Jump to content

Recommended Posts

Hello,

 

I was thinking about how Sony could implement a feature seen on the Pentax cameras that mitigates star streaking by making use of IBIS. This is a hasty write up and mockup of my idea. I would like to know if you think that critical parts are missing or simply wrong. Thanks in advance

 

Ben

 

 

-----

 

 

Star Tracker

 
App uses IBIS to increase maximum exposure time before stars begin to streak.
 
Usually the exposure time one is able to achieve when doing astrophotography is depending on the focal length. The longer the focal length the shorter the exposure. One guiding rule for example says 500s / focal length (ff equivalent) gives you roughly the maximum exposure time before stars start streaking. These rules fall apart more and more as highe pixel densities in modern sensors show streaking faster.
 
Since those Sony Alpha series cameras that feature IBIS have no built in GPS there are no means of orientation and sophisticated guesstimates as to how the sky will move during the time of exposure. It would however only take 2 calibration frames taken i.e. a minute apart and marking 2 Stars in both frames in order to calculate how the IBIS would need to move.
 
Having two x and y coordinates for 2 points in time will enable you to derive lots of information:
  • sky movement in x-axis / time
  • sky movement in y-axis / time 
  • sky movement in angle / time
  • agnostic from whatehever lens and sensor size combination you are using.
 
Combined with the knowledge about the maximum physical movement capabilities of the IBIS-system, the app should be able to calculate a maximum exposure time before the stars start streaking.
 
App flow:
 
  • instruct User to dial in an exposure that enables them to pinpoint stars without streaking and to frame the picture (final!)
  • start calibration
startrackermockup.jpg
  • calculate all necessary values
  • let the user choose:
    • exposure time up the calculated maximum
    • fitting ISO and aperture
    • amount of images taken in a row
  • instruct user to not move the camera
  • start of the image aquisition after a short delay
 
 
Bonus/convenience features for users that might not be as accustomed to astrophotography:
  • automatic star detection, reusing code from panorama stitching
  • automatic blending of all aquired images to enhance snr ratio
  • automatic exposure with presets like “ettr" for widefield pictures (details in the star cores can be lost) or “conservative” for dso pictures (need more highlight lattitude)
Link to post
Share on other sites

I doubt there's enough range of motion in the sensor to get much benefit in this application. Cool thought though. A more likely function would be some sort of multiple shot stitching where star motion is corrected for, noise too. 

 

 

 

Thank you for your response.

 

Aligning and stacking multiple shots in post is rather easy although it can be very tedious once stack more than a few images.

 

May I ask what makes you think that the IBIS has not enough range of motion? Is the IBIS in Pentax cameras more capable in terms of how high focal lengths they support ?

Using IBIS instead of a proper equatorial mount has of course diminishing returns as you increase the focal length but if the Sony IBIS can help me shoot my 500mm mirror lens handheld at 1/50th of a second instead of 1/500 of a second, it would at least yield some benefit. Or to put it more generic - I think that the benefits and limitations of ibis for handheld shooting will correlate to the benefits and limitations when using it on a tripod for astrophotography.

 

But thanks again, I guess I will add some lines for expected limitations and reasons for bad calibration like lens distortion, too near calibration stars etc.

Link to post
Share on other sites

Here's a video showing the motion of the IBIS in an A7R2. As you can see, the tilting motion, which would be required to track stars, is really quite minimal.  That portion of motion related to the tilting that the sensor can therefore correct is necessarily also quite minimal.  What's enough for hand-holding a large lens is really quite a bit less than what you'd need to correct for the motion of the Earth.  You can manipulate your own sensor to get an idea of just how much play there is in there, at your own risk. 

Anyhow, you might be able to track a star through a degree or two, which depending on the focal length might be a minute or so of exposure time. In other words, the benefit is quite minimal, especially when compared to digital stitching of multiple images instead.  

 

You've identified an easily solvable problem - correcting for star streaks in an app.  No need to overcomplicate the solution.  If Sony made an app for it, I'd probably buy it.  Also, I wish they'd let you take a still instead of just a video with the existing star streak app.  

Link to post
Share on other sites

As you can see, the tilting motion, which would be required to track stars, is really quite minimal.  

 

If by tilting you mean pitch and yaw of the IBIS then this would lead to a undesired field curvature. The Pentax models only have x,y and roll and they perfectly manage track stars. Please see technical specifications ("(...)allowing the sensor to oscillate in three directions—horizontally, vertically, and rotationally. ").

 

This app would not be able to substitute a proper tracking mount but it would give something like 2-4 stops of light. If only working with a tripod I went up to 1200 Pictures (all in raw) in order to get some faint detail at only 135mm.

multiframepsp.jpg

It is  a misconception that astrophotography only uses very long focal lengths. It might be true for planetary or very deep sky objects like the horse head nebula in the constellation of Orion. Andromeda is almost too big at 200mm, Bernards loop can be framed at 50-90mm, Orion holds lots of details at 200mm and during milky way season you are very well suited with 24mm or wider. All these uses cases would massively benefit from such an app. Stacking alone often has to fight against the noise floor of the sensors and is very ineffective at a certain point. Combining Sensor stabilized pictures and stacking will/would go a long way.

Link to post
Share on other sites

  • 2 weeks later...

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
×
×
  • Create New...