Add getters for camera parameters #1419 #2724

Dtsitos · 2025-12-17T11:26:54Z

Describe your changes

This PR adds read-only getters to the camera API to retrieve derived world-space
camera parameters (azimuth, elevation, and distance).

The values are computed from the camera position and focal point and do not
modify camera state. This enables external applications and bindings to query
camera orientation consistently without duplicating math.

No existing behavior is changed.

Issue ticket number and link if any

Fixes #1419
#1419

Checklist for finalizing the PR

I have performed a self-review of my code
I have added tests for new features and bugfixes
I have added documentation for new features
If it is a modifying the libf3d API, I have updated bindings
If it is a modifying the .github/workflows/versions.json, I have updated docker_timestamp

Continuous integration

\ci fast

…ickly for further processing

…n to compute the distance from the camera position to the focal point

github-actions · 2025-12-17T11:27:02Z

You are modifying libf3d public API! ⚠️Please update bindings accordingly⚠️!
You can find them in their respective directories: c, python, java, webassembly.

mwestphal · 2025-12-17T12:23:44Z

Hello @Dtsitos

You do not seem to have used #2470 as a base, is that intended ?

Dtsitos · 2025-12-17T14:41:41Z

Hello there @mwestphal and thank you for the quick reply,

I implemented the getters using the camera’s position and focal point with standard mathematical functions (sqrt, atan etc.) to provide world-space azimuth, elevation, and distance. This makes the code more simple, self-contained, and does not rely on renderer-specific environment vectors. Based on the issue description I think this is more desirable and less complex both to maintain and to understand.

If this is alright with you I will start working on the bindings ASAP.

mwestphal · 2025-12-17T15:27:28Z

Hello there @mwestphal and thank you for the quick reply,

I implemented the getters using the camera’s position and focal point with standard mathematical functions (sqrt, atan etc.) to provide world-space azimuth, elevation, and distance. This makes the code more simple, self-contained, and does not rely on renderer-specific environment vectors. Based on the issue description I think this is more desirable and less complex both to maintain and to understand.

Indeed, that looks fine, but I defer to @snoyer and @Meakk for the actual implementation review.

shouldn't the method be called simply getElevation though ?

If this is alright with you I will start working on the bindings ASAP.

Lets add test first :)

snoyer · 2025-12-17T15:41:25Z

does not rely on renderer-specific environment vectors

It does rely on the assumption that the environment's horizontal plane is XY and Z is up, which is not always the case in f3d because the user can provide a custom up direction vector.

mwestphal · 2025-12-17T16:34:51Z

does not rely on renderer-specific environment vectors

It does rely on the assumption that the environment's horizontal plane is XY and Z is up, which is not always the case in f3d because the user can provide a custom up direction vector.

Right! It couldnt be that simple :)

mwestphal · 2025-12-22T07:48:35Z

Need any help @Dtsitos ?

Dtsitos · 2025-12-24T16:37:40Z

I am sorry for not responding earlier I was packing for vacation. I have made progress and I think my solution is correct now. I will try to send it here when I get back home in three days. If I am causing a delay in the project and someone else wants to solve it feel free to unassign me and thank you for your co-operation.

mwestphal · 2025-12-24T16:59:11Z

No worries! enjoy your vacation :)

Dtsitos · 2025-12-30T15:53:26Z

I think my solution should be complete now at least mathematically. I did work on the previous attempt another user made and mostly improved by avoiding code reuse with the helper function and doing the floating point check before normalising. Also I added some basic tests. If this solution seems alright I will try to work on the bindings. To be totally honest I have never worked with bindings before but I started reading up on that today.

mwestphal · 2026-01-01T09:40:14Z

library/testing/TestSDKCamera.cxx

Nice, do you think you can look at this test and add similar test cases ?

https://github.com/f3d-app/f3d/pull/2499/changes#diff-811eba52fd9718f76325db6bf27ed2cbb299ba716b9a44b36ef17259caa03bf9

mwestphal · 2026-01-01T09:41:00Z

I think my solution should be complete now at least mathematically. I did work on the previous attempt another user made and mostly improved by avoiding code reuse with the helper function and doing the floating point check before normalising. Also I added some basic tests. If this solution seems alright I will try to work on the bindings. To be totally honest I have never worked with bindings before but I started reading up on that today.

Indeed, it looks fine :), lets focus on improving the testing first, but you can work on the bindings as well.

mwestphal

LGTM, please rebase on master and also push a to a feature branch instead of master, even on your own branch.

mwestphal · 2026-01-01T09:42:28Z

library/private/camera_impl.h

  class internals;
  std::unique_ptr<internals> Internals;
+  void getPositionToFocalVector(vector3_t& vec) const;


Suggested change

class internals;

std::unique_ptr<internals> Internals;

void getPositionToFocalVector(vector3_t& vec) const;

/**

* Add some doc

*/

void GetPositionToFocalVector(vector3_t& vec) const;

class internals;

std::unique_ptr<internals> Internals;

mwestphal · 2026-01-01T09:43:20Z

library/public/camera.h

-
+  /** Return the camera azimuth angle in degrees */
+  [[nodiscard]] virtual double getWorldAzimuth() const = 0;
+  /** Return the camera elevation angle in  degrees */


Suggested change

/** Return the camera elevation angle in degrees */

/** Return the camera elevation angle in degrees */

snoyer · 2026-01-02T04:54:43Z

library/src/camera_impl.cxx

+  vector3_t cross;
+  vtkMath::Cross(right, horizontal.data(), cross.data());
+  double sign = (vtkMath::Dot(cross.data(), up) >= 0.0) ? 1.0 : -1.0;
+
+  double angleRad = vtkMath::AngleBetweenVectors(right, horizontal.data());
+  return sign * vtkMath::DegreesFromRadians(angleRad);


can that be simplified with vtkMath::SignedAngleBetweenVectors?

snoyer · 2026-01-02T04:54:57Z

library/src/camera_impl.cxx

+  double dot = vtkMath::Dot(view.data(), up);
+  dot = std::clamp(dot, -1.0, 1.0);
+
+  return vtkMath::DegreesFromRadians(std::asin(dot));


can that be written in terms of AngleBetweenVectors?

Dtsitos · 2026-01-02T11:57:26Z

Good evening (in my timezone at least) and happy new year!! I improved my documentation and my implementation for Azimuth based on the comment from @snoyer. I will start working on testing now. I made a new feature branch on my local repo as you requested but I am not accustomed to github and not sure how to change the pull request to not pull into f3d-app:master, could you help me with that?

mwestphal · 2026-01-02T12:04:55Z

Good evening (in my timezone at least) and happy new year!! I improved my documentation and my implementation for Azimuth based on the comment from @snoyer. I will start working on testing now. I made a new feature branch on my local repo as you requested but I am not accustomed to github and not sure how to change the pull request to not pull into f3d-app:master, could you help me with that?

I'm afraid you need to create a new PR, there is no work around for this :)

Dtsitos · 2026-01-02T12:13:04Z

Will do!

mwestphal · 2026-01-02T12:22:34Z

Superseed by: #2767

Dtsitos added 6 commits December 13, 2025 19:48

added distance, azimuth and elevation API for returning the values qu…

935740b

…ickly for further processing

Updated implementation to not use M_PI and added comments to camera.h

c589cd3

Merge branch 'f3d-app:master' into master

8ae12b9

Clarified naming for new functions and added a private helper functio…

a54c71e

…n to compute the distance from the camera position to the focal point

Merge branch 'f3d-app:master' into master

613f282

Merge remote-tracking branch 'origin/master'

1c7c5ac

Dtsitos added 3 commits December 30, 2025 17:21

Changed getters to be enviromentally aware

af3407a

Merge branch 'f3d-app:master' into master

03173e2

Added tests for three new getters

3f8671b

mwestphal reviewed Jan 1, 2026

View reviewed changes

mwestphal requested changes Jan 1, 2026

View reviewed changes

mwestphal reviewed Jan 1, 2026

View reviewed changes

snoyer reviewed Jan 2, 2026

View reviewed changes

mwestphal mentioned this pull request Jan 2, 2026

Add getters for camera parameters #1419 #2767

Open

5 tasks

mwestphal closed this Jan 2, 2026

	/** Return the camera elevation angle in degrees */
	/** Return the camera elevation angle in degrees */

Uh oh!

Add getters for camera parameters #1419 #2724

Add getters for camera parameters #1419 #2724

Uh oh!

Conversation

Dtsitos commented Dec 17, 2025

Describe your changes

Issue ticket number and link if any

Checklist for finalizing the PR

Continuous integration

Uh oh!

github-actions bot commented Dec 17, 2025

Uh oh!

mwestphal commented Dec 17, 2025

Uh oh!

Dtsitos commented Dec 17, 2025

Uh oh!

mwestphal commented Dec 17, 2025

Uh oh!

snoyer commented Dec 17, 2025

Uh oh!

mwestphal commented Dec 17, 2025

Uh oh!

mwestphal commented Dec 22, 2025

Uh oh!

Dtsitos commented Dec 24, 2025

Uh oh!

mwestphal commented Dec 24, 2025

Uh oh!

Dtsitos commented Dec 30, 2025

Uh oh!

mwestphal Jan 1, 2026

Choose a reason for hiding this comment

Uh oh!

mwestphal commented Jan 1, 2026

Uh oh!

mwestphal left a comment

Choose a reason for hiding this comment

Uh oh!

mwestphal Jan 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mwestphal Jan 1, 2026

Choose a reason for hiding this comment

Uh oh!

snoyer Jan 2, 2026

Choose a reason for hiding this comment

Uh oh!

snoyer Jan 2, 2026

Choose a reason for hiding this comment

Uh oh!

Dtsitos commented Jan 2, 2026

Uh oh!

mwestphal commented Jan 2, 2026

Uh oh!

Dtsitos commented Jan 2, 2026

Uh oh!

mwestphal commented Jan 2, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

mwestphal Jan 1, 2026 •

edited

Loading