Will small layer heights (0.07mm or 0.05mm) work well with larger nozzles (0.8mm)?  

Tuning matters less with vase mode but can still be a factor

Posted by: @thyraz

[...] Isn't adjusting over-/underextrusion vs adjusting the line width the same in vase mode (once we passed the first few solid layers)?

The end effect is the same, yes. The difference is that, with an accurate extrusion multiplier, the slicer can ensure a consistent volumetric rate is applied as opposed to just squirting out whatever amount of filament is actually passing through the nozzle. Flow consistency is very important for exterior surface finish so the more closely we can help the slicer model reality, the better the job it can do producing gcode that yields even results.

Outside vase mode, line width most likely adjusts the amount of extrusion and the horizontal distance between the lines.
But there are no other extrusion lines next to each other in vase mode.

Extrusion width is one of 3 parameters (width, height, speed) that determine the volumetric rate of plastic being pushed through the hotend. Push too much and you get skips and jams. Push too little and you get underextrusion and gaps. Here again, the more accurately the slicer settings reflect reality, the closer printed results can be expected to resemble our slicer results.  There are no adjacent perimeters in vase mode, true, but if the amount of pressure is inconsistent, there can be variations that can pile up and cause those blobs you often see with low layer heights. 

So the whole reason why we can print with a larger width compared to the nozzle diameter, is that we over-extrude.
Which means, that we based on the distance to the line below (= layer height) extrude more material than we need to get a line with the same width as the nozzle diameter.

I'm not sure exactly what you mean, but so long as your extrusion width is within ~200% of your nozzle width and you're using quality nozzles (the B dimension on the E3D nozzles), you can get reasonable results. If you go beyond that 200%, filament can expand above your layer height (it squishes out past the neck of the nozzle) and start to pile up. If you are using wider extrusion width, that extrusion multiplier is important to make sure you don't extrude enough filament to go beyond that dimension.

That means, that more material is pressed to both, left and right which results in a thicker line.

Just to be clear, a wider line... unless you exceed the nozzle dimensions in which case, a line with rough edges.

But the more you do that and the thinner the layer height is, this might get problematic (how much oozes to the left, how much to the right).
So maybe this really results in the problem I see.
It looks like the layers aren't perfectly aligned.
Like a tiny, but very irregular z-wobble with a rough finish.

That and possibly a bit of filament piling up and catching the nozzle or creating little blobs. Again, that low layer height really magnifies any issues. It's a very very tiny gap between the nozzle and the extruded filament, and extruded filament does expand.

@Joan could be that this gcode increases the problem.
If my thoughts above are correct and over-extrusion and line width is the same, this might result in something about 0.84mm instead of 0.8mm.

Some non-zero amount in any case. Again, the idea of the calibration is to reduce such deviation. Setting your actual filament diameter and extrusion multiplier helps the slicer out.

But I don't think there's a magic number where it starts to completely fail and before that everything is completely fine.
This will be more of a soft transition.

Definitely a range, IME. I just want to give the slicer as accurate a base to calculate from as possible rather than try to work around variables. I can say that higher layer heights are easier to work with, and there's a point of diminishing returns when you start to go below 0.1mm.

So if I have to reduce the extrusion multiplier by a lot, I will end up with thinner walls than 0.8mm.

Hmmm... No, more like if you calculate the correct (reduced or increased) multiplier AND filament diameter, the slicer is more likely to produce gcode that produces a 0.8mm wide extrusion on your print that corresponds to the 0.8mm wide extrusions you're seeing in the slicer.

Might be interesting how far I need to reduce it and then measure which width is possible with 0.07mm layer height and a 0.4mm nozzle.

You can just enter some higher and lower values and observe the difference, sure. I'll eyeball it on occasion. I do like to spend 5-10 minutes with a new type of filament doing some quick tests and adjusting my settings though.

@bostro I already printed a lot things without vase mode with 0.05mm layer using this filament.
As the surfaces and the seals of my last prints have been fine, it shouldn't be completely wrong.

If it's working, then great! I'm just noting that thinner layers can be achieved with a larger nozzle, but a bit more tuning may be required.

But as stated above: If i understand vase mode correctly, overextrusion should't be that big problem in vase mode.
As there are no already exisiting layer lines on one side of the nozzle, this shouldn't result in the typical over-extrusion effects.

Within the confines of the nozzle dimension and flow variations, correct. I'm just trying to reduce variables and guesswork.

Thank you also for your great website, I already read a lot of the articles in the last 2 months when I received my first printer (Prusa Mini+).

Glad it's helpful. I'm a big fan of larger nozzles, so please don't take any of this as discouragement. I've been doing a lot of testing using a single, wide perimeter to quickly produce prints that are as strong as prints made with multiple narrower perimeters. "Big chonk" printing is fun.