Engraving software comparisons...Duotones

[Joseph Tovar]

For anyone interested in a comparison of a few different filters/scripts which convert photos to duotones for engraving. Attached is a tile sample I created using 3 different methods. I don't have Photograv so I didn't test it. This was done with the cheap white tiles from HD and some Black Oil based paint.

1)Photoshop Andromeda Screens filter
2)GoldMethod Action Script for Photoshop
3)Photoshop Alpha Engraver II filter

By the way, do you notice the verticle lines on some of the engravings? Is this normal?

[Andrea Weissenseel]

My preference is the one with the Gold Method, the face and the background comes out best - I have Photograv but I prefer the Gold Method for my photo engravings.

I don't think that the vertical lines are normal. My Laser did that too in the beginning, and they replaced the drive on the X-axis and its gone since.

Andrea

[Joseph Tovar]

I agree with you. I think the GoldScript looks the best. My wife likes the Engraver one, but to me it looks a little blurry up close.

As for the verticle lines, they don't appear on any other type of engraving, only the tiles. Do you think it could be the settings? For my 35W system, I'm using 100%P, 15%S, 500PPI, and Density 6.

Thanks for the comments....

[David Harvey]

I think the Goldscrpt looks the best too. It's more even toned than the others. The lines?...maybe it has something to do with the angle of the screens...were they set to 0 degrees?

[Andrey Anfimov]

Gold Method are making sharpening of picture.
Make some sharp before Engraver II and you will have the same result as G.M.

[Dan Hintz]

Joseph,

In this case, the GoldScript looks significantly better than Andromeda (the Engraver isn't too bad, though)... in an earlier posting, the Andromeda plug-in looked significantly better than the GoldScript. Interesting...

For your settings, you're running at 500dpi vertically, but 1,000dpi horizontally. It's something to consider when it comes to banding issues...

[Dave Johnson29]

For your settings, you're running at 500dpi vertically, but 1,000dpi horizontally.


I am with Dan on this one. I have not seen banding since I set the same dpi for X and Y that the laser is set for or vice versa. It was so bad sometimes it looked like it had been printed on graph paper.

[Joseph Tovar]

Hi Dan,

You are right about seeing the Andromada look better on the other post. That was using laser tile, so there was no paint added. Maybe on the side by side sampe I did this time(using plain white tiles), I may have let the paint dry a little to long before wiping off and so the Andromeda pic held the paint better and looks a bit darker?

It's hard to try to get the pics exactly the same, since the filters and scripts do different things and at different resolutions...

For example, I started out with a 300dpi picture. I wanted to turneit grey scale then auto contrast before trying any methods, but the Engraver filter needs it to be RGB or something with multiple color layers otherwise it doesn't work. That filter doesn't seam to care about the dpi.

As for the Andromeda filter, I can convert to greyscale, auto contrast, but the filter complains that the dpi is too low. I was given a tutorial by ULS which says to make the photo 1000dpi before applying the filter. So, I apply the 1000dpi, bring it in, then convert back to 300 dpi....This might be why the quality is off compared to the others. My whole point was trying to bring them all into Corel at 300dpi for engraving.

Lastly, with GoldScript, I didn't do anything. I left it color and ran the script.

So..even though I started with the same pic, it could be because each method, needs a bit of tweaking to get them to work properly. This could be the discrepency between the images...

As for the banding, how can you tell it's 500dpi Vertically and 1000dpi Horizontally? Once I get it into the ULS driver, I just set it to 500ppi and let it do it's work. Is this something to do with the "Density" control? Is each level a different PPI along a different axis?

Lastly, I understand dpi when it comes to the graphic and sending it to an inkjet printer, but how does the dpi - ppi conversion work? Besides working with a nice hi-res photo from the start, is there a recommended dpi setting for sending it to the laser? If I use 500ppi on the laser, should I send it over at 500 dpi...?

Thanks for all the suggestions!

[Joseph Tovar]

I am with Dan on this one. I have not seen banding since I set the same dpi for X and Y that the laser is set for or vice versa. It was so bad sometimes it looked like it had been printed on graph paper.

Hi Dave/Dan,

I must have been typing up my other post as you were typing yours...I actually asked in my post above, but I'll ask again. How do you make sure the X/Y ppi settings are equal? Is it the ppi and density settings? I only have 1 place to put in a ppi value in my driver.

[Dave Johnson29]

How do you make sure the X/Y ppi settings are equal? Is it the ppi and density settings? I only have 1 place to put in a ppi value in my driver.

If the laser only has a single DPI setting then adjust the picture DPI to match that in Corel Photopaint before working on the pic. make sure both X and Y match what you are about to burn. I guess you can do that in Photoshop but I have not used it for a looooooong time.

Basically if your pic is at 135dpi then the laser should also be set at 135dpi.

[Dan Hintz]

I only have 1 place to put in a ppi value in my driver.
It's one of several minor annoyances I found on the ULS machines. You are allowed to set different resolutions in the X and Y axes, but they don't make it obvious that's what you're doing.

The ppi setting is dpi along the X-axis, and the image density slider determines the dpi along the Y-axis. The six image densities correspond to: 75, 150, 250, 333, 500, and 1000 dpi. Write those values down, they're not in the manual... yet another minor annoyance that's easily fixable, yet has never been done.

[Joseph Tovar]

...The six image densities correspond to: 75, 150, 250, 333, 500, and 1000 dpi. Write those values down, they're not in the manual... yet another minor annoyance that's easily fixable, yet has never been done.

Hi Dan,

I had email ULS last night about a cause for the lines and I also asked them about the Density setting this morning(before I saw the responses SMC responses). Here is what I received from them...

"..So if your image den is set to 6 then your ppi should be at 1000 PPI, Den 5 is 500, 4 is 333, 3 is 250, 2 is 200 and is 1 is 166.."

Everything seams the same except for the lower settings(1 and 2)
Do you think the Density settings are different for different systems? I don't I have ever used any setting less than 4, so it might not even matter.

So I guess the lesson is...match the DPI of the image to my PPI and density settings in the driver and it should be fine.

Thanks again for all your help!

[Dave Johnson29]

The six image densities correspond to: 75, 150, 250, 333, 500, and 1000 dpi. Write those values down, they're not in the manual... yet another minor annoyance that's easily fixable, yet has never been done.

Hey Dan,
If that is an annoyance do not hold your breath for it being fixed soon. My 1992 ULS does the same. Must be a desirable design feature. :D:D

[Richard Rumancik]

It's one of several minor annoyances I found on the ULS machines. You are allowed to set different resolutions in the X and Y axes, but they don't make it obvious that's what you're doing.

The ppi setting is dpi along the X-axis, and the image density slider determines the dpi along the Y-axis. The six image densities correspond to: 75, 150, 250, 333, 500, and 1000 dpi. . . ..

Dan - I just want to understand what you are saying - do you not like the feature of different dots per inch in vertical and horizontal axes, or are you just annoyed in the manner that it is implemented?

To me, the ability to set different dpi's would seem to be a good thing. The reason is that extra columns of dots in the vertical direction are essentially "free" i.e. it won't take extra laser time. If you double the number of "rows" then you double the laser time.

Many lasers only let you change both simultaneously. But it would seem that if I were plotting line art (filled vector shapes, text etc.) , I could get a better looking image by doubling the horizontal dpi while holding the vertical constant. At least I visualize in my mind that it would look better. Not sure about photos though, or how it would handle any pre-defined bitmaps.

What does seem odd to me is having the "density" setting instead of actually allowing you to set your own values. But my guess is that some people have been confused by it and so they changed the user interface to "simplify" it - at the expense of having real control of the laser. Too bad. Seems like it could be useful if implemented well.

The only other unfortunate thing is that they use the term ppi. Most other lasers use this acronym for pulses per inch which is a different input parameter. So here is a recipe for lots of confusion between owners of different brands.

[Scott Shepherd]

I can't be certain, but in speaking to Dan, and reading this post, I believe his frustration is not in the ability to control the X and Y individually, but rather 2 things (at least), that make it hard to understand. 1) You get a slider bar that goes from 1-6 with no mention of the actual value of the DPI, and two, it's not in the manual. So there is no way for the end user to know what 1,2,3,4,5, or 6 means or how it corresponds to anything else. I believe he (and me!) believe it should simply list the values. If not listing the values, give me a page in the manual I can print and stick on the wall so I can know what I'm setting each time.

It's the user interface verbage that's the problem, not the feature. I use that feature almost daily to fine tune results. It's a WONDERFUL feature and allows me to do jobs that I could not do when I did not have control over them independently.

[Dan Hintz]

Steve nailed it... as pretty as the ULS interface appears at first glance, it's not nearly as user-friendly as it could be. Somewhere around here I keep a list of my annoyances with the GUI, but I could recreate most of that list without too much trouble just because the things really annoy me!

Being able to modify the dpi in both axes independently makes for a more powerful machine, but I see no valid reason for not giving me full control. Don't allow me to type in a value between 1-1000 in one axis, and then limit me to six pre-determined values in the other. If you must limit me, then don't make it a slider with symbols that have no direct bearing on the setting, and certainly don't display symbols that don't follow the nomenclature you set up for the other axis! Both axes use stepper motors of the same type, so there's no need to limit me to those six values... by doing so, you usually limit the values that can be chosen for the other axis, as well, diluting the capabilities. If people (e.g., beginners) often choose the same value for both axes, have a checkbox that locks the two axes to the values selected from a single slider or entry box. It's not friggin' rocket science!

Just once I would love a ULS rep to say to me "We understand you have some ideas that may improve our user interface, and we would loveto have a look at them." But they won't... because that's how companies operate. Allowing outsiders to make suggestions would make them look weak, like they don't have the proper expertise in house to make a GUI... problem is, it's obvious they don't. You should never design a professional product's interface without using a professional. Yeah, it's better than a DOS command line, but it frustrates the user nonetheless.

[Dan Hintz]

"..So if your image den is set to 6 then your ppi should be at 1000 PPI, Den 5 is 500, 4 is 333, 3 is 250, 2 is 200 and is 1 is 166.."
My values came from the local rep, Paul, who has been with ULS since (almost) the beginning...10+ years with them, so I'm told. Steve and I highly value his advice and experience, but that doesn't mean he can't be wrong, and I've seen the ULS tech support team be wrong on more than one occasion. I don't know who is correct for the last two settings, but it would be too hard to find out... print a 1" black square and count the dots under a microscope. Tedious, but it works.

[Doug Griffith]

...they don't have the proper expertise in house to make a GUI... problem is, it's obvious they don't.

I have never personally used the ULS driver but the thing about any interface is that usability is subjective. What works for one may not work for another. It's possible they dumbed down their features to keep it less complex for entry level users. Maybe there should be an advanced users preference. Select that option and get all the bells and whistles. If it is just considered ugly or clumsy or unintuitive, that is very subjective. I personally like the Adobe Ilustrator interface while Corel Draw's makes me cringe. I'm sure others think the complete opposite. Neither of us are wrong.

[Dan Hintz]

Doug,

Your points about the subjective side of things are certainly true, but the points I made about dpi aren't really subjective. If ULS expects the user to understand dpi settings in the X-axis (as evidenced by the ability to type in a specific value from 1-1000), to "dumb it down" with a slider (and one that uses non-standard nomenclature, to boot) on the Y-axis serves no purpose other than to confuse the user. I have no problem with a user-selectable "difficulty setting" or checkbox to turn certain advanced features on/off, but this isn't one of those cases.