Following recent discussions about application defaults with other users, I removed most of my user defaults to test whether or not McIDAS-V displays features in data by using just the min/max of the image. I suspected that if there was a bad pixel in the image, little would be seen in scene features until the unit was changed. Here is one case, where I loaded a time series, and the range of entire set varies from 0-316.7, due to the fact that the "Most Recent Image" is still empty. This could also happen if the imagery data wraps (the pixel saturates and then the radiance wraps to the cold end). Normally, I would set a user default with range 180-320 using an Inverse Gray Scale color table. This generally is a good range to see features in the data. Is it possible that user's might be looking for the bad pixels and want to see the entire range, of course, but I think there is a good argument for being able to see the features in the data.
If an application default is set to display this imagery, it still would be possible to report the min/max range in the legend label. I would think that similar defaults would work for the other IR bands. I generally use tighter ranges on some of the other IR bands, but the point it to be able to see features in the imagery when the data is first displayed for all users. Individual users could override the ranges and change color tables with the user defaults.
We had some follow up questions for you on this request.
1. Considering an IR brightness temperature, are you more interested to see a static range of values displayed by default (like 180K-320K) for all satellites? Then the min/max pixel value would be listed in the legend.
2. Or are you considering something more complex where there is an algorithm in the code that would inspect the data and disregard bad pixels or blank images in a time series to calculate the "best" range to display the data. (and list the min/max in the legend)
Considering option #1, you can consider that this can be done already by making a general parameter default for all IR temps (like you mentioned), then the min/max reporting would need to be developed.
Considering option #2, much more development would be needed with such an algorithm.
For the IR bands, it does not need to be a complicated algorithm, it could just be a range which is most likely to show some features in the data. In most cases, standard min/max works well, except in the case shown above and when bad pixels are present. The idea is to get McIDAS-V to show a reasonably nice image for everyone, regardless of familiarity with user parameter defaults. However, it would be nice to know when bad pixels are present, thus the request to report the min/max in the "extra legend labels."
I believe we'd add a system parameter default for a specific range of temperatures along with developing the best way to report the min/max of the image/imagery to the user.
