With this update I'm starting writing in english, as I promised.
We can start with this object, is a vase found on thingverse, I have added a fibonacci color gradient in openscad to test the swiching color.
I printed this at perimeters 65mm/sec and infill 0, 3bottom layer, it's reduced by 50% from the original thing.
It took 1h and 12 min to comoplete the print.
I have used the basket and an optimized "tool change gcode" in order to have the "next color" ready when the print request to change color.
At the moment it's impossible to change the color on the same position while the printig goes on, because different colors have different characteristics (diameter, viscosity ...temp of extruding)
I have tested this with 2 colors coming from 2 different suppliers, the result of the test is a not good final printing! and you can easily see on the printing when the colors have changed, because some times the material is not enough and some other times it is too much... ( I think that's because the characteristic of the 2 filaments from 2 diffentent suppliers are different!|)
So... No more problems ! Use The Basket!
I don't want you to be unable to use 2 different (or more in the future) colors from 2 or more different suppliers!
You have 2 colors... you can always use the F.U.M. (FalaUnionMono) !
...I have improved the tool change gcode, so when the printer request the filament to change, the FUM retracts the previous filament and when the next filament comes inside the hot end, the FUM is over the basket, and it over goes the 3mm the normal lenght of change.. and after a little retraction goes restarting to print!
(all this takes 6 seconds... versus the 5 seconds if you change the color on the same place...)
In a print like this I have noticed that the requested color arrives with a dealy of 2 layers, that is because the melting zone is quite big and I'm using a 3mm filament at the moment, but you can arrive at some mm if you reduce the temp and the printing speed...and increase the extra extrusion of next filament into the basket..!
I think about reducing this transiction delay in the next improvements...