feature req: Add accurate estimates

[mark-hahn]

I find the Cura estimates are usually accurate to within a minute. OctoPrint estimates are way too long, sometimes 200% wrong, until there are only 5 to 10 mins left. And there is no estimate at all when the print is started, which is when you need it the most.

Can anyone steal the Cura code? This ETA would be great if it gave useful info.

[pablogventura]

Thank you very much for your feedback!
Estimating the time of an impression is a difficult task. Sometimes it depends on configurations that are not in the gcode, but in the firmware of the printer (acceleration, jerk, etc.).
My plugin uses the time calculated by OctoPrint. I wouldn't know how to use the time calculated by Cura here, but any idea would be very welcome!

This topic is discussed in this issue of Octoprint: OctoPrint/OctoPrint#270

[mcarpenterjr]

I have recently begun using Octoprint and I have noticed there are differences in either the estimated time or the print time between these 3 scenarios; Printing locally from Octoprint, Printing from the SD Card from Octoprint and Printing directly from the printer's SD Card.

Typically printing locally from Octoprint takes the longest/has the most inaccurate time estimate. Printing from SD through Octoprint takes close to/is reasonably accurate when estimating time and printing directly through the printer from the SD card seems to be the most accurate.

I haven't really taken the time to focus on actual print time's these are things that I ultimately suspect. What I may do when I have some time is run a 3DBenchy using each method and independently time the execution and compare actual runtime with the estimated time.

[mark-hahn]

Just compare the times to Cura. It has been dead-on in my experience.

…

On Sun, Mar 4, 2018 at 7:35 PM, Mark Carpenter Jr ***@***.***> wrote: I have recently begun using Octoprint and I have noticed there are differences in either the estimated time or the print time between these 3 scenarios; Printing locally from Octoprint, Printing from the SD Card from Octoprint and Printing directly from the printer's SD Card. Typically printing locally from Octoprint takes the longest/has the most inaccurate time estimate. Printing from SD through Octoprint takes close to/is reasonably accurate when estimating time and printing directly through the printer from the SD card seems to be the most accurate. I haven't really taken the time to focus on actual print time's these are things that I ultimately suspect. What I may do when I have some time is run a 3DBenchy using each method and independently time the execution and compare actual runtime with the estimated time. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#1 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAxhv0_cQTNMI3w4lVx5sRPqDvL6PdOxks5tbLKZgaJpZM4RKXkc> .

[mcarpenterjr]

I just ran a print that Cura estimated at 4hrs and when I uploaded and printed locally from Octoprint it ran closer to 6. I have some suspicions and I'm going to dedicate some time to figure out what's going on, it might be an Octoprint thing or it could just be me.

[mark-hahn]

It might be that my settings are lucky.

…

On Sun, Mar 4, 2018 at 8:00 PM, Mark Carpenter Jr ***@***.***> wrote: I just ran a print that Cura estimated at 4hrs and when I uploaded and printed locally from Octoprint it ran closer to 6. I have some suspicions and I'm going to dedicate some time to figure out what's going on, it might be an Octoprint thing or it could just be me. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#1 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAxhv_hBsgYu5yBNNZZX3TgnN2QXPLnHks5tbLhFgaJpZM4RKXkc> .

[mcarpenterjr]

This evenings print, Cura Said 4.5 hours,
Octoprint Says...

[mark-hahn]

Notice that octoprint says 3.5 total and then says 5.5 while printing. Cura's 4.5 estimate is halfway between. My experience is octoprint estimates too long and then shrinks the time as it gets near the end until it matches cura. ​