A videotape accompanies this article.
Methods for digital video recording, storage, and communication of movement disorders†
Article first published online: 8 OCT 2001
Copyright © 2001 Movement Disorder Society
Volume 16, Issue 6, pages 1196–1200, November 2001
How to Cite
Jog, M. S. and Grantier, L. (2001), Methods for digital video recording, storage, and communication of movement disorders. Mov. Disord., 16: 1196–1200. doi: 10.1002/mds.1199
- Issue published online: 29 NOV 2001
- Article first published online: 8 OCT 2001
- Manuscript Accepted: 8 MAY 2001
- Manuscript Revised: 21 MAR 2001
- Manuscript Received: 2 OCT 2000
- MPEG video format;
- video capture;
- streaming media
Video technology has now reached a level of sophistication that allows easy digitization. Digital video can be easily edited, reproduced, incorporated into databases, and posted on intra- and Internet sites for clinical use and demonstration purposes. Numerous methods exist for the production of digital video. This article synthesizes and simplifies the available methodologies in order to easily choose the technology that is the most appropriate for the movement disorder specialist's end use. Depending on available resources, issues such as cost, ease, and time to conversion are discussed. In addition, our experience with the use of one of the methodologies is briefly presented. © 2001 Movement Disorder Society.
The recording and storage of video information on patients with movement disorders is not a new concept. In fact, many movement disorder experts regularly videotape their patients. This practice is extremely useful for maintaining a profile of disease progression over the years. It is also a valuable instructional tool. Videotapes are used by Movement Disorders as a regular form of communication and education.
Many movement disorders clinics are gradually converting patient text information into easily searchable databases that reside on networked servers. Integration of digital video information into a password-protected clinical database would allow easy cataloging of patient videos and provide random access to these videos from any networked computer. Since many databases (such as ours) are already “web-enabled,” the data can be accessed from any computer that has the ability to access a web-browser. Securely stored patient videos can therefore be reviewed, along with other clinical information, in the clinic, a feature unavailable with videos stored on tape. Videos can also be stored on CD-ROM, DVD (digital video disc), or other storage media with long shelf life. They can also be posted on teaching web sites (see our web site at http://mdc.lhsc.on.ca/) for secure dissemination to physicians and other health care providers.
In order to achieve these goals, videos need to be available in digital format. An expanding and confusing number of technologies are becoming available which may make it difficult for the movement disorder specialist to decide on the appropriate tools for their own use. This article discusses the options that are currently available and reviews the features of each. Simplicity of use, time required, cost, retrievability of videos, and video parameters such as resolution, file size, and compression are the important issues that ultimately determine the choice of the technology. In addition, interface with existing clinic analog video libraries and technology is of great importance.
METHODS AND DISCUSSION
Video technology can be broadly divided into analog and digital format. The methodology is reviewed in these two broad categories. The first section reviews analog video and its drawbacks for use in the analog format. The second section is divided into analog recording and subsequent digital conversion (A-D) vs. direct digital recording (DDR). The advantages and disadvantages of these systems are discussed. The use of a tripod is recommended to reduce jitter inevitable in handheld recording devices.
Analog recording and playback devices are in common use. Handheld, light, and popular 8-mm and VHS-C cameras, and heavier and less favored regular VHS cameras, are all easily available. Most movement disorder clinics record patient videos in this fashion. Tapes are cataloged with patient information written either on the back of the tape or in an accompanying text database. In addition, patients have access to analog cameras and send videos to their physicians in this format. The journal Movement Disorders also receives and distributes analog video. Once recorded in analog format, this high-quality video can be archived and retrieved at any time if needed, but with a limited shelf life.1
Despite ease of use, low cost, and ready availability, analog tapes have numerous disadvantages. First, analog recording cannot be compressed. Therefore, there is a limited duration of video that can be stored on a tape (usually 2 hours per tape). This leads to a clinic accumulating many tapes over the years with patients recorded sequentially. Our clinic for example, records approximately 15 tapes per year. Second, analog videos are not “random access.” In order to quickly review a patient's video during follow-up from several prior visits requires time-consuming searching of numerous tapes. As the clinic video library grows, tapes may not be physically located in the clinic, making access even harder. In addition, the mechanical wear and tear from playing, rewinding, and forwarding can damage the tape. Since this is the only copy of the recording, once damaged the data is not recoverable.1
Eight-mm tapes require an expensive (∼$1,500 CDN), not easily portable player. Therefore, the composition of a video presentation with many different patients is tedious, requires dubbing of the video onto a VHS tape, and inherently reduces video quality. The final VHS tape is also not random access and the video cannot be easily incorporated into multimedia presentations. Slides have to be paused and video projection has to be started in the middle of a presentation.
There are two main options available, A-D and DDR.2–4 The major difference in these two systems is at the stage of video capture. Further modifications to the video such as editing, compression, and conversion to a universally playable format for incorporation into a database or posting to a website are required in both systems.
Analog recording with subsequent conversion to digital (A-D)
Most clinics already own an analog camera and no new investment is required. Subsequent to recording, videos are digitized and stored onto a computer by video capture hardware easily installed in a standard computer.
Video capture cards.
This technology has evolved rapidly. Expensive, high-end cards (e.g., DPS Perception, Avid system) can provide features such as online, nonlinear editing and special effects but these are not necessary for the regular user. Complex editing of video, analog, or digital is time-consuming and is best avoided. Since current video compression technology allows large amounts of good quality video to be easily stored (see below), editing to reduce video file size is not an issue. In addition, most capture card vendors provide excellent accompanying editing software. If necessary, more sophisticated editors such as Adobe Premiere® can provide many special effects and are cheaper than a hardware solution for editing (such as the Avid nonlinear editing system). Video capture cards are now able to provide full screen, full resolution (1028 × 1028) video capture at fairly low cost. We currently use a Maxtor Marvel® capture card with the provided capture software ($300 CDN) that provides this degree of resolution. In addition, it doubles as the computers' video display card. Considering that a good video card without capture abilities costs ∼$150, integrated capture/video cards offer value for the money. Other manufacturers (and there are many) such as ATI “all in one” or Intel's Smart Video Recorder video capture cards provide reasonable resolution at an even lower cost. The captured video is usually in a universal AVI (Audio-Visual-Interlaced) format that can be opened by any standard video card and player running Windows, Macintosh, or Linux/Unix or other operating systems.
The speed and quality of capture does not determine the time taken for digitization. As long as the hard drive is large enough to store the digitized video a 2-hour tape can be captured in exactly that much time, irrespective of the computer speed. Computer bus and processor speed allow easier, higher frame rate capture, improving video resolution. Our computer is a standard Pentium 600® with a 100 MHz PCI bus with LX chipset, 128 Mb of RAM and a 20 Gb IDE hard drive running Windows 98®. Since the capture cards already have a video display integrated, a separate video card is not required. Due to faster seek times, an ATA or SCSI hard drive is recommended.
Direct Digital Recording (DDR)
Digital video recorders are rapidly becoming available. Although expensive, these cameras have the advantage of recording in digital format onto tape. Recording directly into the digital format preserves image quality. The video output from this tape (DV or mini DV) can be played directly on the television using the camera as a playback device or be transferred to a regular VHS tape. Similar to analog cameras, these cameras use CCD (charge-coupled device) technology and can also store large amounts of still images. The image resolution is at or below 1 megapixel as compared to the digital still cameras (3.3 megapixel and higher). Higher-end cameras have an analog input, allowing them to be used as “digitizing devices” to convert previously recorded analog video into digital format, although conversions of large libraries would cause serious wear and tear on the camera.
Transfer to computer from digital camera.
Digital video is still on tape and therefore not random access. Files are large and require transfer to a computer for further processing before use in a clinic database or for streaming over the Internet. This transfer can occur over a regular slow, parallel port .01 Mega bits per second (Mbps), a universal serial bus (USB) port (1.5 Mbps), or the Institute of Electrical and Electronics Engineers (IEEE) 1394 port technology (up to 100 Mbps).4 This technology, also known as FireWire™ (Apple computers) or iLink™ (Sony Electronics), now available for PCs, can theoretically transfer data at up to 400 Mbps and may be able to go even three times faster in the near future. The transfer of digital video requires this speed, given that a 60-minute DV tape can store 11 gigabytes of data! IEEE 1394 technology can transfer broadcast quality, dual channel, full motion (30 frames per second) video with CD quality sound. Although some digital cameras come with a 1394 port, most computers, televisions, and other devices do not, and computer manufacturers have been reluctant to put this port on every computer. A special hardware card is therefore required for the download to the computer. New iMacs® do have such an input connector. This transfer of digital video to the computer for further video editing is faster than the A-D transfer.
There are some drawbacks to DDR from a movement disorders perspective. Since the videos are still on tape, they essentially have the same drawbacks regarding lack of random access and wear and tear as analog videos on tape. Therefore, incorporation of videos into databases requires further processing by transfer to a computer with special hardware. In addition, most movement disorder clinics already have large video libraries. Although some digital cameras have analog inputs they, are not intended for the digitization of large analog video libraries that many clinics possess. Given the expense of a digital camera, need for a computer with special hardware, difficulty of interfacing with large clinic video libraries, the faster transfer capabilities (several minutes as compared to 2 hours for A-D conversion) is the only immediate advantage of digital video.
Further processing and playback
Subsequent to computer transfer, further rendering is required before videos can be easily stored. These steps are common to A-D and DDR systems.
The AVI or DV file format video is large, with 2 hours of full-screen, full-resolution capture requiring up to 12 gigabytes! It is necessary to compress these files to effectively store them. Most capture cards such as ours also have built-in hardware and software to export video into other formats such as MPEG5–8 (Moving Picture Experts Group) 1 or 2. Although the IEEE 1394 technology allows rapid transfer of DV, additional software (e.g., Cleaner 5®, LSX-MPEG 2 Transcoder ME®) is needed for conversion to MPEG 1 or 2. These formats reduce the file sizes significantly (usually between 7–10× compression), while maintaining excellent resolution. Unfortunately, MPEG 1 and 2 files are still relatively large. In addition to storage difficulties (2 hours of video is ∼1.2 gigabytes), these large files do not transfer well across the Internet (see videos on the accompanying VHS tape). Another popular format is the Quicktime® movie format. Since compressed video is harder to edit, we suggest that the AVI file be used if extensive editing is required. New MPEG editing software is also rapidly becoming available.
In order to reduce file size and especially the ability to easily transfer video over the Internet (streaming multimedia), two main software tools have become available: Real Media Producer® and Microsoft's Windows Media Encoder®. Both of these software packages can be easily downloaded from their respective websites and produce videos from AVI format into either the Real Media format or MPEG4-ASF®9, 10 (Advanced Streaming Format) format.
We use the Windows Media Encoder® to render videos into MPEG-4 ASF® files with parameters of 30 frames per second, bit rate of 250 Kbps, and an image size of 480 × 320 and CD-quality sound. At these settings, 1 minute of video generates ∼2 Mb of data, allowing ∼360 minutes (∼6 hours) to be recorded onto a single CD-ROM. Multiple bit rates are available from 25 Kbps to 3 Mbps. File size is clearly dependent on the bit rate and our experience has shown that higher bit rates do not provide better quality of video. In fact, a lower bit rate of 150 Kbps is also reasonably good (see accompanying VHS tape).
An additional advantage of the Windows Media Encoder®9, 10 is its ability to capture live source. Thus, it is possible to entirely eliminate the step of capture into AVI and then subsequent rendering into MPEG-4 ASF. The entire 2-hour tape can therefore be captured uninterrupted followed by subsequent separation into individual patient segments for storage. We currently follow this procedure and are able to transfer a full 2-hour tape to the database within 4 hours. Software capable of directly rendering DV to MPEG has to be purchased separately.
Storage and Playback
The freely downloadable Real Media® and Windows Media® players can play the videos on any standard computer. All Windows® computers have Windows Media Player and Real media player can be downloaded from the Real Media website. A standard video card can play these at any screen size, including full screen, and no special hardware is required.
MPEG 4 files are small enough so that several hours of videos can easily be stored on a CD-ROM.11–13 CD writers with software are easily available. A standard 32× CD-ROM can play these videos without any frame loss. While it is true that DVD can store larger amounts of data (up to 17 gigabytes), DVD writers are very expensive and, unlike CD-ROM, DVD file format has not yet been standardized by the industry. In addition, DVD media is expensive ($20 vs $1.00).
Web-based streaming multimedia—hard disk storage
The ASF or Real Media file format allows video streaming over any network. Streaming technology is a new file encoding and transfer protocol that allows the user to view multimedia files over the Internet without having to wait for the entire file to download. Therefore, 2 or 3-megabyte files which, over a 56.6 K modem, would take half an hour at best to download, can be viewed instantaneously. Streaming of the videos across the Internet requires them to be stored on a machine that is set up as a server. A Real Media Server® is required at additional cost to stream real media video files while Windows Media® files can be streamed from a variety of different server platforms. We have successfully streamed Windows Media® files using the Windows Media server and Unix, Linux, Windows NT® and Lotus Domino® servers.
Once the files are placed on the hard disk of a secure server, the user can view them easily through the Internet using a browser. Our current clinic and video database library uses a Windows NT® platform running the Lotus Domino Server®. Videos are cataloged in a password-protected secure web-enabled video database. This secure database is behind a “firewall,” allowing only those with access to see the videos. In our particular case, the entire clinic, including patient information, is also gradually being made available through a web-accessible database. This allows us to view all of the pertinent clinical patient information along with the videos over the Internet in a secure fashion, from the clinic at our institution to a clinic anywhere in the world.
We have tested the ability of these files to stream to computers using fast (such as with cable) and slow (28.8K or 56.6K modems) connections. These videos can be run with minimal difficulty over both these connection speeds. In addition to the computer playing the videos, Internet connection speed becomes important. Clearly, if a user expects to see video streams over the Internet it is suggested that a proper, faster Internet (such as a T1 or a T3 connection, DSL, or cable) connection be used. Users cannot expect to get this kind of multimedia information to their computer over the Internet and still use a modem. Nevertheless, with connection speeds of 100 Mbps and higher (broadband) coming in the near future, video transfer over the Internet will no longer be an issue. Our patient videos have been successfully streamed across the Internet to New York and even New Zealand.
Patient videos in digital format have become an easy and necessary tool for movement disorders. Storage and playback in analog form has limitations. Simple analog recording, storage, and playback are extremely easy to perform, but videos are quite difficult to easily access. Tapes are also sensitive to damage, a problem shared by the DV format tapes. Digital video format is, therefore, in our view extremely important. Direct digital recording is clearly an exciting development but currently has limitations. This technology requires investment into a fairly expensive digital camera. The recording format is still tape and effective use of the digital video requires further compression before it can be put onto a CD or into a database. The transfer requires a computer with IEEE 1394 hardware for reasonably fast download. Preexisting analog clinic libraries cannot be used until they are digitized, which is not easily done even with high-end cameras that have A-D capability.
Analog recording with subsequent digitization is cheap, provides good quality, and interfaces well with preexisting clinic libraries. Most clinics already have an analog camera and a PC. The capture hardware/software can be installed on a standard PC, is excellent, and very affordable with digitization taking only an hour or so longer than the direct digital recording format. Capture software that compress video during capture is freely available. Table 1 summarizes many of these parameters and serves as a guide for the user to decide the most appropriate form of recording and storage for their clinic.
|Time required||In-clinic only||Extra (3–4 hours)||Extra (2–3 hours)|
|Recording media||Analog tape||Analog tape||Digital tape|
|Interface with preexisting analog clinic video library||Excellent||Excellent||Difficult|
|Cost||None additional||• Pre-existing camera ($0)||• Digital camera (∼$2,000)|
|• Standard computer with CD-writer||• Standard computer with CD-writer|
|• Capture hardware ($400); AVI/MPEG software included||• IEEE 1394 hardware for transfer|
|• MPEG/AVI conversion software|
|Random access||Poor (tape only)||Excellent (after digitization)||Excellent after transfer to computer and compression|
|Compression||Not possible||On-the-fly||Excellent after transfer to computer and compression|
|Database/Internet||Not possible||Excellent||Excellent after transfer to computer and compression|
LEGEND TO VIDEOTAPE
A sample of videos in various formats is provided. The video was captured on a standard Sony 8-mm camera and subsequently rendered. Segments are labeled according to the video file format. Videos in MPEG 1 format at the standard bit rate (4 Mbps) and MPEG 4 format in four bit rates (150, 250, 500 Kbps, and 1 Mbps) are provided. Note the slight differences in the quality. The comparable digital file size differences are large.
- 1Magnetic tape storage and handling. Washington, DC: Commission on Preservation and Access; June 1995..
- 5MPEG “what is.” Moving Picture Experts Group. http://whatis.com/mpeg.htm
- 6Overview of the MPEG-4 Version 1 standard. International Organisation for Standardisation. http://wwwam.hhi.de/mpeg-video/standards/mpeg-4.htm
- 8The MPEG group home page. http://drogo.cselt.stet.it/mpeg/
- 9Streaming media information and home page. http://members.tripod.com
- 10Windows Media Technologies http://www.microsoft.com/windows/windowsmedia/en/download/default.asp
- 11Readability or when is a CD's life over? Permanence and handling of CD's. http://www.kodak.com/global/en/professional/products/storage/pcd/techInfo/permanence3.shtml
- 12So how long can CD's last? Permanence and handling of CD's. http://www.kodak.com/global/en/professional/products/storage/pcd/techInfo/permanence5.shtml
- 13Ensuring the longevity of digital documents. Sci Am Jan 1995;42–47..