As I wrote in my last post there are problems when you have to scale down imagery significantly. So I am working on a new map mosaic experiment to see what I can do with it.
Starting from four tiles at 0.4 metre resolution i.e. 1 pixel measures 0.4 metres on one side, I am scaling each of these tiles to double in each direction so they have four times as many pixels in total.
The key will be the assumption that Qgis will be able to load these much larger tiles and display them as well as the originals, we just have to tweak the jgw file (world file) to say the pixel size is half of what it was previously and then they should draw in the right place
I have to recreate the mosaic from scratch i.e. I can’t use my previous mosaic in any shape or form unless I scale up all the historic aerials as well and that will lose significant quality but as they are only background since I have these completely new ones that are the most significant that could be an option (starting from the old mosaic and arranging the new tiles around it) but I have decided to start from scratch and build a completely new one with only Retrolens coverage.
So far the results look really good in Gimp and I am banking on the larger tiles looking really good in Qgis because this has been my experience with Qgis and map mosaics to date, basing on experience of using Christchurch City 0.075 metre LDS contemporary coverage and the higher resolution Christchurch retrolens stuff as Qgis handles this well.