Crazy report of image size

Hovering on the glasses icon produces the surprising news that my image on POTN is bigger than the one I uploaded. POST 18844867

Downloads as 1600x1280 on my end, which is what the software is telling me.
I wonder if the info you saw was somehow cached?

CyberDyneSystems wrote in post #18844944
Downloads as 1600x1280 on my end, which is what the software is telling me.
I wonder if the info you saw was somehow cached?

I don't know. Now it looks like 1600 to me, too. One possibility is that this spectacle prescription is even more out of date than I thought.

.

I, too, see it listed as 1600 by 1200 when I click the glasses icon.

.

OhLook wrote in post #18844867
That looked so much better before I reduced it to 1024 px and the grass got all blurry. Let's try this again:

Hosted photo: posted by OhLook in
./showthread.php?p=188​44867&i=i152938295
forum: Still Life, B/W & Experimental

EDIT: The second version does look better, but the strange thing is, the glasses icon says it's 6000 px wide. What I uploaded is 3000 px wide; I checked it again.

EDIT EDIT: And what's even stranger is that the original is only 3750 px wide.

.

I don't know what happened about those numbers. Anyway, here's a related question. I usually reduce images in landscape format to 1024 horiz. Any viewer would think I can't focus, as they always come out less sharp than before reduction. If reduced less or not at all, they take inconveniently long to upload and to copy to backups, and they use more storage space. But soupy images are disappointing. What's a good compromise, and is anything gained by reducing to a dimension greater than 1600?

Although I too rarely actually do it, the general consensus is that after resizing one must sharpen again.
Of course, the sharpening method/settings differs for a 1280 resize vs. the method/setting you used for 3500 etc.
And this is where I tend to get lost myself, usually ending up over-sharpened.

https://fstoppers.com …pening-your-images-228316

CyberDyneSystems wrote in post #18845452
Although I too rarely actually do it, the general consensus is that after resizing one must sharpen again.
Of course, the sharpening method/settings differs for a 1280 resize vs. the method/setting you used for 3500 etc.
And this is where I tend to get lost myself, usually ending up over-sharpened.

https://fstoppers.com …pening-your-images-228316

Thanks for the tips. Sharpening feels like cheating in the first place because it substitutes for doing things right the first time (focusing perfectly, holding still enough), but I suppose this is misplaced guilt.

The simple program that I most often use has just one setting for sharpening. I'll leave refinements to the PP specialists.

Asking again: since 1024 px loses so much quality, what's a good size, and does going over 1600 help at all?

Personally, I don't think that size has any bearing on how sharp an image ends up looking. You may also want to look at your jpeg compression setting however, if that's set too high (or a lower number quality setting) that really could rob you needlessly of sharpness regardless of image pixel size.

OhLook wrote in post #18845480
Sharpening feels like cheating in the first place because it substitutes for doing things right the first time (focusing perfectly, holding still enough), but I suppose this is misplaced guilt.

yes it is.

if it makes you feel better, sharpening is not a result of the digital age, http://www.photocritic​.org …ng-photos-in-the-darkroom

It has nothing to do with cheating and everything to do with reproducing the original capture on film/sensor on other media. That said, there is also an artistic element to sharpening a reproduction of the original image. But that also applies to color correction, saturation, contrast and even cropping. If one has no interested in applying their own vision to a photo, then sharpening, tone adjustments or cropping should be avoided, but if you are interested in making good looking images then use the tools you have at your disposal ... just as photographers have done for the last century.

One can use those tools with an eye towards faithful reproduction or taking a more artistic approach, either way they are all important tools and avoiding them completely is silly.

CyberDyneSystems wrote in post #18845752
Personally, I don't think that size has any bearing on how sharp an image ends up looking.

Well, the more I reduce an image before posting it, the fuzzier it gets. The one I linked to in this thread shows the difference.

Where would I find it?

Left Handed Brisket wrote in post #18845858
yes it is. . . . It has nothing to do with cheating and everything to do with reproducing the original capture on film/sensor on other media.

I know, thanks.

OhLook wrote in post #18845480
Asking again: since 1024 px loses so much quality, what's a good size, and does going over 1600 help at all?

If you upload a "bigger than 1600" image with upload tool's default sharpening and downsize to 1600, the result is of very high quality.

If you want to compare results and pick your favorite workflow, you can upload with forum's upload tool and then save the image (win: right click, save image as: change extension to .jpg), change settings, upload again, save again. As much as you like. The image is not saved on forum until you submit the post. This way you can try every possible setting and compare all in Photoshop easily.

HOSTED PHOTO
please log in to view hosted photos in full size.

HOSTED PHOTO
please log in to view hosted photos in full size.

OhLook, why not go back to the thread you started about this in 2017? All the answers are there.

In short: sharpness is the level of edge contrast (acutance). The size of the pic doesn't matter. After downsizing your image always apply some sharpening because downsizing softens the edges a bit and you need to compensate for that. Has nothing to do with how you focused in camera (assuming you focused right of course) but it is a result of the downsizing algorithm.

Detail is about resolution. The higher the resolution the more detail. And the bigger your image the more detail you will see. The smaller your image, the more detail you lose. Compare it to the human eye. We are able to see more fine detail in something up close than at some distance.

I think it's always better to do your downsizing and your sharpening yourself and not have it done by a website, but try Pekka's advice. It might work well for you.

Pekka wrote in post #18846047
If you want to compare results and pick your favorite workflow, you can upload with forum's upload tool and then save the image (win: right click, save image as: change extension to .jpg), change settings, upload again, save again. As much as you like. The image is not saved on forum until you submit the post.

Thank you!

Levina de Ruijter wrote in post #18846071
OhLook, why not go back to the thread you started about this in 2017? All the answers are there.

I didn't understand everything that was said in that thread. One thing I remember is the advice to upload at 1024 px (longer side), but I'm not satisfied with the resolution thus obtained.

"The size of the pic doesn't matter" but "downsizing softens the edges"?

Why not sharpen in accordance with the size I plan to reduce the image to instead of reducing first and sharpening later? In Preview, reducing size makes the image tiny. To see it at its final size, I have to close the image file and reopen it. I also suspect that sharpening twice makes artifacts, although I'd need to test this further. It did strange things on one image.

OhLook wrote in post #18846101
"The size of the pic doesn't matter" but "downsizing softens the edges"?

No, it doesn't matter, as long as you sharpen after downsizing.

No, it doesn't work like that. You cannot sharpen to compensate for something that hasn't happened yet. You end up over-sharpening your images which will introduce artefacts. If you want to (globally) sharpen not more than one time, then do it after downsizing.

Actually, a lot of people sharpen three times. You can still read this advice all over the internet. First round: to compensate for the AA filter in camera. Second round: sharpening of the edited image, sometimes globally, sometimes only selectively, sometimes both. Third round: to sharpen after downsizing. Multiple rounds of sharpening don't introduce artefacts. Not if you do it right.

Levina de Ruijter wrote in post #18846121

OhLook wrote in post #18846101
"The size of the pic doesn't matter" but "downsizing softens the edges"?

No, it doesn't matter, as long as you sharpen after downsizing.

If your SOOC image (i.e., a large one) is sharp, and you reduce it by half and the resulting image (i.e., a small one) is soft, then image size did matter. Are we talking about the same thing?

My laptop shrinks images to fit the screen. I often edit a 3750 x 3000 original, but it isn't nearly that big when I work on it. In fact, it's about the size it'll be when posted at POTN. Given that it is, whatever sharpening it gets ought to remain in a later viewing at that size, just as the white balance and other variables stay the same. I don't see why doing things in a different order changes the result. When you sharpen, you're just changing the colors of some pixels, right? Then, when you downsize, you lose pixels, but the surrounding ones are supposed to change again to preserve the detail in the larger version?