If you’re in edit mode, select a vertex, and then press the N key, you see the Transform Properties “dialog” (is it called a dialog in Blender?). On that dialog, you see the x, y, z coordinates of the vertex, but only to 3 decimal places. For example, x = 0.022. Two questions:
Is there some way to see more decimal places?
Even if they don’t show, are more than 3 places being taken and used internally if I actually enter more (even though they don’t show)? The display shows the value rounded to 3 places, but internally is it still the more I entered?
It seems that in some manner more places actually exist for each object, because I can physically move the vertex so it is actually sitting between x = 0.022 and x = 0.023.
Even worse, the View Properties dialog shows the 3D Cursor coordinates only to 2 decimal places.
I’ve looked for an option somewhere to change the number of displayable decimal places, but it looks like this is not available. Yet for some high precision modeling, say of a delicate machine part, it would be convenient to see and work with more decimal places. It’s possible to work without this availability, but sometimes really awkward.
In truth, in the “real world” of Blender users, I doubt that more than 3 places really will matter ever. I understand that. Nevertheless, I am curious about this.
Maybe it’s saying “no” to my first question, but I’m not a programmer enough (certainly not with Blender) to be sure. Even if it remains tedious to do, I do hope the answer to my Q2 is yes.
Given that Blender doesn’t generally use specified units of measurement, you can make you machined parts more accurate by using a different scale when modelling. If, for example, you decide One Blender Unit = 10 metres, then 3 decimal places isn’t very small but if you decide One Blender Unit = 1 millimetre, then 3 decimal places becomes incredibly precise.
I think float’s decimal accuracy is six digits. You can easily prove this by for instance adding printf’s to suitable spots in the source. To add more accuracy to input you need to naturally edit the UI code handling the numinputs.
I think you can access the values via Python too. So that might be an easier route.
I haven’t started learning to use Python in Blender yet, but will someday. Thanks for the information. In the meantime, I’ve been doing a couple of experiments, and it looks like one can actually enter several places of precision into the numerical interface textboxes, and the entered values seem to be stored internally. However, if you later click on a textbox to see the full value, it’s not available in the interface, having rounded a value to only 3 places (but, the internal number remains valid). Once you exit the textbox, however, then the rounded value in the interface becomes the new internal value. A little weird, but interesting.
Figured I’d throw an example out that I ran into the other day. I was building a sewer ceiling that was supposed to fit a 512 square unit area in unreal, and none of the ASE exporters have built-in scaling options. For some reason I’d built the thing to be 6x6 instead of four or eight or something reasonably divisible. Scaling the whole thing by 85.333 (in the N window) reported extents of 511.xxx, but when I threw on some more digits it eventually clicked over to 512.000.
Yeah, I’m finding examples like that also. It has prompted me to write up a short tut or doc that discusses this stuff a bit, containing some specific experiments one can do to try to understand this more (for those who might want this much detail). I’ll post it as soon as I can get it done. It’ll happen within 2-3 weeks, I suppose, depending on my mood and the number of Shiner Bochs I’ve been drinking.