We use cookies to personalize content, interact with our analytics companies, advertising networks and cooperatives, and demographic companies, provide social media features, and to analyze our traffic. Our social media, advertising and analytics partners may combine it with other information that you’ve provided to them or that they’ve collected from your use of their services. Learn more.
Phil, very hard to tell from most posted photos, but viewing your results I think your focus may be off. If you did not “fix” your lens in place after focusing, it may have gotten moved or bumped, or your Z axis height is not reset to the original focus height By “fix” I mean prevent the lens from being turned with a dab of hot melt glue on the threads to the laser housing. I use a piece of wood cut to 3 inches long to set my laser to a repeatable distance above the wood surface so I do not have to fiddle with refocusing all the time.
Poplar engraves very well, and I think you should be getting better results with it.
John, I was thinking the same thing. I wasn’t too sure but the focus looks to be slightly off to me as well. Of course I could never get the stock lens to focus better than that anyway. I had to swap mine out. If my machine wasn’t down I would offer to run one of those images for you Phil and see what it looks like on my machine.
Also the hot glue trick saved me headaches on several occasions.
I am currently using the ebay one at the recommendation of @picengravertoo. I love it. I can get a much better focal point with this one. Don’t get me wrong, the Jtech lenses are top quality. I just found it overly difficult to focus the oval shaped beam. Mostly due to my not so stellar vision.
I’ve swapped mine out for that eBay one too but currently waiting on a replacement as the one they sent must have the order or direction of the lenses wrong. I get about a 1”-2” wide rectangle, only when the lens is falling out is it about 3/4” wide. Hope this replacement works but it’s been slow arriving and they didn’t provide tracking. I only went with the eBay one after seeing it recommended on here a few times, didn’t know about the JTech one
I’d say about 3.5” as I set the distance between the lense and the wood to 3”. I recall trying around 4” but nothing closer. The lense fully screwed in the rectangle was about 2” wide and when all the way out but just holding on by a thread it was near 3/4”. I was never able to get a fine spot size. Seller said that they’d send a new one since that was not correct. I didn’t try installing it backwards as the slot looked like it’d catch on the spring and cause damage.
I found that I could focus much better on black like aniodized aluminum then wood. Not sure if that would help you verify the focus. I did find that I could do better than what I had been thinking was great.
@DavidLovely If you get a chance try setting the distance between the housing and the wood to 3 inches. I had some confusion regarding this as well. The lens will only be in by a couple of threads, hence the need for hot glue to hold it in, but you should be able to get a very fine dot. When I took mine down for upgrades I was running at .006 resolution.
@PhilJohnson Are you running yours on a 45? Since the Jtech lens projects in an oval, if the oval isn’t parallel to the engraving angle it can cause what look like like a focusing issue.
Dithering is just B&W dots, no gray. Stay with grey scale. Try .006 pixel resolution, but no smaller I think. If the lines touch, image will get way too dark IMO. Also, bump up the min depth some.
Phil, having zero experience on laser etching, let me comment on the photography side.
The big difference between the picture you tested and the picture you wish to print like lies in the midtones; or the lack thereof in your case. There’s two reasons that directly come to mind.
(a) Lighting. Yours looks frontal and harsh (i.e. small source). The Christmas baby’s is off-camera (looks like Rembrandt lighting) and using a soft source (large softbox or octabank). This produces a very rich tonal gradation.
(b) Post-processing. Again, having not seeing the output picture but just judging from the etching, yours looks like very high contrast. I suspect your historgram looks like a comb instead of a continuous curve. A comb with most data cramped on the two far ends of the spectrum. It’s very easy to get carried away with post-processing and start losing levels, especially if you process at 8bit/channel. A histogram that looks like a comb (i.e. has lost levels) produces posterization when printed.
Having said that, everything is a matter of preference. There’s nothing wrong with high contrast, if that is what you are after. But it seems that in this example you are not.
Just keep in mind that from an image processing aspect, this is a 3 stage process:
capture
processing
output
Tweaking with the output settings will not give you back any midtones if they have been lost in the processing stage or not even recorded in the capture stage.
PS. You do not need to go buy an octabank if you wish to experiment. Try using a northern window.
Your capture is very rich here (clouds are the largest softbox you can wish for!)
Your first monochrome retains most of the luminance channel, nothing critical seems lost (the water reflection registers as a specular highlight in our brain, it does not bother).
Your second monochrome starts losing some relevant highlight details (cheeks). If you were rim lighting an athlete, it would look cool. If you were shooting a bride, her dress would look like paper and she would not be happy with you.
It’s tricky cause it seems you want to light up the eyes (midtones), but end up lifting the highlights too. Here’s some ideas:
If possible, correct at the source. E.g. next time you shoot something similar try using a reflector in front of your subject as fill-in.
If you have a camera that shoots RAW, you will be processing 16bit/channel (and be blown away by the tonal detail, not to mention the dynamic range and the lost highlight reconstruction possibilities) and then output your final image at 8bit/channel and do not touch it anymore.
If (8bit) post processing is your only option (cannot re-shoot, no RAW), try curves. Gently. Put a pin at the last quarter of your diagonal (to protect the highlights), maybe another one close to the bottom left to prevent your shadows from washing off and then try pulling the tones you wish to correct slightly (up in this case). I haven’t used gimp in a while, but I believe there must be a tool to hint you which exact tones you need to target (placing the colorpicker on the eyelids for instance).
Here’s another triplet of bullets (what’s happening today, everything I write comes in a triplet of bullets) in ascending order of difficulty but also of output quality and possibilities.
The idea behind the last two is the creation of masks based on the luminosity information of the image itself. Masks created this way are self-feathering (freaking genius).
There are a couple of different things at play here:
Encoding: The files stored on disc for 8 bit are usually in sRGB space, not linear - that means “128” isn’t 50% gray (it’s closer to 180). If the engraving software doesn’t deal with this you might have to de-gamma them first.
Burn: not all materials will darken linearly with the application of heat, and not everything will darken the same way as anything else, either. You may need a different “pixel brightness to burn power” curve for each type of material you use. If you dial your settings in on birch ply, you may need to re-do them for cherry, alder, maple, slate, card stock, etc.
Lots of materials don’t darken at all over the first few percent of the laser output, then darken rapidly for a bit of the range, and then the burn rate slows down after that - you’ll get the best results if you can figure out a good min and max for power first, then dial in the curves after that.
What you are describing here is the very essence of color management. This is how a CMS system creates input and output ICC profiles for different devices (cameras, printers, monitors…) . By feeding the device to-be-profiled with a known set of values and then measuring their output and interpolating everything in between.
Only in that case this is done by software and the measurement typically involves a colorimeter (for incident measurement - e.g. for monitors) or a spectrophotometer (for reflective measurement - e.g. for prints).
From the discussion here I imagine that the “profiling” in laser etching is done manually by the user and the measurement tool is the eye. This is pretty imperfect as our brain is biased by environmental factors (light and color temperature of the room we sit in, adjacent tones, …)
A burned matrix of grayscale patches on wood falls in the reflective categorey, so I imagine that a spectrophotometer could still be used (at least in theory). If you can write the profiling software and then enable color management (i.e. support of “wood” icc profiles) in your burning software, then you might be up to something very useful (and profitable). Can I keep 1% for the original idea? Phil won’t ask for a cut, he’ll be happy to provide the testbed.
