When to fix errors in an accepted manuscript?

In our newest "how can we help you?" thread, D writes:

I recently got a paper accepted for publication, and the editor sent my manuscript to the typesetting office. Since then, I noticed a few typos and minor mistakes that I had somehow missed until now (and so did my reviewers). Since I still have not received the proofs, do you think it is all right to contact the editor and let them know I would like to fix these mistakes? Or is it already too late for that?

A couple of readers submitted responses. An Editor writes:

The time to fix this is in the proofs. When you get your proofs read them with care. There is a chance that some of the mistakes will be caught in the proof stage, by others involved. But if they are not you will get a chance. Further, as an editor of a journal, I read the proofs as well, and catch yet-uncaught errors. But wait for the proofs.

Mike Titelbaum added:

Presumably once the piece is typeset they will send you proofs. Are the typos minor enough that you could fix them in the proof-correction stage?

It seems to me that Mike's question here is potentially important. One should ordinarily not make significant substantive changes to a final manuscript without running them by the Editor, right? But this doesn't sound like what the OP is describing. If it's just a matter of a few typos here and there, that's what proofs are for fixing!

Comments

[D]

OP here. Many thanks to Marcus for posting my question and to everyone who answered.

The changes I intend to make are fixing a few minor typos, and also, I wish to slightly alter two sentences (without changing their meaning) in order to improve clarity.

[Tim]

Wait until the proofs. It sounds like OP wants to make small changes that the editor doesn't need to be bothered with. So wait until the proofs to make them. Besides, sometimes the typesetting process introduces new errors. So you'll have to check the proofs anyway.