Hi, there. I'm very new to the world of vectors and have managed with GIMP so far in my life, but now I am designing a brochure and the images I'm importing just are not working out so I'm taking a crack at this for the first time. Here's my main question:

Can you create a vector file from a pre-existing bitmap file (I think that's the right terminology, from what I've read so far in the tutorials)? I.e., many of the files I'm importing in my brochure are photos, in JPG format. One of them is also a screenshot of Googlemaps, also saved as a JPG. I know that to vectorize a bitmap file I have to do something related to tracing, but that sounds incredibly difficult when we're talking about a photograph as opposed to the animated lines of a Googlemap section. Is this the way I should be going about this? Or should I try to get closer to the source (i.e. my screenshot could also be saved as a BMP...) and somehow import it or save it as a vector format that way?

Also, I know that for most professional printing they require 300 DPI resolution. I know that means dots per inch, but how do I achieve this? If I'm working in a vector file, will that automatically give me good enough resolution, since it's scalable? How would I retain high resolution from a digital photo, for example, if I'm resizing it so it's only a few square inches so it will fit on a brochure?

Thanks so much!!

"Trace bitmap" will automatically create a vectorised version of a bitmap image. You can read about this in the Inkscape Guide. This feature is really only effective for vectorising logos or creating a stylised effect. It should work well with Google Maps but probably won't work well on your photos unless you are after something stylised.

When working with Vector you don't need to worry about DPI until you print. The important thing is to make sure your document has the correct "real-world" dimensions.

When including a Bitmap, the number of pixels will not be modified. But how large all small you make the bitmap on the canvas will effect how large all small the pixels will be on the printed page. So making sure you bitmap has a lot of pixels and making sure you don't make it too large on the canvas will ensure the pixels will remain small on the printed page.

If you're getting a professional printer to print the brochure you'll probably need to provide the brochure as a PDF. I have no idea what happens to bitmap when the SVG is saved as a PDF. It's possible it is embedded without any changes, or it's possible it is re-sampled to a specific resolution. Does anyone know?

Okay, that's helpful, thanks. In the meanwhile, I've been playing around with Trace Bitmap. It seems to be sort of working for the Googlemaps image, but only if I set it to "multiple scans" and "colors," and have been seeing how many scans I can do before my computer freezes! It looks like the max you can input is 256, but that seems pretty taxing on the system. I've successfully tried it at 80 scans, but the resulting image is still really blurry. Which is fine I guess, since I don't care if this looks EXACTLY like Googlemaps. How do I go about manually tracing it from here, to get the image sharper?

So given that I can't vectorize digital photos using Trace Bitmap, do people just generally input them into their print documents as-is, and assume that the photo was taken at a high enough resolution to make it come out clear? Is there any format that would work better for this, i.e. is JPG preferable to GIF, TIFF, etc.?

Thanks!

Hi ristretto6,
This may be a no-no, I'm not sure. But I found with careful selection, I could separate the scans (although when I did this, I only used 8, not 80 scans). And I actually found one of them was pretty darn close to what I wanted. Although I did still have to do some editing to make it sharp. I don't think you can get away from having to touch up a bitmap trace in Inkscape. And I don't think there's any other way to make it sharper, other than editing the nodes and/or handles. But I could be wrong about the latter.

The more scans you allow the more object that will be created. It might look more like the original but if you intend to edit it, you've got a lotta lotta objects to tweak. With something like Google Maps you really only want one object per colour. Note that Trace Bitmap isn't perfect--you're never going to get precise vector version of it.

Yes, people just put photos in as is. If they are low resolution they might resize them in another program first. The best format for your bitmap is the original format so long as you never edit it. If you need to edit it and resave it don't use JPG. PNG is a good format as it's not lossy.

Got it, thanks. So if I do a trace of the Google Map, for example, how do I go in and edit it afterwards? I.e. could you point me in the direction of what tool to use, what I do with it, etc.? Do I literally manually trace over the existing lines and then somehow make the blurry background disappear when I'm done?

And then what is the best format to save it in when I'm finished? If I just keep it as the default Inkscape file type, can I just import that directly into Scribus and that will give me the clearest import?

Thanks

If you haven't done all the exercises in Help, there are two in particular that may give answers: Help/Manual/Quick Start/Exercises 3 & 4 (logo: paths and tracing).

Not a graphic artist, I'm just starting with Inkscape and found those examples very helpful.

Also, Help/Tutorials/ tutorial 4 on tracing is more helpful after doing the two exercises.

I totally forgot about the included tutorials.

Also note that I believe the resulting trace will be in a group. You can ungroup the objects to edit their nodes.

Quick Start -- Whoa!
I never noticed that before...OH that is going to be SO helpful!!

micro, I wonder if it would be helpful to newbies if that chapter was specifically noted in the Inks Resources forum, in the Official Inkscape Resources topic. (Considering myself still a newbie, I know it's going to be a boon for me.) I know it's at the top of the Index of the Guide, but I guess I've never happened to look up anything under 'A'. So I've breezed right by it so many times Or if you don't think it's appropriate for that topic, maybe I'll post a new topic about it.

Well anyway, just a thought

Okay, thanks. I'm going to play around with some of those tutorials and see if they help. In the meantime though I'm still having some major photo importing problems. My original photos are in .jpg and .bmp formats, and pretty large. I'm sure they'd print great if printed at regular size, but when I scale them down to about 275 pixels wide to fit in the brochure I'm making, they print terribly. Here's what I've been doing:

- opening the .jpg or .bmp in GIMP
- scaling to the desired size
- saving as a .png

I've tried the opposite order; saving as a .png and then scaling down, and either way I have the same problem. Is there no way to scale down a digital photograph so that it prints well at a smaller size? The .bmp files I'm using are screenshots, so obviously starting pretty big as well, but they look like photographs when they're at their original size so I'm hoping there's a way to scale those down as well while retaining the good resolution.

Any more feedback on this would be much appreciated!

Thanks,
-lindsey

You didn't say what makes it look bad when printed, but I'll assume you're seeing pixels.

When you resize the bitmap you want to make sure you have around 200 pixels for ever inch of the image that will be printed. So if you're printing an image so it's size on the page is 4 inches by 4 inches, then the image needs to be 800 pixels by 800 pixels. If it has less pixels than that then you'll start to see the pixels on the printed page.

Hmm, that's interesting. So is that what DPI means--is dots per inch the same as pixels per inch?

Okay so I have the original photograph open in gimp--it's a JPG, and 600 x 450 pixels. When I go to scale image, I see that the X and Y resolution is set at 96 pixels / inch. But if I set it to 200 pixels/inch, and then rescale the image so that it's 300 x 225 pixels, it looks very pixellated (yes, that's what I meant by "bad" ) when I import it into Scribus. If I do the opposite, and rescale it to say, 3", the imported image is still very grainy, but bigger. This doesn't make any sense to me because I think 3" is about the real-life size that I want this image, and the 300 pixel image looks okay visually in GIMP. Sorry, I know this might be more of a Scribus question but they don't seem to have a forum and I know you guys here know your graphics! Does this have to do with how I'm importing it, or the way I'm tweaking it in GIMP?

Thanks!

Did micoUgly create that sticky explaining image resolution yet? Probably not.

Anyway, your correct that dots as it's used in DPI can also be refered to a pixels.

DPI only effect how big the image will appear in the real world. A 200x200 pixel image will look the same size on your computer screen regardless of it's resolution (DPI - I'll use the terms "resolution" and "DPI" interchangeably). You can increase or decrease the resolution without affecting the number of pixels. You'll only see the effect of resolution when you print it because your screen can only display a finite number of pixels and it can't make these pixels bigger or smaller.

Some software will let you work on your image in "real world" units. In this situation the software understands that you want a 2x2 inch picture and if you increase the DPI value it will increase the number pixels to ensure that it still prints at a size of 2x2 inches. The image will then get bigger on your screen.