Repository navigation
Add fixes from Eli's comments - #22
Alex-Broughton wants to merge 4 commits into
Conversation
4b92569 to
d7a4e88
Compare
4a1d567 to
a8b3a7b
Compare
a8b3a7b to
1123462
Compare
czwa
left a comment
czwa
left a comment
There was a problem hiding this comment.
These are almost all clarifications/suggestions. No need to accept my suggestions if you don't like them.
Thank you for fixing my typos and bad sentences.
| Blue boxes denote static ISR steps that apply pre-computed calibration inputs; orange boxes denote steps whose application depends on the science exposure itself (for example, overscan subtraction, dithering, or saturation-footprint growth); white boxes mark the raw input and post-ISR output. | ||
| The dashed optional step applies pre-computed per-amplifier offsets when enabled.} | ||
| A dashed border marks a computation rather than an artifact correction (integer-to-float conversion, dithering, overscan statistics, and variance plane estimation). | ||
| The amplifier-offset step remains optional and applies pre-computed per-amplifier offsets when enabled.} |
There was a problem hiding this comment.
Amplifier offsets are a computation. It's a linear solution to the set of edge differences to find the per-amp offsets needed to minimize those differences. Maybe "The amplifier offset step remains optional, as it is an empirical calculation to improve background continuity" or something like that?
| \subsection{ISR Pipeline Overview} \label{sec:isr-science:overview} | ||
|
|
||
| The LSSTCam has a particular photon transfer chain and therefore a particular ISR correction pipeline. The following outline follows Figure~\ref{fig:isr-pipeline} step by step. | ||
| The LSSTCam has a particular photon transfer chain and therefore a particular ISR correction pipeline. This pipeline takes in a set of calibrations and a raw image, and it outputs a science-ready image, with complete with a mask plane, provenance, and metadata. |
There was a problem hiding this comment.
Double "with". "with a mask plane, variance, provenance, and metadata"? Or we already mentioned variance and mask above, so "science-ready image, with provenance of used calibrations and metadata including useful statistics"? The double "with" is the only correction I think this needs, these are just suggestions.
The next sentences cover metadata and provenance, so maybe ignore most of this comment.
| \item \textbf{Apply gain.} | ||
| Per-amplifier gains derived from the PTC (\S\ref{sec:processing:calibration:ptc}) convert pixel values from ADU to electrons. | ||
| Electron units are the native domain for deferred-charge modeling, brighter-fatter correction, and variance plane estimation, all of which were calibrated assuming a linear conversion from measured counts to stored charge. | ||
| Electron units are the native domain for deferred-charge modeling, BFE correction, and variance plane estimation, all of which were calibrated assuming a linear conversion from measured counts to stored charge. |
There was a problem hiding this comment.
I think this is the first time this appears in the paper, so "brighter-fatter effect (BFE) correction". Ignore if I missed one in my search.
|
|
||
| Each subsection will describe the calibration product, its construction algorithm, and the corresponding ISR correction task. | ||
| We may also include a table of the data product and the name of the associated calibration task. | ||
| Every correction in the ISR sequence described in \S\ref{sec:isr-science} requires a calibration data product that characterizes the instrumental effect being removed, measured in advance from dedicated calibration exposures rather than from the science image itself. |
There was a problem hiding this comment.
"Nearly all"? As an example, dithering doesn't.
| For LSSTCam DP2 calibration production we fit both the temperature and temporal coefficients as well as the scattered-light offset $O$ on the spline. | ||
| In addition, any hardware sensitivities are absorbed by nuisance parameters that we parameterize as first order linear corrections to $\mu$ before the spline is evaluated: | ||
| \begin{equation} | ||
| \begin{aligned} |
There was a problem hiding this comment.
If we do both, then maybe this should just be a single equation \mu_f &\rightarrow \mu_0(1 + \alpha T_i) (1 + \beta t_i)? Then there are only two \mus to consider, so we have a pre-correction \mu_0 and post-correction \mu_f.
No description provided.