I’m a bit confused because AutoMetrics generates different results every time I run it, depending on the side bearings values before running the tool.
When I start with fixed side bearings of 100, the outcome is different than when I start with all side bearings at 50.
I would have thought that, if the glyph shapes are used for calculating the side bearings, the results should always be the same, because the glyphs don’t change, no matter what side bearings I start with.
Now when I look at my “M” for instance … sometimes after running AutoMetrics the LSB is bigger than the RSB, sometimes it’s the other way around.
I don’t quite know how to decide which result is “right”.
It should provide equal results. We are about the release an update which contains improvements to Optical Metrics. I’ll send you a link to it so you can give it a try.
Aweome, thanks a lot for this! I’ll try out as soon as I can! 
@Erwin: I just realized that the link you sent me was for FontCreator 11.0, but I still have 10.1. Is there gonna be an update for this version, too?
Also, I tried the update version on my laptop (which is 64 bit, so I could install it there) and ran AutoMetrics on an installed font, just to see what it does.
What’s different than in my version is that, if I set fixed side bearings for all glyphs before running the tool, the value doesn’t matter. The results will be the same (I tried with 50, 100 and 200). But if I run AutoMetrics several times in a row, the results still do vary.
I’m sorry but we never update older versions. It would take way too much time and effort.
I think the algorithm will indeed show minor differences, but shouldn’t show too much differences.
That’s what I thought.
The thing is, I don’t really know what differences would be considered minor or big. I have no experience in font making whatsoever, though I’m trying to learn as much as I can. But I’m still just a beginner.
When I purchased FontCreator I chose the Professional Edition first and foremost for the AutoMetrics and AutoKern features. Now I’m not sure how much I can really rely on them.
What irritates me is that, depending on what side bearings I start with before running AutoMetrics, sometimes the LSB end up being bigger, sometimes the RSB.
For instance my “M”:
Starting with fixed side bearings of 100 for all glyphs, AutoMetrics will make the side bearings 158/132. If I run it again it’s 143/158. That seems weird to my amateur brain. I would have thought it does matter which one is bigger and that there are some basic rules about this.
I’m totally uncertain about how to interpret this and would greatly appreciate any tips.
I can give some more feedback, but I would first like to know this:
Are these values a result from running the latest unpublished version 11?
No, these are from 10.1 because this is what I have to work with for now.
We are aware the algorithm isn’t perfect, so we improved it in the upcoming release. However it is still not perfect, but a little closer.
Most font designers tend to keep left and right side bearings equal for horizontally symmetrical glyphs. FontCreator uses a slightly different approach, so this is not always identical. Version 10 does generate really good side bearings, and the upcoming release is even better. But even then, it is always wise to manually inspect and if needed adjust the results.
Oh, I’m certainly not aiming for my font to become perfect. That’s impossible, considering my lack of experience. However, I’m a perfectionist nonetheless (however contratictory this might be, lol), that’s why I try to get it as good as even possible.
Really? I guess then I must have only looked at fonts where this is not the case. But good to know that there’s not really a “right” or “wrong” here.
Yes, of course, I’m aware of this, but the better the foundation I can start with, the easier it will be for me.
Will there be a notification on the forum when the new release is out? I do consider getting an upgrade to 11.
Hello again! I just meant to ask if the updated version will be released soon.
I decided to buy the upgrade to FC 11 and now I’m just waiting for this so I can keep on working. 
We released the final private build yesterday and had to fix one more issue. We’re now updating the user manual, so we intend to release it tomorrow morning!
That’s great! Thank you!
We’ve just released the promised update!
Awesome! Thanks a lot for letting me know!
I’ve been playing around with the AutoMetrics in the new version, but something doesn’t work right.
When I run AutoMetrics it now messes up my composites. Before running it, every accent is nicely centered (or in its place).
After running the tool they no longer are.
This can’t be undone either:
I tried running Autometrics with fixed side bearings and set them all to 100. The composites were messed up, but the original letters in the glyph edit windows looked fine.
But when I hit Undo, the line of the right side bearing will jump back to where it was before, but on the left side there are now two lines, one at 100 and one at the original value. And the composites are still messed up, even after hitting Undo. I have to close the font and re-open it to get things back to normal.
In 10.1 that didn’t happen. No matter how often I ran AutoMetrics, the accents always remained where they were.
That sounds like a bug to me. We’ll look into this.
Okay, thank you.
It seems version 10 failed to set the left and right side bearings for composite glyphs. We fixed that with version 11, which is why it behaves differently.
The undo issue is indeed a bug which will be solved with the next upcoming release.
Hm, but if “differently” means that AutoMetrics now moves the diacritic marks of my composites, then I have a problem. Either I make all composites simple again to avoid that, or I have to manually fix all diacritics afterwards. Neither option really makes sense to me.