Vprism: Video data analysis software

Andy's Main Page

Note: This page was originally written in 2002. At that time, vPrism was available from Lesson Lab. It ran on Mac OS9. The company appears to have changed hands since then, and is not offering a current version of vPrism software. This is a disappointment because at the time vPrism was made, there was no other video data analysis software that could come close to its power and ease of use.

Apparently other people feel the same way. Of the few pages on my humble website, this page gets more hits than any other. People from all over the developed world stop by here. I wonder what you are looking for, and whether you are finding it or not. I'm sorry to say that I'm not aware of any modern video analysis software that has the same capabilities as vPrism. If you find something, I'd really like to know about it. Please contact me: andyjohnson at bhsu.edu.

I have not contacted Lesson Lab (the company that used to sell vPrism) about their current products. This page is left here in case you have access to vPrism, want to know about it, or wish to contact Lesson Lab.

The big idea:

Vprism uses digitized video. It coordinates movie files with a database file containing transcripts and code information. Vprism allows you to create transcripts and event codes which can be exported for analysis and reporting. It also can export video clips that contain both the video and text information. I had my video on a CD in MPEG. Vprism relied on quicktime, so basically anything that quicktime could play (in OS9) you could use in vPrism.

Vprism makes multiple, flexible connections between the transcript and video files on a computer.

There are advantages to using video in digital format. For one thing, you don't need a VCR and monitor to analyze your data. (But you do have to get your video into digital format, no longer the challenge it used to be.) Also, having random access to video allows jumping to any point in the video very quickly. Combining speedy random access with lots of ways to link text with the video fundamentally transforms the character of video. It "civilizes" a wilderness of video information. For instance, you can jump from the beginning of an hour-long segment to the end in less than three seconds, you get right to the spot you wanted, and the appropriate passage in the transcript is automatically shown.

Here is a screenshot of vPrism's main window.

Steps in using vPrism

• First, the video has to be converted from videotape into digital format and saved somewhere. A single CD-ROM can hold up to an hour of video in MPEG-1 format. Other file formats and storage media can be used as well, but you have to be able to play it on OS9, don't forget. Vprism can treat large chunks of video on a disk or multiple disks as a single unit or as a series of smaller pieces.
  
• Second, you use vPrism to connect the video to events and/or a transcript. Vprism has features that support transcribing. It has no speech recognition. You have to do the typing. This is still much easier than using a VCR. Surprisingly, the audio system on computers are often better than those on TVs, so using digital video sometimes makes it easier to interpret speech on noisy video. Also, segments can be replayed over and over again very quickly, which also aids interpreting unclear speech. You can also import preexisting transcripts if you massage them into the right format (tab delimited text). You also have to link each line of the transcript to the appropriate time in the video. This last step is easy to do and a lot of fun.
  
• Third, some kind of analysis involving coding is done. Vprism allows practically unlimited coding on many segments of video. Again, event codes link to the video (or to the transcript) so that a particular event can be located, played and replayed very quickly and easily.
  
  Note: Vprism stores all of its transcript, timing, and code information in a single database file. This allows searches for particular text in any of the transcripts, or searches for a particular coded event in any of your analyzed videos. Vprism does not, however, have the capability for sophsticated analyses of text like other text-based qualitative analysis software. If you want to do this, you can export transcripts and event codes as text files for use in other programs.
  
  Vprism allows reordering of subsets of the coded video. For instance, (once the video is coded) one can very easily compile a list of particular events of interest (such as all of the events belonging to a single code) and view them in any order for comparison.
  
• A final step, if you want to present your video to others, is to compile segments of video and export the compilation as a stand-alone video file. This makes for very easy video editing.

Saying goodbye to vPrism

Vprism was originally created for the analysis of over 250 hours of classroom video for the TIMSS project. There were attempts to sell the software as designed - as an analysis tool. That's when I learned about it. I was very happy to use it, but it is no longer updated and marketed by "Lesson Lab". Lesson Lab has become a subsidiary of Pearson Education. They seem to offer some kind of web player service showing videos of classrooms, but they do not mention vPrism on their website anymore. Nevertheless, here is the address:

LessonLab Inc. 3330 Ocean Park Blvd. Suite 115 Santa Monica, CA 90405 (tel) 310-581-2300 (fax) 310-581-2002 http://www.lessonlab.com/ info@lessonlab.com

Vprism Minimum System Requirements:

Mac OS: Apple Power Macintosh G3 or G4, system software 8.5 or higher,
266MHz or faster, 96MB RAM, Quicktime 4.0.
(Performs well on an iMac, G3 Powerbook, or iBook.)You also must have some method of digitizing and storing your video.

Even older option for video data analysis:

, produced by Jeremy Roschelle, links transcripts in a computer file with video on tape. Again, it runs on OS9 only.

Cvideo makes multiple connections between a text file on a computer and a videotape in a special VCR.

To use this software, you have to have a VCR with "Control L" or "LANC" capabilities, and an older Macintosh with a serial port. CVideo does not offer coding capabilities. It uses VHS tapes and cleverly relies on the time counter on the VCR, so finding a segment requires winding the tape to the right spot. CVideo is freeware, but I no longer know where to find it. It also requires a serial connction from your computer to your expensive VCR. This technology seems to have disappeared.

There are other software tools for video data analysis but it's been a long time since I studied them. The last time I looked carefully (around 2003), Vprism, and then Cvideo, were the best video analysis tools for education purposes. Surely there must be something better by now. Please let me know.

Disclaimer: This document was constructed and posted by Andy Johnson as a public service. It is is provided solely for information, and represents no obligation on the part of LessonLab, Inc. or of Andy Johnson. For further information, contact LessonLab. 4/20/07 , Andy Johnson