How to calibrate x, y and z?

This the same exact number I’ve obtained when doing my tests. I hope they had X-Y calibration settings in the slicer instead of the Scale feature.

1 Like

So I have to apply 100.5% on x and y … and how about the z? How do you set it?
Vertical holes are almost ellipsis.
. Thanks

I found no cure for Z. When using the scale feature, it will either add a layer (typically 0.2mm for me) or nothing and 0.2mm is too much for small objects. For example, my 20mm object goes from 19.92 to 20.12. I guess Z scaling could be used on tall objects if the print is too short by at least one layer height.

Circles not round is another issue. There are threads about it.

1 Like

For really coarse adjustments you could still use the “Scale [S]” feature on Z, but you will always end up with whatever the incremental nature of the layers will give you. Smaller layer heights will give you a better approximation. See if that helps for top surfaces.

The slight un-roundness of holes on the sides of bodies is IMHO within the usual tolerance of what I’m used to seeing in FDM. If I ever observed it, I attributed it to overhang-drooping in the upper section of the circle. Smaller layer heights may improve this as well.

Also, whenever I have holes that need a very tight fit, I design the part in a way that allows me to drill-out the holes to the desired dimension.

I did a test.
Z was supposed to be 50. It is 49.85
Pressure advance calibrated at 0.02

Now i got this.

I used adaptive layer. 0.5 adaptive and 5 radius.

What’s wrong now?

1 Like

Looks like a poorly / none calibrated material. I suggest first calibrating the filament, to ensure a reliable base variable.

In addition, the messurement is still in tolerance for additive fdm manufacturing.

I did pressure adv test and temp test.
Am I missing some ohter test?


Flow rate is an important test for print quality,

1 Like

exactly the same number i get! so its a common issue

1 Like

Try slowing the travel speeds to max 75 and see if speed affects the print dimensions.

been having the same problem. ive gotten a .1 different between x/y since ive started using this machine anyone figured it out yet??

Hello there,

before you say such harsh things, maybe get some knowledge about shrinkage first?

Your 30x30x30 CAD model surely is just a body, without any material attached to it. Right?

Thermo plastic materials all have a shrinkage factor. If you dont include that into your dimensions, then of course, you wont get 30x30x30. Your printer and Bambu Studio are not at fault here. It is you, the user, who has to be aware of that.

The industry compensates for material shrinkage and the home users have to do it as well…

Soundwave, please don’t take it personally but the above statement is a favourite among the armchair experts in the 3D printing community. I see it repeated all over the place without any actual explanation or details, just a throw away comment that people keep repeating. It’s become a folk law of 3D printing community. However, there are many of us with personal experience that know 3d printing dimensional accuracy can easily achieve +/-0.05mm in most cases. Another that is often repeated is that high-end printers only advertise +/- 0.2 mm in their datasheets but understand that is a far higher standard than a simple hole being 6mm or 5.8mm, they are talking dimensional accurency between any two features being within 0.2mm across a whole print bed which maybe 2-3 times the size of the x1c.

Shrinkage is just one factor affecting dimensionality in 3D printed parts however there are many others. The amount of shrinkage can also be affected by certain geometry.

For the case in this forum topic we are talking holes. Holes are special in that they exhibit shrinkage more than other geometry. This has been theorised to be caused by multiple reasons, here are the two I think make sense but might be wrong.

  1. A circle is usually draw with either lots of very small linear segments or an arc, in either case the effect is the same, a circle who’s circumference shortens will pull inwards. Think about a rope that goes the whole way around the earth (ignore mountains, assume a perfect sphere), how much longer do you need to make the rope to lift it one meter/foot off the ground around the whole earth? 2 * Pi meter/foot or 6.28 meter/foot. (Small changes make a big effect, in the cases of holes the id is not supported, so nothing is stopping it pulling inwards, unlike an outside radius which is less affected by this phenomenon).

  2. When printing a circle, the nozzle is constantly dragging inwards, as it moves around, some part of the motion is always inwards.

If you look at SuperSlicer, lots of work has been done on achieving software based dimensional accuracy. However, for some reason the developers at PrusaSlicer believe dimensional accuracy is something you should design into your models/designs, rather than be fixed by the manufacturing process or software used during the manufacturing process. Which personally I disagree with. If a hole should be 6mm in size, why should I put 6.2mm on the design to account for the manufacturing process (different to accounting for required tolerance, eg fitting two parts together), this should be taken care of by the software or tooling used in the manufacturing process. If you believe the software shouldn’t do clever things when it can, then you need to get rid of the 3d printer and start whittling your models by hand. I’m sure high end slicers take shrinkage into account based on geometry.

There are three key ways you can handle hole shrinkage (in the bambu eco system).

  1. Use Classic wall generation instead of Arachne (Arachne is the default in BambuStudio: Quality → Wall Generator). This by itself will reduce the error to be about 0.1mm instead of 0.2mm

  2. Compensation (OrcaSlicer: X-Y Hole Compensation) Luckily hole shrinkage doesn’t seem to be directly proportional to hole size. It seems rather constant, not perfectly but close. In my experience it seems to be around 0.2mm

  3. Use a different geometry for circular holes. One example is a polyhole, use a polygon instead of a hole. SuperSlicer has a feature to do this automatically for holes, which I’ve recently ported to OrcaSilcer from SuperSlicer (waiting on pull request being approved/merge and released see: (Ported automatic hole to polyhole conversion from superslicer by lovelytwo · Pull Request #2336 · SoftFever/OrcaSlicer · GitHub)). You might think this would make ugly holes, but in practice as long as your holes are bigger than 3-4mm, the number of sides to the poly is higher enough you don’t notice, also you can do fancy things like offsetting the polygons at each layer, which reduces the polygon look further.

The top right hole is an 8mm poly hole, and the bottom left is 10mm regular hole.

Extra reading about Polyholes and the 3d printing hole problem.

In regard to difference dimensional accuracy in x-y, my suspicious is this is most likely caused by belt tension, but is the next thing I need to look into.


0.15mm difference in Z could very well be caused by the adaptive layer heights. I’m not sure the slicer accounts for what the actual dimensions should be. I know you need to take care if you need it really precise even without the adaptive layer height.

For example if your model is supposed to be 50.01mm tall but you print with 0.2mm layer height, you’ll end up with 50.2mm print I think it always rounds up. With 0.5mm layer height, who knows what can happen, I never printed that tall (I see absolutely no point in high layer heights, you are likely to hit flow rate limit so often that it will start affecting the surface finish)

Hi, could you tell me what was the problem with the square dimensions, please? I have the same issue. The square is about 0.2mm smaller than a 3D file. Thank you

I really can’t believe there is no setting for step precision for x and Y. I had my previous Makerbot dialed into with .02 or .03 max. I am getting over a 2mm discrepancy on a 240mm part which is a lot and parts aren’t fitting. I was able to do some basic math ratios and dial it in. There should def be a setting for that, I’m sure it’s capable. Getting a great finish though.

Are you referring to X-Y compensation for dimensional accuracy? In Bambu Studio they refer to it as X-Y compensation. Hole is defined as dimensions that are within an enclosed space and contour is for outer perimeter.

Also, what is your objective? Is it dimensional accuracy or two parts fitting together? What test are you using to determine that your accuracy isn’t where you need it?

What is wanted is the ability to calibrate X, Y and Z steps as one can do on most other printers. This is so that if you print, say, a 50mm cube that it is very close to 50mm in each direction. There is no way to do that on the Bambu printers at present. That said, my own prints are better than 99% accurate in size, so I have not found a need for this.


You can’t calibrate steps, because those are set correctly.
If you are getting incorrect dimensions, then you have a problem to solve somewhere, but you should not compensate for it with steps. On CoreXY, incorrect steps would result in a diagonal skew, not just one dimension being off… I think?
It is possible that you have loose belts.
It is possible you have a backlash somewhere. (Those two things could manifest mostly the same).
It is possible slicer puts more material on one side and not the other, because it needs more walls for overhangs or thickness.
It is possible the geometry shrinks in one dimension more.