Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
02-04-2020, 02:02 PM - 1 Like   #1
Veteran Member
Leumas's Avatar

Join Date: Aug 2016
Posts: 454
Horizon Correction Ruins Composites

Not sure if anyone has mentioned this, so I thought I would.

When doing composite shots a long while back I couldn't figure out why they were coming out so blurry. Turns out the culprit was horizon correction.


So quick tip for composites

02-04-2020, 03:31 PM - 1 Like   #2
Digitiser of Film
Loyal Site Supporter
BigMackCam's Avatar

Join Date: Mar 2010
Location: North East of England
Posts: 20,574
QuoteOriginally posted by Leumas Quote
Not sure if anyone has mentioned this, so I thought I would.

When doing composite shots a long while back I couldn't figure out why they were coming out so blurry. Turns out the culprit was horizon correction.


So quick tip for composites
Interesting... I don't quite understand why that should be the case.

I assume you're talking about the K-1's auto horizon correction where the sensor's SR mechanics are used to rotate the sensor? If so, are you continuing to shoot with SR enabled (such that it might interfere or reach rotational limits in combo with the horizon correction)?

Do you have an example you could share of the source images? Also, what software are you using for the composites?

I'll be honest, it doesn't sound like horizon correction is causing this in isolation, but that there may be a combination of factors. At least, that's my starting point...
02-04-2020, 04:22 PM   #3
Otis Memorial Pentaxian
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 42,007
QuoteOriginally posted by BigMackCam Quote
Interesting... I don't quite understand why that should be the case.
I can imagine, but only if the camera were being jostled by wind or ground movement with an inadequate tripod such that an adjustment were applied some frames and not others. My understanding is that the horizon correction is pretty coarse (single degree increments).


Steve
02-04-2020, 04:35 PM   #4
Digitiser of Film
Loyal Site Supporter
BigMackCam's Avatar

Join Date: Mar 2010
Location: North East of England
Posts: 20,574
QuoteOriginally posted by stevebrot Quote
I can imagine, but only if the camera were being jostled by wind or ground movement with an inadequate tripod such that an adjustment were applied some frames and not others. My understanding is that the horizon correction is pretty coarse (single degree increments).
Perhaps... but even then, the compositing software should be dealing with the rotational differences (the ones I've tried do - at least for HDR and panorama composites).

I wonder if the OP is shooting raw then applying good lens corrections before creating the composites? Or are these in-camera JPEG composites?

There are all sorts of factors coming to mind, but rather than second-guessing, I trust the OP can give us a bit more detail on exactly how he was shooting and creating the final images...

02-04-2020, 06:13 PM   #5
Otis Memorial Pentaxian
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 42,007
QuoteOriginally posted by BigMackCam Quote
There are all sorts of factors coming to mind, but rather than second-guessing, I trust the OP can give us a bit more detail on exactly how he was shooting and creating the final images...
The EXIF makernotes might hold some clues:

Pentax Tags | LevelInfo


Steve
02-04-2020, 08:27 PM - 1 Like   #6
Pentaxian
photoptimist's Avatar

Join Date: Jul 2016
Photos: Albums
Posts: 5,113
Hmmm...

The problem is plausible with two mechanisms of action:

Accelerometer noise: If the K-1's accelerometer has any noise, frame-to-frame differences in the estimated direction of gravity will create frame-to-frame differences in the "corrrected" rotation. To a first approximation, accelerometer noise levels of about ±1/5000 of a G (or worse) will create about a 2-pixel span of blur (or worse) in the corners.

Tripod vibration: Even a fraction of a millimeter of modest frequency tripod vibration (wind, etc.) would show up in the accelerometer data as varying shift in the direction of "down" and trick the horizon correction. With horizon correction turned off, those submillimeter vibrations wouldn't affect the image. But with horizon correction turned on, they would.

P.S. One of the sublime mysteries of the universe is that gravity and acceleration are indistinguishable. Accelerate the camera by just 0.01G to the left and the electronic level will be absolutely convinced that the camera is tilted by 0.5° and do a horizon correction that moves all the corners 40 pixels. (I was convinced my camera was broken when I tried to take a photo from an accelerating aircraft, and the electronic level said the camera was was hugely tilted even though the horizon looked level in the viewfinder.)
02-04-2020, 08:37 PM   #7
Veteran Member
bertwert's Avatar

Join Date: Jun 2015
Location: Golden, BC
Posts: 15,173
QuoteOriginally posted by photoptimist Quote
P.S. One of the sublime mysteries of the universe is that gravity and acceleration are indistinguishable. Accelerate the camera by just 0.01G to the left and the electronic level will be absolutely convinced that the camera is tilted by 0.5° and do a horizon correction that moves all the corners 40 pixels. (I was convinced my camera was broken when I tried to take a photo from an accelerating aircraft, and the electronic level said the camera was was hugely tilted even though the horizon looked level in the viewfinder.)
This has never occurred to me, however it makes perfect sense and now seems obvious.

I still took my camera and slid it around a table while watching the electronic level...

02-04-2020, 09:11 PM   #8
Otis Memorial Pentaxian
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 42,007
QuoteOriginally posted by photoptimist Quote
Tripod vibration: Even a fraction of a millimeter of modest frequency tripod vibration (wind, etc.) would show up in the accelerometer data as varying shift in the direction of "down" and trick the horizon correction. With horizon correction turned off, those submillimeter vibrations wouldn't affect the image. But with horizon correction turned on, they would.
The issue for the OP is related to composites. Whether the source frames are soft is not known, though a failure of accurate registration due to spurious horizon adjustment may very well do so for the final merged image.


Steve
02-05-2020, 09:10 AM   #9
Pentaxian
photoptimist's Avatar

Join Date: Jul 2016
Photos: Albums
Posts: 5,113
QuoteOriginally posted by stevebrot Quote
The issue for the OP is related to composites. Whether the source frames are soft is not known, though a failure of accurate registration due to spurious horizon adjustment may very well do so for the final merged image.


Steve
Indeed.

However if the OP is using the camera's interval composite function, then there's no registration process. Interval composite acts just like multi-exposure mode. It makes no attempt to align the images to each other when combining them.

If the OP is using some PC-based stacking program, then that app would presumably attempt to align the images. In such cases, I have a hard time thinking of a way that horizon correction would degrade post-process image stacking.
02-05-2020, 10:15 AM   #10
Otis Memorial Pentaxian
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 42,007
QuoteOriginally posted by photoptimist Quote
Indeed.

However if the OP is using the camera's interval composite function, then there's no registration process. Interval composite acts just like multi-exposure mode. It makes no attempt to align the images to each other when combining them.

If the OP is using some PC-based stacking program, then that app would presumably attempt to align the images. In such cases, I have a hard time thinking of a way that horizon correction would degrade post-process image stacking.
Yep...hence Mike's initial question regarding what software was being used.

BTW...like @bertwert, I just had to test out the lateral acceleration = tilt phenomenon; what a surprise.


Steve
02-06-2020, 03:24 AM   #11
Veteran Member
Leumas's Avatar

Join Date: Aug 2016
Posts: 454
Original Poster
QuoteOriginally posted by BigMackCam Quote
Interesting... I don't quite understand why that should be the case.

I assume you're talking about the K-1's auto horizon correction where the sensor's SR mechanics are used to rotate the sensor? If so, are you continuing to shoot with SR enabled (such that it might interfere or reach rotational limits in combo with the horizon correction)?

Do you have an example you could share of the source images? Also, what software are you using for the composites?

I'll be honest, it doesn't sound like horizon correction is causing this in isolation, but that there may be a combination of factors. At least, that's my starting point...
No software, I'm talking about in camera composite mode.
And yes I'm referring to the in camera horizon correction. My use case was no SR, and tripod.
02-06-2020, 04:01 AM - 1 Like   #12
Digitiser of Film
Loyal Site Supporter
BigMackCam's Avatar

Join Date: Mar 2010
Location: North East of England
Posts: 20,574
QuoteOriginally posted by Leumas Quote
No software, I'm talking about in camera composite mode.
And yes I'm referring to the in camera horizon correction. My use case was no SR, and tripod.
Interesting... In which case, I'm leaning toward thinking that minor camera / tripod movement may have been sufficient to cause horizon correction adjustment during the sequence of shots, and the coarse nature of the adjustment(s) - as advised by @stevebrot above - resulted in one or more frames not being perfectly aligned with the others. It seems like the horizon correction feature is better used for single shots, or with the camera on a highly-stablised mount and tripod.

Given you were shooting from a tripod, a one-time visual alignment or use of the camera's horizon level indicator would seem to be a better approach. Live and learn, eh?

Thanks for posting about this. I'm sure it'll help some of our members
02-06-2020, 07:00 AM - 1 Like   #13
Master of the obvious
Loyal Site Supporter
savoche's Avatar

Join Date: Jun 2011
Location: Lowlands of Norway
Posts: 18,309
QuoteOriginally posted by Leumas Quote
When doing composite shots a long while back I couldn't figure out why they were coming out so blurry. Turns out the culprit was horizon correction.
Good tip, and not at all obvious until you think about it (as illustrated by the above exchange). Thanks for posting!
02-06-2020, 07:25 PM   #14
Veteran Member
Leumas's Avatar

Join Date: Aug 2016
Posts: 454
Original Poster
QuoteOriginally posted by BigMackCam Quote
Interesting... In which case, I'm leaning toward thinking that minor camera / tripod movement may have been sufficient to cause horizon correction adjustment during the sequence of shots, and the coarse nature of the adjustment(s) - as advised by @stevebrot above - resulted in one or more frames not being perfectly aligned with the others. It seems like the horizon correction feature is better used for single shots, or with the camera on a highly-stablised mount and tripod.

Given you were shooting from a tripod, a one-time visual alignment or use of the camera's horizon level indicator would seem to be a better approach. Live and learn, eh?

Thanks for posting about this. I'm sure it'll help some of our members
Thanks, hopefully spare some headache for a few


QuoteOriginally posted by savoche Quote
Good tip, and not at all obvious until you think about it (as illustrated by the above exchange). Thanks for posting!
Yeah, it was frustrating and time consuming to figure out. I was trying to capture a water scene at the time and had gone through multiple sets, all with the same problem until I turned horizon correction off, then presto! Tack sharp after that...shame I had lost most of my good light by that point. haha
02-17-2020, 03:34 AM - 1 Like   #15
Senior Member




Join Date: Oct 2015
Photos: Gallery
Posts: 166
I came to the same conclusion about horizon correction being the problem but in my case it was a timelapse that was ruined. Its an interesting feature in theory but in practise its best just kept turned off.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
camera, composites, correction, dslr, horizon, horizon correction ruins, photography
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Horizon Correction wings Pentax K-1 & K-1 II 23 10-05-2019 08:10 AM
People Composites for fun Drinkkeri Photo Critique 3 12-21-2018 11:38 PM
What was the first Pentax camera with Horizon Correction? bladerunner6 Pentax DSLR Discussion 10 08-23-2018 11:06 PM
K-7 Horizon correction. LeDave Pentax DSLR Discussion 2 12-11-2009 11:41 PM



All times are GMT -7. The time now is 10:50 PM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top