I have been experimenting with using the Windows 10 default photos app to "improve" the appearance of source images when presented in reports or other printed material. It's fairly simple to do any of the following (and certainly quicker than in a more complex program such as Photoshop Elements)
Crop to eliminate black borders (all borders or just where there is excessive blank space in the image).
Crop pre-1911 Census images to remove the reference number included in the scan, which can be rather variable in its positioning, and even rotated up to 1871, so they look more like full page 1911 images.
De-skew images that are clearly not straight.
Crop down parish registers to a single page rather than the random mix of single and double pages that different scanning exercises has produced.
Adjust brightness and contrast to enhance legibility of faded documents, particularly the pencil 1841 Census and some parish registers.
The results look a lot better, although for Parish Registers I restrict processing of early free-form registers to just simple cropping of excessive borders and de-skewing. However, the purist in me feels just a bit uneasy with "tampering with the evidence", and the fact that any processing of a lossy format such as jpeg will result in some loss of detail. I guess one answer might be to keep two copies of images, an original "research version" and an "improved" version for presentation, but that does seem overkill for hobbyist use.
Yes, editing a lossy format such as JPG is a problem.
One workaround is to use say the fhugdownloads:contents:irfanview|> Utility ~ IrfanView and convert JPG images to PNG that are not lossy and edit those in Irfanview, which is free and quite easy to use for the basic editing you described. FH is happy with PNG format files.
The Photos app will open PNG files but seems to immediately convert them to JPG format.
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry
One way of retaining origin data for an image (e.g. TNA refs) is to record it in the IPTC data, which certain image formats can include. That way it stays with the image file as meta-data. Of course, you need appropriate software - I use XnView for Windows, which can access and update it.
I personally would really not want to cut off the TNA references - though if the TNA references is in the IPTC data (because I put it there) this seems less of an issue. It just goes against my grain to throw anything away. Even filenames (that I nearly always over-write) can tell me things that I might otherwise lose - like if I've just browsed a "film-strip" to find something and want to find it again, knowing that it's image 236, say, is very useful.
I did try trimming the black-space round census images once - and discovered that with my default save parameters, I was actually making the file-sizes bigger, which wasn't the intention at all.
I'd be a bit wary of de-skewing where the original images are poor quality - any manipulation of a lossy format such as JPG will result in loss of information - hopefully not noticeable but....
One advantage of putting the source citation in the metadata is that it doesn’t have to be modified to accommodate limitations in file names. For example, RG10/1234 folio 56 p7 is perfectly acceptable in metadata but not as a filename.
If you only need to 'crop' source images, then I suggest you don't edit the image at all.
Instead use the Link to Detail frame in FH (similar to Link to Face for head & shoulders shots).
Then you retain the original JPG, but only the 'cropped' frame of interest appears in Reports, etc.
(Does not help with skewed or faint images.)
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry
I don't include images of sources, to avoid Copyright and Terms of Service issues (Crown Copyright applies to a lot of stuff in the UK, but commercial ToS often constrain reproduction of downloaded images on the Internet). I do include complete transcripts (produced using as much image enhancement as necessary) plus exact references to the image I used (website, census reference etc.) so that people can check my transcript at my site of choice or anywhere else they have access.
With images, you run into the problem of having to argue your case if the images were not downloaded from a commercial site but they think they were. ("Unnamed commercial site: those are our images! You can't publish them". "One place study: no they aren't: we photographed the originals at X archives and obtained permission to publish from the original author, being the Church of [Wales] in this instance. "Commercial site; we'll sue". "Us: No, really, we have permission -- our images, not yours". A resolution is still pending.
Indeed - my images are solely for personal use or sharing with immediate family, and I never republish elsewhere. I can see how this arises - there is a world of difference between a non-exclusive permission to use and reassigning copyright ownership. The company I used to work for always retained ownership of copyright of anything we published externally, and there was an office full of presumably well-paid lawyers to enforce it!
trevorrix wrote:I don't see a need to crop or alter original source images such as census. Digital storage is cheap these days.
True but printing thick borders of black ink for backgrounds is perhaps messy and expensive. Then again, using the FH facility to crop the image with a frame should(?) generally suffice there.
And then there's also the need to straighten many images that have been poorly scanned by data providers which I would consider to be a necessary usability enhancement; as opposed to directly editing an image, which suggests making some alteration to the actual image content.
