Month: April 2021

Posted on

JPG vs. Raw

JPG looks good in typical situations

It seems like deciding on jpg vs. raw formats for our images is a problem for some photographers. I’m not sure why. Maybe it is lack of knowledge or maybe because it is sometimes discussed in almost mythological terms. Jpg and raw are just 2 ways of saving our images. Each is good for some things but there are tradeoffs to consider. It is just technology, not magic.

Image formats

When you take an image on your digital camera, each manufacturer has their own proprietary magic they do on the bits coming off the sensor. This lets them tune their image to meet their goals. If you shot the same scene with different cameras you would notice subtle differences – slight color balance differences, slight variations is tonal contrast, different handling of shadows and highlights, etc. These are usually small, but they give a camera it’s unique character.

But we need to consume these pixels in our image processing software. So there needs to be standardized ways of storing the images and reading them in our computer. These are file formats. There are 2 main choices.

Jpg is an industry standard format. The format is very widely understood and used. All images, once converted to jpg, are compatible.

What we call raw files are really proprietary file formats created by different camera manufacturers. Image processing software, like Adobe Lightroom, has taken the responsibility to be able to read the files written by virtually all camera manufacturers. For instance, I shoot Nikon, so the images LIghtroom reads and handles have the “.nef” extension. Lightroom knows how to interpret this and convert it to editable pixels.

The key thing here is that these raw files all contain roughly the same information, but are not directly compatible. Thankfully our software handles the differences gracefully.

Technical details – jpg

The term jpg, more precisely “jpeg” is derived from a standard created by the Joint Photographic Experts Group. The name jpg is an abbreviated acronym.

The problem was that digital files are very large, this made them consume lots of disk space back when disks were small. It also used up lots of bandwidth transmitting them back when internet was slow and much more expensive (anyone remember dial up modems?).

The jpg standard is based on some brilliant insights on human perception to allow encoding image files so they look good but are much smaller. The underlying principle is that humans are more sensitive to variations of tone (luminance) than they are to color (chrominance). The jpg processing reduces the luminance information and greatly reduces the chrominace data to acheive reductions of about 10x typically.

In general, transforming an image to jpg is a multi step process. It involves a transformation where the luminance and chrominance information is separated. Then the chrominance information is downsampled, or reduced. Then there is a grouping of data into blocks and a process called discrete cosine transform is applied to the data blocks. This transformed information is quantitized and encoded. Finally the data is written out in a defined format as a jpg file. It is not at all necessary to know these details, just that the data in a jpg file is far removed from the original pixels that came from the camera.

It is a lossy compression technique. Yes, it throws away a lot of data. This is one of the big tradeoff points of jpg. But a fringe benefit is that the image is made to look “nice”. The result is pleasing to most people without further processing

Technical details – raw

These files are called “raw” because they contain minimally processed data from the camera sensor. They are absolutely not ready to be viewed or processed. Some people describe it as a digital negative. Conceptually this is pretty good way to help us think about it, but it is not a valid description. The data is not negative and it is not viewable. It might be better to think of it as exposed but unprocessed film.

To follow this metaphor, a raw image processor like Lightroom “develops” the image and makes it viewable and editable.

Why raw? It captures and beings into the computer all the data that the camera sensor was able to record. It has the full range of color and tones. Nothing has been eliminated yet.

In addition, the raw format has not had any lossy compression applied. Nothing is thrown away or reduced. Because of these things an image from a raw file requires manual editing to complete it. Sometimes a lot of editing.

Tradeoffs

So jpg is made small as possible and generally nice looking as soon as you see it. You can immediately look at it or send it to someone or post it to your social media. Yes, some information has been intentionally eliminated, but that is not important to most people. If you don’t notice it then it must not matter.

On the other hand, if you want to make a large print of a jpg you may see noisy patterns that are euphemistically called “artifacts”. This might be mitigated with clever software, but your mileage may vary.

And there is an editing danger you need to be aware of: every time you save a jpg file it goes through the transform process to reduce data. So every time you edit it and save it you lose information and introduce more artifacts. If you want to edit a jpg always save the edited file in a lossless format, like psd or tif.

The raw files are usually very large. On my current main camera a typical raw file is 50-70 MBytes. A high quality jpg of the same resolution is around 4-5 MBytes. So, 10 to 1 or greater differential. And the raw files require an investment of time and training and tools to process them into a respectable state.

But, and this makes up for everything, the raw file preserves every bit of information that we can wring out of the sensor. A modern sensor is marvelous and enables very aggressive processing. The raw format contains the full resolution of the pixels. It is not limited to 8 bit data like jpg. I often do things with the image data that I could not have envisioned when I took the original photo.

Different needs

When would you want to use one vs. the other? Well, if I was shooting a wedding I would probably use jpg. Say I come away with 3000 images. I would want to be able to scan through and see good views of all of them so I can quickly pick the 100-200 best to share with the client. If I did my job well the images should not need much editing. I would not have time to process this many raw files.

Also, if I am shooting snaps of my family that is a time for jpgs. And if I was on vacation and just shooting travel photos for memories that is good jpg territory. I guess if my memory card was nearly full and I didn’t have a spare I might switch to jpg to keep shooting a few more frames. I try to prevent that from happening.

For me, any other time requires raw files. It is my go-to choice. I know I want to process the images heavily. I am not afraid of the techniques. Given the choice I will always want to retain the maximum information and resolution possible. This given me the flexibility to make massive changes or change my mind and go back to re-process the image for a different look

I tried to present a very neutral view of the tradeoffs of the 2 formats. I can sympathize that the choice is hard for some people. For me, it is straightforward. Use jpg if I am taking shots of people and I am confident it will need little processing. Otherwise, definitely raw.

The image with this article is a jpg. It looks fine for this application.

Posted on

Look Sharp

Sharp example: good sharpness and wide DOF

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.

Posted on

Pixel Damage

Old image barely salvaged from poor Photoshop editing

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.

Posted on

Light

Bad light photo

Maybe it seems silly to talk about light and photography. It seems obvious. But light is one of the things many photographers obsess about, worry about, plan around. Good light, bad light, golden hour, etc.

As photographers we need to need to be very aware of light. We cannot make photographs without light. The light we have at any given time strongly influences the pictures we make. Let’s talk about awareness of it rather than trying to tell you what light you should or shouldn’t use.

“Good” light

Good light. Ah, the holy grail. Many people search for it all the time. I know photographers who will not take their cameras out unless the light is “right”. Sometimes, I confess, I do it myself. You know, its a blizzard out, I won’t bother. The light isn’t right.

This very elitist view is unfortunate, but people come to the attitude honestly, because that is what many instructors teach. They say you have to research a location, find the exact right time of year and angle of light for a particular landscape subject. Then hope the weather cooperates on the one hour window have you allowed yourself to shoot your subject.

And “golden hour”, the prime time for all outdoor photographers. Many people are taught that it is worthless to even try to shoot after the sun has been up an hour and until an hour before it sets. Learn to think different. You are needlessly limiting your opportunities.

If your thing is shooting portraits perhaps you prefer an overcast, soft light day. This makes gentle, even, predictable light for excellent results. No doubt, but how many of those ideal times do you have? And how many good opportunities do you miss because the light is not exactly the way you want?

“Bad” light

Photo instructors teach, or at least imply, that there is bad light that should be avoided. The harsh light of midday is a prime example. It is made to seem that no self respecting photographer goes out to try to shoot when the sun is high overhead. The shadows are harsh and the light is flat and boring. At least, that is what they say.

Or maybe it is a uniformly overcast day with a bright but featureless sky. Terrible we are taught. There is little tonal separation and the sky is flat and boring. You can’t shoot good landscapes then.

Or after sunset when there is no direct light and the exposure times are getting long. That is another time people pack up their equipment to leave.

Or even if you are shooting in midday (shame on you) but you forgot to being your 11 stop neutral density filter to being the light levels down enough to do a 10 second exposure of that waterfall to streak the water like you intend. That must mean the light is bad. Or…

Adaptable

OK. If you have held on this long you will probably get that I’m suggesting that light is not good or bad, it just is. Use the light you have. Embrace it and figure out the best way to use it. In most of the examples I cited above the photographer had a fixed expectation of what they wanted to see and shoot. If the light did not match their expectations, it must be “bad”.

Most of us do not have the resources of, say, National Geographic funding our shoots. If we take a trip to a location we have been wanting to photograph, we can’t just hunker down and wait it out for a week or 2 if the weather is not what we wanted. We have to be flexible and adaptable.

If we get to our location and it is closed for some reason, we cannot change that. But we can find something else maybe even more interesting. Same with the light. Figure out what works right now. What you end up with may be better than what you planned.

Try practicing this flexibility. It is a great creativity exercise. An attitude and practice is what Jay Maisel calls “going out empty“. That is, leave your expectations at home. Just wander around and learn to see things that are interesting. Things the light works for. Things that excite you. In doing this you develop the skill to be able to work with what you have before you. To let the conditions, including the light, guide what you do. But no matter the conditions, to be able to make interesting pictures.

Street photographers are probably better at this than most of us. They have to be adaptable. Light not working here? Move. Can’t find the subject you had in mind? Get interested in what is there. They are used to letting their creativity guide them to good shots.

Awareness

I’m afraid too many instructors train their students to be internally focused. To have a preconceived idea of what they are going to shoot and to reject anything else. There are certain times when a fixed idea of what you want is required, but for most of us, it does not have to be the normal pattern.

I encourage you to instead be aware of the light, of the surroundings, of the activity or the scenes around us and flow with it. Use your creativity to make something interesting out of what you find rather than coming home disappointed because you did not find exactly what you intended.

Take the photo at the top of this article for example. Midday, flat light, hazy, featureless sky, not what I would usually look for. But I think this works. 🙂

Maybe it is too far of a stretch, but it seems to be an attitude of gratitude. Be thankful for what you have and learn to find surprises that delight everyday. Who knows where it will take you?