Why my stereo rms increase with using increase focal length?

At first i was using 50mm focal length and get rms around 1
then when i use 75mm focal length the rms increased to 3
WHY pls

Have you recalibrated used cameras after you have changed their focal lengths?

yaya i have recalibrate using the new focal length, btw in theory the rms will be decrease if increase focal length right??

Hi,
How many grid do you use to calibrate your camera?

in theory, longer focal length means smaller field of view, which means less area of every lens is used, which should decrease lens distortions.

however, if the lens conforms to OpenCV’s lens distortion model, regardless of the severity of that distortion, the reprojection error should be tiny in both cases.

an RMS of 1, or even 3, sounds like you have very bad cameras or something else is going on.

please post the pictures you used for your calibrations. consider using dropbox/google drive/some image pastebin site.

hi, im using asymmetrical circle grid of 11x4

okk i will post the calibrations picture, sorry for didnt provide enough information

Hi, sorry for late reply here are the link of my calibration photos for 50mm lens

https://drive.google.com/drive/folders/17sGn5oIXMayMvUoddP7OdEBABxS0_YZt?usp=sharing

for70mm lens it not with me right now.

For 50mm ,although the rms=1.x, but still i can get good disparity map

but for 70mm i though would be reduce the rms but it increased to 3.

i have capture images with different angle and position,distance for calibration

i have check that all calibration photos have same orders when i call drawchessboardcorner function

can you suggest me more as my rms cannot decrease to below 0.5

what code do you use for calibration?

i did single calibrate for both single camera only then pass to stereocalibrate

What support do you use for grid? Is it really a plane?

yes i paste my grid at my transport card which i can sure its flat

I use an aludibond grid (4 mm). It is flat.

if i wants to determine little height different of my object like different in mm, any suggestions on my lens or coding??

i have check that all calibration photos have same orders when i call drawchessboardcorner function

yaaa same order is correct