Acceptable Sharpness?

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Deanimator wrote:The full size image was resized -50%. ... I believe that the crop is from that image.
OK, good. So then it looks like your procedure for showing an "actual pixels" crop from some image is now working properly. Now all that's needed is to make that an actual pixels crop from a camera resolution image, and we'll all be comparing the same sort of oranges.
The full size image was resized -50%. I thought the original would be too big to upload. I know it would be elsewhere.
Just because an image can be uploaded, doesn't mean that it should be uploaded. At some sites, you can upload a large image and the hosting site will both a) keep track of the original resolution image, and b) automatically generate resized copies that can be served out as appropriate. At photomacrography.net, that doesn't happen. Instead, the photomacrography.net forum software will accept large images but will immediately downsize them to 1024 pixels and throw away the originals.

So you have to ask yourself, do you really want to hand over control of the downsizing to some unknown piece of code? Or would you rather downsize it yourself so that you retain control over what gets served back to viewers?

For myself, the answer is simple: do the downsizing myself, and upload to the forum only images that are already 1024 pixels or smaller, with a file size of 300KB or shorter. That way the forum will serve back exactly what I uploaded.

The same idea is described like this in our Posting Guidelines:
One caution: there is a feature that at first glance looks like it may be useful, but actually it is not. If the image on your PC exceeds the maximum image size for the forum, then the forum software will automatically and silently resize it for you. Don't be fooled! The forum software will indeed resize an overly large image, but the quality is very poor. It is much better to have the image sized properly before uploading.
But whichever way you go on that decision, when we ask for "actual pixels", we mean the same size pixels that the camera recorded.

--Rik

Deanimator
Posts: 870
Joined: Tue Oct 23, 2012 7:01 pm
Location: North Olmsted, Ohio, U.S.A.

Post by Deanimator »

Well, here's the original image with no resizing, as well as a crop:

Image
Image

Deanimator
Posts: 870
Joined: Tue Oct 23, 2012 7:01 pm
Location: North Olmsted, Ohio, U.S.A.

Post by Deanimator »

The same subject with 20 micron steps.

Much more contrast in post.

Image

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

You/we would need to see "actual pixels" again to compare, to see if the step size change made a difference.
Post in the highest jpg quality you can within the 300kB limit, these are much less.
If in photoshop you export "for web"/"for web and devices"/"(Legacy)" - names vary, it'll tell you the file size in advance. It only needs a representative section, such as you posted last time but as a separate image.

I thought the previous image didn't look super sharp, but then with this objective, diffraction is playing a part. It wouldn't be "pixel-sharp". The subject isn't the easiest for telling what detail might be there. Fine sandpaper, pencil-lead scrapings, butterfly wings, semiconductor wafers, laser-printed paper, and more have been tried! The last of those is certainly good. Address labels and bank statements with personal text (or barcode/dot patterns) can be sources.
Chris R

Deanimator
Posts: 870
Joined: Tue Oct 23, 2012 7:01 pm
Location: North Olmsted, Ohio, U.S.A.

Post by Deanimator »

ChrisR wrote:You/we would need to see "actual pixels" again to compare, to see if the step size change made a difference.
Post in the highest jpg quality you can within the 300kB limit, these are much less.
Here's a < 1024 crop:

Image

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

By comparison it looks sharpened - which I guess means it's sharper, assuming you didn't!
The one with the dot is the latest. I've increased the contrast of the earlier one to roughly match..
Image
Chris R

Deanimator
Posts: 870
Joined: Tue Oct 23, 2012 7:01 pm
Location: North Olmsted, Ohio, U.S.A.

Post by Deanimator »

ChrisR wrote:By comparison it looks sharpened - which I guess means it's sharper, assuming you didn't!
The one with the dot is the latest. I've increased the contrast of the earlier one to roughly match..
Image
It was sharpened to the same degree as previous versions, but appears sharper. Apparently the smaller steps did make at least some difference.

Going from continuous light to flashes made the most difference.

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I generally use a rather small step, partly for the edge of sharpness and partly for the odd flash failure - you can miss one frame without it noticing.
Chris R

Deanimator
Posts: 870
Joined: Tue Oct 23, 2012 7:01 pm
Location: North Olmsted, Ohio, U.S.A.

Post by Deanimator »

Something with more texture, a piece of leather:

Original and cropped:

Image

Image

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Sharpened or unsharpened?

--Rik

Deanimator
Posts: 870
Joined: Tue Oct 23, 2012 7:01 pm
Location: North Olmsted, Ohio, U.S.A.

Post by Deanimator »

rjlittlefield wrote:Sharpened or unsharpened?

--Rik
Sharpened +75.

The combination of flash, reduced step size (20 microns) and increased sharpening seem to have made a lot of difference.

I've got a 300+ image stack of an emery board running in Zerene right now. I forgot to put my diffuser over it, so there's no telling how it'll come out.

Post Reply Previous topicNext topic