Photo editors are angry. They are angry and frustrated. I have seen and talked to a lot of photo editors these past weeks and I hear the same thing over and over again.
The digital evolution has not made their lives easier. They are tired of these little shop websites that are quirky and impossible to navigate. They are frustrated of not being able to find the images they need because they encounter resistant technology created by amateurs. They are upset by the level of incompetent platforms they encounter.
Everyone is a web expert these days and everyone has a take on what a website should do. There are more search experts out there than there are results for “photo” in Google. The promise of an easy, simple and user friendly experience has turned into a nightmarish crash and freeze experience. At best, the results of a search returns very poor images. If it is not the wrong images that pop up, it is of a poor quality photograph, at best extremely well key worded. And they do not want to see that. Neither do they want to spend hours trying to figure out how to download the image they need.
Promises. Promises of a fast search, of a breezy summer day in digital heaven, of such a convenient way to reach photo editing nirvana, all have been broken. Broken by photo agencies managers that, in a sure way to save money, have given the most important task of their company to a college kid, or their nephew. Or some French/Russian/Asian/British guy that speaks very quickly and sounds very intelligent because you cannot understand his accent. Worse even, some, mostly photographers themselves, have taken up the challenge and said, “I can do that !!”. And they proceed in creating some of the worst looking, less effective tools the industry has ever seen.
This is the same industry that screams when amateur photographers dare to walk on their turf, yet they have no problem, themselves, as amateur programmers, to design their own websites. And it makes image buyers extremely angry and frustrated.
No photographers believes they can fly a jet plane after a few lessons yet they have no problem convincing themselves that they can build a state of the art digital licensing platform. No photo agency manager believes week end photographers should be allowed to license their images yet they have no issue competing with professional web designer and software engineers. There is a far cry between knowing how to license images and how to build a website. And that cry, I hear it everyday, comes from the users : the photo editors.
Drop the code, drop that book on HTML and advanced search. Drop even the idea that your website will be better than your neighbors. Because, even in the remote case you will succeed, you will still need the appropriate content to make it interesting. Isn’t that why you are in this business to start with ? Licensing compelling content ? Hire professionals, real professionals and gracefully admit you cannot do or control everything. It is not because you figured out how to make your brand new Canon Mark III to work that you know anything about digital technology. Spoiler alert: Canon made it easy for you to understand how to make their camera work.
Go back to what your really know: creating and licensing the best images that the world has ever seen. Before your clients loose faith in you.
NO SENSE: I have read a lot about some new web 2.0 visual search aggregating website called xcavator. If so called journalists had done a little research, they would have discovered 2 things: it is nothing more than an open source application called Imgseek available for free here that anyone can install on their database wrapped into a web2.0 candy bar wrapping. Furthermore, if you play around with it, you can see that it only barely does a thumbnail search and sends you back to the original website should you need to download a hi res. And in some case, not even to the image itself but to the search page where you are forced to redo your initial search. Not quite sure how this is useful for anyone.
If you want to beat the system Don”t follow it.
you are all stuck on the same page.
I thing the best thing for everyone would be a good electrical stick .
And then You just open the door and move one foot in front of each other and open wide your eyes and you will see lots and lots of images that no one can see just you.
Harry Maconeil the system breaker
Dear Sir,
Responding to your comments about xcavator:
First of all, it is not Imgseek. xcavator is powerful, proprietary visual search technology developed by CogniSign (www.cognisign.com). Second, we do not want or need hi-res images to perform this visual search, which is fantastic for a search portal. The site’s intent is to allow a user to browse quickly for similar images, using both tags and visual search, and thumbnails allow us to do everything faster. The user quickly finds the best photo to meet his needs, and double clicks on the thumbnail image to go to the stock photo site to view and buy a high res version. Third, as you find photos you like, you can save them to Faves and Email them to yourself or others for later reference. If you’d like to better understand the design intent of the site, please check out the Video Intro link.
A stockfotofinder with a visual search. Still, anyone can have the same result with a Imgseek, an free open source app. another of those web 2.0 extravaganza…
Paul,
I respectfully disagree with your comparison between xcavator and imgSeek.
imgSeek compares a selected image with EVERY image in the data set. Because of that it is very slow. For the collections measured 1 mln and more it is useless. xcavator uses indexed search and scalable architecture, which can handle billions of images.
Desining xcavator we had an objective of combining visual search with seemless browsing experience. We believe that our solution with tag cloud, color picker (which we call The Orb), dragging and dropping images in the Image Search field creates an environment where you can look for the right image without struggling with the interface. And, indeed, the interface was designed in web 2.0 style, but it was rather a mean than the objective.
The search functionality in xcavator is quite complicated, and most of users completely miss its most powerful features. We explain these features in our Video Intro, but very few, as it turns out, make an effort to watch it.
Let me explain how search in xcavator.net works.
It consists of 4 components:
tag-based search. It is a typical feature in most of image search engines, which is a kind of Google-like search but in a smaller scale. Rarely, some search engines add semantic processing to the tag search; they perform slightly better than the rest. Presently, we do not have this feature.
color-based search. This kind of search is also pretty common, since it is easy to implement. The issue here is how you define distance between colors and how you define the dominant color for a picture. We believe that we did a very
good job in this respect and our solution better. Check our color search: the pictures selected by color look very natural. Also, we believe that our color picker (The Orb)
is a nice feature from the design standpoint.
image-based search. You may download or drag image into search image box, and it automatically finds similar images. This search is similar to imgSeek search. In the respect of its quality it may be slightly less accurate than imgSeek, but it can work with huge data sets whereas imgSeek is limited to small data sets.
It should be mentioned that previous 3 search components are tightly integrated and color or image-based searches run as fast on the whole data set as on any subset defined by tags. This is a very tricky task. If it is not done right, some searches may take very long time for large collections of images.
interactive image-based search. If you want to compare xcavator with imgSeek, than you should compare imgSeek with our interactive visual search component, which beats imgSeek in every aspect. First, you may decide what is important and what is not for your search. You may pick points or draw lines in the chosen image, and the search will be performed based only on the selected areas. For instance, you may look for a house on a grass field, and ignore whether the sky blue or gray. Second, the matches found are tolerant to the location, scale, rotation of an object of interest. That is, a big image of a house in the previous example (suppose, that the house is located in the right part of the chosen photograph) will be matched with a smaller image of similar house in the left side of a photograph. imgSeek does not have such a capacity: its algorithms require the house
to be of the same size and located in exactly the same part of the photograph.
The described interactive image search component is very powerful, but it is slow: roughly speaking, as slow as imgSeek. There is no way to match a chosen image with the whole image set. Instead of making my post even longer, I refer you to our Video Intro which describes in detail how to restrict the image set and make our interactive search highly efficient.
Best,
Lenny Kontsevich @xcavator.net