Day 12 of using the trial version of LB and I’m finding a lot of great features.
I’ve noticed something odd. LGB is faster.
At first I thought it was the cold weather we’ve been having (unheated workspace) but I began to wonder so I decided to test things out.
I used;
The same art
The same piece of 3mm plywood (stored indoors)
The same power, speed settings in both softwares
I set the power pretty low on purpose to exaggerate the problem.
S-Value Max set to 1000
Air assist on
Constant power on both
The problem is that amount of passes differed to finally cut all the way through. LB took 3 more passes to fully cut thru the ply.
My next thought was maybe I was hitting a glue pocket or a knot, so I did the same test a few more times in different areas of the same piece of plywood, all with the same results as the first test.
Is there something different in LB that I’m not aware of that is causing this to happen?
It could be variable power. LaserGRBL does not have it. It is always in the Constant Output power mode.
I know you said you had Constant in both, but there has to be a difference somewhere. LaserGRBL is often recommended as a test comparison to Lightburn.
The first thing is to check the gcode. You can export the generated gcode from LB and compare it to the code generated by LaserGRBL. Check the differences.
Just FYI: LaserGRBL has both modes, it just calls them M3 and M4, which are the firmware commands.
Both G-Code files have S550, but LightBurn uses the GRBL-M3 dialect:
Machines set up with the GRBL-M3 profile don’t support Variable Power mode.
Setting both programs to match the value of GRBL’s $30 parameter and using the same GRBL dialect will help equalize the results. The GRBL-M3 dialect prevents varying the power, which may trip you when you want different levels.
The only difference I see between the G-codes is the M5 command, which should not affect anything. I don’t see why the laser should have different power output between both projects.
The only thing I would change is that you selected the grbl-m3 profile in LB, which is wrong. You should use type “grbl” for your laser. Otherwise, dynamic power is not available at all.
That is not the pass thru I was refering to. If you import a GCode file, you can only run it, not change the original design.
Export in mm, import in inches?
SVG is not an image file. It is a vector file. JPG, PDF, BMP, and so on are image files. NC, CNC, GC, TXT, GCODE and so on are machine program language files. All completely different processes. This may be where you are having difficulty.
Yes a vector file is an image file just like a raster file .
I’ve been a graphic designer since the 70s and programing since the 80s. I started with Corel Draw version 1 in 1990 then a few years later I switched to Aldus which then became Adobe. Uploaded my first app to Google play early 2015.
Everyone was new at one time. That is not a fault. I was thinking you were in one mindset based on experience and meant no disrespect. Q&A is all part of the diagnostic process.
No, they are not the same…
When scaling a vector graphic image, the graphic primitives that make up the image can be rendered using geometric transformations at any resolution with no loss of image quality. When scaling a raster graphics image, a new image with a higher or lower number of pixels must be generated.
What this is saying is that a scaled up image file has larger pixels, not greater resolution. Raster is just how the image is captured or displayed. TV’s use raster scanning. Oscilloscopes use vector scanning.
But did you export as Inches? Your numbers are approximately a 25.4:1 ratio.
Do not settle. Keep going at it until it makes sense. You owe that to yourself!
Even your explanation proves my point that a vector file is an image file. You are just confusing yourself without truly understanding the basics, what is being written or said by others.
FWIW, LaserGBRL only works in mm and only saves gcode files in mm. When I imported the .nc file from LGB into LB I know for a fact it was set in mm.
To confuse you even more a vector image file can contain a raster image. Test it out yourself. Open Inkscape or Illustrator then import a JPG file. Now save that file as a SVG, EPS or AI (adobe illustrator file)
Here’s a tip for you, if a AI file was created in Illustrator and you can’t view it just change the extension to PDF. This may or may not work if the AL was created in any other software.
It appears we are dancing around semantics here. For example…
This implies that a GCode file that can create an image is also an image file itself. If you save a JPG as an SVG, that file is converted. If I take that SVG, load it into Lightburn and create a GCode file, and burn an image, that does not make the last file an image file. We went from image to vector to control language, three (3) completely different file formats.
To be more exact, a vector file can contain an image file. This is true, but the image file cannot be recovered from the vector file. The image portion is converted and becomes part of the vector pathing.
I conceed you can call any file that recreates an image as an image file. But I prefer to make the distinction between image creating file and image format file. The first is an artistic definition. The second is a computer format definition.
I like this tip because I do not use Adobe Illustrator. Other software creating .ai files may not export using the same format standard. If these .ai files CAN BE imported into Illustrator, this implies Adobe exports using various formats, maybe defaulting to PDF. Again we get into the topic of file formats and conversions.
It was around Posting #13 that this thread went off-topic. We were discussing the difference in cutting results between LaserGRBL and Lightburn, plus an imported SVG size conversion. I did not see where either of these questions were actually resolved.
There should be no difference in cutting results between Lightburn and LaserGRBL if both are using the same settings.
If you are trying to compare two GCode files, there should be no attempt to edit or modify either one in either software package. Introducing new variables invalidates the test.
Summary: I am still wondering why you got different results.
