NASA - National Aeronautics and Space Administration Follow this link to skip to the main content + Contact NASA
Kepler Guest Observer Program

Guest Observer Proposal Preparation - Target Pixel Masks
  1. STANDARD PIXEL MASKS
  2. CAN MODIFICATIONS BE MADE TO A STANDARD PIXEL MASK?
  3. WHY WOULD I MODIFY A STANDARD PIXEL MASK?
  4. CAN I PROVIDE AN ALTERNATIVE CUSTOM-MADE PIXEL MASK?
  5. WHY WOULD I NEED A CUSTOM PIXEL MASK?
  6. WHY DOES A REALLY BRIGHT TARGET NEED SPECIAL HANDLING?

Standard Pixel Masks

Standard pixels masks are sufficient for the vast number of proposed targets. A standard pixel mask is created by the Kepler Project based upon KIC parameters and instrument models:

  • Kepler magnitude (Kp)
  • location on the focal plane (to account for a spatially varying point spread function)
  • the distribution of surrounding sources
  • the background signal at the source location

The Kepler pipeline is designed to maximize the signal-to-noise ratio of the light curve by optimizing the number of pixels in the light curve extraction aperture. Each mask therefore comprises three regions:

  • the optimal aperture for photometry.
  • a halo, normally one pixel in width, that provides margin for targets moving within their apertures.
  • one extra column outside of the halo to capture the effect of charge undershoot across the detector.

Figure 1: On the left, a typical Kepler target image comprising 37 pixels. The right-hand image contains a bitmap that describes the employment of each pixel in the target mask. Black pixels are not collected, yellow are halo and undershoot pixels, white are included in the photometric aperture.

The Kepler spacecraft does not have the onboard storage for 160,000 unique pixel masks. There is instead an onboard lookup table containing an optimal set of 1,024 unique pixel patterns. Target pixel masks are assigned autonomously to be the smallest lookup mask which fully contains the required pixel set for a target. The brighter the target, the larger the pixel mask.

Figure 2: The relation between Kepler magnitude and target pixel mask size for Guest Observer targets proposed in Quarter 6. The break in the curve at Kp = 11 is the magnitude at which stars begin to saturate on the detector.



Can modifications be made to a standard pixel mask?

Yes, provided modifications are justified within your proposal.



Why would I modify a standard pixel mask?

The Kepler Magnitude in the KIC is too faint

Some Kepler targets can vary greatly in brightness over time. The value of Kp in the KIC may not be suitable to optimize signal-to-noise when the target is at its brightest. Proposers can provide an amended Kp in the magnitude column of their proposal target table. A KIC magnitude will be overridden by a brighter value provided by the proposer during target management.

The target is an extended galaxy

Proposers may request pixel halos around their extended targets. A halo is an extra buffer of pixels around the standard mask. Add the word "halo" to the comment column in your proposal target table and the number of pixel halos 1-4 requested.



Can I provide an alternative custom-made pixel mask?

Yes, provided this need is justified in your proposal. Instruction for creating a custom mask are provided here. Proposers need only provide a custom mask once their target has been accepted by the GO Target Allocation Committee. For proposal purposes, users need provide only celestial coordinates and a Kepler magnitude estimate in the Target Table.



Why would I need a custom pixel mask?

Custom masks are required for targets that are not listed in the Kepler Input Catalog. These sources will generally be faint, Kp > 17. A custom target file will also be accepted from extended target programs when the default magnitude-dependent pixel halo applied to an extended target under-captures the target.



Why does a really bright target need special handling?

Bright targets saturate the detector, resulting in charge bleeding. standard pixel masks are inefficient at capturing these targets because the default to the next largest pixel pattern in the lookup table onboard the spacecraft. This will be huge cover mostly background pixels around the targets and be wasteful with thousands of valuable pixels. Custom masks will also default to the next largest pixel pattern in the lookup table and the same problem occurs. The solution is to create a "dedicated mask". These are uploaded directly to the spacecraft lookup table and shaped precisely to match the shape of specific bright targets on the detector. An example is provided in figure 3. The two brightest accepted targets (green circles) in figure 2 were accepted as dedicated mask targets. Note the order of magnitude decrease in the required pixels compared to their standard mask sizes (red circles). The drawback to dedicated masks is that they consume valuable resources onboard the spacecraft. In order not to impact the pixel efficiency of the exoplanet survey, only 7 dedicated masks are commissioned each quarter. Dedicated masks are therefore a limited, very-competitive, resource. The results obtained so far from these bright sources has been spectacular.

Figure 3: Left - The quarter 4 Full Frame Image in the region of the saturated star RR Lyr (Kp = 7.9). This plot is a product of the FFI viewer KEPLERFFI. Right - The pixels captured by a dedicated mask placed over RR Lyr during quarter 8. Pixel value increases from blue to red. White pixels were not collected. This plot was created by the mask inspection tool kepmask.

The Kepler spacecraft does not have the onboard storage for 160,000 unique pixel masks. There is instead an onboard lookup table containing an optimal set of 1,024 unique pixel patterns. Target pixel masks are assigned autonomously to be the smallest lookup mask which fully contains the required pixel set for a target. The brighter the target, the larger the pixel mask.

Dedicated masks will not be allowed for long cadence targets. Pixel inefficiency is a hit that will be absorbed into the long cadence pixel budget provided the science justification is a strong one.

Dedicated masks will be constructed for short cadence targets because the pixel inefficiency of standard masks is unacceptable to the short cadence pixel budget. Short cadence targets brighter than Kp = 8.0 will be considered a dedicated mask request.

The proposer need do no extra work for proposing bright targets. The construction of dedicated masks is performed by Kepler engineers during target management.


Questions concerning Kepler's science opportunities and open programs, public archive or community tools? Contact us via the email address.
FirstGov - Your First Click to the US Government
+ Freedom of Information Act
+ Budgets, Strategic Plans and Accountability Reports
+ The President's Management Agenda
+ NASA Privacy Statement, Disclaimer,
and Accessibility Certification

+ Inspector General Hotline
+ Equal Employment Opportunity Data Posted Pursuant
to the No Fear Act

+ Information-Dissemination Priorities and Inventories
NASA - National Aeronautics and Space Administration
Editor: Martin Still
NASA Official: Jessie Dotson
Last Updated: Jan 6, 2012
+ Contact NASA