Why process digital images?

Here you can discuss topics such as focus stacking, stitching and other techniques that relate to the processing of micrographs.
Post Reply
Message
Author
User avatar
gekko
Posts: 4701
Joined: Sun Oct 12, 2014 7:38 am
Location: Durham, NC, USA.

Why process digital images?

#1 Post by gekko » Mon May 30, 2016 2:31 pm

There are sometimes comments made about post processing digital images that seem to question the the need or desirability of such activity, and impying that posting jpeg images straight of the the camera, presumably just downsized to fit the forum image size, would result in truer images. I would like to try to explain why, as far as I know, the practical aspects of digital imaging makes this not so.

I think it would be too much to expect that the resolution you see through the eyepiece can be maintained when the digital image is reduced to a small fraction of its original resolution after it is downsized to fit the monitor screen. The original resolution seen and recorded by the camera is typically currently in the range of 12 to 24 megapixels. On the screen it is less than one megapixel. One needs a much larger, much higher resolution monitor to display the digital image at nearly sufficient resolution in order to approximate the view through the microscope; most of us don't possess such a monitor. In addition, the digital image is a result of digital (discrete point) sampling of the non-discrete analog image by a finite set of pixels at a finite spacing. In order to avoid artifacts due to aliasing (and this is true as well for audio digital sampling, or any digital sampling of a continuous signal), a low pass filter is used that reduces the image sharpness, hence the need for sharpening (either in post processing or in the camera itself). Digital cameras that produce jpeg (or tiff) image files will have sharpened those images as part of converting the digital data into an image. Point-&-shoot cameras generally tend to produce over-sharpened, over-saturated, contrasty images. The better cameras have provision for the user to adjust the amount of sharpening, contrast, and saturation. RAW data are not sharpened and must be sharpened when converted to an image if they are not to look blurry. The act of downsizing the image (from the original multi-megapixel size to less than one megapixel or so necessary to fit on the screen) itself results (in addition to the loss of resolution) in further blurring of the image that requires sharpening. All part of digitally sampling an analog signal using a less than infinite number of pixels spaced at more than an infinitesimal distance apart.

We must remember that in the days of B&W film, we had the choice of many different types of film, each producing images of a distinctive character. Sharpness, contrast, and grain were controlled by the choice of film (Panatomic-X, Plus-X, Tri-X, etc.), through chemistry, e.g., by using a high acutance developer, and by choice of an agitation regime while the film is being developed. During printing, unsharp mask (a term borrowed into the digital process) was also used to enhance sharpness. Contrast (which can also affect the perception of sharpness) was "manipulated" through the choice of film, developer, development time and temperature, and, during printing, by the choice of contrast filter or the contrast grade of printing paper. The introduction of color (positive and negative) films multiplied the choices available, and in addition to the differences in resolution, and gradation that existed between different B&W films, they presented very significant differences in their images, including saturation and color rendition: Kodachrome images looked very different from Ektachrome or Fujichrome or Agfachrome.

At least that is how I see it.

_________________
http://s1070.photobucket.com/user/micritter/profile/
https://vimeo.com/user30158560/videos

JimT
Posts: 3247
Joined: Fri Oct 24, 2014 1:57 pm

Re: Why process digital images?

#2 Post by JimT » Mon May 30, 2016 2:56 pm

I agree completely. Enhancing the image to bring out the most information available in post processing is to take advantage of all the tools at one's disposal. One can always do a before and after for comparison if desired.

User avatar
gekko
Posts: 4701
Joined: Sun Oct 12, 2014 7:38 am
Location: Durham, NC, USA.

Re: Why process digital images?

#3 Post by gekko » Mon May 30, 2016 3:33 pm

Thank you, JimT.

User avatar
75RR
Posts: 6261
Joined: Sun Oct 12, 2014 2:34 am
Location: Estepona

Re: Why process digital images?

#4 Post by 75RR » Mon May 30, 2016 4:06 pm

I also agree.
The position that an unvarnished digital image straight from the camera is somehow natural is simply not true.
Additional Post-Processing in order to provide viewers with the best representation of the subject seen through the microscope is not only perfectly valid but is also many times necessary.
Zeiss Standard WL (somewhat fashion challenged) & Wild M8
Olympus E-P2 (Micro Four Thirds Camera)

kinase
Posts: 66
Joined: Wed Jul 29, 2015 3:18 am

Re: Why process digital images?

#5 Post by kinase » Mon May 30, 2016 5:15 pm

I don't know why people get upset by post processing, and never think about preprocessing either. Even Old darkroom stuff had lots of burning and dodging, and even cloning in the instance of some old Nazi photos. I really miss the medium format camera I used to shoot with but the film and the developing are to much now, and I even did the black and white myself.

It's even weirder that with the confocal I use I can very easily hide or show things depending on the laser power and detector voltage and a bunch of other settings. Even if a fluorescence signal is barely there, I can show it and then eliminate the noise created by doing that. Plus deconvolution software is intense computer editing.

User avatar
gekko
Posts: 4701
Joined: Sun Oct 12, 2014 7:38 am
Location: Durham, NC, USA.

Re: Why process digital images?

#6 Post by gekko » Mon May 30, 2016 6:15 pm

75RR and kinase: many thanks for your comments. I think there is an implicit trust that the camera records exactly what it sees, i.e. the TRUTH (as 74RR implied) and, therefore, any modification of the image, by definition, results in an untrue representation of the object. The thought that the digital camera does not record actual images but rather a matrix of numbers which are then converted automatically into an image by an algorithm developed by engineers to give what they believe to be an image of an "average" scene that would please most people may be a bit uncomfortable.

billbillt
Posts: 2885
Joined: Mon Nov 10, 2014 10:01 pm

Re: Why process digital images?

#7 Post by billbillt » Mon May 30, 2016 7:00 pm

This worry(complaint?) seems to surface a few times a year.. Myself, I feel anything done to an image to "help" it is fair game.. I guess I fail to see why it is frowned upon by some people..

BillT

User avatar
vasselle
Posts: 2763
Joined: Sat Oct 25, 2014 5:32 pm
Location: France

Re: Why process digital images?

#8 Post by vasselle » Mon May 30, 2016 7:17 pm

Bonjour à tous.
Personnellement je photographie toujours en RAW pour les photos au microscope,et je les développes avec le logiciel Lightroom 5.
Car le RAW permet une plage dynamique plus importante.
Car il travaille sur 12 ou 14 Bit alors que le JPG travaille sur 8 Bit.
Et il permet de plus facilement rattraper les hautes lumières et les basses lumières,de corriger une balances des blancs mal faite,et c'est non destructifs pour la photo avec ce mode qui est le RAW.
Car le JPG malheureusement,plus on touche a image et plus il y a une destruction de celle si.
Le JPG je ment sert juste à la fin une fois que ma photo et terminée.
Cordialement seb
Microscope Leitz Laborlux k
Boitier EOS 1200D + EOS 1100D

User avatar
zzffnn
Posts: 2711
Joined: Sat Jun 20, 2015 3:57 am
Location: Houston, Texas
Contact:

Re: Why process digital images?

#9 Post by zzffnn » Mon May 30, 2016 8:11 pm

I don't have an issue with digital processing. Most people in this forum are pretty moderate/conservative when it comes to digital enhancement, which is a good thing to me.

Confocal needs processing, for example.

Also camera capture is sometimes less sharp than eyepiece view and digital enhancement brings photos closer to eye view. Nothing wrong with that.
Selling my Canon FD 200mm F/2.8 lens

apochronaut
Posts: 2733
Joined: Fri May 15, 2015 12:15 am

Re: Why process digital images?

#10 Post by apochronaut » Mon May 30, 2016 10:27 pm

If you are talking about me, Gekko, you have missed my point. Firstly this is a microscope forum, not a photography forum. We all know there is another one dedicated to photomicrography, so that is a good place for those who are very interested in many aspects of the photomicrographic art to be expressive.
However, there are a lot of individuals on this forum who are just feeling their way around a microscope and want some support, with which to improve their microscopic skills. I'm sure there are a lot who could care less about photomicrography. I for one am in the latter camp. When I share images, it is to share the subject matter, for the wonder of that subject matter. Photomicrography, is simply a tool to capture an event witnessed through a microscope. Digitally enhanced images can provide a clearer representation of that event for sure , and be kind to the person who posted them but they might not be so kind to those who are trying to work there way through the myriad of possibilities, that produced those images. Often, no one ever explains what they did to the image, or heaven forbid, show how terrible the image was before they, frigged with it. If it needs repeating, very little of the time, looking through a microscope is image captured. I'm pretty sure most people using a microscope are interested in learning to use a microscope to best advantage. Some don't even have photographic capability. Looking at artificially polished images, that in no way resemble what was actually seen, might be beneficial for photomicrographers but has somewhat less value to microscopists, irrespective of their experience level.
Astronomer's will spend an hour gathering enough light to photograph a star and go " look at that star" ! Well, no one actually ever saw the star but the pictures are great. Surely, microscopy is different than that.

User avatar
mrsonchus
Posts: 3582
Joined: Tue Feb 03, 2015 9:42 pm
Location: Cumbria, UK

Re: Why process digital images?

#11 Post by mrsonchus » Mon May 30, 2016 10:44 pm

As anyone that's seen my pictures will know, I nearly always use pp to 'improve' my images - the idea usually being to try to present something as close as possible to what I see through my 'scope to the viewer, with the emphasis on 'information' or you may say 'details' of tissue, although of course other facets are very important such as colour.

The idea of 'seeing the truth' through anything, my own eyes included, seems to me at least to be a touch nonsensical.... Not bad, just of little meaning to me personally.

I also (often) go far further into the realms of trying to drag information kicking and screaming from such an image using all sorts of weird and wonderful methods.... I for one consider most of my own images to be impressions of the tissue rather than 'truth'.

Here are a few examples, a straightforward 'enhancement'... sharper, cleaner and better balanced that that which emerged even from my Canon 18mp - true to colour also (as seen through my 'scope of course) - the purple isn't a 'true' image of the tissue either! :D
ws_orig_palisade.jpg
ws_orig_palisade.jpg (179.37 KiB) Viewed 5734 times
and a few mysterious shenanigans later bring the tissue, to my eyes & taste anyway, into the impression of more 3D tissue information...
ws_enhanced_palisade.jpg
ws_enhanced_palisade.jpg (153.65 KiB) Viewed 5734 times
Here's a manipulation that I think gives an increase in cellular 'spatial arrangement' information in 2 dimensions - I personally find this to be yielding additional information compared to the original image...
ws_enhanced2.jpg
ws_enhanced2.jpg (240.16 KiB) Viewed 5734 times
Now, to go quite barmy with effects I create images for chromosome 3D enhancement (I hope...) that again I certainly think adds a wealth of information to the original picture and definitely allows me to get a 'feel for' the finger-like chromosomes and their relative 3D arrangement/s as they go through mitosis...

I love these two images... (I measured the chromosomes to be about 0.7µ diameter from one of my 1µ sections a while back now)
The first is of condensed chromosomes within a cell, sectioned it appears 'through the top' of the 'chromosome basket' as I like to think of it.... To me this gives a very good idea indeed of what's going on in the cells and the structural nature (at a basic level of course) of the chromosomes - they remind me almost of SEM images that I see online....
ws_enhanced-chromosome-bask.jpg
ws_enhanced-chromosome-bask.jpg (125.62 KiB) Viewed 5734 times
and this one to me really helps to visualise the 'finger-like' structure seen as the chromosomes part from their 'sister' duplicates during division seen in my other images not altered in this way...
ws_enhanced-chromosome.jpg
ws_enhanced-chromosome.jpg (197.08 KiB) Viewed 5734 times
For my money, pp is a very valuable tool indeed in microscopy - the purist will think otherwise and of course has an equally valid point, just different. :)

Just by 2p - worth... :)
John B

JimT
Posts: 3247
Joined: Fri Oct 24, 2014 1:57 pm

Re: Why process digital images?

#12 Post by JimT » Mon May 30, 2016 10:53 pm

Apochronaut, you also raise some good points. For a beginner who looks at the many very good images posted here it could be discouraging. “It doesn't look like that in my microscope!”

Might be good if we all at least mentioned that the image was enhanced in post processing when we do it. All of mine are.

On the other hand nobody would care about Hubble images if they were not enhanced.

As I said on an astronomy forum 10 years age, good topic to discuss over a few beers :)

User avatar
mrsonchus
Posts: 3582
Joined: Tue Feb 03, 2015 9:42 pm
Location: Cumbria, UK

Re: Why process digital images?

#13 Post by mrsonchus » Mon May 30, 2016 11:05 pm

Agree with you 100% Gekko, implicitly also. :)

Of course the aim may even be to present an image more 'true' to that as seen through the 'scope in the first place? Surely a good way to share the experience; without pp the 'scope-view and captured image can be very different in my experience, then perhaps I just need the crutch of pp to make up for my lack of ability?

As for helping beginners, which included myself a little over a year ago so I think I have a little insight into the experience of starting from nothing and being nurtered, helped and encouraged by the fine fellows of this forum - the best bet is not to shroud everything in technobabble but to provide a record of what was done and how it was done, preferabley with images - sound familiar?

:)
John B

User avatar
gekko
Posts: 4701
Joined: Sun Oct 12, 2014 7:38 am
Location: Durham, NC, USA.

Re: Why process digital images?

#14 Post by gekko » Tue May 31, 2016 12:16 am

Thank you all very much for a very fascinating as well as illuminating discussion, with the many varied and valid points of view. The world would be a terribly boring place had we all had the same view on, and the same purpose in microscopy or photmicrography.

User avatar
mrsonchus
Posts: 3582
Joined: Tue Feb 03, 2015 9:42 pm
Location: Cumbria, UK

Re: Why process digital images?

#15 Post by mrsonchus » Tue May 31, 2016 4:33 am

gekko wrote:Thank you all very much for a very fascinating as well as illuminating discussion, with the many varied and valid points of view. The world would be a terribly boring place had we all had the same view on, and the same purpose in microscopy or photmicrography.
Well said - that's the whole point - difference doesn't equate with enmity - ever in my book my friend. :)

That said, it would surely be both informative and indeed a privilege to actually see some of the 'ultimate truth' images discussed here for us all, myself included, to better understand this side of the opinion-divide? I've posted many that give example to the pro-pp side after all......

Any chance of the produce (assuming it exists of course) of the 'purity' technique/s to be posted for us to actually see and maybe judge for ourselves rather than just be coached on their supremacy? Come on - show us some images, we've read enough rhetoric re their status as the holy-grail of imaging through the 'scope... Unless of course their representation on the forum would compromise their quality....
John B

Post Reply