Calibration in the slicer is a great idea - lets make it faster

[julie777]

During the last two days I ran all the calibration for one filament. It took many hours, and that is on a P1P. There has to be a way to speed it up.

My first thought is based on my experience as a software developer and a troubleshooter. Instead of looking at everything break the problem into parts and see which part has the problem (binary search).

What if a test such as the flow rate test, which has two passes that each take a very long time to print and print many objects, could be reduced to more steps with much fewer objects. I am not sure that three per pass will work, but it will make the idea clear.

Currently pass one prints nine samples for analysis. If only three were printed it would do 2 things.

• reduce printing time by 2/3rds
• make it easier to compare the samples because the differences are larger and easier to spot. (I had a very hard time choosing the best print when they were all so close together.)

Choose the worse end to discard and the next step prints 3 more, but these are bounded by the best two.
Again saving time and making comparison easier.

(I know you are probably thinking about all the calculating that you will have to do. That is addressed in another discussion about making the calibration process easier. #701 )

Currently there are 19 test pieces printed. I think it might be possible to reduce that by at least 1/3rd and maybe up to half using the above method.

Some of the other tests might be able to use a similar strategy.

Instead of just implementing tests that someone out there has done, let's take time to think about it and how it can be done more efficiently.

Another example is the max flow rate test. The default is to print from 5 to 20 by .5. That is a lot of steps and the test uses a very long path. Is that many corners necessary, is that long of a straight part necessary, should there be different kinds of corners?

I can see using just three flow rates per print would massively reduce the printing time, and allow narrowing in quickly. (I plan to try this and report back on this one.)

Maybe it takes four points and not three to make a decision. We won't know until we try it.