The new iPhone: it’s a smartphone with a camera, Jim, but not as we know it…
My first analysis of the iPhone 8 experience has to be about the camera. I tried the terminal out briefly during a meal with some friends over the weekend and without a doubt, the camera completely stole my attention.
I took a few pictures of my friends at the table — discretion prevents me from sharing them, so instead I have included some demos from Apple’s website, but the effect was exactly the same, and can also be seen clearly in this article: a totally different experience to any other camera, smartphone or otherwise. This is the first time I really get a sense of the power of a series of algorithms trying to understand what I want to do, identifying the photograph better, measuring and evaluating the conditions, and allowing me to achieve the promised effect.
But is it photography, you might ask? It’s certainly another way of approaching it. With any other camera, obtaining a good bokeh, a properly vignetted background or effective lighting of a face are not techniques within everybody’s reach, requiring careful measurement and a reasonable understanding of concepts such as depth of field. Applying our experience with a conventional camera to a smartphone is complicated needs some adaptation, and doesn’t always produce the best results. Without a doubt, smartphone cameras have improved a lot, but so far they have largely been about capturing the moment rather than creating something in its own right. In a restaurant, when I have photographed the label of a bottle of wine for later use, I could blur the background, or use a flash at close range to vignette it roughly to avoid reflections that would spoil the image, but until now, that was about as good as it got.
Now, anyone, absolutely anyone, can take a picture and, simply by playing with a menu of five options, switch between natural, studio, outline, stage or black and white. The effects are dramatic, impressive, even somewhat exaggerated, but without a doubt, they serve their function: pretty soon we’ll be tired of these types of effects. The final image also stores the resulting parameters so that they can be retouched, which is pretty much the same as using filters.
But what is arguably most interesting is the way these effects are achieved: we’re no longer talking about parameters such as brightness, contrast, saturation, sharpness, etc., or blur, tilt-shift, vignette…but instead, the use of a series of algorithms that interpret the image and manage the parameters to obtain a certain result. The camera goes from taking an image, to “understanding” that image, to “knowing” what a face is, what parts are in front or behind in the image, what areas the user might be interested in highlighting, blurring or hiding. After just a few minutes playing around with the camera I have to admit I was captivated. It’s as if I was using computer instead of a camera — which in fact I was, because that’s what a smartphone is these days — with a sophisticated editing program that also makes decisions for me to achieve the right results.
Obviously, the iPhone 8 camera is not perfect. The vignettes are sometimes a bit extreme, with loss of detail, such as a nose in a profile or any shape that stands out a little from what the algorithm interprets as a face, and the effects are applied in a somewhat more erratic way in group photos, and there will undoubtedly be situations in which the application just gets it plain wrong, producing strange results. But the impression, even after a few minutes use, is impressive: an algorithm is able to evaluate an image and decide how to apply some of the most popular effects means that millions of people will use it because it produces incredibly professional results. In some ways this is not dissimilar to what Instagram achieved with its filters, or at another level, Camera + … after properly filtering a photo and feeling like an ace photographer, then sharing on the social networks and basking in the dopamine produced by each Like. This takes the whole thing to a new league, thanks to the application of machine learning to the process.
A well-designed, eye-catching development that, if you have had any contact with the development of machine learning algorithms, will open your eyes. Coming soon to a screen near you.
(En español, aquí)
