Photoshopping....too much?

[Sure Squintsalot]

Outside of scientific papers, microscopy images are rarely seen without some postprocessing. Even image stacking qualifies as photo manipulation (of a type). Add spot removal, de-fringing, contrast enhancement, sharpening, and color saturation and it's easy to post-process your way to a "too perfect" image. More than a few scientists have even been called out on their heavily PP'd images, prompting forced explanations and the creation of publishing rules when edited photos are involved.

But where do you draw the line? Spot removal? Color correction? Sharpening? Some photographers have no rules other than getting a "best", most aesthetically pleasing image; we've all seen those photos. Purists, on the other hand, will present an image out of the box, letting their subject stand on its own merits. Most of us are somewhere in between, but knowing where to draw the line is tricky.

Here are two images of some wet-mounted larvae found in a birdbath. Both are 8 image stacks taken in brightfield are processed identically other than the lower has an added dropped shadow:

Screenshot 2023-05-06 132323.jpg (139.02 KiB) Viewed 4248 times

No subject details have been added to the lower image. A few questions:

• Does the shadow add anything to the photo that shouldn't be there?

• Does your expertise in microscopy change how you see such photos versus how a non-expert would?

• Do you ever get suspicious of those "perfect" microscopy photos? (Nikon Small World, I'm looking at you!)

• What are your limits when post processing your images?

On the other hand, some people just don't know when to stop.

Screenshot 2023-05-06 132518.jpg (73.45 KiB) Viewed 4248 times

[Alexander]

Digital cameras produce raw data files. Every picture that comes out of a digital camera is processed and manipulated. "Out of box" just says the processing was left to some automated software in the camera rather than doing it by themselves in Photoshop. Compare the same picture processed in the camera using different picture styles.

Doing it manually in Photoshop carries the risk to overdo it. Common faults are over-sharpening and overdoing color saturation.

Most of the stunning pictures even in publications like Nature are not "out of box". To give you an example:
https://de.wikipedia.org/wiki/Fluoresze ... tCells.jpg
It is impossible to see that pictures through the eye-pieces of a microscope. The emission of DAPI is in the same color range as the excitation of FITC. The picture is obviously layered in Photoshop. The blue, green and red parts are from different pictures and mounted together. Looks great but is synthetic and far away from "out of box".

[MichaelG.]

Digital cameras produce raw data files. Every picture that comes out of a digital camera is processed and manipulated. "Out of box" just says the processing was left to some automated software in the camera rather than doing it by themselves in Photoshop. Compare the same picture processed in the camera using different picture styles.

That’s a very important point … but I would just like to mention that it particularly applies to general-purpose photographic cameras.
Dedicated ‘scientific’ microscope cameras have little or no automation … making it much easier to extract quantitive data.

One of several reasons why [to those looking for ‘attractive’ images] they appear over-priced.

MichaelG.
.
.
Readers may find the [2019 vintage] document by Louis Keal of interest:
See my Dropbox link in this thread: https://www.microbehunter.com/microscop ... eal#p67953

[Hobbyst46]

It is impossible to see that pictures through the eye-pieces of a microscope. The emission of DAPI is in the same color range as the excitation of FITC. The picture is obviously layered in Photoshop. The blue, green and red parts are from different pictures and mounted together. Looks great but is synthetic and far away from "out of box".

Yes, and many high-resolution fluorescence images are produced not by cameras but rather by photomultipliers, so the raw images are basically colorless, and color is
added by the processing software and can be arbitrarily chosen.

[Phill Brown]

Unless the RGB screen is calibrated it's another factor.

[blekenbleu]

• Does your expertise in microscopy change how you see such photos versus how a non-expert would?

There is a very real effect sometimes called "simultaneous contrast",
where discernment of light detail is confounded by a dark background and vice-versa.
That is one reason why gray photo mat is popular for framed photographs.

To some extent, your application of dropped shadow has that effect.
FWIW, I attempted to crop those two images and display their difference using ImageMagick:

Code: Select all

magick top.jpg bottom.jpg -compose Mathematics -define compose:args="0,-1,1,.5" -composite diff.jpg

.. perhaps those crops are misaligned:

diff.jpg (68.51 KiB) Viewed 4134 times

Increasing gamma and reducing brightness helps perceive detail in upper image, IMO.

[imkap]

Editing is a part of the process. It is up to the author's taste how much and in which direction. You could leave it to the camera processor if you like the results.

I like the image without the shadow more

Digital photographers use software and analog (film) use a darkroom.

[KurtM]

^^^ I agree with imkap.

[einman]

I agree-I prefer the image without the shadow.

[Sure Squintsalot]

Most of the stunning pictures even in publications like Nature are not "out of box". To give you an example:
https://de.wikipedia.org/wiki/Fluoresze ... tCells.jpg
It is impossible to see that pictures through the eye-pieces of a microscope. The emission of DAPI is in the same color range as the excitation of FITC. The picture is obviously layered in Photoshop. The blue, green and red parts are from different pictures and mounted together. Looks great but is synthetic and far away from "out of box".

Post processing is something that is done AFTER an image is aquired by a user, it isn't something that is done by the image aquisition hardware/software and most microscopy images in scientific papers are presented with little post processing; those that are PP'd, come with written caveats in the offered data, or in the captions. This is required protocol that I've seen come about in the early 2000's after software was developed that could detect bogus, photoshopped images in scientific papers; here's a good paper on the topic from 2004: https://www.researchgate.net/publicatio ... nipulation

Screenshot 2023-05-07 221519.jpg (123.44 KiB) Viewed 4056 times

There is a very real effect sometimes called "simultaneous contrast",
where discernment of light detail is confounded by a dark background and vice-versa.
That is one reason why gray photo mat is popular for framed photographs.

To some extent, your application of dropped shadow has that effect.

That's interesting though despite seeing gray used as background canvas for any number of photoediting and displaying softwares, I have never, personally, seen gray matting used to display images in, say, a gallery or museum. Maybe the effect becomes a problem in transmitted light viewing vs. reflected light viewing. That's a cool image map, BTW!

Editing is a part of the process. It is up to the author's taste how much and in which direction.

Are they under any ethical obligation to inform the viewer of their editing?
I guess I'm asking where, in your mind, does "photo-journalism" end and "art" begin and whether or not we should note the difference in our images.

[Scarodactyl]

I don't think basic post processing requires any particular disclosure. It's only important if it's important, which is going to be contextual. As for misconduct what we actually see pursued are crude cut'n'paste jobs on western blots and other outright fabrications which add central details which are not present in the original data. Sharpening, focus stacking or enhancing contrast are unlikely to even enter the conversation.

[imkap]

I guess I'm asking where, in your mind, does "photo-journalism" end and "art" begin and whether or not we should note the difference in our images.

Well if you added an organ to a rotifer, probably you should not claim it is a mutation...

Post processing is needed because camera can get things wrong by itself, e.g. it can completely miss the white balance, the colors, etc. That's why we have RAW, so we can make a photo look like we think it should, more natural or more appealing. Maybe you'd like to emphasise a feature by certain image manipulation, or delete distracting dust or debris etc.

[MichaelG.]

Most of the stunning pictures even in publications like Nature are not "out of box". To give you an example:
https://de.wikipedia.org/wiki/Fluoresze ... tCells.jpg
It is impossible to see that pictures through the eye-pieces of a microscope. The emission of DAPI is in the same color range as the excitation of FITC. The picture is obviously layered in Photoshop. The blue, green and red parts are from different pictures and mounted together. Looks great but is synthetic and far away from "out of box".

I can’t really see any problem with that one ^^^
It is what it is : a beautifully executed ‘visualisation’ of a structure that would otherwise be incomprehensible.
… and the accompanying text gives a good description.

MichaelG.

[blekenbleu]

I have never, personally, seen gray matting used to display images in, say, a gallery or museum.
Maybe the effect becomes a problem in transmitted light viewing vs. reflected light viewing.

Many art works will have been created with e.g. white surround and/or spot illumination anticipated.

That's a cool image map, BTW!

In that example, pixels are simply arithmetic differences between RGB component values, with 50% gray added;
non-50% gray values correspond to differences. If source images were perfectly registered,
that would indicate drop shadow affected more than background values.

For some difference images, more representative changes involve converting values to e.g. CIE Lab,