Several ways of getting a map to use within
M4OPS are described
in another post. Taking photographs is one of them, which is what this post is about.
It is particularly important to ensure that you are
allowed to photograph the maps by both the copyright owner and the place they are kept, and that the practical arrangements (such as tripod, lights, space) are acceptable to your hosts eg the library or archives.
Although we want to end up with a single image it is often best to start by taking
a series of overlapping images. Peter can then
stitch these together as part of the technical process of putting the map into
M4OPS. More technical aspects of photographing maps can be found
here.
In principle we want an image that is:
- as clear and sharp as possible, even when enlarged
- as undistorted as possible
- as easy to process as possible
In practical terms when photographing a map this means that:
- although the normal default jpg/jpeg format is usable, it is better to use one of the lossless formats, such as PNG or TIFF
- (I am told that even better would be a raw format but this makes the processing too complicated)
- there is no need to be concerned about file sizes at this stage as the process to get the map into M4OPS will take care of any issues - it is better to go for the highest resolution and we can always reduce it by post processing
- most modern digital cameras, including many of those in modern smart phones, are up to the task, but some are better than others
- the optimum area of a map to take with a typical 8 MP camera in any one shot is about 20-30 cm (7-12 inches) wide (my 8 MP camera is 3504 pixels wide so this gives 292 dpi)
- make sure the camera is set to its largest size of recording pixels (many cameras have the ability to reduce the number of pixels in order to economise on storage)
- 300 dpi is sufficient, although anything over 150 dpi is usable
- it is best to take shots in some order (eg top left to top right, then down a bit and left to right again etc) as this makes the organisation of stitching easier
- it is best if shots overlap by 10-30%, with more being needed if the map is fairly empty
- do not get any background (eg floor, table) in the shots as this will encourage false matches
- it is best to use a method of stabilising the camera, such as a 'leaning tripod' (see photos attached to the technical post)
- the camera should be as perpendicular to the map as possible so, if the map is flat on the ground or a table, use a (double) spirit level on the camera so it is horizontal in both dimensions
- to avoid camera shake, it is best to use a shutter release cable, or else a delay timer
- as far as possible, all shots should be from the same distance to the map
- it is likely that you will be within the focussing range where it is worth turning on your camera's macro mode
- the lighting should be as good and even as possible, if possible from above without throwing shadows, and as consistent as possible between shots
- if you are not using a raw format, it can helpful to specify the same "white balance" across all shots, instead of leaving it to auto - either calculating it using a white or neutral grey surface, or choosing a preselected white balance settings
- note that during the post processing, that Peter does, some issues that apply to all the shots of a map (eg there might be a red tinge on all of them because of the light) can be corrected during or after the stitching
For displaying maps online we do want very accurate copies and, although good results have been obtained with amateur photography, it is possible that paying for professionally done
scans will be the best way to go for some maps. It is also worth bearing in mind that if a single, poor photograph of a map is all that is possible then it may be better to put this on
M4OPS than have nothing at all, until better images can be obtained.
