Exporting Images from Lightroom

Why is the concept  of sizing or setting the aspect ratio of images for print so difficult for some people to grasp? I don't get it. So here's an attempt to address the issue in a way that might help people understand it better.

Pixels-per-inch or Dots-per-inch

This comes up OVER and OVER and OVER again and people do not understand it.

The pixels-per-inch (ppi) box is only useful to people who specify their image sizes in INCHES and not in pixels. If you tell Lightroom (or Photoshop or any other program) that you want a picture to be 10 inches wide and 300 ppi, that means you are telling it to give you a 3000 pixel wide image. If you want  you can do that, but you could also tell it you want to print a 100 inch wide image at 30 ppi and you'll get EXACTLY THE SAME IMAGE. Or you could just tell it to give you a 3000 pixel wide image and it doesn't matter what you put in the ppi box! Here:

One of these images is 819 pixels x 1024 pixels at 1 (yes, that's ONE) ppi. The other one is 819 pixels x 1024 pixels at 1,000 ppi. Can you tell which is which? No you can't. THEY'RE THE SAME. Keep reading...

Added after the fact: This post is getting a LOT of hits because this question seems to come up a lot on Facebook. Both the Aspect Ratio and the Print Resolution questions. If you find this useful, please consider clicking the donation request at the bottom – purely optional but it would help! At least, I would ask you to subscribe to my newsletter (it's free and you can unsubscribe with one click if you must). Here's the link  

Now let's look at the ASPECT RATIO, or in plain English, the SHAPE of the picture.

Suppose this is the original picture:

This image is copyrighted and the property of the author. If you want to use it for any purpose whatsoever, you need written permission. Email me. If you want a print, I'd be happy to accommodate you at a very reasonable price. Please don't rip it off.

It's ASPECT RATIO (that's the technical term: maybe it's easier to understand the word "SHAPE") is 2:3. What does that mean?

For every 2 units high, it is 3 units wide. When it came out of the camera it was 4912 pixels high x 7360 pixels wide. That's out of a D800 so it's 36 million pixels if you multiply it out. If it came out of an old D70 it would be 2000 x 3008 pixels: only 6 Megapixels, but exactly the same shape — 2:3*.

* for all intents and purposes. I know neither one is EXACTLY 2:3 but we're splitting hairs here.

A Canon 60D has a resolution of 3456 x 5184 pixels. Do the math: guess what the shape of the image is? That's right, 2:3. Pretty standard in the DSLR world. Your camera might be different but let's go with this example.

OK, what if you wanted to print this picture without cutting ("cropping") anything off the height or the width of the image? What shape would the paper you're printing on have to be?

If you said 2:3 then you're staying with me! Can you think of some examples of paper that is that ratio? Look at this list:

• 4" x 6" (10 cm x 15 cm)
• 5" x 7" (12.5 cm x 17.5 cm)
• 8" x 10" (20 cm x 25 cm)
• 8" x 12" (and so on...)
• 11" x 14" 
• 12" x 12"
• 16" x 20"
• 16" x 24"
• 20" x 30"

Which ones are the right shape? Do the math, I'll wait right here...

If you said the 4x6, the 8x12, the 16x24 and the 20x30, now you're cooking!

So what happens if you try to print that picture on an square piece of paper, say 12"x12", for example. One of three things:

1. the image can completely fill the paper, but some of the width will get cut off. In other words, you're printing the full 12" height, but there's only room for 12" of width so the other six inches are going to get cut off (remember your picture is 2:3 shape, so if it's 12" high it's going to be 18" wide).





2. you can squeeze the 18" width of the picture into the 12" width of the paper, but then you'd be squeezing the height too, so the height will only be 8" high and there'll be a 2" band of unprinted paper above and below the picture, or






3. you can squeeze the width of the picture without changing the height, but this distorts your picture (you'd have to get fancy in Photoshop to do this, though)

Get it? So it's your choice how you want the image to be printed (nobody ever uses the third choice), with some stuff getting cut off or with some unprinted paper. Make sense?

So what should you do to make the picture fit the paper? It's up to you, of course. Chances are it isn't going to get cropped the way you like it by the printer, so you should decide in advance and change the shape of the picture to match the shape of the paper. That's called "CROPPING" and you do that in Lightroom in the Develop module. 

The red arrow shows you where to click to get the crop tool. The blue arrow shows where to click to get the dropdown menu that you see with the different shapes 

So do the math – it's basic grade 4 arithmetic – and figure out what aspect ratio you want. If you don't see it there, you can click "Enter Custom" and put in your own numbers, so if you want your own special size, that's where you can do it.**

** This isn't limited to printing, you know. For example, the shape of the header at the top of this blog is 1280 x 350 pixels. If you look closely at the list, you'll see "3.657 x 1" which is that shape. You don't actually have to do the math, you could just enter 1280 x 350 and Lightroom will do it for you!

Added Later:

A good workflow practice in Lightroom would be to create a  "Virtual Copy" of the image if you're going to produce different crop shapes. Then work your magic on that copy without affecting your original. You can't rename the virtual copy without renaming the original but you can put something in the "Copy Name" field in the metadata, like "square version" so that you can search for it later. 

Summary:

So now you should understand why, when you try to print your 2:3 ratio picture on a piece of 5x7 paper, some of it is getting cut off. The only way to prevent that is to print it like the second example above. Or you can decide to cut it off wherever YOU want by cropping in Lightroom. So far so good?

Now let's look at RESOLUTION.

You're probably looking at this article on a computer screen (OK, maybe on a tablet or phone, but let's go with the easy example!)

When you set up your computer, you chose the screen resolution (or, let's face it, you accepted whatever you got when it came home from the store). Depending on what kind or shape of monitor you have, you might have selected 640x480 pixels (if you did, you're older than me. And trust me, I'm OLD!). Or 1600x900. Or 1280x1024. Or some other number. My 26" Samsung monitor is set for 1920x1050 pixels.

Now just to confuse you further, my "26-inch" monitor is really only 23½" wide. The 26" is a diagonal measurement. But let's divide that 1920 pixels into 23½", we get about 81 pixels per inch. The universal standard for computer monitors and the Internet is 72 pixels per inch but who's counting? Close enough.

You could quibble with me and tell me you have a fancy 'Retina' display or a giant monitor, or a $500 video card that lets you see 4000 pixels on your screen but hey...

Now go to the top of this article and read what I said about the resolution of the owl picture as it came out of my camera. Wow. 7360x4912 pixels! If I tried to look at every pixel of that on my computer, it would be dripping out of my monitor and down both sides of my tabletop! So guess what? I CAN'T look at them all. I can only squeeze them into the 1920 pixels (wide) that are available to me.

OK, now go find a magnifying glass. I'll wait right here. Are you back? Now look really closely at your computer screen. See how it looks like a bunch of little square dots? Those are called pixels.

So if I want to output a picture for use online or just to be viewed on a computer or tablet, I don't need it to be any bigger than about 2000 pixels (wide). Otherwise it's a waste anyway, you can't see the in-between ones. Since you probably aren't looking at any pictures at the full width of your monitor, even that big is a waste. Make them smaller!

Facebook is special. They have some nasty compression algorithms that mess up pictures. If you research it, you'll find that the best size to make a picture, to optimize how it looks on Facebook, is 2048 pixels. Don't even think of asking why.

Your computer is smart. You don't have to tell it how many pixels per inch to make the picture, that's determined by the computer and the monitor. Even if you tell it you want 10,000 pixels per inch it can't do it! It's going to change it to 72 ppi (Pixels per inch) for you. The important thing for you to decide is how many pixels (wide, for example) you want the image to be. Half a screen width is about 800 pixels. What about if you only output your picture at 200 pixels wide, and then tried to blow it up to full-screen (your browser can do it, for instance)? Then it's going to look all unsharp and fuzzy because there's not enough information to exactly light up all of the pixels. Get it?

So if you're outputting a picture for the internet, somewhere between 800 and 2000 pixels is best. Of course the bigger you make it, the bigger the image file size becomes, the slower it is to load, etc. Your call...

I choose to make my pictures 1280 pixels wide for this blog. Here's what my settings look like in the Lightroom export module:

The internet, and your computer monitor, all work in the sRGB colour space. If you set it to something else, your colours are going to look funny. And generally you're going to want JPEG image files, the internet likes them the best. 

Understand that your files inside Lightroom are probably not going to be that size. For that matter, they won't necessarily be JPEG's, they may not be in that colour space... they're your NEGATIVES. You need to EXPORT the pictures from Lightroom to use them for anything, and this is where you get to tell it what kind of files, what size, etc.

Now what about for printing?

Go get your magnifying glass and a photographic print. Do you see a bunch of little pixels? (If you do, you're not looking at a photo print, you're looking at something reproduced on a printing press, not an inkjet or giclée printer). Your printer has a bunch of little nozzles that spray ink on the paper but in a really teeny-tiny pattern. Some of the better ones might spray up to 1400 ink dots per inch! (DPI stands for "DOTS per inch". PPI is "PIXELS per inch". One's for ink, the other for screens).

I'm not going to get into how the ink spreads as it soaks into the paper, or how it blends together with nearby dots. Suffice it to say: the universal standard for not being able to see the little individual dots is 300 DPI. Some people use 240 DPI, some use 360... but let's call it 300 to be safe. If you have less information than that the in-between dots are going to take on an average value, not the right value, so your picture will get fuzzy.

Here's the same owl shot but I saved it at only 128 pixels wide then blew it up. The same thing happens if you don't have enough pixels to print! And you need a lot more for printing than you do for displaying on-screen! 

Doing the arithmetic, if you want a sharp 8x10 print, you are going to need 300x8 dots by 300x10 dots, or 2400x3000 pixels. For a 16x20, double that. For a 20x30, you need 6000x9000 dots. That's more than most cameras can do but Adobe, who makes Lightroom and Photoshop, and a bunch of other companies know how to intelligently upsize your pictures. The person who REALLY knows how is the company that you're sending your images to print! Always ask them what they want when your image is undersized.

Where were we... by now, you understand that the really important number is how many pixels (or dots) your picture contains. Just like your monitor, that printer will change the number of dots per inch to whatever it has available. If you give it only 1500 dots and you ask for a 10" print, you're only going to get 150 dots per inch (which surprisingly, isn't bad most of the time!).

If you give them too many pixels, don't worry! The printer is smart (not necessarily the person pushing the buttons, the machine!) and it will figure out what to do with the in-between ones, just like your monitor does if you give it an oversized picture. The files might be too big, though!

Why does Lightroom have that dreaded resolution box I marked with an "X" above? It's for people who would prefer to output pictures of a certain size in inches (or cm) and then you have to tell it how many pixels per inch to render. If you do the math yourself, and specify the image size in pixels, that box doesn't matter.

Here's one of my typical settings for going to print (it does 14" on the long side at 300ppi). Note that most printers like JPEG/sRGB as well, but talk to your print shop!

So get in the habit of figuring out how many pixels or dots you want. Let the computer do the rest. Or, choose the size of the picture you want in inches (or cm) and tell it 72 ppi for screen or 300 ppi for printing.


BIG SUMMARY.

In Lightroom, first choose what SHAPE you want the image to be using the crop tool in the Develop module. Next choose what SIZE it should be in pixels wide x pixels high, using 72 ppi for use on screen or 300 ppi for use in print in order to make your calculations.

If you're going to be printing different sizes and shapes, it's probably smart to create separate output files for each version and virtual copy is one way of doing that and staying organized.

Make sense?

Check out my weekly blog, The Faczen Image! Click the "Newsletter" button if you want to be kept up to date. Try it. I dare you.

— 30 —

Lightroom Performance and Previews

This article was written in response to a request from the Admins on the Facebook "Lightroom Q&A" group. 

---

If you appreciate the effort, please have a look at my main weekly blog which contains my "sporadic musings and compelling images" as well as photography and post-processing tips and techniques. All designed to give you ideas! By hitting the "Newsletter" button at upper right on the blog, you'll get an email note when a new post appears, as well as free access to my eBooks and other resources.

---

Note: I use Lightroom 5 (version 5.7 at this writing) and Photoshop CC 2014. Some of these functions also work in earlier versions of the programs. I run on a PC, not a Mac but they're almost the same in Lightroom.

Lightroom sometimes appears to respond and process images slowly. Some of that is due to your preview settings, which you can change to match your needs and available hardware.

Right after importing new images, people sometimes complain that when Lightroom opens a RAW file, first it looks sharp, then it looks ugly for a while, then it gets sharp again, especially when viewing the image larger than grid view. What's happening?

That exact sequence really only happens on newly imported files, where Lightroom hasn't yet had time to generate the standard previews. First thing you see is the JPEG view that is embedded into the RAW file by your camera. That gives way to the unprocessed RAW file while Lightroom churns away in the background to generate the 1:1 preview. Finally you'll see the image in all its glory, including edits you've made inside Lightroom.

You may also see a delay in seeing this full quality image when you view an image in a larger view (like Loupe, or in the Develop module, or especially when you zoom in to a 100% view). Again, Lightroom needs the 1:1 preview and may ask you to wait while it's being generated.

Importing new images

In the process of importing new images, Lightroom's overall performance might appear to be slow because it's working in the background, using computer resources. One of those things is generating the full 1:1 preview you use whenever you zoom in on an image. You can improve that performance by delaying when that happens (you can do that at any time from Library → Previews in the main top menu). Here's what the upper right corner of your LR screen looks like just before you tell it to start importing:

You have four options here. These are the settings I use. I always make a second copy of all images to a different hard drive – in this case the internal drive in the computer (the originals are stored on an external drive). I want Smart Previews for all my images so I can edit them when that external drive that contains the actual images is not plugged into the computer. I don't tell it to block suspected duplicates because in my workflow, I format all my camera memory cards after the pictures are safely imported: if I forget, that's my problem!

I tell Lightroom to build 1:1 previews for all images during the import process. If I wanted to make importing faster, I would select "Minimal" previews or one of the lesser sizes. But I don't shoot NFL football games where I only have minutes to process and cull thousands of images during halftime. When I do an import, I generally plug the card into the computer, start it going, then pour myself a cup of coffee (or something stronger!) and give it time to get it done. 

Hard Drive Space

Now let's look at what those 1:1 preview files look like.

Previews take up some hard drive space. But not that much, in the grand scheme of things! Remember that Lightroom is just a glorified database: it does NOT contain your actual images, just pointers to tell it where to look for them, just like that card file in the old public library tells you which shelf to look for when seeking a specific book.

This Lightroom Catalog is my main one. It contains 63,423 images as I write this, many of which are 40 Mp giant RAW files from my D800 (not that it matters what size the images are to the catalog!). If you ever want to know WHERE your LR catalog is, you can bring this dialog up by clicking "Catalog Settings" on the "Edit" menu in the Nav bar at top left (may be a bit different on a Mac). So I have almost 3 Tb of images (3000 Gb) all listed in a 2.47 Gb ".lrcat" file in my internal hard drive.

I don't want to stray too far off-topic here but this is a screenshot of my catalog folder on the hard drive. The Lightroom catalog itself is 2.5 Gb. When you're backing it up (or copying it elsewhere) THAT is the file you need to copy. The two .lrdata preview files are not important to back up, Lightroom will regenerate those. The Catalog Preview file is relatively big: 13 Gb in this case. 

Now my Catalog Previews file is ONLY 13 Gb because I've chosen to manage its size. If I had full 1:1 previews from all 63,000 images it would be much bigger. Here's where you manage that.

In the same place as before (Edit → Catalog Settings), open the "File Handling" tab. You have three choices you can make about your previews. 

The first line is about the SIZE of the preview.

My biggest monitor is set to 1920x1080 pixels. I chose a preview size so I could see ALL of those pixels, so just larger than the screen. If you have a lower resolution monitor, you don't need the preview to be that big.  

You can also choose the preview quality. I find "Medium" to be good enough for me. Remember, you're not changing the image, just how you view it onscreen when you zoom in to edit. When I do a really fine edit, I switch to Photoshop.

Now here's the big space saver. You get to decide how long the huge 1:1 previews stay on your computer.

In my case, I decided I don't need the 1:1 previews instantly available on images I'm not currently working on. So I tell Lightroom to discard them after one week.  

Performance

When you click on an image in Loupe view or Full Screen, or in the Develop or Print or Slideshow or Web module, and especially when you zoom in, Lightroom needs that 1:1 preview. So it takes some time to generate it if you have discarded it. Once it has, it'll keep it around as long as you've told it to in the above menu.

Adobe has a resource page if you want to read more on this subject. It's here: https://helpx.adobe.com/lightroom/kb/optimize-performance-lightroom.html

To summarize: 

If you have tons of hard drive space, go ahead and keep your 1:1 previews forever. If you have limited RAM or CPU speed, you may want to hold off generating them immediately on import. Lightroom can churn away in the background if you tell it to (Library → Previews from the main top menu) or when you decide to look closely at an image. It all depends on your hardware resources and your workflow requirements.

Does that make sense?

---

Please take a moment to visit my weekly blog. I dare you.

— 30 —

Black and White images on Black Backgrounds

Black and White images on Black Backgrounds

A while ago, a British photographer by the name of Antony Northcutt posted some images of black-and-white flowers on a black background. They were outstanding and I wanted to learn how to do that. Fortunately he had written an eBook documenting his technique, which I bought and studied.

Pablo Picasso said that a “good artist copies; a great artist steals”. Don’t take it literally — the great artist uses the work of others as an inspiration, builds on it and makes it his own. I loved Northcutt’s work but not only did I change his technique but also the result as well. It inspired me to create something my own and I hope this does the same for you.

What I love about Photoshop is that there are a million ways to do anything. That means that everyone’s approach is going to be different! In this short tutorial, I’ll show you my workflow and some of the thinking behind it, to give you some ideas.

 This deals with making black and white images of flowers (or other subjects!) on black backgrounds. The key points are subject choice, selection, background rendering, and toning including non-destructive dodging and burning. I edit in Photoshop CC 2014 but most of this works in earlier versions as well.

Please click on any image to view it full-screen.

The original image (note: these images are all screen captures). I chose this image for this tutorial because it would help illustrate my techniques, not because it's a particularly strong artistic image!

Choosing the Subject 

Here’s my weak suit. I don’t have that great an eye. I’m looking for repetitive patterns and not-too-fine details. Flowers are easier because they’re created that way! But they don’t all work, and flora with fine textured edges elude me. Because we’re doing black and white and removing the background, you can choose subjects without too much regard for background. I’m looking for something big enough to fill most of the frame because I don’t want to have to crop too much. And with soft lighting because I want to save those deep blacks for the background.

When I take the picture, I expose for the subject because the background is going to be removed anyway. I want the broadest possible tonal range so of course I shoot in RAW. Sometimes I’ll bracket exposures because colours can fool you and occasionally I might blend exposures in an HDR program as well. The key is to watch your histogram and make sure you don’t blow out the highlights.

This works well on Portraits too!

Initial Processing 

I’m a Lightroom fan. But the toolbox in Lightroom and Adobe Camera Raw are identical so use whichever one you’re most comfortable with. I find it easy to surf through the images, switching from colour to black and white and back again, to help visualize what will work best. Sometimes I’ll crop at this stage, because I know what I’m shooting for anyway. I personally like square images for this approach. At first, my goal is to tone the image so that I have details at both ends – dark and light. Again, ignore the background and just look at the subject! I will be toning in more detail later, so this is just a “rough cut”. Now you’re ready to export to Photoshop.

Another reason to crop at this point is to reduce the file size. As you add layers in Photoshop, you increase the file size exponentially: working with a full-frame image from a D800 as I do, you can get above 1Gb really quickly! 

Once there, my very first step is ALWAYS to hit Ctrl-J, Cmd-J on a Mac, to duplicate the background layer. I want to leave the original layer intact in case I want to go back, and I don’t want to rule out reducing opacity or changing blend modes on the working layer(s). Turn off visibility of the background layer. The next step can be the tough one.

Making the Selection 

Again there are a number of ways to skin this cat. For me, one of the simplest is the Quick Selection Tool, followed by Refine Edge. My goal here is to get all of the flower, I can finesse the edges later. Another easy tool to use is Topaz Remask. Make sure you get it all. If you select too much you can erase it later, but it’s much harder to add something you missed. Using Refine Edge, you can now clean up the edges of your selection. This isn’t a workshop on how to do that: suffice it to say, the simpler the flower, the easier it is. A lacy subject with complicated edges can be very challenging!

Now once it’s selected, ctrl-J (cmd-J) works a little differently: it copies your selection only onto a fresh layer.

Making a black background 

I’ve seen this done a number of ways, but I’m in favour of the KISS principle. Northcutt had you select the layer underneath, create an exposure adjustment layer, slide the exposure down as far as it will go, and you have an almost-black background. There’s one advantage to that method: if you want to keep some vestiges of what was around the subject in the final image. But the problem is, you won’t get a pure black. That makes a huge difference when you go to print. But every artist and every image is different!

Here’s my simple technique: 

• Create a new layer. 
• Fill it with black (alt-backspace on PC, option-delete on Mac). 
• Slide the new layer underneath your selection layer. 

You’re done.

Now it’s time for a little cleanup. You can go into your selection layer and use the eraser tool, but if you make a mistake, it’s hard to go back. Instead, add a mask on the layer and paint on the mask. Anything painted in black will reveal the layer underneath, anything painted in white will hide it – and it’s easy to go back and forth (the “X” key switches the foreground and background colours). Depending on how fussy you are, this can take a few moments, or hours!

Sidebar: "Do as I say, not as I do"! Mea Culpa: when I did this image, it was more about creating this tutorial than a final image so I took some shortcuts. Trust me, it's much better to use a mask because you can fine tune right down to the pixel level. And I would not consider doing this kind of editing without my Wacom tablet and stylus. If you're reading this and you don't have one, it's time to go shopping (Amazon Canada or Amazon US).

The Black and White Conversion 

You could leave your image in colour. You’re the artist! But I like black and white for this, and, well, I’m the artist today! Again, there are lots of ways to approach this. Before we do…

At this point I give up my totally non-destructive approach. I keep everything I’ve done up to now separate so I could go back if I want to, but I personally find it easier to work on a combined layer at this point. So I “Stamp” a fresh combined layer on top of the others, without deleting them. You do that with “Ctrl-Alt-Shift-E” (Cmd-Option-Shift-E). That takes a lot of fingers and I do it often, so I’ve programmed that into my Wacom tablet on a single hotkey. You can make a keyboard shortcut for it too.

Sidebar: You’ve done a bunch of work. Save often as you go along. Save it once early, then just hit ctrl-S (Cmd-S) when you think of it and you’re saving it again. Save it as a .PSD file to preserve all your layers. 

OK now for the black and white. You could just go up in the menu and change the whole thing to greyscale. Boooh. No control! In the layers pallet, create a black-and-white adjustment layer. Now you have a bunch of sliders that let you choose what happens to each colour in the original. Play. This is where the artist in you starts to come out.

Sometimes I’ll use an outside plugin program called Google (Nik) Silver Efex Pro 2 for the conversion. It does much the same thing except you get much more control. Topaz also makes a similar program called “BW Effects”. These are great for adding multiple levels of control.

Sidebar: There are other plugins and techniques that you might want to consider using. Topaz Clarity. Adding detail and sharpening with a hi-pass layer. InFocus. DFine. Viveza. It’s your call… (Here's the link to Topaz Labs) 

Toning your image 

I have two more steps in my workflow to make the image jump out at you. One of them is toning the image using Curves, the other one is to dodge and burn certain areas.

 I’ll next add a curves adjustment layer which adjusts the overall toning of the image. Make the curve “S”-shaped if you want more contrast. Pull specific spots to re-tone specific densities. Now I'm going to “Stamp” another fresh layer because I want to do a non-destructive dodge and burn.

The dodge and burn trick: create a new layer and fill it with 50% gray. Change the blend mode to “overlay”. In this mode, 50% grey has absolutely zero effect on your image. But if you paint in black on this layer, it darkens, or “burns in” what’s underneath it. If you paint white, it lightens it, or dodges what’s underneath. I usually choose a soft brush, turn the opacity down (sometimes really, really low, like 10%!) and I paint. If you paint over the same spot, it gets darker (or lighter). The “X” key switches back and forth. Make a mistake? Switch colours. Or just fill the whole layer with grey again and start over!

As a rule of thumb, I burn in the dark things and textures to make them stronger, and dodge the light things to make them soft. In this case, I also used some local sharpening to bring out the water droplets.

This is what the dodge/burn layer looked like on its own after I was finished painting.

Almost done! 

I added a hi-pass sharpening layer (much the same technique as the Dodge/Burn layer. "Overlay" blend mode, the Hi-pass filter is under "others" in Photoshop, I typically use a value of 4 px). I did a little more dodging and burning. In the sample image, I had included a stem, which I wanted out of focus. I used an iris blur, but it caught the edges of the flower so I added a mask and painted out the areas I didn’t want blurred. Then I decided I did want something other than pure black in the background, so I made a new layer, opened it in Topaz Impression, masked the flower edges then masked the interior to retain all the detail. Done!

At first, I just masked the edges of the flower so they would remain crisp. But as I looked at it more closely, I decided I also wanted the textures in the flower to stand out, so I continued painting on the mask as you see below. 

I was trying to remember which preset I used in Topaz Impression. I think, "Georgia O'Keeffe I" but I'm not sure. That's why you should name your layers as you go along! 

If you want to see what the image looks like matted in a frame, increase the canvas size, fill it with white, draw a keyline and add your signature!

Summary 

The key here is to select a subject with great tonality and texture, but not too much complexity. Visualize it in black and white. Put it on a pure black background, then enhance it with the traditional darkroom tools now translated into Photoshop. Pay attention to detail and your results will be outstanding.

Here's another image for your viewing pleasure. I used much the same series of techniques and finished it with both Topaz Glow and Topaz Impression. This was made from a small section of a larger image; the selection of the flower was somewhat more difficult due to its complexity. The grey background around it is the version I use in Social Media or my normal weekly blog.

Online resources: 

Antony Northcutt eBook
Non-destructive dodging/burning


Interested in doing a workshop on this or another topic? Please contact me by email or visit my website at www.photography.to.

— 30 —

Planning or accident?

Many of the readers of my regular blog are not photographers (they're people who like my photos and hopefully will contact me to say they want to buy one!). The ones who ARE photographers are usually not as heavily into post-processing as I am. However this is directed at those who make pictures.

Good pictures can happen by accident. No question. But sometimes you can help the process along by pre-planning and attention to detail.

I've been preaching about many things, in my courses, in my newspaper columns, in my blog and in club sessions and outings. You know, "work the scene", "know your equipment" (my famous "RTFM" comment), "it's all about the light", "don't use automatic", "envision what you want the image to look like".

There are others out there who insist you can't get a good image without thousands of dollars of lighting equipment, fancy lenses and cameras... (OK, I do have a D800 but a D70 would have made this shot. Even a Canon would. Maybe. {insert smiley face here}).

I'm a planner. I don't expect that everyone is and sometimes it's a detriment: I spend a lot of time thinking about what I'm doing and I lose spontaneity. "Everything in moderation", so maybe some of this will rub off. In this case, I want to go behind the scenes and tell you a bit about the thinking that went into the making of this image.

Steve Hill is the Curator of the Haliburton Highlands Museum and like his predecessor (whom I met in 2008), he likes to play "blacksmith" for his adoring public.There's a Blacksmith shop in the old Homestead exhibition at the museum. The camera club arranged that Steve would be there that day, so this was all planned and he knew he was there to be a model. Nothing candid about this shot, but he wasn't directed to stand 'just so' and do certain things.  Click on the image to blow it up.

When I heard we were going there for a shoot, I had already envisioned the picture I wanted. Not exactly, of course, but back in 2008 I had taken some shots of his predecessor, Tom, in the same shop and I had great success with using off-camera flash fill. I planned to do it again. The hardware setup, for those who want to know, was a Nikon SB600 speedlight with a Gary Fong diffuser attached, fired remotely by using "commander" mode in the D800 camera. Lens was my mid-range 24-120 and I shot handheld at 40mm. Although I tried the 70-200 it was too far away to get the angle I wanted and to include the window.

I thought I needed to have the window in the shot. It was a light source, of course, but I needed to anchor where the light was coming from and I put the flash on the same side, the left, to match it. I stepped to the left (sorry if I got in other photographers' way!) to get the composition right, with Steve not blocking the window. I needed the vertical orientation to include the anvil, his face, the window...

It's important to understand that when you're mixing flash and ambient light, the shutter speed controls the ambient and you control the amount of flash fill with the aperture (assuming you don't fiddle with the ISO). In this case, though, I used the built-in electronics, the camera told the flash how much output to use by measuring the light through the lens (TTL) during the shot. So I had to use "Flash Compensation" to adjust that. The point is, the shutter speed makes no difference to the flash since the burst of light is so much faster.

The ideal way to do this is to shoot some test images without the flash, to get the levels right, then add it in. That's what I did, and I dialed the flash back to -1ev, so it wasn't acting as the main light source. The end exposure here was 1/30 second at f/8, ISO 200. The speed was a bit slow, I probably should have opened a stop or popped the ISO up to get 1/60 but it worked. VR on the lens helps. I locked the aperture at f/8 to get a comfortable depth of field and besides, it's the sweet spot for that lens. Actually, good thing I did, because it was the right speed for the sparks coming off the anvil.

So I actually thought about all this stuff before going out to shoot: I made sure I had the flash and diffuser with me, and the right lens, a tripod (which I ended up not using for this shot). I knew I wanted commander mode, that I would be dialing the flash back and would measure the ambient light before the shot.

I made a couple of mistakes. Two that could have been disastrous and in fact, caused me to only get a couple of usable shots. (1) I had been shooting brackets before and forgot to turn it off. So many of my frames were not exposed correctly and it took me a few minutes to figure out why. (2) Would you believe I left my memory card at home in the computer? All I had in the camera was a 4Gb CF card and I only discovered this later when I tried to take some other pictures and couldn't. The camera said it could only hold 30 images, in fact it held 90 before giving up. (3) I didn't get the right lighting on Steve's beard. Then I forgot about it and I was lucky it worked out.

Post-Processing.

When I got home (and struggled to upload the pictures, I can't find my CF card reader), I zeroed in on this one frame. Here's what it looked like before I did anything to it, except cropping and straightening it up a bit:

So I didn't really need to do too much to it. I tweaked it a bit in Lightroom, I wanted to take the clarity back a little bit to soften it and adjust the white clipping so nothing was blown out, and I added a bit of dark vignetting to keep the eye in the image. Now over to Photoshop.

My first action in Photoshop is almost always to hit ctrl-J and duplicate the background layer. Two reasons: I want to preserve my original image in case I decide to go back to it and I may want to blend all or part of it into the finished image. In this case, I also knew I wanted to treat Steve and the background separately, so the next thing I did was to make a selection of him, the anvil and sparks, and copy them onto a separate layer. Now I could tweak the lighting on his face and eye, add some sharpening to the beard. Like I said, it didn't need much. The joys of getting it right in camera!

Now the background. I cheated. I tried a few tricks, like motion blur, adding texture... even Topaz Simplify. In the end, I opened it in Topaz Impression and looked at a few sample styles. I immediately loved the "Impasto 1" preset, unmodified. However, as I mentioned earlier, the result was a bit too strong so I reduced the opacity of that layer to let the original show through a bit. And the layer with Steve on it is not at full opacity either, to help blend it in with the rest of the image.

I turned the layer with Steve and the anvil on it back on. I corrected a few edge issues and lightened up the anvil, and saved it. Done, except for my "BlogFrame" which is how I like to show my images online. I had written a short Photoshop Action to make that happen: it put a grey frame and dropshadow around the image, prepared a text layer for the title and a separate layer where I could stamp on my signature. Two minutes later: finished.

So what's the message here? I've written up what I thought about and what I did from the initial concept to the finished image. Did I need a fancy 36 Megapixel D800 DSLR to make this shot? No. Pretty well all of the Nikons since the D70 have had all the capability I needed. Did I need a $2000 fast lens? No, I shot at f/8, with one of the least expensive lenses I own (this is an original 24-120, I got it back with my D70 in 2006, everyone says it's the worst lens Nikon ever made. Right.). Did I need a high-tech TTL Nikon Speedlight? Well, it helped. I could have done this with a reflector instead. I did need something to bounce light in from the front. Could you do this with a point-and-shoot camera? Nah...although it could be done by a really experienced photographer with virtually anything.

As a small aside, using commander mode and a remote flash is tricky. You need to practice that in advance so you know how: when I first got there, the remote wasn't triggering. It only took me a minute to realize that although the camera was set up right, the flash sensor wasn't activated.  You need to know your equipment. RTFM.

This is about the process. Hopefully it gives you, the reader, an idea about how I created a successful picture, from planning to execution. Maybe you'll use some of these ideas in your future shoots.

— 30 —

Workflow Tips

I thought I'd share a couple of workflow tips for my photographer friends.

I don't have to do high speed editing like a sports photographer might. I generally don't import thousands of images at a time, but I might do a couple of hundred. This is just a quick overview of my process and I've highlighted a few of the things I think about. This is not a detailed explanation, it takes too long to write that and I don't want to put anyone to sleep. Just some things to think about.

1. Import
I put the memory card in the reader and I have the computer set up to automatically open Lightroom. The import dialog appears, and I've written a preset called "D800 import" that automatically sets up some parameters: where the images are saved, where the second copy (backup) is saved (on a different drive), it adds my copyright info in the metadata, and some really basic settings like sharpening, clarity, vibrance, camera profile, lens correction, etc. I can always change those but most RAW files need something to make them look ok on import. Click "Go" and depending how many images, get a coffee.

2. Flag Status
The D800 gives you huge files. I don't necessarily want to keep them all and some of them are certainly worth prioritizing for the next step in the editing cycle. I blow them up to full size on the big monitor and go through the whole batch, hitting "X" for reject, "P" for pick or just skipping them if I'm not sure. Pictures I want to work on first, I'll hit "6" (puts an "Edit Now" flag on them) and really good ones I'll mark with a 3- or 4- or very rarely a 5-star rating.  I also use "7" on series' of bracketed pictures I will want to merge to HDR.

Next step is the hard-hearted click on ctrl-backspace to permanently delete all the rejected photos from disk. These are pictures you're NEVER EVER going to want to see again.

This whole thing is really a quick process, assuming I've taken that few minutes to let Lightroom generate the  previews before I start. 200 images, maybe 10 minutes.

3. Keywording
This is a good time to do it. These are all potential keepers, and if you ever want to find them again... you can do them quickly in batches and add more later.

4. Editing
I'm not going into detail here. Each picture needs a different treatment. Sometimes you want to do some common edits, especially if a number of pictures are going to be merged together, as an HDR for instance, or all themed to work together for some other reason. Lightroom lets you sync edits across a number of images.

I do everything I can in Lightroom, then export selected images to edit in Photoshop. I use plug-ins a lot: if nothing else they save orders-of-magnitude of time instead of working blindly in Photoshop. My go-to plugins are from Topaz and Nik. I have the full suites from both of those manufacturers. Here are the links:

Topaz

Nik

Step 1 once an image is imported to Photoshop: Ctrl-J duplicates the layer so I have the original to fall back on any time I want to. Step 2 is usually a hi-pass sharpening layer (blend mode to overlay, radius 4px is my go-to setting). The rest varies from image to image. Step 3: Blow the image up onscreen to at least 100% and go around it to find whatever else needs fixing. There's always something!

Once I get back to Lightroom, that's when I crop and straighten (I may have done a preliminary crop before exporting to Photoshop). Final touchup, and as Ansel Adams used to say, a picture isn't finished until the edges are darkened a bit. A really useful tool is the new radial filter in Lightroom that you can use to really make the subject pop.

For display on my blog or on Social Media, I often like to build a shadow frame like the one below. I created an action to do that in Photoshop with one click, then I enter the caption, my signature and tweak it as I see fit before saving it as a separate file. I don't want to overwrite my original. Contact me if you want a copy of the action (works in CC or CS6)

5. Exporting

You do understand that everything I've done up to now is creating the source file, the "negative" as you will, that I archive. What I do with them next depends: are they for use on social media? My blog? For sale as a fine art print? Wallpaper? For the media (newspaper)? You need to export Jpegs for that. So that's what I do next, but only when I want to use them for something. There's no need to save the exported files back into Lightroom: I can always make them again from the source files. 

I set up a bunch of export presets that say where the file should be stored, what it's called, how big it is, whether there's a watermark, etc. Once you've made the preset, it's 'select the preset' then one click to export as many files as you want!

6. Backup

I have Lightroom set to make a backup of the catalog whenever I exit the program. The backup goes on a different drive from the original (you know why. If you need me to tell you why, sell your computer and buy a box of pencils). Do it religiously. 

In my workflow, I sync my external drive where the main images are, to one or two other external drives. The keepers also go on another one which is stored off-site. I do that every month, on the first. Only then do I delete the original backup images I made when I imported them. And I keep a month's worth anyway. You need a backup plan...

Is this the world's greatest workflow? No, of course not. Should you do it exactly like this? No. Did this give you some idea about how you might treat your own work? I hope so!

Be sure to visit my regular weekly blog, interesting images and ramblings from the Highlands! Click the "Newsletter" button (or here) for a weekly heads-up about what's on the newest posting.

— 30 —

Shooting Stars

The Lyrids meteor shower got me interested in shooting stars again. I wanted to catch a meteor! I've been through somewhat of a learning curve, although I've shot stars before. I thought I would save some of you some hassle, and give you some ideas at the same time. These are some random thoughts on the subject, some quick answers and some pitfalls to avoid.

Click on any image to blow it up. I made them all about 1200 pixels wide.

We'll talk about how I captured and made these three images, totally different techniques.

Nikon D600, Nikkor 17-35mm lens at 17mm, 30 seconds at f/2.8, ISO 2500

Nikon D800, Nikkor 17-35mm lens at 17mm, 30 seconds at f/2.8, ISO 800.

Same exposure: Nikon D800, Nikkor 17-35mm lens at 17mm, 30 seconds at f/2.8, ISO 800. But I merged over 50 frames in StarStaX.

First of all, what equipment do you need? A camera with an appropriate lens, a tripod, and optionally but really important (especially for the 3rd shot), a remote cable release that locks. At the risk of restating the obvious, here's what you need to do:

• Find a spot where you have a relatively unobstructed view of the sky and where it's dark
• Dark means, well, DARK. In all three of those pictures, less so in the first one, there's light pollution in the sky. No, folks, that's not the vestiges of the sunset, those last two pictures were shot after midnight and the light in the sky is from the little town of Minden, Ontario, more than 10 km away.
• Put the camera on the tripod. Before you do, focus it to infinity (it's best to know in advance where infinity really is, in relation to that "∞" mark on your lens). Make sure you turn off autofocusing, VR or IS or whatever your camera maker calls it, and switch to Manual exposure. Turn OFF "Long Exposure Noise Reduction" (Lightroom, Photoshop and a myriad of plug-ins do it better). Point it at the sky.
• Attach your cable release. If you don't have one, you'll want to use the self timer in the camera to give it time to settle down before taking the shot. That won't work for the third shot above.
• Crank your ISO up to 4000 or so, shutter speed to about 10 seconds and aperture as big as your lens allows, and shoot a test shot. This isn't a keeper, it's a quick way to see if you've focused correctly, if the camera is level (the D800 has a built-in artificial horizon that really helps!) and if the composition is what you want. Move it around until you're happy.

If you want to shoot a static star field like the first image, you need to remember that, well, the earth moves. The distance that the star moves in your picture depends on how long the shutter is open and the focal length of your lens. The longer the telephoto, the more it seems to move. So a wide angle works best.

There's a little magic formula, called the "Rule of 600". Divide 600 by your focal length: example, a 50mm lens will give you "12" as the answer. But remember that if you have a cropped sensor, you need to divide that by the crop factor (around 1.5?) so that means 8 seconds. If your shutter is open longer than that, when you blow up the picture you will see star trails.

It's good to have something in the foreground. Here I painted the trees with a small flashlight while the shutter was open. I think the bright star is Venus, and the cluster above it is the Pleiades (did you know the Subaru logo is based on them?). D300, 12mm, 30 seconds at f/4, ISO 1600.

If you don't have a big aperture lens, here's where you're going to get in trouble: you're going to have to crank that ISO up really high to get a good exposure! Here's a starting point:

Aperture = f/2.8, Shutter Speed = 30 seconds, ISO = 1600. I'm sure you can do the arithmetic for your combination. See why that f/5.6, 55-105mm lens is going to get you in trouble?

I underexposed the second shot because I didn't want as many stars. Also there was more ambient light in that direction.

If you want to catch a meteor, you need infinite patience (or incredible luck!).

Here's what you need that locking cable release for. Whatever shutter speed you select, click the cable release and lock it. It will keep shooting pictures, one after another, until you unlock it. Go inside where it's warm, pour yourself a coffee (or a scotch), relax and come back later.

I set up the camera behind the house for about 40 minutes, doing a sequence of 30 second exposures which I later blended together in Photoshop (the technique is to stack multiple layers with "Lighten" blend mode. I did it in batches of 10 until I got all 74 images done). No meteors. But I did see 3 things in the images:

This image is 2 frames merged together. According to what I read, this matches the profile of a satellite. The middle is brighter because the sun reflected off a brighter part of the rotating satellite at that point. I was pointing roughly NorthEast which is consistent with the sun setting behind me. The image was shot at 9:53 pm, about 2 hours after sunset.

There are 4 merged frames in this image. The big streak is clearly an airplane, you can see the regular red flashing light and it's a dual trace. There's another really fine trace that I think was possibly another satellite. 

Here's what happens when you merge those images together.

This is the full merge of all 74 images. Pretty, but not what I was hoping for. Too many stars. You can't even see the second satellite. 

Trying to solve the "too many stars" issue. I first reduced the exposure of the individual frames by one full stop, then dragged down the highlights and blacks even more. I exported the images as jpegs (6000px wide) and used StarStaX* to combine them. Still not enough, I opened it in Photoshop and used Topaz Simplify.  This is a "work in process". As soon as I find the right formula, I'll share it.

* StarStax is a freestanding program that works on all platforms. It's job is to automate stacking those images to make star trail shots. You could do a lot of it in Photoshop manually (as I did at the beginning) but StarStaX has some utilities to smooth the transition by filling in gaps (haven't figured that out yet!). It's FREEWARE and you can get it at http://www.markus-enzweiler.de/software/software.html. Besides, after you export the images to Jpeg, it's a couple of keystrokes to do the comp, all at once, unlike the 45 minutes I took to do it in Photoshop earlier.

Here's a shot that failed.

Same settings, but only 30 exposures. I got tired, and wanted to do the car shot at the top, so I interrupted the sequence. 15 minutes is not enough with a wide angle lens. PS: when I go back to redo this, I'm going vertical. Imagine the reflections of the stars in the water!

The other way of doing star trails is one long exposure. But noise is generated by your sensor when it's under power (especially when it's hot), so the quality seems to be better by merging shorter exposures. That said, you can control the exposure more traditionally.

D300, 12mm lens at f/5.6, ISO 100, 6,130 seconds (102 minutes).

D800, 1850 seconds (30 minutes) at f/2.8, 17mm, ISO 100 

The second shot at the top? OK, I'll admit it. The meteor was created in Photoshop.

Use a black layer and a small soft brush loaded with white. Click where the tail should be, hold "Shift" and click where the head should be (keep it horizontal, you can rotate it later). This works best with a Wacom tablet, make it pressure sensitive. Now double the size of the brush, turn the opacity down to half, and paint in the head. Add a touch of Gaussian Blur or it'll just look like a Q-tip! Practice makes perfect!

Now the third shot: the car. I did all the things I said above; several quick test shots at high ISO to evaluate the focus and composition, one of the test shots showed an ugly shadow on the ground (the light on the car was from a mercury vapour street light about 400 m away, the only light around! But the camera/tripod cast an ugly shadow), so I had to recompose. I got into the car after setting things in motion, in order to trigger the interior lights for a frame or 2. Too bright. So I edited those out in the computer and turned on my iPhone while I was in the car.

Here's the picture again. It'll blow up if you click on it. 

To prepare the images for StarStaX, I opened one in Lightroom, took the exposure down a full stop (too many stars), added highlights and darkened the blacks, added contrast and clarity. then synced all of the changes to the other images. I edited about 12 of the 54 images in Photoshop to get rid of two airplanes and three passing cars (the airplanes were easy: healing brush. The cars were tougher, big soft layer masks, 'darken' blend mode). Then I exported the 54 jpegs to a folder and dragged them into StarStaX.

I'm new to StarStaX so I haven't figured the "gap fill mode" properly yet. But when I omitted the bad frames (cars) it left gaps in the star trails, so I fixed them and put them back in. I ticked "comet mode" which is really cool. Once done: too many stars! So I added a layer of Topaz Simplify BuzSim at about 50% opacity to control how many stars in the image.

---

So the above was a rambling mess. I threw all kinds of random stuff at you. But if you go back and read it again, I'll bet it'll save you some time in your star-shooting learning curve!

If you want to do a workshop on this, I'm happy to oblige. We'll spend some time going over the theory, get you set up, then go out and shoot. Then we can come back (next day, not the same night!) and do some computer magic. Very reasonable price. Contact me!

— 30 —