Hi all,
I found a possible motion tracker bug but I don’t know how to report it.
There are huge difference between renderers (blender and cycles) after solving. I new to the tracker and this one may be my fault but I’m not sure. Here is the video of my problem
Try turning off GPU in cycles? Are you using OpenCL or CUDA?
I wasn’t even aware that the motion tracker uses the renderer?
I use CUDA. I am new to tracker and I don’t know all the details.
I found a way to report it. I hope it will be fixed if it is a bug.
Thank you.
Follow the instructions written.
Click the Magnifier button to search for categories, if applicable, to tag the bug being reported (probably not needed, but I think it helps).
This is not a bug. The renderer has nothing to do with camera solving. The fact that you got different results has to do with slightly different positions of your markers in the 2 different attempts. In fact the position of the tracking markers is your biggest problem.
In the video you had put the markers only on a flat plane, and only in the very center of the frame. That’s a problem. Instead you should try to cover as much of the frame as possible, and also different heights, not just a plane. In your example don’t just track the markers in the center of your frame, but instead also track for example the scissors, the pencil, the background objects.
Thanks for the answer Sebastian. I tried it before and got great results. I think at least there are 8 markers in the scene and I 've got 11 markers. There are some youtube videos with little markers and have great results. I also don’t understand why there is a huge difference. 46 and 0.5 difference is huge. Is this normal? Is it may be a marker related bug?
Edit: Sebastian I didn’t know you are a Blender developer. I have another question. Fps differences cause solve errors? I converted my images at 30 fps but in motion tracker I used 24 fps. Is this a problem?
The fps should not matter. The tracker only looks at the single frames, no matter what the fps is.
About the huge difference: The fact that blender was actually able to solve the one shot correctly is actually really surprising and was pure luck. Yes, you can get great tracking results with only 8 markers, but these markers should ideally cover the entire frame, and, most importantly, they must not be just a flat plane, but at the very least one or tow should introduce some depth.
That’s not just a requirement of Blender but of all trackers. See this tutorial about Syntheyes, which describes a similar problem: https://www.youtube.com/watch?v=QmL-UbwfUtY&feature=youtu.be
If you are a cloud member you can check out Track Match Blend 2, which should cover everything you need to know: https://cloud.blender.org/p/track-match-2/
Thank you again, Sebastian 