How to Use the Effects

If you wanna have the best experience using Picture to People, there are some very important rules and tips you should always regard:

• You MUST configure your browser to accept Javascript. If you don't, you will not be able to use Picture to People online effects correctly.
• You MUST upload jpg files. No other file format will be allowed. This format gives very small files in average and can be opened/saved in almost all general use Computer Graphics programs. The result image of an effect is always a jpg too, despite maybe the browser can let you save the image in other formats.
• There was a big effort to make this site work correctly in Firefox 2/3 and Internet Explorer 6/7. It was a very boring and bothering task. People make Internet standards, but they are not followed, so software development is a mess. Some (not exhaustive) tests showed Chrome behaves in a very similar way to Firefox, so you can try to use P2P with it. All other browsers were not tested and are not recommended. Unfortunately, Internet Explorer doesn't work correctly with some CSS features, so Firefox 3 is the more suitable browser concerning P2P site appearence.
• Despite Firefox makes a better work than Internet Explorer concerning page rendering, some deep tests showed Firefox is very worse concerning the cache engine. Sometimes Firefox shows an old (cached) page/data even if it should refresh (request again) the such page/data. This problem can affect Picture to People behavior in some specific situations. If you have this kind of problem using P2P with Firefox, consider disable Firefox cache (using "about:config") or use Internet Explorer.
• The "back" button of Firefox can have some bad consequences never seen in Internet Explorer (its very similar to the cache problem, but disable the cache doesn't solve it). Some strange behaviors concerning this can be solved with a simple explicit page refresh.
• Professional level Computer Graphics effects don't come easily. In general, P2P procedural effects are not very fast, cause they need very time and memory intensive calculations, what it's intrinsic to them. Usually they take from some seconds to kind of one or two minutes to be calculated, depending on the effect chosen and the input image.
• After you ask for an effect calculation, respect the warning message that say to you WAIT FOR THE RESULT. You probably will get your result in 1 or 2 minutes at most, so be patient. If you click the "back" button of the browser, you will lose your request and will have to start again.
• No input or output image is saved in server hard-drive, what is a very important project decision concerning eficiency and security. For all effects, you will receive the result image in a simple empty page got from a stream, what makes possible to send the final image to you without recording it in file system. Save the result image in your disk as you want using the browser. After so, you can go to another P2P page using the "back" button of the browser or your bookmarks.
• Concerning the text effects, you should know that a lot of fonts are not perfect, because they have wrong, lacking or inconsistent data in their ".ttf" files. Some possible consequences are: 1) font has no bold or italic representation; 2) font gives clipped letters if you use bold or italic; 3) font doesn't have information about how to draw some characters; 4) font can't be properly rendered in a certain size; 5) any other bad behavior. If you get some strange result using a font, very probably it's not a P2P error. Try the same effect with some other fonts and probably you will have a correct result.
• There are rules concerning the maximum size of the input images. The reason is very simple: avoid processing overhead in server caused by just some few requests. This way, everyone will be able to use the effects without harming the other users.
• Uploaded photos can be at most 1200x1200 pixels. Any image with more than 1200 pixels in width OR height will be refused and the effect will not be calculated.
• Text effect output images can be at most 1200x600 pixels. If the text/font/size combination you chose doesn't fit these dimensions, your text will be renderized clipped. Be aware: the text is renderized centered in final image.
• Uploaded images to be used as patterns can be at most 400x400 pixels. This size is not big, but there is a good reason for that. Since you can upload any weird image to be processed, keep its size not large helps to avoid some bad behaviors for some effects that don't converge in some situations.
• The maximum upload size is 200KB. It's needed to control the individual traffic so everybody will be able to use the P2P project. This size seems few, but it's pretty enough. Almost always a file of a photo with 1200x1200 pixels in jpg format is smaller than 200KB. Anyway, if you need, increase the compression level of your jpg software before saving the photo to upload.
• The bigger the image you use as input for an effect, the more time its calculation will last. Don't use large images (photos) if you don't need they large.
• Several text effects don't work well with some kinds of fonts or some sizes (usually the 'small' and 'tiny' sizes). Anyway, every available font can be used at least with some effects giving good results. You need to use the effects to understand their behavior and find the effect-font combinations you like most. As a general rule, there are some effects that simply don't work with thin typographical styles (the reasons involve a lot of Math). When it's the case, be sure to choose a wide style font to get better results.
• Why don't P2P effects have configuration parameters in the interface? Picture to People effects have its parameters configured to offer the best results in average, concerning the more common types ("types" in a Math sense) of images. Configure the effects is not easy and anyway usually you would not get better results using a different configuration. Beside this, some effects tend to not converge depending on the set of parameters you use. Anyway, you can interfere with effects results according to explained in next topic.
• One easy, understandable and still useful way to change the final result of an effect is make some smart preprocessing of the input photo. Changes in bright, contrast and histogram, for example, are very simple to get and can improve drastically the final result depending on the chosen effect and the result you are looking for. The more you use the effects, the more confortable you get about what you need to make to achieve the result you want.
• Why do P2P text effects have a preview that shows just the text without any effect? This preview is very fast to show and let you have an idea about how your text looks like according to the font, size and style you chose. It's very useful also to show if your text fits the limit width of 1200 pixels (if not, your text will appear clipped).
• Uploaded pattern images should be black and white. If you send a gray of colorful image, it will be accepted and preprocessed by P2P libraries. Anyway, if you wanna have some control about the result (choosing the threshold for black and white transformation), you should change your image to black and white before submition.
• Sometimes you can get very nice and unexpected results mixing individual effects. Think about using the output image of an effect as the input image for another one if you feel it could give you a beautiful art. It's always interesting use the imagination and try something new.