpi. |
600 dpi Theoretically, a printed photo contains information at about 300 dpi. By scanning at 600 dpi, intricate details in the colors show up better. |
1200 - 6400 dpi Negatives are often pretty small, and therefore need a higher resolution. 120 film is rather large, so it would scan fine at 1200 dpi. 110 film is even smaller, and is probably better at 6400 dpi. The goal is to scan at the resolution that works best for the intended use. |
Color: | 24 bit color minimum. I might go up to 48 bit color in some cases, but the difference is rarely noticeable in general digital use. NEVER scan in black and white or grayscale, since it can lose subtle shades. Exception: pages of books / typed documents. In some cases, I am finding it beneficial to select an additional option called "Color Restoration" on the scanner, especially when scanning old film and negatives. Even if I scan black and white film, I will do it in such a way that it still saves as 24 bit color. |
||
Filetype: | .pdf or .jpg It is sometimes useful to scan books or multi-page documents into a .pdf. Lately, however, I have been scanning as .jpg since my computer system will show image thumbnails that make it easier to find a certain scan. |
.jpg or .tif .jpg is great when scanning for use on websites and other general digital uses, just be sure to have compression set to as little as possible for the best general use scan. If the goal is to print the photo, switch to a lossless filetype such as .tif. I find it best to scan in bulk as .jpg, and scan individual things a second time as .tif only as needed. |
My default settings when scanning a mixture of documents and photos are 600 dpi, 24 bit color, and .jpg filetype with as little compression as possible. I generally only change from this when scanning negatives, a photo I want in a high resolution .tif, or a sequence of typed pages from a document or book. Keep in mind, storage is cheap, but scanning twice takes twice the time.
In the time since I purchased my scanner in 2017, I have scanned thousands of photos, negatives, and documents, and experimented with different ways of sharing them on my website. One of my more recent changes was that I stopped cropping out the borders before the scan, and instead leaving a border around the entire image, thus showing the edges. I believe this looks far better than cutting off the edges in many cases. As a result of this change in my method, I still need to rescan quite a bit of material as time permits.
There are exceptions and times when cropping out the border at the time of the scan can be a good idea. Slides and negatives definitely need to be cropped for the scan. Photos too, can sometimes benefit from cropping before the scan, but I find it usually better to leave the cropping until the editing stage after scanning. This way any identifying information written on the borders is not lost until I am ready to display the edited photo with proper identification.
I also often scan the backs of photos since those too often contain identifying information. My opinion is that if anything is written on the back, no matter how insignificant, I should probably also scan the back.
Recently, I have been scanning with the 'Color Restoration' setting turned on. This seems to work best with film, but sometimes it helps with prints. The 'Color Restoration' setting attempts to correct cases where the original colors have faded or shifted. Often this color shift goes to red, but different films can shift differently. I have seen red, green, and blue color shifts, and this scan setting, while not perfect, definitely helps. I have seen one or two cases where the 'Color Restoration' setting shifted the colors the wrong way, but it was a very uncommon occurance.
Having seen the improved results with this setting, my goal is to make new scans of all the slides and possibly film negatives that I previously scanned. Isn't it funny how a simple project suddenly grows in size?
I used to always scan film with the scanner set to Color Film, even if the film was black and white. However, scanning as black and white film can have its uses. I discovered this while making fresh scans of the Stuiber Negatives. Rather than have all my images showing how the film faded over the years to varied amounts of sephia, using the Black and White Film setting on the scanner gets me better results with all the images back to their original untinted black and white. However, I still turn on the 'Color Restoration' setting, and save the scans as 24 bit color images.
Another aspect of the scanning process that I have been experimenting with is 'Backlight Correction.' This setting has the effect of lighting up shadows and darker parts of the scan. Sometimes this helps, especially if faces are in shadows, or the image is very dark to begin with. However, improvement is hit or miss, and it feels like it tends to lose information when it shifts the light levels. This loss is most noticeable in the lighter parts of the image, where the 'Backlight Correction' moves the lighter areas of the image closer to white. In cases where I think 'Backlight Correction' might be useful, I start by scanning at the lowest level, and make a scan at each additonal level that doesn't appear too washed out. I can then look through the scans, and select the best result.
The final step in the scanning process, at least when scanning film and slides, is to leave information about the scan with the film. This information serves as an identifying link between the physical film and the digital copy. I initially only handwrote on a small scrap of paper the number I associated with a roll of film among my scans. As of 2022, I am printing out small dated information cards that provide not only enough to find the roll of film among my scans, but also contains information about the film and the scans themselves. This way I can determine if the film has been scanned, and others can know to contact me to obtain the scans if they need them.
I, unfortunately, am uncertain how to do this with prints. I dare not actually write on (and damage) prints, and very few prints are actually stored in a way conducive to keeping additional papers with them. I may create a solution in the future, but at this point I am leaving prints without any information that they have been scanned.
Perhaps the most tedious part of preparing images for my website or integration into the family tree is editing. I like my content to fit seamlessly into the website, and that requires a perfectly white background. However, it seems impossible to create a perfectly white background while scanning. Often the background on the scan will be off-white, and perhaps include scratches and specs of dirt and dust. These are all very visible against a perfect white border, and do not meet my standard of professionalism. To fix this, I need to edit the background in such a way that it removes the dust and scratches, and fades the off-white to perfect white I use around photos on this site.
The precise multi-step process I use (best with steps 2-7 saved as a macro):
Sometimes I need to make adjustments to this process, particularly if a print was glued to a black paper in a binder or for whatever reason will not easily have a good fade to white. In these cases I make the selection as tight as possible, consider skipping any expansion or limit it to only a few pizels, and then feather the selection by about 2-3 pixels. This will get me a good outline, but there will not be any shadow.
I may occasionally do additional edits to images, but lately have found it better to make sure I'm starting with the absolute best possible scan rather than trying to make adjustments after the fact. When editing is complete, I save the image as a new file. This way I can always go back to the original scan if I feel a need to do so.
Another thing I sometimes do is find records like census pages, where it is optimal to highlight the parts containing the information to share. For these, I will temporarily create a second layer, select the regions of the image I want to highlight, fill with a green shade ("37, 195, 93", or "#25c35d"), and then reduce the opacity of that layer to 35% before I flatten and save the image.
This page is as much for me to remember how I am doing things as it is for others who ask me for advice on scanning their own photos. As I continue to scan and prepare more images for my records and this site, my processes continue to evolve as I experiment to improve my output. These improvements will be added here.
© 2023, by Ben Hinks