[Hover on a thumbnail to see the caption;  Click on a thumbnail for a larger picture]

Photos from TravelLodge lounge on 2005jul20:

Photos from MunicipalRd-39E on 2005jul21:

Photos from Brokenhead Wetlands on 2005jul21:

- a desire to salvage something from that badly blurred caterpillar photo led to my first attempt at computer sharpening;  the "enhanced" version was made with Gimp-v1.2.1 Script-Fu:Alchemy:UnsharpMask with size=50 opacity=99;  that was my 2nd try, after finding that the default settings, 5 and 50, did not do nearly enough.

- I then tried opacity=100 with several mask-size values on a few more "soft" photos, finding that some look overdone with any mask-size over 5, while others keep improving up to a size of 20;  darker photos seem able to take a bigger "dose", and white features seem most prone to becoming overdone;  in my examples, the most troublesome photo involved fine white "hairs" against a dark background.

- I also tried USM-sharpening on some reasonably sharp photos, finding they too can be improved by a moderate dose of sharpening.

- my version of Gimp also offers Filters:Enhance:UnsharpMask, which has PhotoShop-like options, namely: Radius, Amount, Threshold;  ImageMagick has all those plus Sigma;  redskiesatnight.com/Articles/IMsharpen has info on the units used by those 3, and since I can find nothing useful on the units employed by my old script-fu plugin, I'll switch to the Filters:Enhance way of invoking USM (UnSharpMask) for future sharpening experiments;  (a) my first guess that Filters-USM with 5, 1, 0 would be much like Script-Fu-USM with 5, 100 did NOT work out, since the Filters-USM clearly does more sharpening for those parameters;  (b) the Filters-USM with 1, 1, 0 seems much like Script-Fu-USM with 5, 100;  (c) the Filters-USM with 5, 1, 0 seems much like Script-Fu-USM with 20, 100  (I did these comparisons using a sharp synthetic image of size 540x540 having a small number of discrete colours).

- several articles have been written on "smart" sharpening to avoid increasing any noise, e.g. http://gimpguru.org/Tutorials/SmartSharpening2/, but that involves too many steps for me to do manually, though I may try it in a script.

- [Update] I've since learned that Gimp's Filters:Enhance:UnsharpMask often produces better results with a much smaller first-param and somewhat larger 2nd, such as 0.1, 1.8, 0 or thereabouts.  However that's for a photo with only a mild case of softness.  For very soft photos, the first-param needs to be larger, and I have used 3.0, 0.7, 0 to good effect.  (When 5.0 as first param is not enough the photo is probably best discarded;  also I've never used more than 2.0 for the 2nd param.)

- My advice:  find the smallest first-param (radius) for which USM produces significant improvement (use 2.0 for 2nd param while doing this);  then starting at 0.5 for 2nd param, increase until it looks overdone, back off a bit, and then you're at the best settings.

Photos from the Casino in Scanterbury on 2005jul21:

Photos from MossSpurRd on 2005jul21:

Photos from Contour on 2005jul21:

Photos from PTH15 on 2005jul21:

Orchid species+varieties observed 2005-July21 - from notes by Doris Ames
	Blooming	Non-blooming
Road 39E:
	G.tesselata	C.maculata var.maculata
	S.lacera
Brokenhead Wetlands:
	P.obtusata	C.reginae
	M.unifolia	C.parviflorum var.makasin
	P.dilatata	C.parviflorum var.pubescens
	G.repens	C.arietinum
	P.huronensis	C.trifida
	S.romanzoffiana	C.maculata var.occidentalis
Moss Spur Road:
	P.lacera
Contour:
	G.tesselata	P.hookeri
	S.lacera var.lacera	C.acaule
	S.lacera var.gracilis
PTH#15:
	P.huronensis	L.loeselii
	S.romanzoffiana	P.ophioglossoides
		C.reginae
		C.parviflorum var.pubescens
Note that Spiranthes lacera var.gracilis is a new variety for Manitoba; it was tentatively identified at both the 39E and Contour sites by Bob Ferry.