Flight recap from Project: Traveler/NSTAR


Zack Clobes W0ZC
 

I uploaded a few photos from yesterday's flight.



As I dug into the photos from that RasPi  camera, I'm seeing a lot of lens distortion (chromatic aberrations) which is quite disappointing. The aerial photos that I uploaded are un-edited and you can see hints of it if you look at the saturations levels in the middle of the photo vs. the outer edges.

This is the camera sensor/lens that I was using:
 


Zack Clobes, W0ZC
Project: Traveler
www.projecttraveler.org

Join us on Facebook for the latest information:



Project: Traveler is a research project of Custom Digital Services, LLC.


Hank Riley
 

Zack,

It's certainly apparent what you're talking about and it would be disappointing to anyone.

Using the commonplace meaning of chromatic aberration (of two distinct types), what you have is not chromatic aberration.  I'll get a better look tonight but some of the images look as if they're suffering from big differences in brightness, the effect of which is similar superficially to "lens flare" but not the same.

Regarding "lens distortions," this phrase is usually used to refer only to departures from rectilinearity, so it's a geometric or topological deficiency.



On Sunday, July 12, 2020, Zack wrote:

I uploaded a few photos from yesterday's flight.


As I dug into the photos from that RasPi  camera, I'm seeing a lot of lens distortion (chromatic aberrations) which is quite disappointing. 


Steve G8KHW / AJ4XE
 

Hi Zack - Ive not tried (or even seen) that particular pi camera before but I see its based on the 5Mp OmniVision OV5647 sensor - that was the sensor used by the original pi v1 camera.  Raspberry pi replaced the v1 camera with the 8Mp v2 around 2016 because OmniVision had stopped production of the OV5647 in 2014.

Yet since that time there are many camera boards based on that sensor - some as cheap as $4.   I suspect that the OV5647 is now being pirated.  Cameras based on that sensor seem to perform adiquatly on the ground - but when you get them airborne and bright light they suffer with what I call purple haze (very jimi hendrix).  I think your pictures are suffering form a variant of that rather than chromatic aberration.    Maybe something can be done with filtering or different configuration - but I have yet to find it.

The original genuine V1.3 or V2 camera (based on a Sony Sensor) seem best for HAB - although the V2 does need some re-focusing (its set focus is about 2m).  There is also a 12MP v3 camera out - but I haven't tried it yet.

    Steve


On 12/07/2020 20:30, Zack Clobes W0ZC wrote:
I uploaded a few photos from yesterday's flight.



As I dug into the photos from that RasPi  camera, I'm seeing a lot of lens distortion (chromatic aberrations) which is quite disappointing. The aerial photos that I uploaded are un-edited and you can see hints of it if you look at the saturations levels in the middle of the photo vs. the outer edges.

This is the camera sensor/lens that I was using:
 


Zack Clobes, W0ZC
Project: Traveler
www.projecttraveler.org

Join us on Facebook for the latest information:



Project: Traveler is a research project of Custom Digital Services, LLC.

Virus-free. www.avg.com


Zack Clobes W0ZC
 

Yeah, I was just doing some more digging. The Ras Pi forums are referring to it as a "lens correction" issue. Looks like it's a defined problem, although the solution may involve some hand-tuning.


Zack 


On Sun, Jul 12, 2020 at 3:39 PM Hank Riley via groups.io <n1ltv=yahoo.com@groups.io> wrote:
Zack,

It's certainly apparent what you're talking about and it would be disappointing to anyone.

Using the commonplace meaning of chromatic aberration (of two distinct types), what you have is not chromatic aberration.  I'll get a better look tonight but some of the images look as if they're suffering from big differences in brightness, the effect of which is similar superficially to "lens flare" but not the same.

Regarding "lens distortions," this phrase is usually used to refer only to departures from rectilinearity, so it's a geometric or topological deficiency.



On Sunday, July 12, 2020, Zack wrote:

I uploaded a few photos from yesterday's flight.


As I dug into the photos from that RasPi  camera, I'm seeing a lot of lens distortion (chromatic aberrations) which is quite disappointing. 


Steve G8KHW / AJ4XE
 

Actually that does look more like the problem.

    Steve


On 12/07/2020 21:43, Zack Clobes W0ZC wrote:
Yeah, I was just doing some more digging. The Ras Pi forums are referring to it as a "lens correction" issue. Looks like it's a defined problem, although the solution may involve some hand-tuning.


Zack 

On Sun, Jul 12, 2020 at 3:39 PM Hank Riley via groups.io <n1ltv=yahoo.com@groups.io> wrote:
Zack,

It's certainly apparent what you're talking about and it would be disappointing to anyone.

Using the commonplace meaning of chromatic aberration (of two distinct types), what you have is not chromatic aberration.  I'll get a better look tonight but some of the images look as if they're suffering from big differences in brightness, the effect of which is similar superficially to "lens flare" but not the same.

Regarding "lens distortions," this phrase is usually used to refer only to departures from rectilinearity, so it's a geometric or topological deficiency.


https://en.wikipedia.org/wiki/Chromatic_aberration



Hank
___________________________________________________________


On Sunday, July 12, 2020, Zack wrote:

I uploaded a few photos from yesterday's flight.


As I dug into the photos from that RasPi  camera, I'm seeing a lot of lens distortion (chromatic aberrations) which is quite disappointing. 

Virus-free. www.avg.com


Hank Riley
 

Often making the image better looking makes the lens/sensor defects more tolerable.

Nothing fancy in the way of image software was employed here.  Freeware from something like 14 years ago.  Global brightness, contrast, saturation, and sharpness were adjusted.

Probably overdid the darkening, but it demonstrates what can be done.  With something fancy like Photoshop or equivalent, the pink problem could be selectively worked upon.

I think the RPi lens/image sensor system looks really awful, at least the one you got, since this pink haze is happening barely after leaving the ground.  Looks like it's maybe all due to the sensor, not the lens.  Shouldn't have to rework this kind of simple picture; the same thing could happen on the ground at the beach.

Inline image


Zack Clobes W0ZC
 

Yeah, and i intentionally didn't do any post-editing on these photos so as to not mask the problem, etc. Normally I would touch them up a bit before posting.

I'm torn, as I think the optics are decent on this camera and (with processing) would be better than the old Canon that I have CHDK running on. Having the Pi really opens up a whole new world when I'm working with a school to let the kids do some programming on it. But if I can't get a steady supply of these things then I don't know if it's worth pursuing this option. It was nice because I had maybe $80 is the whole solution and could easily reduce that by going with the Pi Zero for instance.


Zack



On Sun, Jul 12, 2020 at 4:11 PM Hank Riley via groups.io <n1ltv=yahoo.com@groups.io> wrote:
Often making the image better looking makes the lens/sensor defects more tolerable.

Nothing fancy in the way of image software was employed here.  Freeware from something like 14 years ago.  Global brightness, contrast, saturation, and sharpness were adjusted.

Probably overdid the darkening, but it demonstrates what can be done.  With something fancy like Photoshop or equivalent, the pink problem could be selectively worked upon.

I think the RPi lens/image sensor system looks really awful, at least the one you got, since this pink haze is happening barely after leaving the ground.  Looks like it's maybe all due to the sensor, not the lens.  Shouldn't have to rework this kind of simple picture; the same thing could happen on the ground at the beach.

Inline image


James Hannon
 

I have heard that cameras lacking an IR blocking filter will have a pink cast.

Jim Hannon

On 7/12/2020 4:10 PM, Hank Riley via groups.io wrote:
Often making the image better looking makes the lens/sensor defects more tolerable.
Nothing fancy in the way of image software was employed here.  Freeware from something like 14 years ago.  Global brightness, contrast, saturation, and sharpness were adjusted.
Probably overdid the darkening, but it demonstrates what can be done. With something fancy like Photoshop or equivalent, the pink problem could be selectively worked upon.
I think the RPi lens/image sensor system looks really awful, at least the one you got, since this pink haze is happening barely after leaving the ground.  Looks like it's maybe all due to the sensor, not the lens. Shouldn't have to rework this kind of simple picture; the same thing could happen on the ground at the beach.
Inline image
--
WB0TXL
WEB Page http://www.ocrslc.net/
Blog http://jimhannon.wordpress.com
CoCoRaHS station IA-LN-7
42,11.90N,91,39.26W