# Stereo Calibration: Very high radial distortion coefficients

**URL:** <https://forum.opencv.org/t/stereo-calibration-very-high-radial-distortion-coefficients/24517>\
**Category:** Uncategorized\
**Tags:** calib3d\
**Created:** [May 26, 2026, 12:18am UTC](https://forum.opencv.org/t/stereo-calibration-very-high-radial-distortion-coefficients/24517 "2026-05-26T00:18:56Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![rrahman1215](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/rrahman1215/32/13707_2.png) [@rrahman1215](https://forum.opencv.org/u/rrahman1215)\
**Post date:** [May 26, 2026, 12:18am UTC](https://forum.opencv.org/t/stereo-calibration-very-high-radial-distortion-coefficients/24517/1 "2026-05-26T00:18:56Z")

</div>

I am getting very high radial distortion coefficients after calibration which is unusual for the type of lense I’m using. When I try to reconstruct speckled squares/rectangles I get ~10% compression in X (horizontal) direction. However, the Y direction measurement is very accurate (\<1% error). I am trying to build a 3D Digital Image Correlation setup and using two Tokina atx-i 100mm f/2.8 FF Macro Lenses with Megaspeed MS55K high speed cameras. The calibration file and images can be found in the link below. I’d appreciate any suggestions for overcoming this problem.

> **[Google Drive: Sign-in](https://accounts.google.com/v3/signin/identifier?continue=https%3A%2F%2Fdrive.google.com%2Ffile%2Fd%2F1zO2hffZTU5o-TEGkz_B4KcF7AjDjj41E%2Fview%3Fusp%3Dsharing&dsh=S-887621269%3A1779785623106744&followup=https%3A%2F%2Fdrive.google.com%2Ffile%2Fd%2F1zO2hffZTU5o-TEGkz_B4KcF7AjDjj41E%2Fview%3Fusp%3Dsharing&osid=1&passive=1209600&service=wise&flowName=GlifWebSignIn&flowEntry=ServiceLogin&ifkv=AWa2PauWi71KmeK829Qf8g5FRvixBnwGyDvQgmWqJ-pvRvRbzL77OR9k_xuggjWiUXXD4MF9hbQvfQ)**
>
> Access Google Drive with a Google account (for personal use) or Google Workspace account (for business use).

---

<div class="post-metadata">

**Author:** ![crackwitz](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/crackwitz/32/14_2.png) [@crackwitz](https://forum.opencv.org/u/crackwitz)\
**Post date:** [May 26, 2026, 8:52am UTC](https://forum.opencv.org/t/stereo-calibration-very-high-radial-distortion-coefficients/24517/2 "2026-05-26T08:52:53Z")

</div>

> Drive

inaccessible

bad luck with calibration usually comes from bad data.

low repro error is necessary but not sufficient. if it’s high, the data is junk for sure. if it’s low, the data can still be junk.

> **[Calibration Best Practices](https://calib.io/blogs/knowledge-base/calibration-best-practices)**
>
> Accurate calibration is of key importance for performance in most machine and computer vision tasks. The following lists our best practices which we have found through extensive experimentation and theoretical considerations.  Choose the right size...

---

<div class="post-metadata">

**Author:** ![rrahman1215](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/rrahman1215/32/13707_2.png) [@rrahman1215](https://forum.opencv.org/u/rrahman1215)\
**Post date:** [May 26, 2026, 6:11pm UTC](https://forum.opencv.org/t/stereo-calibration-very-high-radial-distortion-coefficients/24517/3 "2026-05-26T18:11:30Z")

</div>

Sorry for the inaccessible link earlier. Can you please try this link

> **[Google Drive: Sign-in](https://accounts.google.com/v3/signin/identifier?continue=https%3A%2F%2Fdrive.google.com%2Ffile%2Fd%2F1zO2hffZTU5o-TEGkz_B4KcF7AjDjj41E%2Fview%3Fusp%3Dsharing&dsh=S-887621269%3A1779785623106744&followup=https%3A%2F%2Fdrive.google.com%2Ffile%2Fd%2F1zO2hffZTU5o-TEGkz_B4KcF7AjDjj41E%2Fview%3Fusp%3Dsharing&osid=1&passive=1209600&service=wise&flowName=GlifWebSignIn&flowEntry=ServiceLogin&ifkv=AWa2PauWi71KmeK829Qf8g5FRvixBnwGyDvQgmWqJ-pvRvRbzL77OR9k_xuggjWiUXXD4MF9hbQvfQ)**
>
> Access Google Drive with a Google account (for personal use) or Google Workspace account (for business use).

There is definitely something wrong in the images. However, I can’t figure out the problem. The calibration software can detect the intersection points. The repro error is 0.3 px.

---

<div class="post-metadata">

**Author:** ![crackwitz](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/crackwitz/32/14_2.png) [@crackwitz](https://forum.opencv.org/u/crackwitz)\
**Post date:** [May 26, 2026, 9:19pm UTC](https://forum.opencv.org/t/stereo-calibration-very-high-radial-distortion-coefficients/24517/4 "2026-05-26T21:19:12Z")

</div>

yep, you’re showing the board head-on. that’s the issue.

tilt it out of plane.

and try to get the board into the corners of the view.

I linked to a best practices article on [calib.io](http://calib.io). review it once more.