Ed Schlotzhauer

An artists journey

Category: Mastery

  • Look Sharp

    Look Sharp

    No, I’m not giving advice on fashion trends. You probably wouldn’t want to follow my lead. But I can talk some about image sharpness. Photographers often obsess over getting the sharpest possible image. Today I want to give an overview of the factors that make an image look sharp and some that make it not sharp.

    Sharpness chain

    I described the transforms in the image capture process as the sharpness chain. Physically and logically there are several components that light has to go through before we have an image on our screen to view and edit. In may be more precise to describe this as the “unsharpness” chain, because unfortunately, every step along the way degrades the image to some extent.

    Digital camera loss of sharpness chain

    The original image is, by definition, “perfect” since it is the original. The light then goes through a filter (if you use one, I usually do), the lens itself, the Bayer filter to do color separation, the sensor chip, various processing stages in the camera hardware, and the raw conversion. I include the raw conversion here because the image is not editable until this has been done. There is no gain at each of these stages. This means that each stage degrades the image.

    This is not to be discouraging. Modern cameras and lenses are fantastic. “Fantastic” means they degrade the image less than ever before in history. This is not a bad state of affairs. If you are using excellent equipment all along the chain you can achieve some great theoretical results.

    Focus

    Oops, I said theoretical results. What I mean is that under perfect conditions the camera system can produce excellent results. But we may not always apply the best techniques when we are using the equipment. There are many things we can do to make the image sharpness worse.

    Focus, for instance. My eyes are getting old and weak. I usually rely on the camera auto focus system. And these do a great job now. But did I move the camera after focusing? Did I focus on the right part of the composition? Is the light level bright enough to allow the camera to work properly? Was it properly locked down on a good tripod to keep things rigid?

    Motion

    Another problem is camera shake. Pixels in modern sensors are so tiny that very little motion can smear light over several pixel sites. Yes, my camera has internal image stabilization, but this does not entirely compensate for bad technique.

    Way back in the film days we used a rule of thumb of 1 over the film speed to estimate the minimum shutter speed. That is, if using 200 ISO we should be able to shoot at 1/200 of a second and be able to maintain adequate sharpness. Sensors are so fine pitched now that I think the rule should be around 2-3X the ISO to be conservative. So at ISO 200 I should probably shoot at 1/400 to 1/800 second handheld to get good results. Best to always use a sturdy tripod.

    Another common problem is subject motion. This is when the subject is moving relative to the frame during the time the shutter is open. If the subject is moving “enough” you end up with a blurry streak. If this was not the intent you were after, it is an error caused by bad technique. You have to get the shutter speed up enough to “freeze” the subject.

    It is an internal fight with me to make myself raise the ISO speed enough to get the shutter speed I need. I have years of history that images were too noisy unless I stay down around 100 ISO. But with modern cameras it is much less of a problem. My default ISO is usually 400 now. I know that I can go to 3200 and still get good results in many situations. I just have to make myself do it. When I don’t I often get blurry images.

    Diffraction

    One of the things we worry about a lot is depth of field (DOF). This is sort of an illusory concept. It is an attempt to quantize how much of the area from foreground to background is in focus. The reality is only a very small slice is actually in focus. But DOF describes how much is in “acceptable” focus. But acceptable varies with taste and application. There is no official definition of DOF.

    One way we try to cheat the system is to stop the lens down more to increase DOF. It sort of works. It seems to work. But it is not free. Going to a 2 stop higher f-stop number, like f/16 instead of f/8 means that you are letting in 1/4 the light (it’s logarithmic). It also means you are incurring diffraction effects.

    Diffraction is a complex phenomenon. I will just say that at physically smaller apertures (say f/16 and smaller) the perceived sharpness of your image decreases. So don’t just automatically slam your aperture to f/32 to always maximize DOF. It has downsides. Most lenses have a “sweet spot” around 2-3 stops down from the widest aperture. If you have a great f/2.8 lens it probably has optimum sharpness at around f/5.6 to f/8.

    Diffraction is a real phenomenon of physics and I see it all the time. Don’t let me scare you, though. It is one of the tradeoffs. As an experiment sometime put your rig on a tripod and shoot a spread of the same scene at, say, f/5.6, f/8, f/11, f/16, and f/22. Don’t change the focus point. When you examine the images on your computer at at least 1 to 1 size you will see a fall off of sharpness at f/16 and smaller. On the other hand, the perceived DOF increases at the smaller apertures.

    Trading off DOF and diffraction effects is just one of those balances that photographers have to be able to make automatically. It’s all an artistic judgment. No right or wrong.

    Sharpening

    Regardless of how good or bad your equipment and technique is, at the end of this chain you are now in your computer looking at the image. What can you do?

    First off, expect your image to look blurry when you first view it. What?? I paid thousands for this equipment and it makes blurry images? Yes. if you shoot raw images (always shoot raw unless you have a very good and specific reason not to), almost no processing has been done on it when you first see it on screen.

    All those steps in the sharpness chain guarantee that is seems less sharp to you than you expect. Don’t worry. If you have done your job well you have good data to work with. We can do wonders to increase the perception of sharpness.

    It is the “edges” in your image, the transitions from darker to lighter, that give the perception of sharpness. We have many tools and techniques these days to increase the contrast of these edges.

    Lightroom tools

    If you work in Lightroom, as I do, (or Camera Raw, the equivalent) the Presence section has 2 magic tools: Texture and Clarity. Clarity is a bigger hammer. It increases edge contrast overall. It can really make an image seem to pop.

    Texture is fairly new. It is kind of like Clarity, but gentler and more selective. Increasing Texture concentrates on mid range edges. That is, it ignores the most contrast and least contrast edges and enhances the middle ones. This is a subtle and more fine-grained control. It is a welcome addition to the tool kit.

    Then for finishing an image there are the traditional Sharpening controls in the Detail section. This lets us tune the overall effect by controlling the amount, radius, and detail of the sharpening while being able to use the mask control to adjust the area it is applied to.

    These Lightroom controls are often all that is required to achieve great perceived sharpness. The more I learn the more I am able to completely finish many images using only Lightroom.

    Photoshop tools

    Your workflow or preferences or image needs may take you to Photoshop, the traditional big gun for image processing. There are several tools and techniques that can be used to increase perceived sharpness.

    My go-to tool for Photoshop sharpening is the Smart Sharpen filter. This gives marvelous results and lots of control. It even effectively lets us use Blend-If to selectively fade the sharpening application to highlights and shadows. It is a great tool. And yes, you can go crazy and make the image look horrible, too.

    Another traditional filter is Unsharp Mask. I won’t try to explain why blurring can cause the image to look sharper. It is one of the great mysteries of photography. Maybe a future article. Anyway, this is a software simulation of a technique used by film people to increase sharpness of their prints. It works well. It has somewhat less control than Smart Sharpen, but it is good.

    Then there is the HIgh Pass filter. You almost have to be an engineer to understand the concept, but basically it increases the contrast of the tones at edges to make the image look sharper. It is a very old tool, but it works great for some things.

    There are many possibilities in Photoshop, but I will stop with the Sharpen tool. It is a tool, not a filter. It is brushed on. This lets you brush a sharpening effect very selectively where you want it. It works, but be careful. It is a destructive tool.

    Perception is reality

    There are many options to use and most of them can be combined in various ways to meet your needs. But in the discussion, I kept talking about the “perceived” sharpness. This is the reality of our imaging world. All those stages in the sharpness chain lose quality. The operations we can do in software can make our image look very good. But all these tools we use are trying to simulate what the original scene or our creative vision looked like. All operate on the principle of enhancing edges to make the image look sharper.

    These operations do not actually make the image sharp. They make it appear sharp to the viewer. Maybe it is too fine a distinction. For most of us, all we care about is that it looks good.

    If an image is actually out of focus or blurred badly from camera shake or subject motion we cannot make it perfect. Yes, AI is getting better all the time, but it can’t really make something out of nothing.

    The good news is that these days we have excellent tools for controlling perceived sharpness and making our images almost as sharp looking as we wish.

  • Pixel Damage

    Pixel Damage

    Our images are precious. They are our vision, our creation. We need to treat them with care. Photoshop makes it too easy to damage your pixels. But with some training we can learn to avoid the damage.

    Photoshop and Lightroom Classic are the main editing tools I use and am familiar with. I acknowledge that there are other good tools, but I don’t use them. LIghtroom (I will just call it Lightroom because I think Adobe’s branding scheme is dumb) has the distinct advantage of being totally non-destructive. It is impossible to do any edit on a RAW file that damages or destroys the original pixels. This is a huge win.

    Unfortunately LIghtroom does not have the ultimate power and fine-grained control of Photoshop. So it is often necessary to take images into Photoshop to finish them. But Photoshop is a power tool. As with most any sufficiently powerful tool, it can be dangerous, even magical. I still see instructors training people to edit in ways that damage pixels. This is counter productive and seldom necessary.

    Photoshop can do any amount of damage you want to an image

    I love Photoshop. It is one of the finest pieces of software I have ever used, and I speak as a long time software architect and developer and long time Photoshop user. But is is dangerous. It will freely let you do anything you want to an image.

    Many of the tools in Photoshop operate directly on pixels. You can edit, delete, modify, blur, sharpen, recolor, or paint on your pixels. These operations are destructive after saving and closing the file. That is, you can never get back to the original pixels.

    This amazing power is a two-edged sword. It gives you total freedom to do anything you can imagine, but you can find out later that you cut off your foot in the process. I have images that have been severely damaged in the past because of a lack of sophistication in my editing techniques. I destructively modified the original file and saved it. Even though I have better knowledge and technique now, I cannot go back to start over with the original data.

    If you use Photoshop seriously you are continuously learning new and improved techniques for doing your work. This means you sometimes change your mind and want to go back and modify images you have worked in the past. But if you have painted yourself into a corner because of poor technique, you may not be able to do that.

    I hope to encourage you to learn that there is a better and safer way. One that lets you do anything you want in the confidence that you can change or modify anything in the future.

    Non-destructive editing

    Non-destructive editing is a holy grail of many of us who use Photoshop heavily. It is based on some fairly simple principles that are easily learned. It works very well, is no harder to do, and leaves us able to change our mind about an image any time in the future.

    This is not a Photoshop non-destructive editing tutorial. It cannot be in a short blog. I hope to motivate you to consider a more powerful way of using the tool and give you a few hints of what to pursue in your training. Dave Cross is a great instructor to learn from, as is Ben Willmore. They both have their own tutorial programs or catch them on CreativeLive.

    So here is the quick cheat sheet: use smart objects, adjustment layers and blending modes. Avoid stamps, layer merge, erase, rasterize, and flatten. There, all you need to know. 🙂

    Avoid these

    I suggest you should avoid using stamps, layer merge, erase, rasterize, and flatten. Probably I hit one or more of your regular tools. Sorry. But every one of these permanently commits the edit state. Each one is unalterable once you save your file.

    If you go back to your image a few months later and decide you had too much contrast in a certain area it is very hard to change it. You have to re-select, re-mask, try to make the changes without damaging the rest of the pixels. You are also doing a whole new edit and the result has now permanently changed the pixels to be the way you see the image right now. A few months from now…

    The Stamp specifically

    A favorite technique in many workflows is the stamp. You know, the “hold down the entire left side of the keyboard and E command. This is not the same as the clone stamp tool. The stamp avoids the problems of destructive edits, right? Well, sort of. Yes, it builds in “frozen” points that capture all the edits in an image below it and allows changes without destroying the underlying layers. That is good.

    I hope to convince you you can do better, though. The stamped layer is a roadblock in the editing flow. It marks a point where you can’t go back. When you inevitably decide to edit a layer below the stamp the edit is not reflected up to the stamp and above. You have to delete the stamp layer, recreate it then try to remember what you did to it before and re-make those edits. You may or may not remember what all you did. Adopting a non-destructive workflow avoids this problem.

    Another issue with the stamp is that it makes a copy of all your pixels. File sizes are growing almost unmanageable and the stamp makes it worse. I have many files that must be saved as psb format because the size exceeds the 4GByte limit of tiff. Large file sizes make for slower file open and save, slower editing, the need for lots of RAM, and the requirement for lots more disk space (plus all the backups; you backup religiously don’t you?).

    Use these

    Some things to get in the habit of using are smart objects, adjustment layers, and blend modes. Getting comfortable with these powerful techniques can have a dramatic effect on your editing. A major characteristic of all of them is that their settings can be modified any time and they do not alter pixels, just the way they look.

    Smart objects are your friend. They allow you to wrap a certain state of an image in a container and use it non-destructively. That is, it is in a protective bubble that prevents any operations from the outside that damages its contents. And the smart object can be opened and edited in any way at any time in the future. So it can be changed at will. All edits to a smart object automatically flow back into the file you are using it in. Sounds like magic. Until you get comfortable with them, it kind of is. Good magic.

    Adjustment layers are a simple concept that has been in Photoshop a long time. The subtlety is that there are adjustments that alter pixels but there are also adjustment layers that put a transparent sheet over the image and do their changes to that. Always use the adjustment layer. It is lightweight (doesn’t make a copy of the pixels), can be changed at any time, and can have a mask to restrict its effect to selected areas.

    Blend modes come in 2 types, the blend modes on a layer or brush and the blend-if controls to feather things between layers. I don’t have the space to go into them, but they can have a very beneficial impact on your editing. And they are forever changeable. And they do not grow your file significantly. All good things you will like.

    Don’t paint yourself into a corner

    This non-destructive workflow is all about not painting yourself into a corner (sorry if that is an American idiom that doesn’t translate well). It means not trapping yourself down a path you cannot recover from, a position where you cannot escape. This flexible way of working allows you go go back to any stage of your work and make changes. If you need, you can even strip off all the edits you have made and start over from the original pixels. Everything is preserved.

    No more unrecoverable originals.

    When you get comfortable with this way of editing you will not go back. For me, when I consider doing something that permanently alters pixels something stops me. It feels dirty or wrong. I can always find a non-destructive way to accomplish what I want and it is just as fast and easy.

    I find that when I come back to an image after a period of time I often want to make changes. Sometimes small tweaks but sometimes a complete reinterpretation of what I want the image to be. A non-destructive workflow allows me the freedom I want to be able to do this. And I never go down a path I cannot recover from.

    I consider that working in this way is a sign you are well on your way to Photoshop mastery, if there is such a thing.