Creality Slicer

[rmazzi]

I though it would be best to start a new thread on this topic. I had posted to a differrent forum Ender 5 S1 - Will it be “orphaned”? but the topics seems better matched to this forum.

I have an Ender 5 S1, which does not have an available profile in Cura 5.3.1. When i got started with the printer, i used the Mac version of the Creality Slicer and got good results, but the program randomly closed after I sliced and saved a file. This concerned me, so I set up a profile in standard Cura using the setting suggested in the MwT video. The result was a much longer slicing time, with the printer pausing frequently in the center.

Then I tried to use the Win version of the Creality Slicer and found two versions. One with E5S1 in the file heading, and another without it. Naturally, Creality does a lousy job with information on the releases, so i picked the E5S1 version and tried it out. Its performace was poor as well.

I compared the Gcode files for the various attempts and found a couple of differences. First, the G29 command seems unnecessary and slows down the print by several minutes at the front end. It seems the E5S1 remembers the stored mesh reliably. The second different seems to be in the G0 command early in the code. For the Cura 5.3.1 version it was set with F6000. For the Win version, it was set for F9000, and for the Mac version (fastest estmated slicing time), it was set for F15000.

I have been unable to find any online sources that clearly describe if or how this setting effects the print speed. I know the feed rate and flow rate setting are % based, so infer that this % is applied to the F value? Does anyone know how this works or if this is a red herring?

Since I only have Creality printers, the creality slicer would be a solution, but i don’t really understand how it is different than cura other than statements on the web that say it is a cura branch from Creality and it is “optimized” for their printers.

[Geit]

In my opinion profiles are overrated. For my printers there was never a proper “profile”.

I installed cura, set up the printer dimensions, fixed the Nozzle size and done.

I guess the main issue with your setups on different systems is, that there is no proper centralized source code repository for the config files, so the windows version, is different from the mac version and the linux version.

Someone changes the default movement speed and boom you have F9000 on that version, while this change did not apply to other versions. Of course this is only an assumption. Even so a logic one, especially when not cross compiling all versions on the same system.

The whole profile stuff is quite useless and not worth to deal with. Take a printer that matches your printers dimensions and you are good to go. If the movement speeds are to fast, change them.

You need to change them anyway in the future. As soon as you modify something, like e.g. a new fan duct, you need to change the tool head size. Start and end gcode is not complicated. It is just defining how to home, performing a bed level, turn heaters on and off and so on. Again you want to change those settings eventually to make the printer behave before and after each print. e.g. i have my printer pushing the print bed to the front after homing, so I can easy reach the printed object. Once you made a change.

A profile only helps once.On the printers first usage. Any change you make later on gets lost with every update of that very same profile, so you want to create your own one, to avoid these unwanted config resets.

Of course you can install pre-configurated files from Chep or DrVax, but in the end you have settings activated without even knowing. Again you will run into problems, because your printer got optimized for a use case, which infers with your use case. There is no “turn this on and the printer prints perfect” option. You need to decide which options to use on a model basis. Starting with the optimal orientation ending with print speed and type of support and infill.

The default settings in Cura are working with any printer. All it requires is to learn what option to tweak to get better/faster/stronger results. Slicers have a 1000 options, but in the end you mostly ever need to change 20 at max.

Speaking of the Fxxxx values. Marlin has build in limits, so you printer is probably maxed out anyway. Especially the F15000 will likely never be used as you printers firmware will use a fixed lower value. The creator of that profile probably just used this high value to trigger the max-speed preset by the manufacturer, as that value is (should be) within reasonable limits and cannot cause any hardware damage.

Another good reason why profiles are kinda useless as speed values are defining the quality of the print and (depending on the hardware) you may not want to max out stepper speeds to gain more print quality.

[rmazzi]

Thanks for the clarity. I have been using profiles and settings interchangeably. Aside from the basics of dimensions, the other variable in the profile is the start and end code. I am still puzzled. The profile, start and end codes do not have anything in them to effect the efficiency of the slicing. Yet, i got very different results between a Generic cura file and the creality file. At least the estimates were different by about an hour. However, I did not do a complete print for the other slicers. Perhaps the estimates are grossly inaccurate. As a note, even with the Mac Based creality slicer file, the print actually finished a good 30 minutes faster than the estimates.

I am concluding that something is different the settings, but i have yet to find it. Any suggestions on where it may lurk or is it possible that Creality does optimize the slicing code itself for their printers? I suppose in theory they could since Cura is open source.

[rmazzi]

Pretty sure i have gone down a useless rabbit hole with this. However I would like to know if I am on the right track with some of the critical cura slicer settings if i want to optimize speed with quality, depenting on the print.

1. Bigger Layer height is faster, but surface is less smooth
2. Infill is a trade off of speed and strength
3. Speed and acceleration settings matter and could be adjusted based on the type of the print. e.g I printed a thin tube 2.4 mm thick with a wall thickness of 0.8 mm and 2 layer thickness, so wall speed was about all that matters on that one. No room for infill
4. Retraction and temps would matter if the print was traveling in mid air to the next point. On my current prints this did not matter since they were basically flat plates.

It seems if I manage these, i am a good way there and can build off of that.

Out of curiosity, my rabbit hole took me to comparing the G codes. How does the FXXXXX parameter get set? Which printer setttings drive those numbers. I noted that they changed periodically throughout the code. The speed settings in cura are mm/s. I thought the F parameters were the same units, so i don’t see how the parameters all range from 1800 up to 15000 yet speed was set at 120 mm/s or less in cura.

[Geit]

You cannot compare the results of different slicers. They may base both on cura, but probably not the same version. Then in addition all settings need to be equal.

The slicer results also depend on the printers firmware. It predicts the print speed on the speed set, but the firmware may reject to speeds.

If Slicer #1 is using F15000 it calculates the print time using that 15000 value. Slicer #2 is using F6000 and using that. The result is a slower print time and the firmware limits to F5000, so the real world result is even slower.

You need to read out the movement speeds of your printer using the gcode dump commands. Then you need to apply these values to your print and move speed settings. Now the estimated print time is more or less correct on all slicers, if the rest of the settings is the same.

AFAIR there are plugins to read the printers config from marlin and apply the print speed limits to the printers config.

[Pigjes]

I would even go as far as to say that tweaking a “printer designated profile” even depends on the printer itself, not just the brand and model, but the one you have in front of you. That really surprised me a lot. My current printer is a perfect example of that, as the 2 before faired well with all the tweaks that people gave online, now it was a search to make it work even.

[Alan]

“Out of curiosity, my rabbit hole took me to comparing the G codes. How does the FXXXXX parameter get set? Which printer setttings drive those numbers. I noted that they changed periodically throughout the code. The speed settings in cura are mm/s. I thought the F parameters were the same units, so i don’t see how the parameters all range from 1800 up to 15000 yet speed was set at 120 mm/s or less in cura.”

When the printer moves an axis from one point to another, it will do so with a speed set by the slicer. Typically denoted as a speed (mm/s). The movement will not be from 0 to 100% instantaneous instead two other parameters come into play, Jerk (mm/s) and Acceleration (mm/s^2).

So, jerk is the speed that the motor reaches as soon as a move is given. For example, if jerk is set at 15 mm/s with a print speed of 60 mm/s, the printer will not do 0-60 mm/s instantly. It will go from 0 to 15 mm/s immediately. Then, after an acceleration, it will reach the set speed as seen in the figure above. Both jerk and acceleration limits are set in firmware and can be set in the slicer. When the slicer and firmware values differ, print time estimates can vary greatly from reality. A desired feedrate may never be reached if the distance is small.

Cheers

[rmazzi]

Thank you all for the education. Based on your input, I went back to square one and found some errors in my settings, corrected them and have a good working profile for my Ender 5