More About Headshots & Portraits & AI

More About Headshots & Portraits & AI

by William Lulow

I have tried to indicate in many previous articles, that there is a special connection between a photographer and his or her subjects when making portraits. The “sitting”  is a kind of special time shared by both because usually, the photographer and subject have never met. The former may know of the latter. He or she may know who the subject is, but personality is a big part of the equation.

I recently came across a sponsored post on Linked In about a service that will take any old selfie you may have or several selfies, for that matter, and combine them using ARTIFICIAL INTELLIGENCE to produce a headshot that they claim would be better than any of the originals. There is no question that AI can combine any photographs, but selfie’s and any other images that weren’t produced by a professional headshot session may lack the important connection that usually results in headshots that sell the person. In addition, haphazard kinds of images with different backgrounds and different lightings may present some problems even for AI. It would, most likely, eliminate any background and create its own. It is certainly remarkable what can be done these days, but we should remember that any image designed to promote someone or something shouldn’t be one made on the spur of the moment, “off-the-cuff”, or even just made up! Since, AI by definition, is not human, it will try to find the best expression, perhaps add a background and maybe even some highlights. But it cannot produce a lighting say, that isn’t there in any of the originals it has from which to work.  And, if all of this is done artificially, it just won’t have a certain ring of “truth” to it. It may produce an acceptable photograph, but is that what you would want to sell YOU? People have been using “acceptable” images for decades but when they see what a professional picture can do to increase results, I think they will be convinced to use real photographs, to which many of my clients can testify.

As I have said often, the photographer’s technique in the studio or even at a location shoot, is what produces that special quality necessary for any successful headshot. I have met actors in the past who have paid $1,000 to have headshots done by pre-eminent photographers. The AI shots I have seen are good, but many lack that special part of an expression, always in the subject’s eyes, that tells you who they really are and what they can do.

I am not sure that any artificial intelligence can be a substitute for this. Anyone can take a picture of someone these days with a decent iPhone and some technique practice. If the intention of the shot was not specifically to make a commercially successful headshot, it will show. The result might be good, but it will still lack something. It is often hard to explain in words just what that something is. But you will be able to see it in the photographs. Look carefully at the subject’s eyes to tell if there was a connection to the photographer! A selfie is just that: a self-portrait! No matter how good you can smile, it can’t replace the human reaction to another person!

Whether it’s a lawyer a businessman or

an actor!

If the connection is there, it will show in the photographs. What AI hopes to do is to be able to combine several images in order to produce one, or even several that will be able to show the person the way they would like to be shown, but it’s hard for a photographer to imagine how that can be accomplished “artificially.”

Producing a “salable” headshot or portrait combines a great deal of photographic knowledge. Lighting, background, pose and the personalities of the subject and the photographer are all key to the success of any image. It’s tempting to offer solutions to this process that seem to be cheaper and easier, but those are not always the benchmarks to be sought in this process. The old adage of “you get what you pay for,” comes to mind when considering this type of “problem solving.”

I wonder how successful these efforts will be, and whether they will be worth it in the long run.

Portraits are a totally different story, as I have mentioned in many previous articles. A portrait is actually more a picture of the individual photographer than of the subject. Witness that many photographers have often photographed the same people, but each is different because of the individual approach of each artist. Factors influencing portraits are: location, time, wardrobe, hair&makeup, mood, lighting and personality of both photographer and subject. Add to these the way in which the final image will be used and you have a formula that any artificial resource would be hard-pressed to try to duplicate.

My 1980s image of radio and TV personality Bob Lape of WABC-TV in New York is one example. This took a combination of lighting, pose and personality to be successful. Bob wrote to me thanking me for being able to achieve an image of him that his wife actually liked!

My image of Jill Harth, well-known New York City makeup artist was used in several publications she authored.

And, my portrait of author Ira Levin, was used for his book and many other types of announcements by his publisher Random House, when the book was released.

So, asking AI to produce images like this, I think, is asking a lot.