Node density of grids to avoid pixelated maps

I made some figures with the remote data sets (the Blue and Black marble mosaics). I started from the lowest resolution (01d) to the highest (30s). Obviously, in the first figures I get a pixelated map. I wonder how can I know in beforehand which resolution should I choose to avoid get a pixelated map (based on its region and width)? Or in other words, how many nodes per inch (or cm) should the grid have? 300 nodes per inch (like it would an image with 300 dpi)?

Maybe it is a bit off of topic but maybe some technical suggestions could be add in the docs. And also a way so GMT automatically select (or suggest) the best resolution given the chosen map scale could be add (like coast -Da). I hope I was clear.

Funny you should mention it. We have a pending NASA proposal (submitted in January) that would do exactly as you indicated, requiring a new GMT defaults GMT_IMAGE_RESOLUTION [300] or similar to guide that decision. Hopefully it will get funded and we can do the work quickly, otherwise it will take a bit longer given all the other tasks.

1 Like

Great!! So I will have to wait.

About the first part. Do you have any tips about the grid resolution in order to avoid pixelated maps?

Given -R -J you can do mapproject and get dimensions in inches, divide by 300 dpi and obtain rough nx,ny estimages, then divide those into the east-west and north-south in degrees to get two estimates for xinc, yinc, then maybe pick the smallest and round to nearest supported increment. May be more involved if oblique projections since you need to pick the right range in longitude and latitude.

1 Like