Which of the following image file formats supports simple animation?

The correct answer is A, Joint Photographic Experts Group. JPEG is named after the committee of the International Organization for Standardization and the International Telecommunication Union that drew up the standard for compressing continuous tone photographs. Its compression is lossy, which means some detail is thrown away for a much smaller file, and it keeps no transparency, so a JPEG logo always carries a background. Option B is wrong because the format has nothing to do with the Java language and predates it in design work. Option C invents an expansion around the word pixel, which is tempting but is not the committee's name. Option D is also invented, and candidates should note that the format is about compression rather than editing. Remember the pair: JPEG for photographs with lossy compression, PNG for lossless images with transparency.

The correct answer is B, MPEG-1 Audio Layer III. MPEG stands for Moving Picture Experts Group, and the group defined three audio layers of rising complexity within its MPEG-1 standard; the third layer gave the file extension .mp3. Its compression is lossy and relies on psychoacoustics, discarding sounds a listener is unlikely to hear, which shrinks a recording to roughly a tenth of its uncompressed size. Option A is invented and no standard uses the phrase multi phase audio. Option C sounds plausible because MIDI really is a music protocol, but MIDI and MP3 are different things: MIDI stores instructions, MP3 stores compressed sound. Option D confuses the format with Microsoft's media player software, which plays many formats but gives its name to none of them. Candidates should also keep the companion full forms ready, since MPEG-2 underlies DVD video and MPEG-4 Part 14 gives the familiar MP4 container.

The correct answer is B, PSD. PSD, the Photoshop Document, keeps every layer, mask, channel and adjustment separately so that the artist can go back and change any of them, which an exported JPEG or PNG cannot do because those flatten the picture into one grid of pixels. Option A, CDR, is the native format of CorelDRAW, a vector drawing package. Option C, AI, is the native format of Adobe Illustrator, which is also a vector program and a common distractor because it too comes from Adobe. Option D, SVG, is the open Scalable Vector Graphics format of the World Wide Web Consortium, written in XML. The simple rule to carry into the hall is that Photoshop is raster and saves PSD, while Illustrator and CorelDRAW are vector and save AI and CDR.

The correct answer is B, SVG. Scalable Vector Graphics stores a picture as mathematical descriptions of lines, curves and shapes written in XML, so the browser redraws it at whatever size is asked for and the edges stay perfectly sharp. That is why logos, maps and icons are kept as SVG. Option A, PNG, is a raster format: it is a grid of pixels, excellent for lossless images with transparency, but it blurs when enlarged. Option C, TIFF, is also raster and is favoured in scanning and printing because it can hold very high bit depths. Option D, BMP, is the plainest raster format of all, storing pixels with little or no compression. The examinable distinction is that raster images are made of pixels and pixelate, while vector images are made of paths and scale freely.

The correct answer is A, size of an object. In computer graphics the three basic two-dimensional transformations are translation, rotation and scaling. Scaling multiplies the coordinates of every point of an object by a scale factor, so the object grows when the factor is greater than one and shrinks when it lies between zero and one; uniform scaling uses the same factor along both axes and so changes the size without changing the proportions. Option B describes translation, which adds a fixed displacement to each coordinate and moves the object without resizing it. Option C is wrong for uniform scaling, since a square scaled equally in both directions is still a square; distortion of shape belongs to shearing or to unequal scaling. Option D is wrong because a single scaling matrix does not move an object's reference point in the way translation does.