Printing through Ocotoprint is about 20% slower
I realized that when I print through Octoprint vs. SD card my prints seem to take about 20% longer than Prusa Slicer estimates. I didn't do an accurate test, but I recall that when printing from SD, the Slicer estimate was very close to the actual print time, but through Octoprint it's ~20% more. Does anyone else have this experience? I'm running Dashboard and DisplayLayerTime plugins (I have a few more, but these are the ones that I imagine could slow things down).
RE: Printing through Ocotoprint is about 20% slower
Easy enough for you to check. Put the same GCode file on Octoprint and on the SD card.
Have a run-off, using a timer. Let us know.
RE: Printing through Ocotoprint is about 20% slower
Only 20%? I am running my first print through Octoprint (Pi zero W) and It's giving me a predicted time of 5 days!!! It's a simple back cover for a project case which I would expect to take a couple of hours at most.
On a different subject, where does Octoprint keep its files? I had hoped to be able to upload to SD card but I cant see how to do it.
RE: Printing through Ocotoprint is about 20% slower
@drpeej
The Octoprint estimated print time is usually way off (unless you using something like the PrintTimeGenius plugin), so you can ignore it. When I say 20%, it's the actual print time vs. what Slicer estimated. I still haven't done a run-off, we'll need to get around it sometime in the near future.
To upload a file to SD through Octoprint, you drag the file to the "Upload to SD" button in the Octoprint windows. However, be warned, it's very slow since the upload goes through the printer's serial connection.
RE: Printing through Ocotoprint is about 20% slower
For putting files on the SD card, you will get far better results running sneaker-net.
Walk to your printer, take the SD card out, walk to your computer, copy the files to your card.
Then reverse the process. Data rate far exceeds what Octoprint can do.
RE: Printing through Ocotoprint is about 20% slower
I've never done a comparison to determine if printing from OctoPrint is slower than from an SD card, but I don't really care. OctoPrint is so useful and indispensable in my workflow I couldn't imagine running my 2 Bear Mk3s printers without it.
However, OctoPrint on it's own is horrible at predicting print times. On the other hand I've found Prusa Slicer to be incredibly accurate with print time predictions. There is a "Must Have" OctoPrint plugin that marries the two. The "M73 ETA Override" plugin uses the Prusa Slicer prediction mechanism M73 GCode statements to override the OctoPrint prediction mechanism.
I see there is now another one called "Prusa (M73 response) ETA Override" as well, but I haven't tried it. I assume it basically does the same thing. The one I have been using has proven to be extremely accurate for me and my prints that typically range between 3 to 36 hours.
This is the one I've been using: https://plugins.octoprint.org/plugins/m73etaoverride/
https://github.com/sysadminsh/OctoPrint-M73ETAOverride
This is the other one that I haven't tried: https://plugins.octoprint.org/plugins/prusaetaoverride/
https://github.com/kanocz/octopi_eta_override
RE: Printing through Ocotoprint is about 20% slower
@mdaneman
Thanks for that. I still don't know where Octoprint puts files while printing them. Presumably it's on the Pi micro sd card somewhere. As I get Pruserslicer to send the G-code direct to Octoprint I dont have any control over it. It works but I'm just curious.
Good news is that my 5 day print run actually took 3 hours which was much as expected! I have tried to install one of the ETA override addins but they fail to install with an error saying PIP is out of date. I haven't a clue what that means or how to fix it.