Time management

Well I am not an amazing player but I hear the advice that many of them give, much of which has been said already.

Don't think long and hard about obvious moves which you already calculated or where there is no real alternative, these should be played fairly quickly.

Learn to recognise the critical position where you have to decide the line you wil go down or commit to a plan of action. You should have put time aside for this point (normally after development when you are formulating a plan)

Think about the general position while your opponent is thinking.

You should know roughly the amount of time you have per move and budget it.

When selecting moves use a process of elimination and when you have eliminated a move as having an unsure resulting position or where you cannot see a definate mate there etc don't dwell so much on recalculating it many times.

Thank you for all these advices. Some of these things I already do and some I know I should do but knowing it is not enough.

Concerning blitz, I play it to get used to play faster in time trouble, and also to get familiar with some patterns, but it doesn't help to play faster before time trouble if I want to maintain a certain quality of play. I should not only try to play faster, but also play better when I play fast.

Also when I'm on increments I still forget about time. That's how I managed to lose on time once. And another time, recently, I was playing an endgame on increments, I made a move, and when I reached to hit the clock, I saw I only had 2 seconds left. I was lucky then, but even though I had been in time trouble for a while, I forgot about it anyway. It also happen to me all the time in blitz. Quite often I don't even notice the warning sound and I'm surprised to discover that I have lost.

The idea to work on structuring my thought process and check my time after I examine each candidate move should be very useful if I can make it an automatism. In general my thought process is way too anarchic, but I find that my best ideas usually comes from going back and forth between variations until I find something that I hadn't noticed at first. I think Blindfool's idea to think about variations as a whole could be very useful here, but I need to spend some time to really understand how I could implement that in my game.
I will also try to set time targets.

Also jposthuma says I should think on my opponent’s time. I already do that, but I have a lot of problems when my opponent plays instantly. The only two games that I lost in the french team championship this year (not sure how I should name that competition in english) were against such opponents. In both cases I outplayed them, played pretty well, and got a winning advantage, and In both cases I reached time pressure even earlier than usual, made a couple of very stupid blunders and lost.
On the other hand, my best games are the ones where my opponent is also slow or the ones I manage to win quickly.