Basic Photoshop for Electron Microscopy 2014 by Eric Miller Microscopy & Imaging Specialist at Northwestern University

8/12/2019 Basic Photoshop for Electron Microscopy 2014 by Eric Miller Microscopy & Imaging Specialist at Northwestern University

1/91

Basic Photoshop for Electron Microscopy

By Eric Jay Miller

Page | 1

Basic Photoshop For ElectronMicroscopy

(Mostly) Simple Techniques to Tune-Up Your ElectronMicrographs with Adobe Photoshop

Authored by Eric Jay MillerNUANCECenter

Northwestern University

http://www.nuance.northwestern.edu/

Eric Jay Miller 2014


2/91


By Eric Jay Miller

Page | 2

Introduction

This paper will discuss several simple procedures that can help to punch yourmicrographs up a notch. These instructions are written in such a way that just about

anybody should be able to pick them and start using them without any previousexperience with Photoshop (although that certainly wont hurt either).

This is not intended to be (nor does the author claim it to be) the end-all and be-all of Photoshop instructions. There are often multiple ways of doing the same operationin Photoshop. These are just a few that work pretty well for this particular application ofelectron micrographs. There might be other procedures that work better for specificapplications (or work just plain better). The reason for the omission of any of thosetechniques is due to the author not having been smart enough to have thought of themprior to writing this.

The program used for this paper was Photoshop CS4, though most olderversions of Photoshop will work perfectly fine with mostof these techniques.

In electron microscopy there are often concerns about the authenticity of animage once it has been processed through Photoshop. These can be valid concerns asthere are a lot of really crazy things one can do with this high-powered program. Themain purpose with the techniques discussed here is to either a) make the information inthe image easier to see, thus yielding more usable information, or b) to make the imagemore aesthetically pleasing. These instructions are geared towards image processingand notimage analysis.


3/91


By Eric Jay Miller

Page | 3

Contents

I. Levels 4II. Simple Pseudo-ColorColor Table 8

III. Simple Pseudo-ColorGradient Map 12

IV. Complex Pseudo-ColorColor Table 18* Addendum to Section IV 31

V. Calculations 38VI. Combining EDS Color MapsCut and Paste 42

VII. Multi-Detector Color 49* Tips & Tricks for Multi-Detector Color 59

VIII. Drawing a Mag Line Marker (Pre-Existing Mag Line) 73IX. Drawing a Mag Line Marker (The SUPERCheating Way) 78

Appendix I The Short Versions 82


4/91


By Eric Jay Miller

Page | 4

I. Levels

The most basic thingone can do to improve the overall look of a micrograph is to adjustthe Brightness and Contrast levels. With Adobe Photoshop, this is a simple, quick and

very accurate procedure.

1.Open your image inPhotoshop.Use:

File > Open

or Ctrl+O

2.Make sure that your image is an 8 BitGrayscale image by clicking on

Image > Mode

Grayscaleand 8 Bits/Channelshould havechecks next to them.


5/91


By Eric Jay Miller

Page | 5

3.An ideal grayscale image has an area thatis absolute black, an area that is absolutewhite and has the maximum shades of grayin between as possible. Even if the image wetook on the microscope is not ideal, we can

usually even it out, at least a little.

Open the Levels dialog box by clicking on:

Image > Adjustments > Levels

or Ctrl + L

4. In the middle of the Levels box there willbe a histogram that shows how the 255shades of gray in your grayscale image are

being used. The extreme left of the histogramrepresents black while the right siderepresents white.Underneath the histogram are 3 smallarrows. These are set points for absoluteblack, absolute white and the shade of grayin between the two.

5.What you will want to do is bring the black

triangle and the white triangle in so that thereis no empty space. In the example, there isan empty space in between the main body ofthe histogram and a peak on the right end.Sometimes a peak on the fringes can beomitted by bringing the triangle past the peakto the main body. But in this case, we willleave it.


6/91


By Eric Jay Miller

Page | 6

6.Next we will set the midpoint. Make surethat the Preview box (which is in the lowerright corner of the Levels box) is checked.Drag the gray triangle back and forth untilthere is a good balance between light and

dark in your image and click OK.

7. If we go back and take a look at the Levelsbox now we can see that Photoshop hasstretched the histogram so that it now usesalmost every shade of gray, whereas before itwas only using about half. Also, the darkestcolor in the unprocessed image was almosthalfway to absolute white and now it is nowclose to absolute black. This will add a lot ofextra definition and visual punch to image.

NOTE:By saving your image, youll make these changes we made PERMANENT! Youmight not want to do that. A good idea is to always keep your original, unprocessedimage, in case you want to do something different to it later on. So once we haveprocessed an image, perform Save Asto save your file as a new file.

File > Save As

Or Shift + Ctrl + S

Then when you close your original file, PS will ask you if you want to save it, SAY NO!You should end up with 2 files, one is the original, the second is the one you did stuff to.


7/91


By Eric Jay Miller

Page | 7

Here is the image before processing. It is not horrible, but it is a little dull.

This is the image after the simple Levels procedure. It now has much more definitionand is much snappier and appealing to the eye.


8/91


By Eric Jay Miller

Page | 8

II. Simple Pseudo ColorColor Table

Lets say we wanted to add a little color to our black and white image. This is a superquick and easy way to get a color image.

1. Well start by opening our image andmaking a copy of it by clicking:Image > DuplicateClose your original image.We dont reallyneed to this, but it will preventanything accidentally happening to ouroriginal image.Once thats done well want to perform theLevels adjustment that we did in Section I.

2.With this technique we are going to give agrayscale image fake color. In the examplethere is a grayscale SEM micrograph of golddust that we will make gold colored.We want to make sure our image is inIndexed Colormode:

Image > Mode > Indexed Color

3.Next click on:

Image > Mode > Color Table

This will bring up the Color Table dialog box.


9/91


By Eric Jay Miller

Page | 9

4.The Color Tablebox shows what color isbeing applied to each of the 256 possiblecolors in the image. If you want, you can clickon each box and specify each color. Thatwould be insane. Instead, well do something

easier.Click the top left box (which should beabsolute black) and hold the mouse buttondown and drag it down to the box in thebottom right corner (absolute white) andrelease the mouse button.

5. When you release the mouse button theColor Picker box will appear. Since blackwas the color at the top of the Color Table in

the last step, well need to start off this stepby picking the color you would like torepresent the darker areas of your image.You can do that just by clicking on it. In thiscase, we will go with black. Click OK.

NOTE:If you picked the light color first when you should have picked the dark color first[or vice versa] it will be obvious pretty quickly as your image will look like a negative(with areas that should be dark being light, etc.). If this happens, just hit cancel on the

Color Table and start over from Step 4.

6.Now choose which color you would like torepresent the lighter areas of your image. Inthis case, we want gold to be colored gold, soa darkish yellow is selected. Click OK.


10/91


By Eric Jay Miller

Page | 10

7.The Color Table will come up again andshow you how your 2 colors have beenstretched to fill all 256 colors. Make sure thePreview box is checked and if you like how itturned out, click OK.

You should be done! Saveit!


11/91


12/91


By Eric Jay Miller

Page | 12

III. Simple Pseudo-ColorGradient Map

The color table in Section II is great for very quick, single-color processing. This methodwere going to discuss is a great quick and not terribly dirty method which can involve

multiple colors.

1. Well start by opening our image andmaking a copy of it by clicking:Image > DuplicateClose your original image.We dont really need to this, butit will preventanything accidentally happening to ouroriginal image.

2.Were now going to switch the color modeof our image from Grayscale to RGB Colorby clicking

Image > Mode > RGB Color

3. Perform the Levels operation on theimage.


or Ctrl + L


13/91


By Eric Jay Miller

Page | 13

4. Now were going to open the GradientMapby clicking

Image > Adjustments > Gradient Map

5.The Gradient Mapbox should pop up andour image is suddenly in whatever awful colorthe default settings are set to. In my case, itsthis dull red. Were going to fix that byclicking that color gradient box thing in the

middle there.

6.Up pops our Gradient Editor now. At thetop we have a number of preset colors wecan try out. I can guarantee you that usingany of the presets will make your image lookawful.Fortunately thats not the point. The presetwill give you a good starting point to workfrom. I want to start off with somethingsimple, like 2 colors. So Ill click on a 2 colorpreset, like that green/red one, and well seewhat happens.

7.As I predicted, my image looks like somesort of Christmas decoration nightmare. Butlike I also said, thats ok.


14/91


By Eric Jay Miller

Page | 14

8. If we take a look at the bottom edge of thenow Red and Green gradient bar, well noticea little red box and a little green box. Theyrecalled Color Stops. If we click on either ofthese boxes, we can select our color. Well

start off by clicking the red Color Stop on theleft. Well then click the box below it that saysColor(and our current color is of course red,so the box will be red).

9.Once we click the Color box, the ColorPicker should open up. Instead of red, Imgoing to hunt around until I find a nice shade

of dark blue that I like. Then click OK.

10.Now well go back and click the littlegreen Color Stopon the lower right corner ofthe gradient bar and then click the (nowgreen) Colorbox in the lower left of thewindow.


15/91


By Eric Jay Miller

Page | 15

11.The color picker comes back and wellpoke around until we have a nice, lightercolor. I just went with a lighter shade ofgreen. Click OK.

12.Now of course if we dont like what wevegot, we can always go back and keepchanging our colors until we have somethingwe DO like.

Anotherthing we could do is move those littlesliders around. If we bring the blue or thegreen box toward the center of the gradientbar, it will intensify that color. The little whitediamond in the middle sets the midpointbetween the 2 colors. (If the grays were welladjusted when you started off this method,you probably wont need to dothis.)

Anyway, if we like what we have, well click OK.

13.And click OKagain.

You may want to do Levelsagain at thispoint. Maybe not.

NOTE:There are a few other ways we can play with our image. One of them is byselecting one of the more complex presets.


16/91


By Eric Jay Miller

Page | 16

As we can see here, we now have 4 colors towork with instead of 2. As per our previousinstructions, we can change each of thesecolors and drag the little Color Stopboxesaround to intensify or de-intensify each of the

colors.You can easily spend all day picking colorsand playing around with what part of yourimage they show up in.

14.Whenever you have something you like,go ahead and Saveit!

File > Save

Or Ctrl + S


17/91


By Eric Jay Miller

Page | 17

Our first image was this perfectly nice B/W offering.

What we got in the end is a reasonable 2 color image. The important part is that it wasFAST!


18/91


19/91


20/91


By Eric Jay Miller

Page | 20

7.We should end up with 3 colorized images.

Now we need to decide which areas of which images we need to cut out.

The sky in the blue image seems to have the largest surface area, so Im going todesignate it as the background image. So all Ill need to do is cut the sun and the raysout of the yellow image and the ground out of the green image and then paste them on

top of the blue image.

8.There are a couple of ways we can selectthese areas. It depends on the amount oftime you have and the amount of effort youwant to put into it.

The bestway is by using the Quick Mask.

We enter Quick Mask Mode by clicking onthe Quick Mask button, or Q on thekeyboard.

9. Well do the selection process in2 stages.

FIRST, we select the Penciltool.

Explanation of what were doing: Bydrawing over an area in the Quick Maskmode, the area will turn red. When we exitthe Quick Mask,the area NOT painted over

will be selected.


21/91


By Eric Jay Miller

Page | 21

10.We can choose the diameter of our pencilby clicking on the little down arrow next to theBrush size indicator in the upper left cornerof the screen.

We need to select a size that will allow us tocover a lot of territory and yet still be smallenough to fit inside where we want to paint.

NOTE: Youll probably need to change thediameter a few times as youre working.

11.Since were on the yellow image, we wantto take out the sun part of this image. Byusing the pencil tool, were going to paint out

ONLY THE CENTERof the sun and the rays,but AVOID THE EDGES.

As we can see here, areas that you selectshow up as red.

This can be a long, tedious process.

NOTE: Keep in mind that if the featuresyoure trying to select are too small to seeclearly, you can always zoom in by selecting

View > Zoom In

or Ctrl + +

You can do this multiple times.

Then, to zoom out, select

View > Zoom Out

or Ctrl + -


22/91


By Eric Jay Miller

Page | 22

12. Once we have the center parts bulkedout with the Penciltool, were going to selectthe Brushtool and use it to go over all of theedges.

And again, you may need to change theBrush size like we did with the Pencilin Step10.

NOTE:Theres a good reason for using theBrushon the edges instead of the Pencil.

The Penciltool will make hard edges around

the area we select. For some things, thatsok. But when we cut this area from this imageand drop it on top of another image, theabrupt change of colors might not look verygood. We can see the Pencil edges wherethings were painted over are pretty obviousand are a bit abrupt.

The Brush tool will give us a softer edgearound our selected area which should blenda little bit with any adjoining color areas.

NOTES 1s NOTE:So I know your questionnow is why dont we just use the Brush toselect EVERYTHING? Well, because of thenon-solid nature of the Brushtool its easy tomiss things with it.

As we can see with this image, I used thebrush to select this area. I went over everypart of it, but you can see some parts are alittle thin. Those thin areas will show up astransparent when we copy and paste thisarea on top of our background. We dontwant that.


23/91


By Eric Jay Miller

Page | 23

NOTES 2n NOTE: There are certainlyinstances where we can use the Pencil toolon everything. It completely depends on howwell defined your image is. This image forexample, has only 2 color areas and the

edges on the structure are pretty nice andsharp, so the Pencil tool was used foreverything and it looked pretty good.

13. Once youre done with all the painting,you should have something that looks likethis. Areas you want selected are red.

NOTE: If you had a major (or minor) screwup while you were painting out your area,youll want to fix it.We can do that easily enough by exiting theQuick Mask (Q) and inverting the selectedarea.

Select > Inverse

or Shift + Ctrl + I

Turn the Quick Mask back on and you canfix any areas you might have strayed outside the edges of your structure.


24/91


By Eric Jay Miller

Page | 24

14. When youre done painting, exit the QuickMask by clicking on the Quick Maskbutton or Qon the keyboard again. Now everything that wasNOTred is now selected.

This is backwards of what we wanted, sowell need to invert the selection.

Select > Inverse

or Shift + Ctrl + I

15.Now well copy the area we selected.

Edit > Copy

or Ctrl + C

16.We said earlier we were going to use theblue image as our background, which meanswe want to paste our just copied parts from

the yellow image into it.

Before we do any pasting, we need to selectour blue background image and switch it fromIndexed Colorto RGB Color.



25/91


By Eric Jay Miller

Page | 25

17.Now with the blue image selected we cango ahead and paste our yellow selection ontoit.

Edit > Paste

or Ctrl + V.

It probably wont be lined up correctly, but wecan leave it alone for now.

18.For now we need to go back to our greenimage and paint out the parts we want fromit.

As before (Steps 9 thru 17) we enter theQuick Mask, use the Pencil for most of itand the Brush tool for the edges, exit theQuick Mask mode, Invert the selection,Copy, then Pasteit into the blue image.

19. So now we have our image with all 3colored areas in it. BUT, theyre a bit

misaligned.


26/91


By Eric Jay Miller

Page | 26

20.Well start aligning them by going to theLayers tab in the lower right corner of thescreen. We start off by selecting the layer wewant to move around.

The most recently created layer should beselected (the one from the green image) andthats as good a place to start as any.

21. Select the Move Tool, or V on thekeyboard.Using the MoveTool, you can click on yourlayer and drag it around the image and

position it where you want.

22. This layer was easy to position, since itshares 3 edges with the background image.(The edges of the layer should snap to theedges of the image.) The next layer will be alittle trickier.


27/91


By Eric Jay Miller

Page | 27

23. Well select the yellow layer from theLayerstab as in Step 20. Because this layerdoesnt fit perfectly against at least 2 cornersof the image like the last layer did, well needto do some fine positioning. To do that, its

easier if we set the layer to 50% opacity, sowe can see it andthe layer behind it.

You can do this by clicking on the Opacitybox there in the Layerstab and either typingin 50 or dragging the slider bar around. (Itdoesnt need to be exactly 50%).

24.We can then use the Movetool to get thelayer mostly lined up, then we can use thearrow keys on the keyboard to nudge the

layer around by single pixels until its all linedup. Then well bring the opacity of the layerback to 100%.

25.Before we save our image, we have onelast thing well want to check. Some of ourcolor layers might be overlapping a bit. Wemight want to check to make sure that theyoverlap the way we want them to. Lets lookcloser at this area on our image where thesun and the ground meet.


28/91


By Eric Jay Miller

Page | 28

26.We can arrange which layer is on top of orbelow others by dragging them around in theLayerstab.

26.The image on the left has the yellowlayer on topwith the green layer behind. Theimage on the right has the greenlayer on topwith the yellow layer behind.The right image looks the smoothest, so Ill stick with that arrangement.

27.OK, weve made it to the end here andDISASTER! We dont like how the colorturned out in one of our areas. No sweat.Go back to the image whose color you wantto change. The area you wanted should stillbe selected. Leave that alone, dont deselectanything. All you need to do is go back to thecolor table, change the colors, copy thealready selected area and paste it into thefinal file. You can then determine if that wasbetter or not. If it was, you can throw awaythe less acceptable color layer.


29/91


By Eric Jay Miller

Page | 29

28.We should have a completed image now,we can go ahead and save it as whatever wewant.

File > Save

Or Ctrl + S


30/91


By Eric Jay Miller

Page | 30

Here we have our original (slightly boring) sunrise over Arizona.

In the end, we have ourselves this nice colorful image that really kind of looks likesomething! (The colors dont always need to be this happy, thats just how I ended upmaking them.)


31/91


By Eric Jay Miller

Page | 31

*ADDENDUM to Section IVComplex Pseudo-Color

So I mentioned at the beginning of Section IVthat there are a number of ways that we

can go about grabbing different areas of our images to throw together later into a finalcolorized image. Here we will discuss some of those other ways. They probably wontbe better than using the Quick Mask, but they may end up being a little faster.

TO START:Well follow the instructions in Section IVup through Step 7. Now insteadof using the Quick Mask, well look at the use the Quick Selection Tool (Section A)and the Lasso Tool(Section B). Well start off with the Quick Selection Tool.

A. Quick Selection Tool

A1. In CS4 the Quick Selection Tool isgrouped with the Magic Wand. If the QuickSelection Toolis not showing, you will needto right-click the Magic Wandto select it.

Once its selected, make sure the NewSelection option is selected on the toolbar.

Also on the toolbar, make sure the brush sizeis set small enough to fit into any small areasyou want to select, but large enough that

youre not painting withit all day.

A2.Lastly before stating, we need to changeall of our already colored images fromIndexed Color mode to RGB Color mode.



32/91


By Eric Jay Miller

Page | 32

A3. Take the Quick Selection Tool anddraw over the areas you want to select. Asyou draw, the selection will flood out acrossyour image until it finds what it perceives asthe edges of your object. This is not always

perfect. When doing this on my image, I triedto just select the sunbeams and ended upwith the whole top half of the image selected.

A4. To fix that, we just go back up to thetoolbar and click on the Subtract fromselection option. We might also select asmaller brush size by clicking the little downarrow just to the right of where it says Brush.

A5. By going back and forth between theAdd to and Subtract from options we can

make sure that only the areas we wantselected are. The speed of this step dependsgreatly on how well defined the area youwant to select is and how nice you want it tobe.

From this point, you can continue on with thenormal procedure lined out in Section IVfrom Step 15onward.


33/91


By Eric Jay Miller

Page | 33

NOTE: Electron micrographs are not oftenknown for their well-defined edges, and thatwill make this technique difficult.

Another major issue you might have in usingthis tool is that in going back and forth

between adding and subtracting differentareas, sometimes youll end up with stray bitsleft selected that you didnt want. In this image for example, we have a big green linegoing through the image. If I pasted this into the main image with the other color layers,Im going to have this green line along the edge of one of my sunbeams. Thats notcool. So I would need to go back through with the Quick Selection Tooland make sureI got rid of those adequately. But sometimes theyre very difficult to see until its too late.

SECOND NOTE: Instead of the QuickSelection Tool, why not use the MagicWand dummy? You CAN!But youll probably

end up with a border of some kind around thearea you selected. (This sometimes alsohappens with the Quick Selection Tool.)I took out the background of this image andyou can see a dark line all around ourstructures. If it were colorized andrecombined with a colorized background, thatline is going to be EXTREMELYnoticeable.

SECOND NOTES NOTE:As another aside,you CAN select large areas of your imagewith the Quick Selection Toolor the MagicWand and then go over all the edges againwith the Pencil or Brush tool. Sometimesthat might save you time, sometimes juststicking with the Pencilis best.

B. Lasso Tool


34/91


By Eric Jay Miller

Page | 34

B1.Now well look at the use of the LassoTool. This will probably be the fastesttechnique and can give pretty decent results.

Select the Lasso Tool.

Set the Featherto ~5 pixels to start. (Thiswill make the edges of your selection areasofter. You might need to play around withthis setting depending on your results.)

NOTE:Another option might be to use theMagnetic Lasso Tool.

With this tool, the selection area willautomatically snap to what the softwaredetermines are edges.

NOTE 2:As cool as that sounds, this might

end up being more of a hindrance than ahelp, depending on how sharp the edges ofyour structure are.

We can see that this selection area will needa lot of cleanup work to make it acceptable.


35/91


By Eric Jay Miller

Page | 35

B2.Take the Lasso Tool, hold down the leftmouse button and CAREFULLYdraw aroundthe area you want to select.When you release the mouse button theselection will snap back to where you started.

So be careful where you do that. We can seein this example, I released the button toosoon and it cut off part of the sun that Iwanted to select.

B3.To select multiple areas ORto fix missedareas like in Step B2, hold down the Shiftkey, then start drawing over the new area toselect.

If youre doing this to fix missed areas, youllwant to overlap the existing selected area asyou draw.

B4.Then when you release the mousebutton, the overlapping areas combine toform a single area.

Start

End


36/91


37/91


By Eric Jay Miller

Page | 37

This is our first image that wecombined using the QuickMask. All the edges look niceand smooth and pretty.

This is our image we did withthe Quick Selection Tool. Theedges are a bit rougher andmore stark. With a quick glanceit seems ok, but upon closerinspection it really doesnt lookthat great. As I said previously,it all depends on how muchtime you have and how muchwork youre willing to do.

Lastly we have the image wecombined using the LassoTool. It looks like theFeathering was set a little bittoo high, as all the color edgeslook very fuzzy. Theres also anarea where the yellow sun and

the green ground overlap thatthe blue sky is showingthrough.

All these issues are fixable bybacking up a few steps.This method with the LassoToolis probably the fastest.


38/91


By Eric Jay Miller

Page | 38

V. Calculations

This process is solely for images that have uneven illumination. This is a commonproblem in many TEM micrographs and can also be encountered in low mag SEM

images. The results of this technique can be very dramatic in repairing otherwiseunsalvageable images.

1.Once your unprocessed image is open inPhotoshop, make a duplicate of it by clicking:

Image > Duplicate

2.Now, make another duplicate by repeatingthe previous step. Close your original image.(You really only need 1 duplicate image forthis, but by making 2 copies and closing your

original, this ensures that nothing willaccidentally happen to it.)You should have 2 images namedfilename_copyand filename_copy_2.


39/91


By Eric Jay Miller

Page | 39

3. Make sure the copy_2 image is selectedand click:

Filter > Noise > Median

4. The Median dialog box will appear. Drag

the slider all the way to the right so that theRadius is 100 pixels. Click OK.

5.Click:

Image > Calculations


40/91


By Eric Jay Miller

Page | 40

6. The Calculations dialog box will open.First, make sure the preview box is checked.Starting at the top, choose copy_2 for yourSource 1 image and check the Invert box.Next, make sure the first copy is your Source

2 image.Now, click on the Blendingdrop down menuand scroll through the different options untilthe illumination in your image appears themost even. In the case of the example it wasthe Vivid Light option.Finally under Result choose NewDocument. Click OK.

This is a subjective, trial-and-error process. You may have to switch the images fromSource 1 to Source 2 and go through all of the Blending options several times before

you find the best result. You can even switch which image is inverted. (This rarely worksthough, but give it a try anyway. As a rule, the image that you performed the Median onshould be inverted.)

7.You should now have a new third image.Now and only nowshould you finally adjustthe Levels. Adjusting them at the beginningof the process can interfere with getting evenresults with Calculations.


41/91


By Eric Jay Miller

Page | 41

Here is the original image. There is a large amount of vignetting around the bottom ofthe image, nearly omitting information.

Here is the image after Calculations. The illumination is now much more even acrossthe whole image.


42/91


By Eric Jay Miller

Page | 42

VI. Combining EDS Color Maps - (Cut and Paste)

If you have a bunch of different EDS color maps and you want to combine them into asingle image, this is one of the ways you can do that. This technique works best if your

images do not have a lot of overlapping areas of color.

1. Open all your EDS color map images inPhotoshop by clicking:

File > Open

or Ctrl + O

2. Were going to create a new image, but wewant it to be the same size as the otherimages we just opened. By following thisprocedure, we ensure that they will be.

Select one of your opened color map imagesand click:

Select>All

or Ctrl + A

A moving dotted line should appear all the way around your image.

3. We now need to copy the image byclicking:

Edit > Copy

or Ctrl + C

Were not going to paste this imageanywhere right now, but by copying it weve

placed the image information on theclipboard. When we go to make a new imagenow, Photoshop will automatically make thenew image the same size as the image on the clipboard.


43/91


By Eric Jay Miller

Page | 43

4. Now we can create a New Image byclicking:

File > New

orCtrl + N

5. Your new image is probably white. Blackmakes a far better background for our colormaps, so we will need to do a fill.

First we will need to make sure theForeground Coloris set to black by clickingthe Default Foreground and BackgroundColorsbutton, which is the small black/whitebox below and to the left of the actualforeground/background color display. Or hitthe Dkey.

6. Select the Paint Bucket tool. (The PaintBucket is often grouped with the GradientTool. If the Paint Bucketis not showing, youwill need to right-click the Gradient Toolandselect the Paint Bucket.)

7. With the Paint Bucket cursor, click onyour new image to make it black.


44/91


By Eric Jay Miller

Page | 44

8.Select one of your color map images.

9. On the left side of the screen, select theMagic Wand tool. (In CS4 the Magic Wandis grouped with the Quick Selection Tool. Ifthe Magic Wand is not showing, you willneed to right-click the Quick Selection Tooland select the Magic Wand.)

With the Magic Wand selected, set theToleranceto 0 by selecting the text box with

the cursor and typing it in. Setting theTolerance to 0 means that the Magic Wandwill only select the exact color you click on.Sometimes that is too stringent. If that is thecase, simply increase the Tolerance leveluntil you get what you need.

10. With the Magic Wand tool, click on anarea in your image that is black.


45/91


By Eric Jay Miller

Page | 45

11. If your image is too small to tell if youreclicking on a black area or a color area, zoomin by selecting

View > Zoom In

or Control (+) +

You can do this multiple times. To zoom out,select

View > Zoom Out

or Control (+) -

12. An animated dotted line should appeararound all the black areas in your image now.

If it does not select ALLthe black areas, holddown the Shift key and select any areas thatwere missed.

13. In reality, we dont care about the blackpart of the image, we want the color selected.So, we will need to invert the selection byselecting

Select>Inverse

or Shift + Ctrl + I

14.We now want to copy the color portion ofthe image by clicking on

Edit>Copy

or Ctrl + C


46/91


By Eric Jay Miller

Page | 46

15.Select your new image and Pastethecopied image into it by clicking

Edit>Paste

or Ctrl + V

Repeat steps 8 through 15 on your remainingimages. I have 3 in my example.

16.You will now need to arrange your layersto find the best order. You can order yourlayers by dragging them up and down in theLayers menu in the lower-right part of thescreen.

17.What does this do? For example, I knowthat my blue layer has a lot of extra noiseacross the rest of the image. Here is what itlooks like with it in front.

Surprise! It looks noisy.


47/91


By Eric Jay Miller

Page | 47

18.By putting my red layer on top, we donthave to see all that blue noise. (You mighteven want to paint the noisy areas black soyoure left with just the large blocks of color.[Do this with the Penciltool before Step 10.]

The noise is not important information in myimage, but it might be in yours. The pointhere is, do what you need to do to get theinformation across in the best way possible.)

Youre done, Saveit!


48/91


By Eric Jay Miller

Page | 48

Here are my original 3 images.

Here is the final combined image.


49/91


By Eric Jay Miller

Page | 49

VII. Multi-Detector Color

Lets say we took several images with different detectors (or different segments from abackscatter detector) of the exact same area. We can take those different images and

combine them with color to get a very special kind of image. Basically what were doinghere is making each detector a light source and assigning a particular color to it. It willmake sense by the time were done.

If you only have 1 SE detector, this procedure IS STILL POSSIBLE! Check out theTips, Tricks and Sugg est ionssection at the end of this section.

ATTENTION:This may look like a long and complicated procedure at first, but its reallyquite simple and once youre familiar with it, can be done very quickly. The timeconsuming part is really taking the images.

1. Open all your images in Photoshop byclicking:

File > Open

or Control + O

2:In this example, Ive got 1 SecondaryElectron Image and 3 Backscatter Electron

Images (taken from different segments of a 4segment detector).

Sometimes images from different detectorswont line up everywhere across the imageexactly. If you have a backscatter detectorwith 2 or more segments, you can take allyour images from it, which this makes thisprocess a bit nicer (because unless thesample is drifting, they should line up prettywell).


50/91


By Eric Jay Miller

Page | 50

3. Were going to create a new image, but wewant it to be the same size as the otherimages we just opened. By following thisprocedure, we ensure that they will be.

Select one of your opened images and click:

Select>All

or Ctrl + A

A moving dotted line should appear all theway around your image.

4. We now need to copy the image byclicking:

Edit > Copy

or Ctrl + C

5. Now we can create a New Image byclicking:

File > New

or Ctrl + N

6. Before you click OK to make your newimage, make sure RGB Color is selectedunder the Color Mode drop box.


51/91


By Eric Jay Miller

Page | 51

7.Your new image is probably white. Thoughit might not be necessary, black makes for abetter background for what were going to do,so we will do a fill (just in case).

First we will need to make sure theForeground Coloris set to black by clickingthe Default Foreground and BackgroundColorsbutton, which is the small black/whitebox below and to the left of the actualforeground/background color display. Or hitthe Dkey.

8.Select the Paint Bucket tool. (The PaintBucket is often grouped with the GradientTool. If the Paint Bucketis not showing, you

will need to right-click the Gradient Toolandselect the Paint Bucket.)

9. With the Paint Bucket cursor, click onyour new image to make it black.

10.Once the new image is created and has ablack background, well then paste the image

we copied back in Step 4into it by clicking:

Edit > Paste

or Ctrl + V


52/91


By Eric Jay Miller

Page | 52

11.We then want to copy and paste the restof our images into this new file, just as we didthe first one.

Im using 4 images in this example, so I have

4 image layers and a background layer.

When youre done copying all your images,you can close them, so that just your newimage is open.

12. Before we go on, we want to make sureall these images are aligned with oneanother. First we need to choose our centralimage that we will align all the other ones to.In this case, Im going to use the Secondaryimage labeled Layer 1.

Start by deselecting the eye icon next tolayer 3 and 4.Now we dont see those layers anymore andthe only one we can see is Layer 2, since its

the topmost layer thats visible.NOTE: If you dont need to align theseimages, skip to Step 15.

13. Select Layer 2 by clicking on it and thenclick on the Opacity box and change it tosomewhere around 50%. (You can drag theslider bar around that appears when you clickthe Opacitybox ORjust type in a number.)We should now be able to see Layer 1through the halfway transparent Layer 2.


53/91


By Eric Jay Miller

Page | 53

14. If you take a look at your image and youcan see areas that look fuzzy/out of focus, orotherwise obviously mis-aligned, well needto line them up.

15. With Layer 2 still selected, select the

Move Tool on the toolbar (or V on thekeyboard). This will allow you to click on thelayer and drag it around. (You can also movethe image around by 1 pixel at a time byusing the arrow keys.)This is where you may notice that someareas of your image will line up while otherswont. If that is the case, youll need tochoose a main area of interest in the imagethat youll align everything to.

16. So once we have that layer aligned, wewill leave the Opacity at 50%. Now wellalign the next layer by deselecting the eye inLayer 2, and reselecting it in Layer 3 (byclicking in the box where the eye should be).

NOTE: Make sure you have the layer youwant to move selected before you startdragging things around.

Youll just repeat this process on each of thelayers until theyre all aligned.


54/91


By Eric Jay Miller

Page | 54

17.Once were done with all of our aligning,well make sure that all the layers are visible.

(Meaning the eye icon is showing next to

each layer. If its not, just click on the emptybox so it appears.)

Now we move on to the real part. Well startwith Layer 4. If we right-click on it well get ahuge pop up menu and well select BlendingOptions.

18. Blend Mode should be Normal. (Youcan change this if you want, but for this type

of operation, Normalis probably best.)If you didnt align the images earlier, youllnow need to change the layers Opacity to50%. (Higher Opacity gives that layer moreintense colors, Lower Opacity gives moremuted colors.)What were interested in now are the 3Channels checkboxes in the middle of thewindow. Were going to set Layers 2, 3 and 4to be different color channels.

19. What we will do is uncheck 2 of theseboxes. For this layer, Im only going to leavethe Rbox checked.You should notice at this point that yourimage went from B/W to 2 crazy colors.

Click OK.


55/91


56/91


By Eric Jay Miller

Page | 56

23.We should now have a really sad looking,washed-out image. Well now save this imageas a new file.

File > Save As

or Shift + Ctrl + S

We want a flattened regular image(.JPG, .TIF, .BMP, .GIF, etc.) without anylayers.

NOTE: If you save the file as a .TIF, PS willask you if you want to keep the layers. Youcan tell it NO 2 different ways. First, in theSave Asdialog box, uncheck the option thatsays Layers. OR after you hit Save, selectthe Discard radio button under LayerCompression.If you Save Asa .GIF, PS will ask you if youwant to flatten the image. SayYES.

24. Well now leave our image with all thelayers alone for now and well open the new

flattened image we just made.

File > Open

Or Ctrl + O


57/91


By Eric Jay Miller

Page | 57

25. Now, once we have a flattened, 1 layerimage, we will do the Levels operation asdescribed in the first section.


orCtrl + L

Now because this is a color image you canalso try the auto functions, they usually workgreat (theyre not always so hot with B/Wimages).

Image > Auto Color

or Shift + Ctrl + B

You should be done! Save it!We may also want to go back to our first multi-layered image and Save it As aPhotoshop file (.PSD). This will save all the layers and stuff so if we want to go backlater and change things up (colors, opacity, intensity) we can.


58/91


By Eric Jay Miller

Page | 58

We started off with 4 regular old B/W micrographs taken of the exact same area, butwith different detectors.

What were left with is this rather colorful, if not slightly unusual, image.


59/91


By Eric Jay Miller

Page | 59

*Tips Tricks and Suggestions For Multi-Detector Color

One of the great and horrible things about this technique is that there are a milliongazillion different options of things we could do to make this final color image. Its great

because we can tweak the image in a lot of different ways, but its difficult becausethere are SOmany options.

A. Opacity

Here are some examples of how, by changing the opacity on the B/W layer in Step 22,we can really control how subtle or in your face the colors are.

30% Opacity will give you some SEVERE colors.

50% is usually a happy medium.


60/91


By Eric Jay Miller

Page | 60

70% give us more subtle hints of color.

This was a heavy metal sample and the BSEsignal coming off of it was pretty severe. As aresult, the colors came out severelysevere.

Not every sample will respond exactly the sameway to every technique. So if you end up withsomething like this, maybe using the SE detectorwith the stage rotation (discussed in Part C ofthis section) might be something to try unlessyoure into this sort of thing. (For whateverbizarre reason that might be.)

B. Taking BSE Images

One of the things we can consider is HOW we take our images. As we talked aboutbefore, I used different segments from a 4 segment BSE detector and the results weregreat. BUT, there are a number of different ways we could go about doing that.

First off, if we take a look at the setup for the detector I used, itactually has 5 segments. There is one extra segment (labeled E)that is off-axis and will give a little shadow to the BSE image(which, as you may know, gives a rather flat-looking image, sincethe detector is directly above the sample).


61/91


By Eric Jay Miller

Page | 61

Turning on all 5 detectors is great for takingyour master image that wont get colored.(On this system its labeled as 3D.)So here we can see our sample, and despitebeing a BSE image, its got a little bit of depth

and definition to it, thanks to the extrasegment.

Now, how should we take our other, more directional BSEimages? We have a few options.The first option, and most obvious thing to do, is to turn on one

segment at a time and take a picture.How many images you take is up to you (if your detector only has2 segments, I guess youre going to take 2 images). I recommendtaking an image from all 4 detectors. You dont need to use themall, but its good to have them to pick and choose from, just incase.

Each image will be fairly directional, so oneside of the image is bright, while the otherside will have a big shadow.


62/91


63/91


By Eric Jay Miller

Page | 63

So how do these 2 different techniques affect the final result? If we work on oursamples exactly the way we described in the instructions through Step 17, this is whatwe have:

First up we have our color image made from the single-segment images. The colors are a little bright and a little

dramatic, but not overly so. The other thing we need to noteabout this image is that it does not have the B/W masterimage on top, like we had discussed doing in theinstructions.If you like the brighter colors, then this is great! Otherwisewell move to the next option.

OK, heres the same image, but with the B/W master imageon top at 50% opacity.The coloring still seems dramatic, but the colors are muchmore muted. Again, this is something you can tinker with.


64/91


By Eric Jay Miller

Page | 64

Our other option was the TOPO mode images. Thedifferently colored areas arent quite as sharp or serious asthe image we got from combining the single segmentimages. (Using the TOPO mode is my preferred method,mainly because Im using all 4 segments at the same time,

rather than just 1. This should result in an image with bettersignal and less noise. But like I said, its just a preferenceand this is totally up to you.)

Another consideration to take into account is thatsometimes not every sample responds the exact same wayto a particular technique. So trying things different waysmightbe beneficial.

One other thing we can consider is whenshould we performthe Levels operation? In the instructions I said to wait until

everything is finished to do them. Here are the single-segment images colorized, but I did the levels on each of theB/W images BEFORE I combined it.The colors here came out pretty serious. (Again, this stilldoes not have the master B/W image on top.) And since thelevels of this image were pretty solid to start off with, theresnot much we can do with them now (aside from maybeadjusting the color balance [Shift + Ctrl + B]).

You may notice that still, given all the many little things we can do from start to finish inthis process, you still will end up with generally the same image. They have theirdifferences, but all the examples we just looked at are not WILDLYdifferent. Now Imnot saying that small details are not important (surely they can be), but as with the othertechniques weve discussed, more time and effort = more better results. (Or moregooder results, whatever.)

C. Using ONLY 1 SE Detector

OK, maybe you have a cheap SEM that only has 1 SE detector and no BSE detector.

OR you took your images using the BSE detector and the image came out lookinginsane. How do you get images from different directions like this only using a single SEdetector?

WE ROTATE THE SAMPLE!


65/91


By Eric Jay Miller

Page | 65

If your microscope isnt SUPERold or SUPERcheap, you will probably have some kindof eucentric rotation option somewhere in the software to rotate the stage around thearea youre currently looking at.

There are a number of things to keep track of when attempting this.

First, this wont work if youre using an upper (through the lens) SE detector. Itsalways looking straight down on your sample, so rotating your sample around isntgoing to do anything. This will only work with a lower (in -chamber) detector.

Second, find an area on your sample that you think will be visible as you rotate yoursample around 360. (That might not be as easy as it sounds. Dont be surprised if afteryouve taken 1 or 2 images, you have to start over in a new area because you lost whatyou were looking at.)

Third, you need a way to center your area ofinterest on the screen as its rotated around.Most SEMs will have something like an areamarker or crosshairs you can turn on that willhelp align your sample in the middle of thescreen. (If your SEM doesnt have somethinglike that, the corner of a sticky note or a pieceof tape stuck on the screen will work just aswell.)In the example were going to look at Imgoing to keep the middle of this pointy thingin the middle of the screen.

Fourth, we need to decide how many images were going to take, and therefore, howmuch to rotate the stage by in between images. 4 images would probably be nice, sowe can have 2 or 3 color images and 1 master image, like we have in the BSE images.(If youre using an SEM that has a through the lens detector.)So we would just rotatethe stage by 90 each time after taking our images. In the example Ive cooked up, Itook 3 images, so I rotated the stage by 120.

At this juncture, we must ask ourselves two very important questions about our SEM.1. Does my SEM have a software controlled stage?2. Does my SEM have a Raster Rotation in the control software that I can

control the exact number of degrees my image is rotated?

If the answer to both questions is YES, then youre in luck and this will be pretty easy ifyou just follow the instructions under Option A.

If you answered NO to either of those questions, youre going to need to follow thesteps under Option Bfor this to work properly.


66/91


By Eric Jay Miller

Page | 66

OPTION A: Software Stage RotationEasy.

1. Center your structure of interest, do all alignments and take an image.2. Enter the number of degrees you want to rotate the stage around into wherever

the stage rotation stuff is located in your software. (This would probably be best if

it was set to Relative rotation and not Absolute.) If your stage has a eucentricrotation function (and it works the way it should), make sure its turned on.

3. Once the stage is rotated, re-center your feature.4. Turn on the Raster Rotation on the SEM and type in the same amount you

rotated the stage (i.e. 120). Your image should then be oriented in the sameway your first image was. (If its not, you probably rotated something in the wrongdirection, like positive instead of negative.) Take a picture.

5. Repeat steps 2-4 again until you have all your images taken. (3 for 120, 4 for90, etc.) Make sure the images are all at the same magnification, or youll be introuble later.

6. At that point you should have all you need to combine your images as per the

regular instructions. But you may need to align the images a little bit.

OPTION B: Manual Stage RotationThis should work pretty good as well, but since it wont be as precise as typing a

number into software, we have a few extra hoops to jump through. 1. Once you center your structure of interest, do all the necessary alignments.

Because of the rotation were doing, were going to end up cropping out a lot ofthe image later, this means well need to take the image at a MUCH lower magthan what you think you need to. A way you can compensate for losing half yourimage later to cropping is by taking a higher resolution images right now. Forexample: if you normally take images at a resolution of 1280x960 pixels, just

bump it up to 2560x1920.2. Take your first image, rotate the stage around whatever amount, re-center your

structure and take another image. Make sure all your images are taken at thesame magnification, or youll be in trouble later.

You should end up with a group of images that look kind of like this:

To make the color image happen, we can follow the instructions as before up throughStep 12.


67/91


By Eric Jay Miller

Page | 67

12A. In our example, weve got all 3 imagespasted into a new file as separate layers. Wehave our correctly oriented image on bottomand only 1 of the 2 other layers is visible andset to 50% Opacity.

12B.With that second image selected, wellneed to rotate it now by going to:

Edit > Transform > Rotate

You can also use the Free Transformtool aswell.

Ctrl + T

12C. If you move the mouse around theedges of your image, your cursor should look

something like this

By holding down the left mouse button anddragging the mouse around you can rotatethe image. You probably wont be able to getit to line up perfectly at first. Just start off bygetting it close.


68/91


By Eric Jay Miller

Page | 68

12D. If you release the mouse button andmove the cursor towards the center of therotated image, it will turn into the Move Tool,which means you can click the image anddrag it around in the X and Y direction. Do

this to get the 2 images better lined up, and ifyou need to, go back to rotating the imageagain.Go back and forth between rotating andnudging the image up/down, left/right, untilyou get it aligned.

At some point you might need to switch fromrotating it around with the mouse and change the angle in the Rotation box by typing ina number. You might have to rotate it in very tiny amounts, so feel free to make thefraction of an angle as small as you want.

12E. Repeat 12A thru 12C on any otherlayers you have. In my example I have 3images, and it should look something likethis.

Theres only a little part in the center of theimage thats useable, so well need to Cropitout. We can do that now, or afterwe colorizethe image. It doesnt seem to matter.

We can now continue with the standard Blendingprocess as described in the normal instructions.Here is what we ended up with. Nifty!

However, this sample, which has a lottopography action going on, makes a goodexample of why we might want to try LESScolor,instead of MORE.

D. Less Can Be More


69/91


By Eric Jay Miller

Page | 69

EX1. I redid the color on this image, but I onlycolorized 2 of the layers instead of all 3, and weend up with something that looks slightly morenormal.It came out kind of green, and of course just by

changing around which Channels boxes werechecked, I could have made it come out purple orblue or red or whatever.I think all those subtle gradations of color you canget with this technique are really fantastic though.

EX2.I liked that green image, but from that greenimage all I needed to do was auto-adjust thecolor:

Image > Auto Color

or Shift + Ctrl + B

And I end up with something that looks like amore sedate version my original image with 3color layers.

EX3. Unfortunately, this technique doesntmake everythinglook fantastic. I had anothersample that I processed. It didnt come out ascool as I thought it would. In fact, I hated it. Imessed around using 3 and 4 color layers ahundred different ways and I never gotanything I liked out of it. It was just way toobusy. There were too manycolors.


70/91


By Eric Jay Miller

Page | 70

Fortunately, before I deleted the images outof disgust, I decided to use only 2 colorlayers instead of 3. Just by doing that onelittle thing, its eyesore status was revoked.

Another thing that was brought out byreducing the number of color layers was youreally got a better idea about what thetopography of the sample is like. With the 3color layers, it was just a big jumble of color.

(This image also has the master B/W imageon the bottom.)

Heres another image I put together with 3color layers. Its ok. But maye I dont want a

crazy color show on my image. (I kinda likethe crazy colors, but thats my personalpreference.)

So I tried putting the image together usingonly 2 color layers and with the Master B/Wimage on top at 50%. I then flattened theimage and did Auto Colorcorrection (Ctrl +Shift + B), then Levels (Ctrl + L).Now, compared to the more colorful image Imade before, it looks pretty dull. BUT, if you

just look at it, on its own, I think its reallyfantastic. Sometimes the colors come out sosublte, its almost like youre looking at ablack and white image. It may even take youa second to realize the colors are there. I think its that aspect of discovery that makesthis image exciting.This is just another option to try in your bag of tricks.


71/91


By Eric Jay Miller

Page | 71

NOTE: When youre choosing from yourdifferent B/W images, does it matter whichones you pick when only doing 2 colors?Sort of.What I think is the best option is to get

images that were taken from opposite sidesof the sample. So for example, I would takeimage 1 and 3 together, but not 1 and 2, or 2and 3. (Now if your BSE detector only has 2segments, then thats what youll end updoing anyway.)By doing this, both sides of the sample willbe illuminated by a different color. (Preferablya lighter one and a darker one.)

E. Change Up the Colors!

The last thing well look at is what we get when we change the color channels around inthe different layers. This will change the overall look of the image a lot more than mostof these other changes weve been talking about.

This first image came out nice, with somesubtle, muted colors. I really like the waythese colors show off the surface structure ofthe sample.

The next image looks total ly different. Wehave vibrant, fun colors that are in broaderrange than the previous image. What did I doto make them different? Simply checking andunchecking different Channels boxes inSteps 19and 20.

Thats it.

1

2 3


72/91


By Eric Jay Miller

Page | 72

In conclusion, the fantastic thing about this technique is not only the ease and speedyou can put these images together, but also the fact that all the tiny details in yourimage are not lost, like they can be when using something like the Gradient Map.

I LOVE THIS TECHNIQUE!


73/91


By Eric Jay Miller

Page | 73

VIII. Drawing a Mag Line Marker(The Cheating Way with Pre-Existing Mag Marker)

So you want to draw a Mag Marker on your image. We have 2 options.

Option A:if you already have a mag marker on your image put there by the SEM/TEMsoftware and you want to draw a different one.Option B:if there is nomag marker andyou know the magnification. Go to Section IX.

WARNING:This section will involve math!

1.Open your image in Photoshop and selectthe Rectangle tool.Rectangles are nice because they alwayscome out perfectly horizontal AND you canchange how thick you want it to be as youre

drawing it.

NOTE:You candraw a straight line using theLine Tool if you hold down the Shift keywhile youre drawing it. But you might not getthe thickness right the first time. You canchange that in the Weight box. But youllneed to be careful to make sure you alsotype in the correct unit, pxfor pixels, infor inches, cmfor centimeters, etc. This will bethe WIDTH if you line, youll draw it as long as it needs to be. A 10 inch wide line isgoing to be quite a bit different than a 10 pixel wide line (and PS seems to default toinches when you type in a number and hit enter here). So you will need to pay attentionto this if youre using the line tool.

2. The shapes toolbar should appear at thetop of the screen. Check to make sureCreate new shape layer is selected. Also,we want to check what color our rectanglewill be. By looking at the box next to where itsays Colorwe can see its set to black. If we

click on this box the Color Picker will open and we can choose our color. White isprobably the best.


74/91


By Eric Jay Miller

Page | 74

3.Were ready to draw our mag marker now.Since this micrograph already has one, were

just going to borrow it by drawing our newone right on top of it. That way, we know itsthe right size.

4.Here is the problem though: where do westart and stop drawing our rectangle inrelation to the existing line? As illustratedhere, we dont want to draw our new markerfrom the inside or outside edges of theexisting marker. It needs to be from thesame side of the first and last hash marks.I know just reading that explanation might not

make sense, but you should be able to figureit out eventually if you stare at the illustrationlong enough.

5. If the existing marker is too small to beable to draw your rectangle over accurately,zoom in by selecting

View > Zoom In

or Control + +

You can do this multiple times. To zoom out,select

View > Zoom Out

or Control + -

6. We can go ahead and draw our markernow by clicking at one end of the existingmarker, dragging the cursor to the other endof the marker and releasing. The layer the

new marker is in should be selected in theLayer tab in the lower right-hand part of thescreen.

An exception to thiswould be if our image isa .BMP file which cannot accept new layers.If that is the case, Save your file in a differentformat, like .TIFbeforedrawing the line.


75/91


By Eric Jay Miller

Page | 75

7. The existing line marker is 5m. Letsassume for a moment that we wanted ournew marker to be a different length, like 2m.With our new lines layer selected in theLayertab, we click on

Edit > Transform Path > Scale

8. A toolbar at the top of the screen shouldappear that gives us a number of options.

Width (W) and Height (H) are currently at100%. To make our line shorter, we justchange the Wvalue. But by how much?Original line = 5m = 100%100% / 5m = 20%20% of original line = 1m.Therefore a 2m line should = 40%.MATH!I enter 40 into the Wbox.

9. To commit the change to your line, clickthe checkmark on the right end of thistoolbar.


76/91


By Eric Jay Miller

Page | 76

10. Select the Move Tool (V on thekeyboard).Using the MoveTool, you can drag your newmarker around the image and position itwhere you want.

11. Weve moved our marker and now weneed to label it. Select the Horizontal TypeTool.

12. A toolbar along the top of the screenshould appear where we can choose all ofour text options. I can see right now I need tochange the text color from black to white.

The size also says 30pt, Im not sure rightnow if that will be too big or too small. Wecan change it later if its no good.


77/91


By Eric Jay Miller

Page | 77

13. When youre done choosing all youroptions we should be able to click on ourimage generally where we want our text to goand start typing.I can see that 30pt was way too big. So I

select my text by clicking and dragging thecursor over it and then selecting a smallersize with the font size drop down menu in thetoolbox.When youre happy with how it looks, clickthe checkmark on the right end of this toolbarto commit the change to your line.Now you can use the Move tool to positionthe text where you want it.

14. If, for whatever reason, you want to edit

your text after youve hit the Commitcheckmark, youll need to select the layeryour text is in in the Layers tab in the lowerright corner of the screen and double-clickthe white box with the Tin it. Your text shouldnow be editable.

15. The only reason to put a new magmarker on your image is because youregoing to crop the old one out. Select theCrop tool and drag the cursor across yourimage to select the area you want to keep.You can adjust the selected area by grabbingand dragging the small gray boxes thatappear on the selection box. When yourehappy with your box, click the Commitcheckmark in the upper right of the screen.

Your image should be ready for saving.


78/91


By Eric Jay Miller

Page | 78

IX. Drawing a Mag Line Marker(The SUPERCheating Way with NOExisting Mag Marker)

CAUTION: This technique ONLY works with Adobe Photoshop. Dont try it with

something else like ImageJ! It totally wont work.

EXTRA CAUTION:This onlyworks if your image HAS NOTbeen resized. (Unless youknow by how much. i.e. If its exactly half the size, reduce your line 50%.)

SUPER CAUTION:This will onlywork with .TIF, .JPG, .BMP and .PNG files. It WILLNOT WORKwith a .GIFfile. You have been warned.

1.Lets say we have an image that has NOmag line marker on it. What do I do?? First,we ask ourselves if we know the

magnification. If the answer is yes, then youcan proceed. If not, you may commencecrying.For this example Im going to use an imagethat DOES have a mag marker on it. This isto prove that this technique actually works.(Because in theory it shouldnt work at allbut somehow it does.)

2. We take the known magnification of ourimage (in my example, its 10,000x.) and we

divide it by 10,000.The answer we get is the distance a lineneeds to be in cm on our image to represent1m.That might be confusing.So, 10,000x / 10,000 = 1This means a 1um marker on our image willneed to be 1cm long.For this example, I need a 5m line, so I justneed to make it 5cm long. Simple!


79/91


By Eric Jay Miller

Page | 79

3. So how do we draw a perfect 1cm longrectangle on our image?First we open our image in Photoshopand turn on the Rulers(if theyre not already)by clicking:

View > Rulers

orCtrl + R

This puts a ruler along the top and left edgeof your image.

4. We then set the units on the ruler toCentimeters by right-clicking on the ruler andselecting Centimeters.

5. In order to draw our rectangle a veryprecise length well need to open the Infobox by clicking

Window > Info

Or F8

The box will show the width and height of ourrectangle in cm as we draw it.

6.Select the Rectangle tool.Rectangles are nice because they alwayscome out perfectly horizontal AND you canchange how thick you want it to be as youredrawing it.


80/91


By Eric Jay Miller

Page | 80

NOTE:You candraw a straight line using theLine Tool if you hold down the Shift keywhile youre drawing it. But you might not getthe thickness right the first time. You canchange that in the Weight box. But youll

need to be careful to make sure you alsotype in the correct unit, pxfor pixels, infor inches, cmfor centimeters, etc. This will bethe WIDTH if you line, youll draw it as long as it needs to be. A 10 inch wide line isgoing to be quite a bit different than a 10 pixel wide line (and PS seems to default toinches when you type in a number and hit enter here). So you will need to pay attentionto this if youre using the line tool.

7. The shapes toolbar should appear at thetop of the screen. Check to make sureCreate new shape layer is selected. Also,we want to check what color our rectangle

will be. By looking at the box next to where itsays Colorwe can see its set to black. If weclick on this box the Color Pickerwill openand we can choose our color. White isprobably the best.

8.Now we can draw our marker. (It doesntmatter where we put it since we can move it

around later.)Keep an eye on the info box and draw yourrectangle out to the correct length. In thisexample, it was 5cm (circled in red).

9.Our newly drawn layer should be selected

in the Layerstab in the lower right corner ofthe screen and automatically namedsomething exciting like Shape 1.


81/91


By Eric Jay Miller

Page | 81

10. Select the Move Tool, or V on thekeyboard.Using the MoveTool, you can drag your newmarker around the image and position itwhere you want.

Then you can put a label on it if you want.The text label process is discussed earlier inthis section in steps A11 thru A14.Your image should be ready for saving.

11.Lastly, I took the line I drew and placed ithere over the existing marker to show thatthis actually worked. We can see that it cameout perfectly.

Ive tried this with numerous micrographs from several different microscopemanufacturers (and even on actual photographic micrographs). It didnt matter whatsize or resolution they were, as long as it was the original non-resized image, nota .GIF, and done in Photoshop, this procedure works. Just dont ask me why.


82/91


By Eric Jay Miller

Page | 82

Appendix IThe Short Versions

I. Levels

Instructions Shortcuts

1 Openimage Ctrl+O

2 Is image 8bit gray scale? Image > Mode

3 If no, change. Image > Mode

4 Open Levels Ctrl + L

5 Make sure Previewboxis checked -

6 Adjust Levels -

7 Save Ctrl + S


83/91


By Eric Jay Miller

Page | 83

II. Simple Pseudo-ColorColor Table


1 OpenImage Ctrl+O

2 DuplicateImage Image > Duplicate3 Close original image Ctrl + F4

4 Switch Image to Indexed Color Image > Mode > Indexed Color

5 Open the Color Table Image > Mode > Color Table

6 Click and drag cursor from top left boxto bottom right

-

7 Select black in the Color Picker. OK -

8 Select the other color. OK -

9 Preview Image. OK -

10 Save Ctrl + S


84/91


By Eric Jay Miller

Page | 84

III. Simple Pseudo-ColorGradient Map


1 OpenImage Ctrl+O

2 DuplicateImage Image > Duplicate3 Close original image Ctrl + F4

4 Switch Image to RGB Color Image > Mode > RGB Color

5 Open the Gradient Map Image > Adjustments >Gradient Map

6 Click the color gradient bar -

7 Select a preset to start from -

8 Select a Color Stop -

9 Click Colorbox to open Color Picker -

10 Choose color. OK -

11 Repeat Steps 8, 9, 10 until all colors are

as desired

-

12 Gradient Editor, OK -

13 Gradient Map, OK -

14 Levels Ctrl + L

15 Save Ctrl + S


85/91


By Eric Jay Miller

Page | 85

IV. Complex Pseudo ColorColor Table


1 Openimage Ctrl + O

2 Duplicateimage Image > Duplicate3 Close Original Ctrl + F4

4 Adjust Levels Ctrl + L

5 Duplicateimage until you have 1 imagefor each color youll be using

Image > Duplicate

6 Set all images to Indexed Color Mode Image > Mode > Indexed Color

7 Open the Color Table Image > Mode > Color Table

8 Click and drag cursor from top left box tobottom right

-

9 Select dark color in the Color Picker. OK -

10 Select light color. OK -11 Preview Image. OK -

12 Repeat 7 thru 11 for each image -

13 Choose which images will have areascut out and which image will be thebackground

-

14 Enter Quick MaskMode on any imageother than the background image

Q

15 Select Penciltool & brush diameter -

16 Draw over bulk areas, avoid edges -

17 Select Brushtool & brush diameter -

18 Draw over edges -19 Exit Quick Maskmode Q

20 Invertselection Ctrl + Shift + I

21 Copy Ctrl + C

22 Switch Background image to RGB Color Image > Mode > RGB Color

23 Paste Ctrl + V

24 Repeat Steps 14 thru 23 on remainingimages

-

25 Select background image with all layerscopied into it

-

26 Select layer to align -27 Set layer Opacityto 50% -

28 Select Move Tool V

29 Left click on layer to drag and align -

30 Use arrow keys for fine alignment -

31 Set layer Opacity to 100% -

32 Repeat Steps 26 thru 31 for remaininglayers

-


86/91


By Eric Jay Miller

Page | 86

33 Arrange layer order -

34 Save Ctrl + S


87/91


By Eric Jay Miller

Page | 87

V. Calculations


1 OpenImage Ctrl+O

2 Make 2 Duplicateimages & close the original Image > Duplicate3 Run a 100 pixel Median Filteron c

Documents

Basic Photoshop for Electron Microscopy 2014 by Eric Miller Microscopy & Imaging Specialist at Northwestern University