My experience with a small home 3D printer has been that that normally, there is tweaking that needs to happen on a per-printer basis to get good output. My feel – and maybe I’m just out-of-date here – is that I can’t just expect to take a 3D model, feed it to any old 3D printer, and get output that’s just fine. Maybe things have changed in the past couple of years.

Paper printers are to the point where one can expect to get pretty much the output one expects by just feeding a file to a given printer.

I’d like to be in a world where I could just go down to the office supply store and use their expensive 3D printer the way I do their expensive paper printer. But my impression is that we aren’t there yet.

My guess is that at the very least, if we were in a world where that happened, there’d need to be printer description files that would list a printer’s and filament’s capabilities very precisely to the point that their output could be reliably and closely predicted for a given model, and that I would be able to see a “3D print preview” of what the printed output would look like in advance.

But last time I was doing much 3D printing, I was still doing things like tweaking nozzle traverse speed and tip temperature manually. And I don’t think that that sort of thing is going to be the user’s manual responsibility in the World Where Everything Just Works Automatically. The software would need to have an accurate picture in advance of what impact those settings would have for a given printer’s environment, without any actual testing for an actual 3D print.

EDIT: Also, my guess is that part of that would need a print to either be done in a temperature-controlled enclosure to reduce impact of the environment around the printer, or at minimum have a printer able to measure ambient air temperature and maybe airflow and use that as an input to adjust print settings – that is, even for a given printer and filament and model where the properties of all three are very precisely measured and modeled, there are still variables that are unknown until print time that can impact output on non-enclosed printers.

EDIT2: I guess what would be necessary is for someone to come up with a standard series of calibration models to 3D print, then for people to build a database of “printer description files” based on what those prints look like that model the impact of things like:

• Printer

• Filament

• Tip traverse speed

• Tip temperature

• Slice height

• Tip blower fan and speed

• For printers without a climate-controlled enclosure, ambient temperature at time of print

• For printers without a climate-controlled enclosure, airflow around printer at time of print

In theory, with enough information, I’d imagine that a printer could do things like use different settings for infill (or any non-visible surfaces that don’t need to be as pretty, just structurally sound) and externally-visible surfaces, which I’m pretty sure wasn’t an option in the software I used when I was last doing prints. Having little threads or rough edges inside infill isn’t going to matter, but having them hanging off exterior surfaces may not be.

It’d be enough information for software to provide an estimate of structural strength, using the above inputs, when the printed object is placed under various types of forces (e.g. crushing force from the top, the side, repeated light force, etc). To estimate how rough an external given surface is going to be. To be able to say how much of an overhang is viable.

But I was still trying to figure out good tradeoffs between print speed and quality and leaving little hanging threads and surfaces sagging and all that manually…I was a long ways from that “plug in a printer description file and the software just computes everything”. The filament that shipped with my printer worked fine with one set of settings with the sample model that came with the printer. What worked fine with that model didn’t work well with 3DBenchy, that famous boat model. New filament that I got required new settings and testing. My first meaningful project had other constraints.