Jump to content

Is the Horizon always straight


Recommended Posts

12 hours ago, geogphotos said:

When standing on a beach looking out to sea the curvature is obvious to the naked eye. The horizon is curved. 

 

Not sure if this helps... https://www.quora.com/At-what-altitude-do-you-see-the-curvature-of-the-Earth

 

Maybe there's a difference between how we perceive the horizon and how the camera captures it? I just checked one of my RAW files taken at 28mm (equiv) focal length whilst standing on Eastbourne beach with the ocean horizon lying close to horizontal across the centre of the frame (so there's no lens distortion that would cause curvature) and the horizon is damn near perfectly straight (not curved). Not quite as extreme as an 18mm lens, but potentially supporting evidence? I've got loads of photos where the horizon looks curved when it's near the top of the frame, but I believe that's due to uncorrected lens barrel distortion.

 

Mark

Edited by M.Chapman
Link to comment
Share on other sites

1 hour ago, geogphotos said:

I've also come across this issue when photographing something closer on the beach such as my children running towards me. If I were to correct the horizon they would be at a tilt!

 

They look fine with simple rotation to me, horizon horizontal and children upright. They're leaning to the left at the moment. I'll post a corrected version here if you're happy for me to do so. I also removed the pincushion distortion that makes the sea look concave (inverse earth curvature??).  That being said, wonky horizons often sell =  artistic effect and can make shot look more "dynamic"  :)

 

Mark

Edited by M.Chapman
Link to comment
Share on other sites

13 hours ago, geogphotos said:

I000014n8gcskDAQ.jpg

 

Obviously can be corrected in Photoshop but crop of my son was published as full page in Sunday Times travel supplement so the wonky horizon wasn't a problem.

 

On this shot the crooked horizon suggests the spontaneity of the moment. It looks fine - and saleable - as it is...

Link to comment
Share on other sites

58 minutes ago, geogphotos said:

 

Please go ahead and post your corrected image. It will be interesting to see.

 

My point is that images don't have to be 'correct' to work and, more importantly, to sell.

 

OK here's a rotated and pincushion distortion removed version.

 

HWEX3_C_rotated.jpg

 

Some content aware fill was required to fill in the blank rotated area.

 

NB. My comments are not whether the corrected version looks better or not. (I like both equally). My point is simply that if the camera is not rotated about the lens axis (i.e. top and bottom edges of the image sensor are horizontal) a distant ocean horizon will always appear horizontal irrespective of where it is in the frame. (Not sure if this still applies if using a tilt shift lens though when the horizon is above or below centre). The horizon may also appear to have some curvature even after lens distortion correction has been applied, if the lens has a wide enough angle view and you're not right down at sea level.

 

Mark

 

 

  • Upvote 1
Link to comment
Share on other sites

With a  background in 3D metrology, the question of ocean horizon curvature in photos continued to intrigue me, so I constructed a more detailed mathematical model than before (see earlier posting). This revealed something I wasn't expecting. The model indicates the curvature of the horizon will also vary depending on where in the image the horizon is positioned in the image when using a wide angle lens. The lowest curvature is seen when the horizon is placed across the centre of the image, and slightly more curvature (can be +30%) will be seen when the horizon is placed at the top or bottom of the frame. This effect is in addition to any lens distortion since my model assumes a perfect lens (with no barrel or pincushion distortion). The results are as follows.

 

Screen_Shot_2018-08-09_at_21.34.30.png

 

 

The model indicates that horizon curvature should be negligible when standing close to sea level with a perfect (no barrel or pincushion distortion) 18mm wide angle lens, but increases with altitude. If the focal length of the lens is increased the horizon curvature in the image is reduced.

 

Hopefully I haven't made any errors. If anyone wants me to run the model with different parameters (focal length, sensor size/pixels/aspect ratio) let me know and I'll post the results here. The geek in me would like to be able to confirm the above with some real photos. Guess what I'll be shooting next time I'm at the coast... :wacko:

 

Mark

(Must get back to Alamy keywording and stop being distracted by the forum!)

 

 

  • Upvote 1
Link to comment
Share on other sites

I would say for seascapes, generally it is best to keep the horizon straight.  But I did a shoot, a while back, on a sailboat and I in some shots I let the horizon go off a little to add some dynamics to the sailing...this photo has sold.  I should reprocess it, I could do a lot better with the raw tools that Photoshop now has (such as the dehaze slider)....this is a 13 year old digital file.

 

caribbean-west-indies-grenadines-sailing

 

 

 

 

Link to comment
Share on other sites

On 8/9/2018 at 21:39, M.Chapman said:

With a  background in 3D metrology, the question of ocean horizon curvature in photos continued to intrigue me, so I constructed a more detailed mathematical model than before (see earlier posting). This revealed something I wasn't expecting. The model indicates the curvature of the horizon will also vary depending on where in the image the horizon is positioned in the image when using a wide angle lens. The lowest curvature is seen when the horizon is placed across the centre of the image, and slightly more curvature (can be +30%) will be seen when the horizon is placed at the top or bottom of the frame. This effect is in addition to any lens distortion since my model assumes a perfect lens (with no barrel or pincushion distortion). The results are as follows.

 

Screen_Shot_2018-08-09_at_21.34.30.png

 

 

The model indicates that horizon curvature should be negligible when standing close to sea level with a perfect (no barrel or pincushion distortion) 18mm wide angle lens, but increases with altitude. If the focal length of the lens is increased the horizon curvature in the image is reduced.

 

Hopefully I haven't made any errors. If anyone wants me to run the model with different parameters (focal length, sensor size/pixels/aspect ratio) let me know and I'll post the results here. The geek in me would like to be able to confirm the above with some real photos. Guess what I'll be shooting next time I'm at the coast... :wacko:

 

Mark

(Must get back to Alamy keywording and stop being distracted by the forum!)

 

 

 

For those who might be interested, here's the geometry behind the model. Obviously not to scale! For diagrammatic clarity I've positioned a huge camera with a massive sensor at a significant altitude, but it shows the basic principle of how a cone of light rays from the line of the visible horizon passes through a lens with a 90 degree angle of view and intersects with the camera sensor to form an inverted image of the horizon.

 

New_Double_Cone_sensor_sphere_text.jpg

 

Mark

Edited by M.Chapman
  • Upvote 1
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.