Repetier Host Mac Time Estimates

[dpmacri]

I'm running Repetier Host on a Mac (version 0.56) and the time estimates it gives when I start a print job seem to be way off. If it says 30 minutes, I can usually expect an hour or more. Has anyone else seen this and is there a known reason? I'm leaving the speed multiplier at 100%, so it shouldn't be due to that (although I would also think that RH should adjust the time accordingly if that's changed).

[Jimustanguitar]

I don't know this for a fact, but I've been told that Repetier calculates its progress based on what line of G-Code it's on. The time that each command takes can vary considerably (hence your observation). It's just telling you how far down the list of instructions the machine is.

[dpmacri]

Hmm, that's kind of lame if true. Seems like since the "speed" of command is know that RH should be able to do a pretty good calculation of the time required. I'll take a look at the source when I get a chance

[Polygonhell]

I'm running Repetier Host on a Mac (version 0.56) and the time estimates it gives when I start a print job seem to be way off. If it says 30 minutes, I can usually expect an hour or more. Has anyone else seen this and is there a known reason? I'm leaving the speed multiplier at 100%, so it shouldn't be due to that (although I would also think that RH should adjust the time accordingly if that's changed).

Yes it's usually at least 2x the estimate.
I suspect all it does is look at the number of lines it's executed and estimate based on the number of lines left, that ought to be pretty good, but there appears to be something off there.
It's not a trivial task to give an accurate estimate because there is printer specific information not available to the Slicer/Host software. When a line of GCode specifies 60mm/s, it's not actually printed at 60mm/s because the firmware controls how the head accelerates to the requested speed.

[Lochemage]

One thing I've noticed in KISSlicer, in the speed settings there is an option where you can input your printers acceleration speed. The tooltip on that option specifically states that it is only used to calculate the estimated time of print. I don't see why this, and possibly other printer settings, can't be added to the host software to help with the estimation (or even have the host take a look at the EEPROM settings for that info). It seems like it could calculate drastically better estimates than it does now.

On top of this, it could further improve estimates by 'remembering' the last 10 or so print results and sample an average error percentage that can be applied to any current estimates. This would make any future estimates smarter.