Editing photos using luminosity selections

Let’s start with a photo in the state we left in the last post. Go into the channels window, and create the Bright 1, Bright 2, Bright 3 and Dark 1, Dark 2 and Dark 3 channels. Each time from now on, when we need to select a certain one, just go back to selections and Ctrl+click on the one you want to select.

There are two ways we can approach editing from here on. Either we want to effect the whole selection, or just paint in a part of it, exactly where we need it. The process is a bit different for both of these, so let’s look at them now.

Global edits on whole selection

When you want to effect the whole selections, the two steps you have to do are:

• Select the luminosity selection you want to use – Go into the channels window, and Ctrl+click on the selection you want to use. Go back to the layers window afterwards
• Add the edit you want to use – Create new fill or adjustment layer, by using the button (circle icon, half white, half black) in the bottom right. You can use any adjustment you need. The selection will be automatically added a layer mask to the new layer and you can then edit your adjustment.

Let’s look at an example to understand this. For this photo, I want to brighten the shadow areas and I want to darken the highlights. Let’s start with the shadows.

1. Go into the channels, and select a Dark selection. The more restrictive selection you choose, the less efect your edit will have. So if you go to Dark 3, or even 4,5 and higher, you effect only the very dark areas in the photo. For this example, let’s choose Dark 2. Ctrl+click on the channel with that selection
2. Go into layers and add a Brightness/Contrast adjustment layer (one can use also other adjustments, like curves or levels to get the same effect). A new layer will be created with the mask applied to it.
3. Now in the properties of that layer, adjust brightness to make the shadow areas brighter. You will see, that the bright areas of the photo are not effected at all and only the shadows are brightened.

Going to the second edit, darkening the highlights, just do the same, just change the selection and the adjustment. For instance, select a Bright 2 selection, create a Brightness/Contrast adjustment layer and lower the brightness. And you have darkened the highlights of the photo.

Don’t limit yourself to just darkening and brightening a certain area. You can use this to change the color balance, saturation, detail, white balance and much more. Just use the adjustment layer you need. Also, one note about changing the brightness. Each time you brighten a dark area or darken a bright area, you photo looses contrast. A good practice is then, that every time you do this, to add a bit of contrast to the same area. If you use Brightness/contrast, add a bit of contrast, in levels, move the mid-tones a bit, in curves, create a S curve. Like this, your edits will look natural and photos will not become grey and flat.

Local edits by painting in the effect

If you don’t want to use the whole selection, the process is a bit different. The steps you have to follow then are:

• Create an adjustment layer – In the layers window, create the adjustment layer you need, and set it to parameters you need. It will effect the whole image, but that is OK for now.
• Add a black mask to it – We need to hide this layer for now. So click on the white mask next to this layer, and press Ctrl+I to invert it to a black mask
• Make a luminosity selection in the Channels window – as before Ctrl+click on the Bright or Dark channel that closely matches the part of the image you want to edit. This will create a selection for you.
• Paint in the edit – now go back to the layers window, choose the Brush tool, white, 0% hardness. Select the black mask you inverted before and start painting in the areas you want the edit to have effect. If the marching ants of the selection are distracting or in the way, you can hide them by pressing Ctrl+H. The selection will limit where you can paint, so you will only effect the area you need to.

Let’s go back to our example. Let’s say I still want to brighten the shadows, but not everywhere. I want to just brighten them around the buildings in the middle of the photo. I will use levels in this example

1. create and edit a new adjustment layer – With the button in the bottom right of the layers window, create a new levels adjustment layer. Brighten the photo by moving the white triangle to the left
2. invert the mask – Select it’s mask, and press Ctrl+I to invert it. Now the adjustment layer will have no effect on the image
3. create your selection – Go into channels and let’s select the Dark 2 one by Ctrl+click on it
4. paint in the effect – go back to the mask, hide the selection with Ctrl+H, and with the white brush start painting around the are you want to brighter. The more times you paint over a spot, the stronger effect the adjustment will have on it. But it’s all limited by the selection, so for instance the bright castle in the middle will not be effected at all.

If you ever add to much, just switch your brush to black color, and paint over the same spot. By switching back and forth between white and black, you can add or remove the effect until you are satisfied with your result. If you see that you selection is too narrow or broad, delete the layer and start again. Don’t forget you can use this with any other adjustment. You can paint in saturation, detail, colors, brightness, contrast and anything else you can put on a layer.

Keeping fewer photos

• Don’t take photos you know you will not like. Sometimes the camera can just stay in the bag. Once you taken enough photos, you just know. Happens to me sometime. I look at the scene, look at the conditions, and just don’t bother. I know I will not like the results anyway. And trying to force it usually does not result in great photos also.
• Don’t take many versions of the same composition. I have a bit of problem with this one sometime, and should try more to fix this. One does not need the same photo over and over. Take one or few, if you want to be sure to have a good one, and move on to a new composition. Like this you maximize you chance to get a great photo. I would make an exception here, for sunrise and sunset photos. The conditions change so quickly then, that even staying with the same composition can produce different results.
• Limit the number of brackets. I tend to say, better safe then sorry, but one can also overdo it. You don’t need 9, you don’t need 7, mostly you don’t even need 5 brackets. If you are not shooting into the sun, 3 brackets is mostly enough. Try a sequence and check the histogram directly on the camera. If you see that you are getting the whole dynamic range in fewer brackets, adjust your settings accordingly.
• Delete technically bad photos immediately. You can’t always tell directly on the camera if a photo is bad. But often there are problems that you can spot immediately, and you won’t be able to fixt them anyway. Maybe you bumped into your tripod. Maybe there was a strong wind. Maybe you used the wrong settings and over/underexposed you shot complletely. Maybe you forgot to turn off autofocus, so your focus is completely off. This and other problems can be seen immediately on the camera, and it’s pointless to keep those photos. Easier to delete them right when you checking the preview.
• Mark the photos you like. I do this usually if I’m doing photos at an event. I tent to go through all right away, as I have to deliver them to a client, and mark the ones I will edit and send. When I’m done I wait for 1-4 weeks and then just delete the unmarked photos. I keep the ones I deliverd, as in my experience, there does not exist a company or a person that does reasonable backups. I had clients coming back to me after years, taht they lost the photos.
• Be decisive. When you look at a photo and you dont like it and don’t need it, delete it. If you are undecisive and not sure, you will probably just delete it sometimes later, or just keep it for nothing. The thing is, the first feeling about it is usually the correct one. You can’t force yourself to like something even if you try.

Topaz Denoise AI

Same as with other Topaz AI applications, the interface is very simple. You load your image and process it. You only have three sliders: remove noise, enhance sharpness and restore detail. They do exactly what the name suggests. There is also a brighten preview option, where you can have it brighten your photos, so you can see the noise better. This has no effect on the image processing, it’s just to make you work simpler.

One nice feature here is the support for RAW images, so you can directly work on the source file. Better to do this before you did other edits on the image. You can save the export as a DNG file RAW afterwards.

But let’s go to the examples. There are three images in every example. The original photo as TIFF (lens correction and chromatic aberrations removed before noise reduction, some brightened a bit to make the noise more visible). Second one is a Noiseware noise reduction with strong noise preset and detail protection set to 6. Last one is the Topaz Denoise AI result, with all the sliders at 0.25.

Please take the comparison between Noiseware and Denoise with reservations. You can set the sliders differently in both, and the results may greatly vary based on photo. There is no way I can show all the possible results here, so this should be taken just as small preview, and one should try both for oneself (there are trials available). I will focus here only on the Denoise results.

Please click on the images to see bigger versions, to see the noise.

Original

Imagenomic Noiseware

Topaz Denoise AI

In this interior photo, Topaz Denoise very nicely removed the noise and added quite nice structure to the pillars. Strangely, which I noticed in multiple photos, some lines caught a bit of a green color tint. You can see it in the middle here and also in the top right corner. Hard to say why this happens. Original and Denoise versions in one photo for comparison here. 

Original

Imagenomic Noiseware

Topaz Denoise AI

Here is a very noisy photo, taken in the middle of the night. Again, you can see how Denoise tries to add more structure to places that it removed noise from. Original and Denoise versions in one photo for comparison here. 

Original

Imagenomic Noiseware

Topaz Denoise AI

This is my favorite result here. The car looks so clean afterwards. All the detail is preserved and overall this is a great result. Original and Denoise versions in one photo for comparison here.

Original

Imagenomic Noiseware

Topaz Denoise AI

This is a different example from the same photo as the previous example. Again, very nice result, very clean, very crisp. Original and Denoise versions in one photo for comparison here.

Original

Imagenomic Noiseware

Topaz Denoise AI

One last example here (again from a car, really have very few photos taken at a higher ISO :)). Again a very nice clean result here. Original and Denoise versions in one photo for comparison here. 

Useless information

Do I really need to know the specs of every camera that is released? Of every lens? Do I even need to know that some camera was released? I have been using Canon cameras for about 10 years now, and I no longer even know what professional cameras Canon offers anymore. It’s useless to me. It will not make my photos better. Rather than reading up on this, I prefer to look at new editing techniques, processes, ways to improve, cool spots to visit, interesting compositions and similar.

Only time I look at this stuff now, is when I know need a new camera. When the shutter count is closing up to the camera limits, it’s time to have a look at something new. But other than that, why?

Gear envy

I know I had this. Seeing other photographers with pricey cameras, all the best gear, can make one feel envious. I felt the same before. But I don’t care anymore. There is a point, where what you have is good enough. The Canon 5D mark II, that I bought in 2012, is probably still good enough for most of what I do. And I still use it. Do I need a 30Mpix photo when doing photos at an event? Do I even need the 22Mpix the 5D mark II gives? It’s different for landscapes, but even there most of times you could get by.

Seeing all the new releases will only make you have more gear envy, that you don’t need at all. Focusing on what you have and how to get more from it, is much more rewarding in the end.

Diminishing returns

I think you all heard of the Law of diminishing returns. Here it would be, that there is a point where spending more on something, will get you almost nothing more. A good example would be tripods. Buying a 200 USD tripod, will give you a much better experience than a 40 USD tripod. It’s can easily be 5 times better. But how much better is a 1000 USD tripod than the 200 USD one? Maybe 2 times better? 1.5 times better? Maybe even less.

This goes hand in hand with gear envy and just useless information about all that is available. Do I need to know that if I spend double what I already did, I get something that is 5% better? No.

Pointless rumors

I hate rumors. And not just in camera news. They mean nothing. Do you know that Canon, Sony, Nikon and others will release new cameras? Of course they will. Will there be new versions of popular gear? Of course there will be. Will the new cameras be better? I do hope so. But do I need to know some random persons idea what they be? Not really. I seen rumor articles based on a single tweet from a new, never before used account. Really. News sites need content, and often they go with anything.

There has not been a big change in photography for a long time. I would count the switch to digital photography as one, but stuff like going mirror-less is just an evolution. You probably can predict every new camera that comes out just by looking at older releases. The changes are quite minor.

Minimalism

Over the last 1-2 years, I have been interested in scaling down. I got rid of a big pile of things. And I’m moving with this approach also to my digital life. I stopped using some social networks, unsubscribed from a lot of accounts on other sites, limited my visits to even more. And information on new cameras and gear just fits perfectly in a category I don’t need and can easily leave out.

Understanding Luminosity selections

In short, Luminosity masks are masks created from luminosity selections. These are selections based on brightness of pixels. The most basic luminosity selections would be then:

• Bright – This selection selects every white pixel by 100%, every black pixel by 0%, and everything in between based on their brightness. So the brighter a pixel is, the more it is selected. So 25% grey pixel would be selected by 75%, 50% grey pixel would be selected by 50% and 75% grey pixel would be selected by 25%.
• Dark – This is the inverse selection to the bright one. A white pixel is selected by 0% and a black pixel by 100%. Everything in between is selected based on how dark it is. So 25% grey pixel would be selected by 25%, 50% grey pixel would be selected by 50% and 75% grey pixel would be selected by 75%

These two selections split an image into two parts. The part that is mostly dark and the part that is mostly bright. A pixel with a 50% grey color, would be right in the middle, selected by 50% in both. I know it’s a bit hard to get the understanding of this, so let’s look at a very basic image, that illustrates this. This is a just a gradient from black to white.

If we do the Bright and Dark selections here, there results would looks like this (this is already shown as a mask, as you can’t display opacity graduation in a selection, white means 100% selected, black means 0% selected).

Bright selection

Dark selection

As you can see on the Bright one, everywhere where it was bright in the original image, there is a shade of grey up to white. On the Dark one, you get the inverse, and it affects the areas that were dark. Let’s looks at one more example, this one with a regular photo.

Original photo

Bright selection

Dark selection

You see the same effect. The Bright selection selects only the bright areas (clouds, snow), the Dark selection only the dark ones (mostly mountains).

There are also Midtone selections, which are created by subtracting Bright and Dark selections from a photo, but I will get to those in a separate article.

Refined Luminosity selections

You often here these selection being referred to as Bright 1, Bright 2, Bright 3 … Dark 1, Dark 2, Dark 3… and so on. In this context, the Bright selection I mentioned would be equal to Bright 1, the Dark to Dark 1. All others are refinement of these selection. By creating an intersection of a selection with itself, you create a new, more restrictive selection. So the higher the number here, the less is selected. On the gradient image I shown you earlier, the restricted masks would looks like this:

Bright 1

Bright 2

Bright 3

Bright 4

Creating Luminosity selections and masks in Photoshop

Ok, now that you hopefully at least have an idea on how luminosity selections look, let’s go into Photoshop and create some.

Open you image in Photoshop, and go into the Channels window. Here you will see 4 layers. The RGB, Red, Green and Blue. We want to work with the RGB one. Hold down Ctrl and click on it. This will create the Bright selection. Click on the Save selection as channel button in the bottom right (white square with black circle in the middle) and this will create a new channel. Let’s rename it to Bright 1.

We will now work with this one. You should still have your selection active. Hold down Ctrl+Alt+Shift (a small X should be next to your cursor) and click on this new Bright 1 channel. The selection will change. Save it again as a new channel. Name the new channel Bright 2.

You can now continue like this and create further, more restrictive selections, Bright 3, Bright 4, … and so on. If you ever loose your selection, just Ctrl+click on the one you want to restrict further, and the Ctrl+Alt+Shift+click on the same one again. Doing so on Bright 1 will create Bright 2, on Bright 2 will create Bright 3 and so on.

Now onto the Dark selection. Go back to the RGB channel and Ctrl+click on it. This will create a Bright 1 selection. Now hit Ctrl+Shift+I on your keyboard (or select Select/Inverse from the menu). This will change the selection to the Dark 1 selection. You can create a new channel from this selection and call it Dark 1. To create all the other ones, just follow the same steps as when creating the Bright selections.

If you want to avoid having to do this all the time, you can make it easier on yourself. Create actions that makes these masks, or buy Raya Pro or TK actions. Both of them create all the masks for you in one button press.

Once you have these channels, just choose one you want to work it and create a selection of it with Ctrl+click. If you want to make it into a layer mask, go back to layers with the selection active, choose the layer you want to use and click on Add layer mask (button looks the same as the one we used to create channels, in the bottom right of the layers window)