Myerson Photo Blog

As much as I love to create Vetta-worthy content, sometimes we have to face the ugly truth that the portfolio needs some filler. Not bad work – I hope we never upload that – just work that is more utilitarian and less creative or artistic in nature. These are the brush strokes, the isolated fire extinguishers, the background textures. There are ways to do these kinds of images well, of course, and that should be the baseline standard. But once you’ve got the pure white background, and effective lighting to create dimensionality and form, the shot is the shot.

That’s what today in the studio will be about. Simple, usable, workhorse images. Maybe even a cliche or two.

500,000 sales on iStock is a phenomenal number, and it takes a phenomenal guy like Sean Locke to get there. Congrats, buddy.

One of the most common rejection reasons on iStockphoto is the “overfiltering” rejection, largely because it encompasses so many possible causes. In this new “Rejections” series, I’ll look at some of the more common causes of the various iStockphoto rejections, and what you can do to avoid the most likely pitfalls.

When it comes to “overfiltering”, one of the things I hate to see is poor editing or cloning. This often happens when the contributor is trying to remove some other rejectable element in an image – a visible person for whom there’s no model release, a logo or other copyright-protected element, or sensor spots, perhaps – but does it in a less than seamless way. Any editing of this sort has to appear natural and realistic, and not look like the file has been edited.

Let’s use this image of a tent in the backyard to illustrate the wrong ways to go about removing material, then we can look at a good solution. As you can see, the bottom corner of the tent door has a manufacturer’s logo and a hand-painted “W1″. The logo is a no-go for iStock, so it will have to be removed. The “w1″ is not really problematic in terms of iStockphoto’s standards, but it’s also not doing me any favors in the image. It’s a distraction, and this would be a more usable image without it, so out it goes.

When it comes to doing it wrong, the three main culprits are The Blur, The Splotch, and The Clone:

The Blur

This is the easiest and least effective way to edit out undesirable elements. The contributor often selects the element (using a mask or lasso selection) and just adds the Gaussian Blur filter. There’s almost no way to make this an effective edit; selectively blurred areas stick out like sore thumbs. Not only does it leave an ugly and distracting region in the image, it also looks completely unnatural. Even more delicate selections than I did here are usually unconvincing. They will, by definition, be more blurry than their surroundings, and more blurry than the lens would have made it. There are really very few applications of this that would look natural.

The Splotch

This is what I call those cases where the contributor opted to cover over the undesirable mark. In general, that’s a fine idea, but The Splotch is a bad way to do it. This is when it’s covered over with just a painted-on single color brush stroke. In the tent example, the areas surrounding the marks we want to remove are mostly monochromatic – the maroon surrounding the “W1″ and the off-white surrounding the manufacturer’s logo. So you might think that a single color stroke could work. But click on that thumbnail above to see how ineffective that is. The big problem with The Splotch is that it makes no effort to replicate the texture of the surface.

The Clone

The Splotch fails because it doesn’t take into account the texture of actual material, so why not use the actual material to cover up the marks? Great idea, and usually when I need to fix an area, that’s my first step as well. But it must be done right. As a click on the thumbnail to the left will attest, it’s very easy to do it wrong. This is more common than you might think. I probably end up rejecting up to five or so images a day for overfiltering of this nature. It comes from a lazy application of cloning techniques. The clone stamp source is set once, and the contributor just click-click-clicks until the offending elements are covered up.

So if that’s how to do it wrong, how do we go about getting it right? The short answer is that there’s no short answer. It really depends on the specifics of the image. For this tent image (the iStock version, by the way, lives here), I did a combination of cloning and healing brush.

Step 1 was some cloning onto a new layer (everything is done on new layers, of course).  The key here is to try to match the lights and darks of the fabric wrinkles. There is not going to be a single clone source that will do it all, so I make sure to continually set new clone sources, taking some highlights from one area, some shadows from another. Don’t be afraid, also, to cover over areas that are actually just fine in the original. Sometimes it takes doing that to make the edits appear to blend in seamlessly. Be on the lookout for obvious pattern repetitions. The crop below shows the results of the cloning step. It’s not bad, but I’ve highlighted areas that are too obvious.

So we move on to step 2. On another new layer, I start using the Healing Brush tool (J). The key to using this tool is to keep a small brush and use multiple small strokes. The tool is most effective when you’re using it away from any hard edges or abrupt color changes. With those guidelines in mind, I did some selective healing to the trouble spots above, and got this:

And there you have it. A much smoother and more seamless edit. The same techniques apply whether I’m removing logos, skin blemishes, wandering background strangers, or annoying sensor spots.

I touched on this briefly in a previous post, but after a recent conversation with a contributor, I decided to expand on this idea. When you get keyword rejections or keyword removals from your iStock uploads it’s usually because of one of three reasons:

1. The term was spam.
2. The term was a “multiple incompatible”.
3. The term was a stretch.

#3 is where I want to spend some time with you today. Stretch terms are those terms that are related to but not conveyed by the subject and context of the image. They usually take the form of concept terms. A real life example to get us all on the same page:

This is one stern math teacher. The kind one used to know in high school after being a little late with the homework.  Not that that ever happened to me, of course.

The terms currently applied to this image:

• Education
• School
• Classroom
• Only Mature Men
• 40s
• Teacher
• Serious
• Mathematics
• Mathematical Symbol
• Whiteboard
• One Man Only
• One Person
• Horizontal
• Photography
• Mature Adult
• Mature Men
• Necktie
• Looking At Camera
• Displeased
• Anger
• Indoors
• Head And Shoulders
• Color Image

What’s not on the list? “Homework”, for one. This image is not about homework, it doesn’t depict homework, and a buyer looking for “Homework” shots would pass over this image in a heartbeat. “Homework” would be a stretch term, and it would be removed by an inspector without hesitation. It’s the kind of stretch term that we see people try to use often.

Enter the concept of Opportunity Rejection. An Opportunity Rejection is a rejection that should spur you on to create and upload something new. By removing the term, the inspector has essentially said to the contributor that this image of a disappointed teacher doesn’t convey the notion of “Homework”. Contributors should take that as a challenge to then create an image of a stern teacher that does convey the notion of “Homework”. By putting the term “Homework” on the original image, the contributor made it clear that he felt there was a market for “Stern” AND “Teacher” AND “Homework”. If there’s a market for it, why on earth wouldn’t he create an image that is undeniably and unequivocally about those things?

The benefit of turning an opportunity rejection into a new separate upload is that you can create a file that is geared specifically to the terms you want to use. You can achieve that Laser Focus I mentioned last week. An image geared specifically to the concept will likely sell better for that collection of terms than will a generalized image. The image above is mostly a generalized image. The teacher is stern, but nothing in the image indicates why that is so, so it’s left open for a number of interpretations, making it a good general option. The flip side of that is that I can’t keyword for every possible interpretation – I can’t drill down to the specifics of “Homework” and “Excuses” and “Poor Performance” and “Class Clown” and every other concept that might disappoint a teacher (again, not that I’d have any first-hand experience with the things that annoy teachers).

I say “mostly generalized”, however, because I did make some specific choices as well. For one thing, he’s a white male adult. So those specifics preclude other specifics. But that kind of thing happens almost any time you use a model. The other specific choice was to make him a math teacher. The formulae on the board behind him put him pretty squarely into the math realm (maybe physics, but I chose math – again, it’s not both). If a buyer needed an image of a disappointed teacher of chemistry, this file would either not work for their purposes, or they’d generalize their search (”Teacher” instead of “Teacher” AND “Chemistry”) and edit the file for their own needs.

Sorry, this is not a post about achieving photographic focus. You and your camera will have to figure that one out without me. This is about achieving content focus in your iStock images.

Content focus is, in short, the subject of your image; the content or concept that the image is meant to convey. One of the most important factors in a stock image is how quickly, clearly and concisely the image conveys a concept. If the image is about a college graduate entering the workplace for the first time, for instance, you’re going to want the image to convey that concept quickly and unambiguously. Because if your image does not, someone else’s will and that’s where the customer will spend his credits.

Because of the way the latest iteration of iStock’s “Best Match” algorthm works (the so-called “BM 2.0″), images will move up the rankings the more clearly they are associated with a given term. Images that are muddled in their presentation or that do not clearly convey a concept will find themselves languishing near the bottom. If you want to appear near the front of a Best Match sort, be sure your file is unambiguous in its meaning.

Lets look at an example. This image is a pretty good match for “Paperwork”, showing up in about the 15th place in a Best Match sort for that keyword as of this writing. It’s also a fine match for “Drowning”, currently occupying the #1 slot there. It also appears in a search for “Emergence”, but while “Emergence” is an appropriate term, this image isn’t really the best image in our database to convey that concept. For exactly that reason, it doesn’t even show in the first 500 results for that term.

The million dollar question, of course, is how to get your own images to move up the rankings? There are a few techniques you can consider:

• Trim the fat: One of the best things you can do is remove keywords that are unlikely to be of any help at all. If I had “Life Vest” on that drowning image above, that would be a prime candidate for removal (and let’s be frank, a term like that should be removed before you even upload). There is no “Life Vest” in the image, and no one searching for “Life Vest” is likely to buy an image that doesn’t contain or even relate to the term.But what about “Photography”, “One Person”, “Vertical”, and those other meta descriptors? No one is likely to search for “Vertical” to buy this image, right, so should that be trimmed? Nope, those kinds of meta terms are fine to keep on because they are usually used in a boolean mode – “Drowning NOT Illustration” or “Paperwork AND Vertical”. Those terms are not subject and concept terms, so they don’t muddy the subject and concept focus of the image.
• Consider Underserved Niches: It’s a whole lot easier to move up the results rankings when there are only a hundred or so images between your image and the top. Lets say you’ve hired a model, and you have a prop closet and wardrobe full of options. You can dress your model in a suit and tie and do a “Businessman” series. Upload that as your concept and you’ll be up against some 77,000 files. Might take a while to climb that ladder. Or you can put him in jeans and a flannel shirt, plunk a guitar in his hand and sit him down by a campfire to make him a “Camp Counselor”. Go search on iStock and see how many files you’re up against in that case. I’ll wait here.Back so soon? Now, it’s clear that “Businessman” is a more common search term than “Camp Counselor”. I’ll grant you that. Is it 20,000 times more common? Let’s look at this from the view point of expected value. This is an Econ 101 concept that seeks to describe the expected potential for a given action. For easy math, let’s first say that the average royalty per sale is $1. Now let us imagine that the search term “Businessman” leads to 100,000 sales per day. That comes to $1.29 per “Businessman” image per day (assuming that sales, on average, will spread out to everyone in the search results. Which they will not, of course. If your image is on page 50+ for that search term, you’ll be very lucky to have “Businessman” lead to any sales at all).

  Now lets say that “Camp Counselor” gets 1% of the search traffic that “Businessman” gets. That would be 100 sales per day for that term*. If there were just 10 matching files, they’d each have an expected value of $10/day. Much sweeter return than the $1.29 “Businessman” average. Of course, this example uses some assumptions. You’d have to make your own assumptions about the numbers in order to determine your own expected value.

• Generic is A Focus: I was recently chatting with another inspector about the notion of a generic file and specific keywords. The file depicted a series of generic spice rack spice jars. The keywords included every kind of spice imaginable. Now, we had to remove all those spice keywords, because the image didn’t actually depict them. The focus of the image is “Spices” in the generic. There are plenty of times when a buyer needs an illustration depicting a parent term in the generic, and not the specific instantiations of that parent. A wise contributor, however, would see the opportunity in this. Illustrate the generic parent term, and keyword it just for the generic, then also create the specific instances of the child terms (the parsleys, sages, rosemaries, and thymeses). The contributor would have to be careful not to make the specific instance illustrations too similar to each other, to avoid the “Duplicate/Serial” rejection, but I think a talented illustrator should be able to do that without much problem.
• Keep The Content Clutter Free As Well: Clutter-free keywording is important, but let’s not forget the content of the image must also immediately convey the concept. Take care to include just those image elements that lead the viewer to that concept, and eliminate everything else that doesn’t support the concept. This isn’t to say that all images must be isolated on white. That would be horrible. Keep in elements that set the stage and make the photo convincing, but if that coffee mug on the table isn’t helping to set the stage or sell the concept, lose it.

If you find that you’d like to improve your Best Match standings, try implementing some of the above. Feel free to drop a line in the comments to let me know what you find.

——————-

* Cry Foul! How can I claim the term Camp Counselor gets 100 sales a day when the few files keyworded with “Camp Counselor” together have a total download count of less than 100? One of the assumptions in the calculations above is that the demand for “Camp Counselor” is 100/day. The existing files in that search may not be suitable for the customers looking for that term, so they may not be buying any Camp Counselor images at all.

Bruce Livingstone, founder of iStockphoto.com, has just announced the publication of a new coffee table book. Published in conjunction with Blurb, the iStockphoto coffee table book features some truly fantastic work from a variety of iStock photographers. It looks like a really slick production. The slickest part, however, is that 100% of the proceeds are going to charity. iStock is donating the proceeds to charity: water, a most worthy organization devoted to bringing clean, safe drinking water to people in developing nations.

• Check out the announcement here.
• Check out charity: water here.
• Check out a preview of the book here.

Here in the Myerson Photo family, we had a fun little 15 minutes of fame recently. My son and occasional model Jonah appeared on a front page ad in the Albany (NY) Jewish World newspaper. It was a half-page, below the fold ad from the newspaper publishers, wishing the subscribers a happy Hanukkah. The photo they used was this one, from my stock photography library:

My parents got their copy of the paper, totally unaware that Jonah was going to be on the front page. Proud grandparents that they are, they emailed the editors and thanked them for using the picture, explaining that the youngster was not just a cute model, but also their adorable grandson. One of the editors got back in touch with them and conducted a short interview for an upcoming story on the local connection to the photo. He and I also spoke briefly. He asked me about that photo, about my stock photography in general, and he asked if I could get a shot of Jonah looking at the newspaper. I sent him this and a few other photos:

It’s been my experience that iStock contributors go through a certain cycle in terms of their upload patterns. I know that this cycle is certainly how my development has progressed, and maybe your experience has been the same. I call it the Quanity-Quality-Quantity cycle.

At first, a lot of contributors kick off their iStock careers with quantity uploads. Maybe it’s because they have a huge catalog of waiting images, or maybe it springs from their enthusiasm for this new thing they’ve found. Either way, it happens that newcomers will often upload a ton of shots in their early days. This often leads to a spate of rejections. Photographers quickly learn that photography is not the same as stock photography, and more importantly, stock photography is not the same as iStock photography.

These initial rejections tend to lead to a learning period. In this next phase, the contributor discovers that his backlog of waiting images may not make great iStock submissions. Or he learns how best to use his camera, lighting equipment, and editing software to avoid the common pitfalls of compression artifacts, exposure problems, and lighting design problems. This period of learning leads to quality uploads (fewer rejections), and often that quality comes with a lower quantity of uploads per week. Higher acceptance rate, no doubt, but fewer images processed.

Once the contributor feels confident in the quality of his work, he often starts to ramp up his submission numbers. Depending on the salability of the images he’s submitted so far, this second quantity phase may actually coincide with a canister change, meaning a higher upload allowance. Contributors at this phase may feel like the “real money” will come in when their portfolio reaches some critical mass level – 100 uploads, 250 uploads, 500 uploads, what have you.

The cycle continues – uploading in quantity usually happens before the contributor makes a new quality leap. When the new quality leap happens, the quantity often trails off slightly (or greatly), as the contributor learns his way around new techniques, new equipment, or his own changing aesthetics. At the higher levels, there becomes very little distinction between quality and quantity, and the contributor reaches an equilibrium point where he can simultaneously try new things, keep his acceptance rate up, and keep his volume of submissions high.

One of the smartest things a new contributor can do, in my mind, is break out of the cycle. Start slow and build the portfolio smartly rather than quickly. At the time of this writing, I have fewer than 500 uploads in my portfolio. I’m not a major player, of course, and don’t have the download volume of some of the top dogs; however I’ve seen many portfolios much larger than mine with far fewer downloads. I think it’s because those contributors spent too long in Quantity mode, and not long enough in Quality mode.

Like any community, the iStock forums have evolved their own peculiar vocabulary of acronyms and terms. Here are some of the more common ones for newcomers and oldtimers alike.

AOTW
Artist of the Week. A weekly honor bestowed upon a contributor who has stood apart from the crowd for some reason. See xOTW.

BM
Best Match. Best Match is one of the search results sort algorithms,  the one that seems to cause the most forum angst. It’s powered by a super secret recipe, the details of which are often the subject of speculation in the forums. Thanks, Karen, for reminding me to include this in the list!

BME (also BDE)
Best Month Ever (Best Day Ever). Commonly seen in the “How were your monthly stats?” threads that tend to pop up as soon as a month ends somewhere in the world. See WME. Edit: Holy cow, I just found out that pal Sean is responsible for coining this acronym! H*ckin’ cool!

CN
Creative Network. Your list of iStock friends, contacts, lovers, and idols.

Crickets
The crippling near-silence that follows a failed attempt at humor, trolling, sympathy, or other forum reaction.

CV
Controlled Vocabulary. This is the lexicon of official tags (keywords) in use on iStock images.

DOTW
Design of the Week. The weekly showcase position in the Designer’s Spotlight. See xOTW.

DL
Download

Ebb and Flow
Sometimes “Uncle Ebb and Aunt Flo” – Just as the tides change, sometimes high, sometimes low, people have found sales in iStock come and go in waves. Unlike the tides, however, there’s no real way of knowing when the current crest or trough will end. Enjoy the ride.

EL
Extended License

ETA
Edited To Add. Fixed a forum post? Add your little “ETA: ” message at the end to let others know what corrections you made.

F5
The F5 key on Windows machines is the page refresh key. iStockers tend to wear the F5 key down to a dull shiny nub when new announcements are on the horizon, when they’re on the verge of a milestone download, or just any old time, really.

FIOTW
Free Image of the Week. Each Sunday, a new image is offered to the community as a free download. To have an image added to the pool of images considered for the FIOTW honors, please send a sitemail to Sirimo. (Vector submissions should go to member Bortonia, and Video submissions to member VCR). Free downloads do not add to your download count for your next canister, but it does earn you some cool visibility. See xOTW.

H*ck
Once, long ago, there was a splash page on iStock that depicted a character uttering the just-past-censors word “F*ck”. Folks felt that, asterisk or not, this may have been too racy, so it was amended to read H*ck. The splash page eventually vanished (was it a 404 page? a contest? I don’t remember…), but the near-expletive lives on.

HHUR
Heart Heart Unicorn Rainbow. Peebert-only.

IOTW
Image of the Week. Each week, the site admins select one of the newly uploaded images as Best-In-Class. On Sunday, the champion image takes it showcase position on the front page. Along with this honor comes a cool little icon, and a week’s worth of front-page visibility.

Lypse
iStockers tend to congregate, descending on a location like a swarm of photographic locusts for the purposes of camaraderie, cameramanship, and consumption of comestibles and … uhhh… booze. Get togethers planned and executed through iStock HQ are known as iStockalypses. Those put on by the community members are known as mini’lypses.

OP
Original Poster, or Original Post (Thanks Mikael!)

Pimp
Generally frowned on in the forums, this is the act of putting one of your own images into a thread discussion (other than into a thread devoted to critiquing, that is). There is a weekly sanctioned thread devoted to themed pimping, however. See PTOTW.

PTOTW
Pimping Thread of the Week. A weekly forum thread (found in the iStockphoto Discussion area) devoted to letting members pimp their favorite images around a given theme.

Totally Touching Floor
As good as it gets.

UL
Upload

Wiki (v)
To wiki a file, in iStock parlance, is to click the “Report Inaccurate Keywords” link to notify the admins of a poorly keyworded image. Wikiing a file puts it into “Keywords Under Review” status, which prevents further wikiing, but does not impact search results.

WME
Worst Month Ever

xOTW
Something of the Week. See IOTW, DOTW, FIOTW, AOTW, PTOTW

I just added a new series to my iStockphoto portfolio: Little Boy Blue.

It’s true what they say about working with children. You’d better be ready to go when they are, because you only get a few minutes before they are ready for something else. Ordinarily in a model shoot – especially with untrained models – I find I have to keep shooting and shooting in order to get the variety of looks and expressions I’m after. This little guy gave me these five expressions over a span of 8 shots. We were done in under 3 minutes, and about 60 seconds of that was my own monkeying with lights. I don’t even get that kind of signal-noise ratio from static product subjects.

Then again, nobody promises to take static products out for ice cream afterwards…