Reverse Vignetting Compensation



Could someone please help me understand how to correct for this behavior?

After asking micasense about this problem, they claim it is a normal and expected response of the sensor. We have analyized other sensors and don’t see this, however, I would hope they would not lie to us. Since they are claiming this is a normal response of the sensor, could Parrot please explain how to do compensation for this?

Thank you

Vignetting Correction Sample Code

Dear Jason,
I copy here the email I sent you regarding the analysis of your data.

Here is something I compiled for you based on your data and our data.

In conclusion, your camera seems normal, at least this is to be expected with extremely low exposure times.
We have known this effect for several months (at least) and decided not to ‘correct’ it since this is not really a problem because no acquisition is actually made in this regime. Please take a look at the images you acquired during your flights and see if the exposure is above 1ms.

jason_franciosa_reverse.pdf (512.1 KB)


So in conclusion sequoia is limited to work with minimum shutter speed at 1/500s (=2ms)
Under cloudy situation and seasonal change it could to be difficult to reach 1/500s … Nice to know ! Should be on your product specification !


Hi @kikislater
If you have time, it would be interesting to check how many of your images have been taken with less than 1/500s exposure time. Since the ISO gain is set to 100, the exposure time should always been above this value in realistic conditions.



It’s not the correct season to check it, but in winter and in Europe, pictures with less than 1/500 should be more that what you think !


I’m not sure I follow you. In winter, the exposure time will increase and go above 1/500s = 0.002s.
We acknowlegde that very bright scenes (exposure time below 1/2000s = 0.0005s ) may cause problems and if you feel like this is standard operating conditions for you, we can discuss how to cope with that.


I’m sorry but in winter there is less light, so exposure will increase due to the lack of light. May be you consider fying above nude crop field, but if you fly over a pine forest of course with no snow exposure time will increase and will be like 1/400 not 1/2000.
With empty crop field, in winter, I’ve never seen exposure time decreasing even in traditionnal photography and aerial photography …


Hi @clement

We see fairly high shutter speeds (small exposure times), here is some data from one of our flights of an orchard.

RED shutter: mean - 2737.68 min - 1444.44 max - 5416.67
NIR shutter: mean - 4080.97 min - 2407.40 max - 5416.67

Here is an image showing some metadata of the red channel photos from this flight, showing:

Image number
Exposure Value
Shutter speed
Ground speed

I remember thinking at the time that the shutter speeds for the multispectral bands were pretty high, and high compared to RGB sensors.

I don’t have access to the images at the moment (I’m travelling at the moment), but will take a look at this particular dataset when I get back to see if there are any of issues regarding vignetting effects with these small exposure times.


Hi @seanmcleod,

The unfortunately named shutter speed is synonymous with exposure time. It is usually expressed in seconds (and traditionally in fraction form).

What units are you using?


Hi @domenzain

My shutter speeds are fractions, so a shutter speed of 2737.68 is a short hand for 1/2737.68.


Hi @Domenzain

Do we need to be concerned with the shutter speeds we’re seeing in terms of vignetting?