-
I want to thank all the members that have upgraded your accounts. I truly appreciate your support of the site monetarily. Supporting the site keeps this site up and running as a lot of work daily goes on behind the scenes. Click to Support Signs101 ...
DPI Question
- Thread starter Deadworry
- Start date
Fred Weiss
Merchant Member
Actually plotters do get described sometimes in resolution terms which are usually around 1000 dpi. Vectors, in and of themselves are device resolution dependent.
Output a vector to a device that will only print at 300 dpi and you get a 300 dpi print. Output it to a printer set for 1440 dpi and you get a 1440 dpi print.
Output a vector to a device that will only print at 300 dpi and you get a 300 dpi print. Output it to a printer set for 1440 dpi and you get a 1440 dpi print.
Bill Preston
New Member
Not so much a reply as another question along the same line.
I only very rarely scan artwork in Signlab5. Cleanup of a ragged image is a bear for me. (Yes I do have the instructional CDs, but that would be too easy.) Basic questions are 1 is a lower DPI and higher magnification of the original better for a relatively clean trace---or, 2, a higher DPI at lower magnification.
At the moment I am working from a digiphoto of lettering. Converted color to B/W, and reversed to a "negative." No real reason for that, except the lettering was then black on white, and easier to see.
Trace is fairly clean. Sent it to plotter and cut in black vinyl. Tomorrow I will weed the cut piece, and edit with an x-acto to smooth out the curves, and whatnot, then rescan that piece, and hopefully get a real clean scan, and a trace that will need no further cleanup.
Opinions please.
bill preston
I only very rarely scan artwork in Signlab5. Cleanup of a ragged image is a bear for me. (Yes I do have the instructional CDs, but that would be too easy.) Basic questions are 1 is a lower DPI and higher magnification of the original better for a relatively clean trace---or, 2, a higher DPI at lower magnification.
At the moment I am working from a digiphoto of lettering. Converted color to B/W, and reversed to a "negative." No real reason for that, except the lettering was then black on white, and easier to see.
Trace is fairly clean. Sent it to plotter and cut in black vinyl. Tomorrow I will weed the cut piece, and edit with an x-acto to smooth out the curves, and whatnot, then rescan that piece, and hopefully get a real clean scan, and a trace that will need no further cleanup.
Opinions please.
bill preston
signguy95
New Member
Bill, I have never used Signlab *But* in flexi this is how I would do it...
1) aquire image
2) in the scanner software, I box out what I would like to scan.
3) select grayscale scanning
4) select 300 dpi
5) click scan
After this flexi imports it into my page...Next is use curve,straighten, and other tools provided by flexi to clean-up the image to the best of my ability...then I am ready for cutting! On clean up I won't spend alot of time *unless* it is a fairly large image, because on larger images it will be more noticeable. Small decals you can get by with less clean up. If you have these kind of tools available it might save time and material (especially in large jobs) in trying to clean up the image.
I am sure someone else here uses signlab and may can give you better advice!
1) aquire image
2) in the scanner software, I box out what I would like to scan.
3) select grayscale scanning
4) select 300 dpi
5) click scan
After this flexi imports it into my page...Next is use curve,straighten, and other tools provided by flexi to clean-up the image to the best of my ability...then I am ready for cutting! On clean up I won't spend alot of time *unless* it is a fairly large image, because on larger images it will be more noticeable. Small decals you can get by with less clean up. If you have these kind of tools available it might save time and material (especially in large jobs) in trying to clean up the image.
I am sure someone else here uses signlab and may can give you better advice!
Fred Weiss
Merchant Member
I'm too much of a perfectionist for my own good and probably work to too high a standard on half of what I do. I don't use SignLab but I'm sure it has "power tools" for cleanup. Anyway here's my approach which very much adapts to the nature of the artwork:
The main tools I use are:
FlexiSign
Adobe Streamline
FontExpert and Typograf
Artwork, for purposes of this discussion, falls into the following classification:
1. Black and White original or either color or grayscale.
2. Strong reference points such as text, borders, underscores and geometric shapes or irregular in nature.
Method 1
A black and white image of a dog. (Little or no visual reference points)
1. Scan in Photoshop at 300 dpi, black and white color mode. Most autotracing applications provide the cleanest and most faithful traces of b/w at 300 dpi. They're fined tuned for that setting. No more - no less.
2. Inspect scan in Photoshop and cleanup any imperfections. Save as TIF
3. Open in Adobe Streamline, set tracing parameters appropriate for the artwork and autotrace the image. Streamline is incredibly fast. Inspect the resulting trace and either keep it or delete and repeat after changing parameter settings. When a satisfactory result is achieved, save as Illustrator AI.
4. Open in Flexisign and use the various power cleanup tools as necessary. Save again as AI.
Method 2
A black and white image with text and a graphic with clean geometric shapes. (Lots of visual reference points.)
1. Scan the image as grayscale in Photoshop for a size of 8 to 10 inches wide or tall at 200 dpi. Touchup very little or not at all. Rotate image to precise horizontal alignment. Save as TIF or JPG depending on my mood. Adjust the opacity of the image to a medium gray to more easily see your vector lines when it is overlaid with vectors later.
2. Identify all type styles used in the graphic using FontExpert and/or Typograf. Install all fonts needed into operating system.
3. Open saved bitmap image file in FlexiSign and size up to about 24 inches wide or tall. Lock the image. We now have a template ready to have a vector image created on top of it. We will follow to a large degree the same approach as was used to create the image originally.
A. Set all type, modify size to match the scan, modify any expansion, condensing, slant etc. that may have been applied to the type originally. Position each line of type to match the baseline of the corresponding scanned type. convert all type to curves. Using the left and right arrow keys, nudge each letter into place to match the position of the scanned characters. Do not move any character up or down even if the scanned is slightly off on an angle from horizontal.
B. At this point start thinking like a draftsman for duplication of the graphic. Look for geometric shapes that are part of the image or that have been cut out of the image. Look for all lines that are on the square or have a parallel relationship to other objects. Use the rectangle, circle, and oval tools to create the shapes, size them and position them on the scanned template. This will provide you with much higher quality results than trying to lay down a path with the pen tool. Let's say we want a circle with a series of 3 horizontal stripes of negative space breaking it up into segments. We could just select the pen tool and start hand tracing it but it would never be as near perfect as if we drew a circle and three rectangles, size and position them, and finally use the rectangles to cutout the circle. And it takes far less time.
C. Use the pen tool to draw Bezier paths for the balance of the graphic as needed. (In our example no hand drawing was needed). Use the weld and cutout tools to combine the shapes to complete the graphic. Use outlining, shadow and distortion tools where appropriate. When complete, save as AI.
D. In the example above, the main circle was cutout by three rectangles across it. The female symbol was created by outlining a circle to create the second circle. The two circles were combined (compounded) so the center is a hole). Next the cross was created from two rectangles. Finally the three objects were welded together. Next it was centered in the segmented circle and sized. Next the symbol was outlined and the outline used as the cutout object from the segmented circle.
Variations come into play when color is introduced but the methodology is the same.
The main tools I use are:
FlexiSign
Adobe Streamline
FontExpert and Typograf
Artwork, for purposes of this discussion, falls into the following classification:
1. Black and White original or either color or grayscale.
2. Strong reference points such as text, borders, underscores and geometric shapes or irregular in nature.
Method 1
A black and white image of a dog. (Little or no visual reference points)
1. Scan in Photoshop at 300 dpi, black and white color mode. Most autotracing applications provide the cleanest and most faithful traces of b/w at 300 dpi. They're fined tuned for that setting. No more - no less.
2. Inspect scan in Photoshop and cleanup any imperfections. Save as TIF
3. Open in Adobe Streamline, set tracing parameters appropriate for the artwork and autotrace the image. Streamline is incredibly fast. Inspect the resulting trace and either keep it or delete and repeat after changing parameter settings. When a satisfactory result is achieved, save as Illustrator AI.
4. Open in Flexisign and use the various power cleanup tools as necessary. Save again as AI.
Method 2
A black and white image with text and a graphic with clean geometric shapes. (Lots of visual reference points.)
1. Scan the image as grayscale in Photoshop for a size of 8 to 10 inches wide or tall at 200 dpi. Touchup very little or not at all. Rotate image to precise horizontal alignment. Save as TIF or JPG depending on my mood. Adjust the opacity of the image to a medium gray to more easily see your vector lines when it is overlaid with vectors later.
2. Identify all type styles used in the graphic using FontExpert and/or Typograf. Install all fonts needed into operating system.
3. Open saved bitmap image file in FlexiSign and size up to about 24 inches wide or tall. Lock the image. We now have a template ready to have a vector image created on top of it. We will follow to a large degree the same approach as was used to create the image originally.
A. Set all type, modify size to match the scan, modify any expansion, condensing, slant etc. that may have been applied to the type originally. Position each line of type to match the baseline of the corresponding scanned type. convert all type to curves. Using the left and right arrow keys, nudge each letter into place to match the position of the scanned characters. Do not move any character up or down even if the scanned is slightly off on an angle from horizontal.
B. At this point start thinking like a draftsman for duplication of the graphic. Look for geometric shapes that are part of the image or that have been cut out of the image. Look for all lines that are on the square or have a parallel relationship to other objects. Use the rectangle, circle, and oval tools to create the shapes, size them and position them on the scanned template. This will provide you with much higher quality results than trying to lay down a path with the pen tool. Let's say we want a circle with a series of 3 horizontal stripes of negative space breaking it up into segments. We could just select the pen tool and start hand tracing it but it would never be as near perfect as if we drew a circle and three rectangles, size and position them, and finally use the rectangles to cutout the circle. And it takes far less time.
C. Use the pen tool to draw Bezier paths for the balance of the graphic as needed. (In our example no hand drawing was needed). Use the weld and cutout tools to combine the shapes to complete the graphic. Use outlining, shadow and distortion tools where appropriate. When complete, save as AI.
D. In the example above, the main circle was cutout by three rectangles across it. The female symbol was created by outlining a circle to create the second circle. The two circles were combined (compounded) so the center is a hole). Next the cross was created from two rectangles. Finally the three objects were welded together. Next it was centered in the segmented circle and sized. Next the symbol was outlined and the outline used as the cutout object from the segmented circle.
Variations come into play when color is introduced but the methodology is the same.
Fred Weiss
Merchant Member
Yup
Fred Weiss
Merchant Member
Quick check of Adobe's site .... it's only sold on it's own for $129.00.
Brochures and trial download are available here
It's really a standalone product that's very good at autotracing and then saving out as an AI file which you can then import into most any application used in graphic design or signmaking. While I mentioned that I do my scanning in Photoshop, that's simply because I like to touch up before autotracing. Streamline supports direct scanning from withing the program.
Its strength if the control over settings, the universal compatability of its output and, IMHO, the superior results it produces when compared to pretty much anything else available.
Brochures and trial download are available here
It's really a standalone product that's very good at autotracing and then saving out as an AI file which you can then import into most any application used in graphic design or signmaking. While I mentioned that I do my scanning in Photoshop, that's simply because I like to touch up before autotracing. Streamline supports direct scanning from withing the program.
Its strength if the control over settings, the universal compatability of its output and, IMHO, the superior results it produces when compared to pretty much anything else available.
Baz
New Member
What a great tutorial Fred 
If your sign shop goes under (God forbid) you should be a teacher.
The way you explained it is exactly how i do my art preparations, but never in a million years could i explain it as comprehensible as you just did.
As far as converting bitmaps to vectors, i'll say that i will use Streamline on 10% of my vector conversion, 90% is built overtop the original scans.
Great step-by-step Fred :U Rock:
If your sign shop goes under (God forbid) you should be a teacher.The way you explained it is exactly how i do my art preparations, but never in a million years could i explain it as comprehensible as you just did.
As far as converting bitmaps to vectors, i'll say that i will use Streamline on 10% of my vector conversion, 90% is built overtop the original scans.
Great step-by-step Fred :U Rock:
Bill Preston
New Member
I wound up going with both the X-acto edit, which got OK results, rescanned, retraced and got fairly good results. Not being especially satisfied, I dug out the instruction CDs, blew off a lot of dust, and went looking for node editing lessons. Found what I needed, and got better results.
Biggest trick was reducing the number of nodes to work with, 'specially when they are crowded together. The tools are fairly similar to "95's" Flexi setup, I guess, and I was able to do what needed doing. There are still a lot of other node editing tools in SL5, and maybe eventually---like when I turn 100, and that isn't that far away--- I might even get halfway good with it.
Thanx to all for the help.
bill preston
Biggest trick was reducing the number of nodes to work with, 'specially when they are crowded together. The tools are fairly similar to "95's" Flexi setup, I guess, and I was able to do what needed doing. There are still a lot of other node editing tools in SL5, and maybe eventually---like when I turn 100, and that isn't that far away--- I might even get halfway good with it.
Thanx to all for the help.
bill preston
geb
New Member
Bill, I have mainly been using Corel trace for my scans, but before I had that I scanned in signlab. I wasn't very good, and found Corel easier.
Anyways, maybe this is the trick your talking about, but if not, in signlab you can double click on an object, to bring up the node edit box, sweep select your object, hit the letter U on your keyboard. This will highlight every other node, then hit delete and half the nodes are gone. I noticed in the past there seemed to be two lines of nodes beside one another after a trace, and using this trick eliminated half the nodes, or in other words one of the lines.
This tip is also on the cadlink website.
George
Anyways, maybe this is the trick your talking about, but if not, in signlab you can double click on an object, to bring up the node edit box, sweep select your object, hit the letter U on your keyboard. This will highlight every other node, then hit delete and half the nodes are gone. I noticed in the past there seemed to be two lines of nodes beside one another after a trace, and using this trick eliminated half the nodes, or in other words one of the lines.
This tip is also on the cadlink website.
George
Greg
New Member
GraphiXtreme said:
Streamline doesn't give you the 50 billion nodes that CorelTrace gives you (at least me anyway). I also have issues with CorelTrace freezing up when an as yet unknown number of nodes exists in the trace. This happens when I ungroup the trace. I don't have the fastest PC here but it does the same thing on a 3.x ghz with 1 GB Ram. Perhaps I've just not discovered all of the proper settings.
The problem with both is that you need a relatively crisp graphic to begin with, not the graphics apparently taken from web pages that are usually handed to me. If you trace a bad image with either of these programs you'll spend more time cleaning it up than you would redrawing it.
95% of the vectorizing I do is manually done using CorelDraw!. I have a tablet but rarely use it, most times I just use the mouse. It is a RARE graphic that is truly "Trace and Cut".
Back to your question...between the two I'd recommend Streamline over CorelTrace.
Fred Weiss
Merchant Member
I have both and I have found Streamline to be superior in most instances to the primary autotracing in Flexi (which is also very good). In it's latest releases, Flexi has added on a few additional choices in autotracing: Bezier, Enhanced Curves and Enhanced Corners. This is what the Help File has to say about each:
I posted earlier regarding the Bezier method. I was alerted to it by Mark Rugen's email newsletter from GiveMeHelp.com. It would appear to be aimed specifically at low resolution bitmaps like Greg was talking about and did an exceptional job on the images I've tried it with so far that are of that type.
The Enhanced Curves and Enhanced Corners choices seemed, to me, to be a tip of the hat to the old Scanvec autotracing that was integral to the discontinued CasMate signmaking program. Those of you who have used CasMate may recall just how exceptional the autotracing was.
If you own a license to Flexi 7.5 and haven't tried these new tracing options, check them out.
I posted earlier regarding the Bezier method. I was alerted to it by Mark Rugen's email newsletter from GiveMeHelp.com. It would appear to be aimed specifically at low resolution bitmaps like Greg was talking about and did an exceptional job on the images I've tried it with so far that are of that type.
The Enhanced Curves and Enhanced Corners choices seemed, to me, to be a tip of the hat to the old Scanvec autotracing that was integral to the discontinued CasMate signmaking program. Those of you who have used CasMate may recall just how exceptional the autotracing was.
If you own a license to Flexi 7.5 and haven't tried these new tracing options, check them out.
vikingarts
New Member
Hi All,
First post here... and new to the forums.
Adobe Streamline can certainly be a useful program. It just takes a bit of getting used to and you need to feel out what images will be a good match for it.
I've tested almost every vector conversion software on the market, as it relates directly to our business.
Streamline is one of the few stand-alone vector conversion softwares available. And the only stand alone product from a major player like Adobe.
Creating sofware to do this work must be a tough nut to crack. It requires some level of subjectivity and objectivity to do this kind of work. Since the amount of points is of major concern to signmakers, this becomes on of the toughest obsticles and the major "fight" with the program: Keeping the curves and lines of the image straight and/or curved and keeping the amount of points to a minimum.
Fred has given some really good tips on using Streamline. I've found that this program works fairly well on images that are:
1. black and white: Streamline does a good job with one color/ high res images. You will have to play around with things like tolerence, etc to make it happen... but it will get you close.
2. Hi-Res: It seems to me that you should try to start with something that looks pretty good to begin with. A hi-res tiff or psd file. Even with this... you'll have to keep in mind that a circle that is put thru Streamline will not be like a circle you would draw in Illustrator (ie. 4 points and a center). You'll almost always have some clean up. And a straight line is rarely the distance between two points. You'll have to get in there and get rid of the extras.
3. Simple images: Streamline works well with fairly simple objects. The more overlapping images or overall difficulty level of the logo... the more cleanup will be required.
In my experience when it comes to color images or difficult images, it seems quickest to recreate the logo by hand. There are a few things that may cause you more time by using a conversion tool than if you were to just start from scratch:
1. Fonts: No tracing software is going to do well with fonts. It's probably best to photoshop these out of there before trying an autotrace software.
2. Trimming: You'll probably end up with some undesirable trims (knockouts, etc). Here is where that hand drawn recreation shines. Will the white be a shape when it comes out of Streamline or knockout? Well.. it just depends.
3. Straight lines and circles: I mentioned this before. Unless the image you are putting in there is just a circle or a line, you'll most likely end up with more points than you expect and less than perfect circles.
4. Full color images: I've just never had any luck with tracing software when it comes to full color images. If you want a cool effect (like with a photo or something) it's great. But if you want something that's smooth, crisp, clean, and cuttable it may take some more time to clean up.
I could probably go on about my experiences with this softare. I do want to note that I am a HUGE fan of Adobe products and work with them through the Adobe Certified Community. I have worked almost exlusively in Illustrator for the last 5 years: usually 10+ hours a day. What a great company in terms of development for any design professionsal!
As for continuing development in Streamline. I don't know. The current version is version 4 which was released in 1997. In that time there have been 5 releases of Illustrator and 7 releases of Photoshop. I believe this only means that creating tracing software to produce output as good as human output remains a very tough niche to fill.
First post here... and new to the forums.
Adobe Streamline can certainly be a useful program. It just takes a bit of getting used to and you need to feel out what images will be a good match for it.
I've tested almost every vector conversion software on the market, as it relates directly to our business.
Streamline is one of the few stand-alone vector conversion softwares available. And the only stand alone product from a major player like Adobe.
Creating sofware to do this work must be a tough nut to crack. It requires some level of subjectivity and objectivity to do this kind of work. Since the amount of points is of major concern to signmakers, this becomes on of the toughest obsticles and the major "fight" with the program: Keeping the curves and lines of the image straight and/or curved and keeping the amount of points to a minimum.
Fred has given some really good tips on using Streamline. I've found that this program works fairly well on images that are:
1. black and white: Streamline does a good job with one color/ high res images. You will have to play around with things like tolerence, etc to make it happen... but it will get you close.
2. Hi-Res: It seems to me that you should try to start with something that looks pretty good to begin with. A hi-res tiff or psd file. Even with this... you'll have to keep in mind that a circle that is put thru Streamline will not be like a circle you would draw in Illustrator (ie. 4 points and a center). You'll almost always have some clean up. And a straight line is rarely the distance between two points. You'll have to get in there and get rid of the extras.
3. Simple images: Streamline works well with fairly simple objects. The more overlapping images or overall difficulty level of the logo... the more cleanup will be required.
In my experience when it comes to color images or difficult images, it seems quickest to recreate the logo by hand. There are a few things that may cause you more time by using a conversion tool than if you were to just start from scratch:
1. Fonts: No tracing software is going to do well with fonts. It's probably best to photoshop these out of there before trying an autotrace software.
2. Trimming: You'll probably end up with some undesirable trims (knockouts, etc). Here is where that hand drawn recreation shines. Will the white be a shape when it comes out of Streamline or knockout? Well.. it just depends.
3. Straight lines and circles: I mentioned this before. Unless the image you are putting in there is just a circle or a line, you'll most likely end up with more points than you expect and less than perfect circles.
4. Full color images: I've just never had any luck with tracing software when it comes to full color images. If you want a cool effect (like with a photo or something) it's great. But if you want something that's smooth, crisp, clean, and cuttable it may take some more time to clean up.
I could probably go on about my experiences with this softare. I do want to note that I am a HUGE fan of Adobe products and work with them through the Adobe Certified Community. I have worked almost exlusively in Illustrator for the last 5 years: usually 10+ hours a day. What a great company in terms of development for any design professionsal!
As for continuing development in Streamline. I don't know. The current version is version 4 which was released in 1997. In that time there have been 5 releases of Illustrator and 7 releases of Photoshop. I believe this only means that creating tracing software to produce output as good as human output remains a very tough niche to fill.