Video Spigot for Windows Product Version 1.0 Release Version 1.3.0.1 ----- Changes for release 1.3 ----------------------------------------------- VideoSpigot for Windows 1. When installing VideoSpigot for Windows, the target system's bus performance is tested. If capture performance is too low, full-field (640x480) capture is disabled and a flag is set in the [SPIGOT.DRV] section of SYSTEM.INI. The flag can be changed to override the disable, enabling full-field capture even if the automatic bus performance measurement implies that full-field capture quality would be compromised. The flag values: FullFieldCapture=0 disable full-field capture FullFieldCapture=1 enable full-field capture FullFieldCapture=2 re-measure bus performance at next startup 2. Addresses in the range F00000 through F1FFFF are not valid for placing VideoSpigot for Windows. Cinepak Codec 1. The compressor will now insert a key frame when it detects a scene change in the source material. 2. Bright fields would occasionally have dark patches in them when the compressed clip was played back. The problem has been fixed. ----- Instructions for Installation of the VideoSpigot Capture Driver ------- 1. Make sure that you have installed your distribution of the released version of Microsoft's Video for Windows. 2. When running Windows, do the following: a. Open up the Control Panel application and invoke the Drivers applet. b. From the Drivers applet's main window select the Add... pushbutton. c. In the Add dialog box which then appears choose the selection "Unlisted or Updated Driver" from the List of Drivers listbox. d. In the Install Driver dialog box which then appears enter either A:\ or B:\ in the edit field, depending on which drive you have inserted the VideoSpigot diskette. Press OK. e. In the Add Unlisted or Updated Driver dialog box which should then appear, you should see the following entries: SuperMatch Cinepak Codec SuperMatch VideoSpigot Capture Driver SuperMatch VideoSpigot Codec SuperMatch YUV 4:2:2 Video Codec Select 'SuperMatch VideoSpigot Capture Driver' and press OK. f. The capture driver will be copied into your Window's system subdirectory and another dialog box entitled, VideoSpigot Device Setup, will appear. g. VideoSpigot uses I/O addresses 0AD6h through 0AD9h. These addresses are active all the time and are not reassignable. While capture is in process, VideoSpigot needs to have access to an interrupt and an 8KB memory buffer in your system. The Scan... pushbutton to the right of the Interrupt Address combo box will start an automatic IRQ scan that will attempt to locate an available IRQ address from among the VideoSpigot choices of 10, 11, or 15. The scan is not foolproof; you must select one of these that has no conflict with other hardware. The Scan... pushbutton to the right of the Base Memory Address combo box will start an automatic memory scan that will attempt to locate available 8KB memory blocks at which VideoSpigot may be installed. VideoSpigot requires an unused 8KB block in the bottom megabyte (addresses 000A0000 through 000FFFFF) or top megabyte (addresses 00F00000 through 00FFFFFF) of your machine's ISA bus physical address space. NOTE: VideoSpigot is a 16 bit device. You must ensure that no 8 bit memory devices reside in the 128KB block in which you have assigned VideoSpigot. The mechanism used by the VideoSpigot capture driver to access the physical address space allocated to the buffer does not require that you exclude the address of the buffer in any EMM386 (or other memory manager) lines in your CONFIG.SYS. Once you have made all your selections, press OK and installation of the capture driver for the VideoSpigot card is complete. -------- Instructions for Installation of VideoSpigot Codecs---------- To fully take advantage of VideoSpigot's capability to capture video data to raw YUV, Spigot compressed, and/or do post-capture compression with the Cinepak codec, you must also install the supplied compressor drivers. To do this: 1. When running Windows, do the following: a. Open up the Control Panel application and invoke the Drivers applet. b. From the Drivers applet's main window select the Add... pushbutton. c. In the Add dialog box which then appears choose the selection "Unlisted or Updated Driver" from the List of Drivers listbox. d. In the Install Driver dialog box which then appears enter either A:\ or B:\ in the edit field, depending on which drive you have inserted the VideoSpigot diskette. Press OK. e. In the Add Unlisted or Updated Driver dialog box which should then appear, you should see the following entries: SuperMatch Cinepak Codec SuperMatch VideoSpigot Capture Driver SuperMatch VideoSpigot Codec SuperMatch YUV 4:2:2 Video Codec Select 'SuperMatch VideoSpigot Codec' and press OK. f. The VideoSpigot Video Codec driver will be copied into your Window's system subdirectory and its installation should then be complete. g. Next, install the SuperMatch YUV 4:2:2 Video Codec by repeating steps b through e. g. Next, install the Cinepak Codec by repeating steps b through e. NOTES ===== There are several factors that you should take into consideration so that you may achieve the most optimal capture rates possible for your system: 1. Disk fragmentation - if your hard disk is highly fragmented then it will take more time to access data within files and therefore reading and writing AVI files will take up more of your system's bandwidth. Because less time is available to the capture process when a disk is fragmented, captures rates will suffer. Make sure that you periodically defragment your disk (several utilities for doing this are available commercially -- PC Tools, WinMaster, etc...). Make especially sure that your disk is defragmented before you create the main capture file that VIDCAP will use during capture. 2. Disk Transfer Performance -- A hard disk which is slow at transferring data will impair capture rates since data will not be able to be transferred to disk fast enough to keep up with the rate at which data is being captured. On these slower hard disks you may find that capture rates actually go up when the data being captured is written to a capture file located on a compressed disk volume (logical disk drives created by utilities such as Stacker that maintain the data on the disk as a compressed data). This is due to the reduced amount of data needing to be written to the disk during the capture process as a result of data compression. 3. System Processor - More powerful CPU chips will allow a greater amount of processing in a given amount of time. Much of the capture process depends on software execution on the system processor, so it follows that more powerful CPU's will yield higher capture rates. This is in fact the case, as you can expect to get higher capture rates on '486 CPU's versus '386 CPU's. 4. System Clock Speed - For the same reasons stated in 3. above higher capture rates will be achieved on systems that have a higher system clock speed. Thus, 50 MHz systems will yield higher capture rates than 33MHz systems which will yield higher capture rates than 25MHz systems. 5. System Bus Speed - Most AT compatibles have adapter card buses which are set at 8MHz. Some compatibles allow the bus speed to be set higher with the tradeoff that some adapter cards may not correctly work at the higher bus speed. If all the adapter cards in your system function correctly at a higher bus speed then you can realize higher capture rates at the higher bus speed. 6. Video Capture Format - During video capture, the VideoSpigot capture driver allows a user to select from among 5 different capture formats: 8 bit palettized (256 colors) 8 bit dithered (256 colors) 16 bit RGB555 (32K colors) 24 bit RGB (16M colors) Spigot Compression (64K colors) YUV 4:2:2 (16M colors) Selection of the YUV 4:2:2 format will normally result in the highest capture rates since the least amount of data manipulation is necessary in this capture format (the native format retrieved from VideoSpigot). However, YUV 4:2:2 format requires twice the amount of file space as 8 bit palettized or Spigot Compression, and on slower hard disks this could result in slower capture rates. You will need to experiment to determine the optimum formats for capture. 7. Video Capture Method - VIDCAP allows a user to select parameters associated with capture in its Capture Video Sequence dialog box. Make sure that you select Capture to Memory. This selection will reduce considerably the degradation that a hard disk will effect on the capture rate. Explanation of SYSTEM.INI Options for VideoSpigot ================================================= The VideoSpigot capture driver retains information about current capture parameters and hardware setups in the Window's SYSTEM.INI file. All options will be recorded in the [spigot.drv] section of the SYSTEM.INI. Current keywords and their possible values are: Keyword Legal Values Description ------- ------------ ----------- CaptureField 0 VideoSpigot will capture only the even fields of an interlaced input signal. For non-interlaced video signals this flag has no effect. 16 (default) VideoSpigot will capture only the odd fields of an interlaced input signal. For non-interlaced video signals this flag has no effect. FullFieldCapture 0 Disable full-field capture 1 Enable full-field capture 2 Force bus performance measurement cycle on next startup of VideoSpigot for Windows. Full-field capture will be enabled or disabled depending on results of bus performance. Hue 0 - 255 Shifts the colors towards green on the low end and towards blue on the high end of the range. The default value is 128. InputChannel 0 Always assume video data is being input on the composite connector of the card. 1 Always assume video data is being input on the SVHS connector of the card. 2 (default) Connector on which active video data is input will be determined via an autodetect algorithm in which the SVHS connector is given priority. Interrupt 10,11,15 Possible IRQ values for the vertical sync interrupt of the VideoSpigot card. LineBreak.1.8 0 - 100 For 1/8 field capture (Width = 80 for NTSC, = 96 for PAL,SECAM): % of scanlines remaining to be read from current frame being captured before capture of next frame is initiated. Default is 20%. LineBreak.1.4 0 - 100 For 1/4 field capture (Width = 160 for NTSC, = 192 for PAL,SECAM): % of scanlines remaining to be read from current frame being captured before capture of next frame is initiated. Default is 20%. LineBreak.3.8 0 - 100 For 3/8 field capture (Width = 240 for NTSC, = 288 for PAL,SECAM): % of scanlines remaining to be read from current frame being captured before capture of next frame is initiated. Default is 20%. LineBreak.1.2 0 - 100 For 1/2 field capture (Width = 320 for NTSC, = 388 for PAL,SECAM): % of scanlines remaining to be read from current frame being captured before capture of next frame is initiated. Default is 20%. LineBreak.Full 0 - 100 For full field capture (Width = 640 for NTSC): % of scanlines remaining to be read from current frame being captured before capture of next frame is initiated. Default is 20%. MemoryBase 0A0000-0FE000 On 8KB boundaries the valid address F02000-FFE000 bases to locate VideoSpigot at. ntscTOF 4-15 The # of lines skipped at the top of the current field being captured when running in NTSC mode. This allows the user to prevent lines of inactive video data from being captured (an example of such non-video data would be closed captioned text prepended to the video data in many recordings). The default # of lines to skip is 11. palTOF 4-15 The # of lines skipped at the top of the current field being captured when running in PAL or SECAM mode. This allows the user to prevent lines of inactive video data from being captured (an example of such non-video data would be closed captioned text prepended to the video data in many recordings). The default # of lines to skip is 15. SpigotTime any Result of bus performance measurement. See FullFieldCapture. The automatic measurement allows full-field capture if SpigotTime is <= 1250. VCR 0 Turn off hardware feature that enables extra gain control for video signals sourced from a VCR. 1 (default) Turn on hardware feature that enables extra gain control for video signals sourced from a VCR. If this feature is turned off, VCR signals may appear to roll. VideoFormat 0 Video data will be captured in the 8 bit Palettized format. 1 Video data will be captured in the 16 bit RGB555 format. 2 (reserved, not used) 3 Video data will be captured in the 24 bit RGB format. 4 Video data will be captured in the Spigot Compressed format. 5 Video data will be captured in the YUV 4:2:2 format. 6 (default) Video data will be captured in the 8 bit Dithered format. VideoStandard 1 Video data is encoded using the PAL video transmission standard. 2 (default) Video data is encoded using the NTSC video transmission standard. 3 Video data is encoded using the SECAM video transmission standard. Width 80 Widths that VideoSpigot can capture 160 (default) when running in NTSC mode. 240 320 640 96 Widths that VideoSpigot can capture 192 (default) when running in PAL,SECAM mode. 288 384