Character Images
Several image personalisation solutions (XMPie uImage, DirectSmile and others) let you create characters from individual images, so recipient names can appear in alphabet soup, cubes, dominoes, etc. While these types of image personalisation effects can look impressive, they usually take a considerable amount of time to create as you need to create individual images for the entire alphabet then mask out the background for each 'character image'.
Well, I've found a nice resource for creating different character image effects. iStockphoto is an online stock photography service with an extensive collection of affordable images. If you search their site for the term 'alphabet', you'll see a large selection of pre-shot images that use different objects to create characters, including coffee beans, shells, sand, cubes and more. Some alphabets even suggest appropriate backgrounds. A handy resource, which could save a lot of time!
Posted on Friday, 20 June 2008 at 12:36 PM | TrackBack: http://www.veedeepee.com/cgi-bin/mt-tb.cgi/155
Comments
Hi Eliot,
Another good resource.
I have only recently realised that Xmpie UImage can use "image fonts" in its image software. I was always under the impression that you were stuck wih manipulating standard typefaces. Do you need to create a "type face" out of your images using something like fontographer first? How would you use the squrae up sand shots you have used as an example, I can't picture how they would be used?
We started off using Directsmile many moons ago and one of the advantages was the use of "image" and "clip" fonts. I have experimented in Photoshop brushes with stroke masks to create a similar "clip font" effect using UImage with mixed results.
While reveiwing the Xmpie Marketplace recently I noticed that "image fonts" had been used to create some of their templates. I am still yet to find a detailed article on how they were created.
Posted by Tim Ball on Monday, 23 June 2008 at 2:42 AM
Tim,
There are a couple of solutions for creating character-based images in uImage and Photoshop.
The first method, as you have already discovered, is to create fonts out of images, or 'clip fonts' as DirectSmile calls them. As you can use any third-party Photoshop plug-in to create your desired effect with uImage, you can take advantage of Alien Skin Splat plug-in, well, up until now. Alien Skin has just announced that they have decided to discontinue this product. I'm not sure if there are other Photoshop plug-ins available to create a similar effect, but if there are, then you can use them.
The second method is to create characters from images, or 'picture fonts' as DirectSmile calls them. This is achieved in a similar method to DirectSmile. Firstly, you need to photograph or create individual 'characters' that you want to make available in your main image, for example alphabet soup characters, or alphabet fridge magnets. This is a time-consuming task, but no more than using DirectSmile. In fact, its a little easier as DirectSmile also requires separate image mask files for each character.
Once you've created your A-Z character set, you need to create placeholder frames for each image by placing one image as a Smart Object in your Photoshop document (using File > Place). You then repeat this step for every position where you want a character to appear. Once you have done this, you can programatically set up your uImage QLingo call to populate each frame with the corresponding character from the recipients name. I've created an example to illustrate this:
Sample 'picture font' image indicated with Smart Object frames
Obviously, to set this up requires QLingo knowledge, and you also need to consider whether you want to align characters left/centre/right, but it's pretty straightforward to anyone who is familiar with working in QLingo.
Hope this helps!
Eliot
Posted by Eliot Harper on Monday, 23 June 2008 at 6:43 AM
The disadvantage of the above approach is that each of the letters must be of about the same width to fit the generic template. You essentially have to design a fixed-width font which is not always desirable.
It would be nice if uImage could utilise some kind of "Image Font" but it doesn't seem to. Maybe something for a future version?
I've created my own work around for this by rendering the variable width letter images into words using a fairly simple Python script (In tandem with the Python Imaging Library). The resulting images can then be used as Smart Object's in a uImage template if they require further wrangling.
Plus I can use this script for future projects with very few alterations.
Ian
Posted by Ian WItham on Thursday, 14 August 2008 at 2:09 PM
Ian, I totally agree. This approach comes with its limitations. I'm impressed with your workaround, but it's probably not a suitable solution for the general uImage user (or for the faint-hearted!).
Posted by Eliot Harper on Thursday, 14 August 2008 at 9:09 PM