Technical requirements. Quicktime MOV format compressed using PhotoJPEG or Motion-JPEG (MJPEG) codecs, with quality set between 70 - 90% or High. Maximum length: 30 seconds. For more info about codecs and formats read on.
Subject and execution. Approach stock videos just as you do stock photos. The first step is choosing a subject or a concept that is stock worthy. Then make sure the camera is stabilized, the subject is in focus and the exposure is set correctly. Lighting, focus and white balance are key. Avoid auto settings on the camera, especially with dynamic settings or subjects, so that the clip doesn’t feature focus hunting or exposure variations.
Avoid static scenes that lack any action. For example a peaceful field on a summer day will be better served by a photo than by 30 seconds of video in which nothing happens. You can pan or zoom in the scene from a detail to another for a more dynamic video. However, if your equipment doesn’t enable smooth camera motions (such as zooms and pans) or smooth focus travels, it’s better to just don’t do them.
Stock video clips also require stabilisation (camera shake is not acceptable) so use a tripod or physical stabiliser; or you can do it the hard way in post production.
Watch out for rolling shutter. This is also called the “jello effect” and causes the image to wobble unnaturally. This is a problem that is quite common especially on older DSLRs, so if your camera exhibits this issue avoid fast moving scenes.
You can even skip the camera. Besides live motion we accept animation (2D and 3D), computer generated motion graphics, stop motion animation, time lapse videos.
Editing. While videos shouldn’t require as much post-processing as photos, some might require slight brightness/contrast adjustments or color grading. Ideally a video clip should only need trimming for removing the beginning and/or the end (when the record button is pressed) or because you need to shorten it to 30 seconds.
Do not upsample your stock videos. A sincere 720p is much better than a pixilated upsampled 1080p.
Audio. Any ambient sound that is relevant for the clip or voices (MR required) are acceptable. But any copyrighted material such as music playing in the background is forbidden and your video will be rejected (that includes the faint Friends theme playing on the TV while you were shooting). If the microphone on the DSLR captured subpar sound because of wind or if there is shuffling from your hands handling the camera or other camera noises, better just scrap the soundtrack.
Etc. For issues regarding copyright, Model Releases and Property Releases the same rules as for photography apply.
There are a lot of pro options: Adobe Premiere Pro, Final Cut Pro, AVID, Sony Vegas etc. These are useful for complex edits, applying effects etc.
Then there are the consumer options but which can handle editing quite well (removing unneeded portions of video from a clip for example): Adobe Premiere Elements, iMovie, Quicktime 7 Pro etc. These can serve the purpose of preparing video for submission very well, provided that the original footage doesn’t need any serious fixing (brightness, noise reduction etc).
For converting videos to PhotoJPEG/MJPEG:
Experiment with compression levels between 70-90% (or between Medium and High depending on your software). Don’t set above 95%, it’s going to be overkill for most videos.
Your workflow should be something like this:
The important thing to remember is that you should not work on a clip that is “low quality” compressed as each edit will degrade it. Think of working in Photoshop and after each adjustment to a photo saving it as a JPEG (didn’t that make you cringe?)
Here is a comprehensive guide for video editing: http://lifehacker.com/5785558/the-basics-of-video-editing-the-complete-guide
A video format is a container for the video information, audio information and other metadata (subtitles, chapters etc). Video information (and audio information) in the container is compressed using a codec. MOV, AVI or MKV are formats which can contain video streams compressed with a variety of codecs such as h.264, MPEG-4, DivX or PhotoJPEG. MKV even got its name from Matryoshka dolls (though it cannot contain more containers).
The quality of the video is related to the codec and the compression it uses.
There are two types of compression: intraframe and interframe. Both are "lossy" compression mechanisms, meaning that information is lost in order to save on space. However one is more lossy than the other.
Interframe means that the compression is dependent on other frames, so this mechanism uses keyframes which are saved as complete frames and then successive frames contain only the pixels that are different from the keyframe. This compression requires more processing power because the computer must process several frames at the same time in order to have a "complete picture" at any point in time. Examples include h.264, MPEG-4, DivX.
Intraframe means that the compression is compressed individually and independently from the other frames. Basically each frame is "key". As you guessed this is the less lossy compression. While it's not as efficient in space saving, it more than makes up for in quality. Each frame is kept as an independent photo and this means very low detail loss, especially in clips with random motion (which are something that interframe codecs struggle with). Also this compression has much lower computing requirements. Examples include PhotoJPEG, MJPEG.
Much lower computing requirements doesn’t sound too important these days with quad-core smartphones, but when editing a video with hundreds or thousands of Full HD frames shaving off any extra processing makes a world of difference. That’s why professional editing software such as Adobe Premiere or Final Cut Pro will convert imported clips to an intermediate intraframe compression (requires little processing power and retains much more quality during editing).
So, to summarize: interframe is lower quality and smaller file, intraframe is higher quality and bigger file.
Further reading on formats and codecs:
Q. My camera saves h.264/MPEG-4 anyway, why should I bother to up-convert to PhotoJPEG?
A. There are levels of “low quality” compressions. The processors in cameras can’t perform strong compressions on the fly, especially when they are being bombarded with 60 Full HD frames each second, so they will save files with less compression so that they don’t need to process them so much. So while the compression is still lower quality than PhotoJPEG/MJPEG, it’s still higher that what can easily be achieved on a computer.
Converting the clips to PhotoJPEG/MJPEG essentially locks the quality at that level. Obviously the quality will not increase, but it will not degrade during processing and editing either (or degrade considerably less).
Obviously this compression is essential for computer generated graphics since there is no camera involved and each frame can retain all the details perfectly.