Improve sample pattern - Githubissues

maresb commented 5 months ago

--sample-pattern 64 on 9mm tape on LabelManager PnP:

--sample-pattern 128 on 9mm tape on LabelManager PnP: (This makes it easy (for me) to count backwards from the 80 and see that there is a faint row 65.)

Horizontal lines now come in fours. (At least for me this makes it easier to count the four lines.)
Horizontal lines are repeated both at the beginning and the end.
The zero-indexed y-value of the last pixel in a box is printed when it fits.
There is an extra width-4 stretch of 1×1 checkerboard.

To accommodate this, I rendered the test pattern as a HorizontallyCombinedRenderEngine. In order to fit the bitmaps into the HorizontallyCombinedRenderEngine, I had to wrap the Image class as a _ImageRenderEngine since the PictureRenderEngine takes a Path object instead of an Image.

CC also @FaBjE for review

tomek-szczesny commented 5 months ago

So this is what it looks like when it is handled by capable hands! :)

I think that counting dots beginning with 1 would make counting easier, and the values more rounded. :) I don't think this would cause any problems in the kitchen. After all the test print is to determine how many pixels are missing on each side.

Other than that, perfect.

maresb commented 5 months ago

@tomek-szczesny, good point! My new favorite Turing test is, "count to 10". :joy:

FaBjE commented 5 months ago

Fine with me 😅 Maybe you can add little howto document so I can link to it from the comment (in the code)?

maresb commented 5 months ago

I need to make another set of photos for my guide after adding the h=64 text.

maresb commented 5 months ago

Partial draft of a markdown doc

maresb commented 5 months ago

@FaBjE, thank you so much for your review, your comments are extremely helpful! I hope to get to it soon. (I'll also try and get to #51 soon.)

In order to be able to describe the general calibration procedure, would you be able to provide a few examples using this new test pattern with your 9mm tape? I'm thinking perhaps one test pattern with 128 px, and then another with 93 px? (My idea is that the top of your tape should be at pixel #91, so with 93 px there should be two rows cut off from the pattern on the top.)

FaBjE commented 5 months ago

@FaBjE, thank you so much for your review, your comments are extremely helpful! I hope to get to it soon. (I'll also try and get to #51 soon.)

In order to be able to describe the general calibration procedure, would you be able to provide a few examples using this new test pattern with your 9mm tape? I'm thinking perhaps one test pattern with 128 px, and then another with 93 px? (My idea is that the top of your tape should be at pixel #91, so with 93 px there should be two rows cut off from the pattern on the top.)

No worries.

I have ordered 6mm and 19mm tape, they are in transit to me. I expect them in about a week. Maybe for the sake of completeness I can print a series (128px test pattern on each tape size). I can do a test-print with 93 px as well. Just send me a list of tape-size / pattern size you want me to test.

Practical issue though, the branch you are working on doesn't support my printer (yet 😉) I can maybe rebase it upon this branch of mine: https://github.com/labelle-org/labelle/compare/main...FaBjE:labelle:feature/addLabelManagerPCII That is just original codebase + my printer ID added. It can´t print proper labels, but it should be OK for the test pattern I think?

maresb commented 4 months ago

I just made another very quick pass on this. Still todo:

[x] Proper hands-on instructions as per https://github.com/labelle-org/labelle/pull/55#discussion_r1650387230
[x] Update the images
[x] Explain how to fix a dirty print head

I just produced a printout with some defects which I believe are due to dust on the print head, so I'll take the opportunity to explain cleaning.

maresb commented 4 months ago

This is now much better thanks to your feedback @FaBjE! I have some structural changes in the works in order to bring develop back into main. If your tapes arrive, then it's probably easiest for now to create a branch merging this into develop so that you can print the updated pattern.

maresb commented 4 months ago

I'm pretty happy with this now. I'm happy to receive any further feedback that anyone may have.

Follow-up work for future PRs:

Add cross-reference to the device specification in the source code once develop has been merged.
Add example photo of test pattern for a tape that is significantly smaller than the print head size.

maresb commented 3 months ago

Merging and releasing this so that we can use it for #62

labelle-org / labelle

Improve sample pattern #55