Page MenuHome

Solve camera motion generates "Solve error: nan"
Closed, ResolvedPublic


System Information
Operating system: Linux-5.1.16-300.fc30.x86_64-x86_64-with-fedora-30-Thirty 64 Bits
Graphics card: GeForce RTX 2070/PCIe/SSE2 NVIDIA Corporation 4.5.0 NVIDIA 430.09

Blender Version
Broken: version: 2.80 (sub 74), branch: master, commit date: 2019-07-11 13:50, hash: rB06312c6d2db8
Worked: (optional)

Short description of error
[Please fill out a short description of the error here]
With five tracks, attempt to solve camera motion generates "Solve error: nan" message

I also saw this error when trying to use the clean up -> clean tracks with error # ( forget the error size I had specified ).

Exact steps for others to reproduce the error
Creates image sequence.
loaded clip into motion tracking
set sequence frames
started creating tracks
jumped into solve camera motion early
Error shows up.
[Please describe the exact steps needed to reproduce the issue]
[Based on the default startup or an attached .blend file (as simple as possible)]

opened a set of PNGs and started creating tracks

Event Timeline

Sebastian Parborg (zeddb) triaged this task as Needs Information from User priority.

The images are not included so I can't really reproduce this.

I'm also a but unsure with the reproduction steps. What do I have to do with the provided file to get it to break?

I just clicked solve camera motion with the settings in the blend file.
Hopefully your system can handle the images ( 324MB ).

Here are the images.

RC2 still has the problem

Sebastian Parborg (zeddb) raised the priority of this task from Needs Information from User to Needs Triage by Developer.

I still can't reproduce the exact error I'm just getting that is can solve the camera trackers (no NaN error).

@Sergey Sharybin (sergey) any ideas?

Yes. It is not possible to solve this exact motion: it is required to have at least 8 correspondences between keyframes (as in, common tracks which location is known).

The issue here is that when you use auto-keyframe it doesn't check for this, which is wrong and i'll fix that.

To make your motion solvable you need to track at least 2 more features.

P.S. Another thing i've noticed here is that you seems to be tracking a flat surface. This doesn't have well defined math solution. Sometimes Blender's solver manages to give decent/expected results sometimes it doesn't. But it's really not something we can address in any way.