It depends. Everybody will tell you that, but there isn’t much good information around on what it depends upon. This is my attempt to say what it might depend on. My list is not exhaustive, but it’s a few things that I have encountered and I share them in the hope that they are useful to you too.
Firstly, if you learn to draw well and sketch as a warm up discipline before you paint, then you can paint faster. That rapidity and certainty of line comes as your develop your facility. There’s no short cut, other than earnestly to pay attention to your drawing and learn to measure! You have to put in the work. You can reach the point where you don’t sketch on the canvas at all, you just set to work painting (although, technically, you sketch with paint and brushes, rather than charcoal – you just refine that sketch into a finished painting, rather than covering the sketch up with paint).
If you are doing a serious commissioned portrait, you may spend eighty to a hundred hours on a single painting like the professionals do, going back to it time and again, reworking things, critically evaluating things and correcting things. If you glaze, you also have to spend the time it takes to wait for the glazes to dry before applying the next one. The technique won’t work, otherwise.
Portrait painters like oils because you can rework the paint for days. That gives you a lot of correction time. I use an acrylic paint with a long open time and which can be reopened with the application of water for that reason, though sometimes I just let it dry. It dries much faster than oil paint anyway.
On the other hand, if you are a student, then I claim what matters more is the number of experiments and mistakes you can make in the time you have. A student artist, like myself, generally has limited time to devote to painting, so every second counts. In that scenario, I think it is more important to study the technique you want to learn from books and paintings, then to give it a go in the studio or class.
You need to fail fast with those time constraints, because spending a hundred hours on a painting that only shows you that the technique you are learning does not work for you is an expensive use of that scarce time.
As a student artist, I think it is important to try as many different brushes, brush strokes, techniques, perspectives, compositions, colour combinations, colour mixes, palette knife techniques, mediums, subjects, canvas types, canvas sizes, concepts and ideas as you possibly can and take note of what works for you and what doesn’t. Learn what you love. So, each painting might only take an hour or so, but each experiment builds your arsenal of techniques.
With all that experimentation and exploration under your belt, when you get your commission to paint a portrait for money, you’ll be able to tackle the task with skill, confidence, ability and a wide range of artistic and technique choices. Those choices are what make you a better painter and your work a better painting.
Then, by all means, perfect the work by taking a hundred hours on it, if that’s what you need, but never forget that painters like John Singer Sargent and Vincent Van Gogh painted quickly, even on their masterpieces.
I am an advocate of painting as fast as you can, while exploring your technique and learning, because getting fussy with the details won’t count for much, if the painting has some other fatal flaw. Of course, fussing with the details is a technique you ought to learn too, but I see so many painters focusing their attention and lavishing their time on the wrong things. Spend your time wisely. Once you know what you’re doing at the easel, then by all means hone and perfect a work that really matters to you. But don’t waste your opportunities to learn and learn and learn.
So, when deciding how long you should spend on a painting, evaluate honestly in your own mind what you are trying to achieve and to learn. If you can’t answer that question, you really aren’t progressing; you’re going through the old, familiar, comfortable, safe motions and you will probably make all the same errors you always make, learning nothing informative or new.
As I said: It all depends.