Please support slicer generated progress marks

zapta

Please allow to use the slicer progress marks instead of the existing 3 types of estimations duet currently provides, often with very large differences between them.

Here is a screenshot from slic3r prusa edition.

Phaedrux

I think there's more to it than just supporting the Gcode. Slic3r uses the marlin motion planner for their time estimates, and they are accurate because it's aware of the speed, acceleration and jerk values used for the Prusa printers.

slic3r doesn't use the RepRapFirmware motion planner and has no idea what your firmware speed settings actually are for the most part, so it's estimates aren't going to be as accurate. So if that's the case, what's the point of adding time remaining tags that won't be any more accurate than the existing DWC estimates?

zapta

@phaedrux said in Please support slicer generated progress marks:

... so it's estimates aren't going to be as accurate.

I believe that with a little bit of creativity it may be accurate. For example, by extrapolating the time so far for the remaining percentage. At the time I played with the estimation logic of sailfish (using S3D percentage markers) and got very good results.

Having displayed three estimations, often with large differences between them is kind of odd

dc42

@zapta said in Please support slicer generated progress marks:

Having displayed three estimations, often with large differences between them is kind of odd

The point about displaying 3 estimations is that when they converge, you have a reasonable idea that the times are accurate. However, I am inclined to remove the estimate based on layer time, because it is rarely accurate.

If you simulate the file before printing it, then DWC 2.0 offers another estimate, based on the simulation time. This is normally very accurate, unless you adjust the speed factor or other relevant parameters while printing so that they are different from the values used while simulating.

I will consider supporting M73 in a future firmware release.

zapta

@dc42 said in Please support slicer generated progress marks:

If you simulate the file before printing it, then DWC 2.0 offers another estimate, based on the simulation time.

Have you considered to use statistical sampled simulation for a prior quick estimate? E.g. sample the file at a few places and extrapolate the overall secs/byte or second/layer ?