Video Capture

Video Capture

Component Video:
- The video signals are kept separate and treated as separate RGB signals throughout the entire recording process.
- Each of the three chrominance signals (RGB) and the luminance signal remain separate even when recorded on the video tape.
- Common component video formats are Betacam, Betacam SX, Digital Betacam, M-II, and the digital format D-1.
- The vast majority of component systems are professional formats, however the component system is starting to appear in consumer level systems with a number of high end TVs and DVD players now using this method.
Composite Video:
- An analog color video signal system that contains all of the color information encoded into one signal.
  - ie. RGB + luminance as one signal
- Typical composite television systems are NTSC, PAL, and SECAM.
- Typical video camera/recorders are:
  - video-8, VHS
  - S-video:
    - A composite format where the luminance signal (Y) and the chrominance (C) signal are kept separate during the encoding and decoding phase. Both signals are combined and occupy the same tracks on the video tape however.
  - Digital video (DVSoft/digital camcorder):
    - field size restricted to either NTSC format or PAL format
- NTSC (National Televisions Systems Committee - 1953):
  - US video format
  - 720 x 486 horiz.lines = 350K pixels;
  - interlaced 29.97fps = 60 fields/sec;
- PAL phase alternating line video format:
  - 720 x 520 lines;
  - interlaced 25fps = 50 fields/sec;
  - in Australia - "PAL B"
- SECAM:
  - acronym for Système séquentiel couleur à mémoire, the color TV system developed in France and used there, in Russia, and in various other countries in Eastern Europe.

AVI - MS Video for Windows audio-video interleaved format:
- VFW Release 1 included:
  - RLE
  - Video-1 compression - good for palletised talking head clips;
  - VFW Release 1.1 (Dec'93) included:
  - Intel Indeo compression good talking head; poor for action; slow in 16bit;
    - slow compression time;
  - Cinepak compression good for action clips; not so good for talking head clips;
    - fast 16 & 24bit playback; v. slow to compress;
FLI - Autodesk Animator format
FLC - Animation format
FLM - filmstrip
Motion JPEG
MPEG:
- Motion Pictures Expert Group compressed file format - 45secs typical video on 1.44Mb?
- advanced JPEG with motion compensation, changes b/n frames in terms of how far each
- object moved ie. intraframe comp. on key frames & the info remaining in delta frames after
- removing interframe redundancies;
- MPEG1:
  - codes video+sound @ 1.5Mbit/sec to allow 70' stereo video on a CD;
  - limited to 150KBps for playback with early CDROMs
  - 325x288 pixels => at full screen mode looks like poor quality VHS video
  - resolution = half of TV (discards half of fields!);
  - predictive DCT-based system:
  - Picture split into 8x8pixel blocks -> DCT coding ->quantification ->Huffman coding
  - Starts with "I" picture (Intra) & processed as above;
  - Next takes 3rd picture "P" (Predicted) & all blocks are cf I to find best
  - match for each block, which can then be assigned a motion vector.
  - The best match block is used as "best guess" input to prediction part of the coder. The difference signal then is processed as above & the final signal with Huffman motion vector signal is stored.
  - Decoder will then generate the middle picture "B" (Bidirectional interpolat.);
- MPEG2:
  - marketed with DVDs as .vob files which normally cannot be copied to disk without DVD ripper software.
  - 720x480 pixels => better quality video than laser discs
  - embedded Dolby AC-3 surround sound audio for 5 speakers & subwoofer
Quicktime:
- Macintosh video sequencing format (.MOV)
VideoCD:
- a VCD has the capacity to store 74/80min (on 650Mb/700Mb CD's) of MPEG1 video (VHS-like quality) and stereo sound
- a VCD can be played on most standalone DVD players as well as any computer with a CDROM or DVD with the help of decoding software
- PAL: 352x288 25fps MPEG1
- NTSC: 352x240 29.97fps MPEG1
SuperVideoCD (SVCD):
- PAL: 480x576 25fps MPEG
- NTSC: 480x480 29.97fps MPEG
DVD video:
- PAL: 720x576 25fps MPEG2
- NTSC: 720x480 29.97fps MPEG2
Digital Video (DV):
- format used in DV video cameras which can be transferred to computer disks directly via a firewire link cable without need for capture card

Video file sizes and transmission rates:

Format (all at 30fps)	Data rate Mb/sec	Compression ratio	File size for 1min video (Mb)
Uncompressed 720x480	20	1:1	1,200
M-JPEG 704x480	3	7:1	180
M-JPEG 352x240	1	20:1	60
MPEG-1 704x480	0.45	45:1	27
MPEG-1 352x240	0.15	138:1	9
MPEG-1 176x120	0.05	400:1	3

Multimedia file conversion:

video files:
- AVI to MPEGI/II:
  - seems the best is TMPEGEnc which is a freeware product, although the Panasonic encoder is especially good for videoCDs although it is not as sharp and poorer color saturation but less blocky
  - see also http://vcdhelp.com which is a great site with lots of information of how to create VCD's and rip DVD's
  - MGI Wave IIII seems to have audio sync problems
  - creating videoCD using TMPEGEnc to get best quality:
    - capture with AVI_IO using StudioVCD and Huffyuv at 768x576 PAL 25fps
    - open in VirtualDub with Swap Fields, Smart Deinterlace Filters & resize to 352x288
    - Frame serve to TMPEGenc
    - load the VCD PAL profile
    - configure: Video/Motion Search Accuracy - High Quality (slow)
    - Advanced/4:3 Display(keep correct AR for full screen) Crop Video (get rid of blurred edge from letterbox) Noise Reduction (if Video Tape)
    - Quantizer Matrices / Use Floating Point DCT
    - 1min takes 5:40 on PIII/800
- Real Media to AVI:
  - there is a freeware program called TINRA available at http://evc.to/tinra
  - as frame rate of RealVideos is variable and AVI is not, audio sync problems occur
  - this program will also convert RealAudio sound files to sound only AVI's which can then be converted to WAV by using a program like AVI2WAV ( http://www.geocities.com/SiliconValley/Way/9629/avi2wav.htm )
  - requires RealPlayer G2 to be installed
- DVD to MPEGII:
  - seems the best DVD ripper is SmartRipper and a close 2nd appears to be DVD Decrypter

Glossary:

Interlaced:
- image frame consists of 2 scans ("fields") each having the alternate horizontal line
Non-interlaced ("Progressive"):
- image frame consists of 1 scan => more accurate especially for motion
Frame:
- A complete scan of an image. In interlaced systems, a frame is two vertical scans (fields). In non-interlaced or progressive-scan systems a frame is completed in one vertical scan.A frame is one complete image of a motion video sequence. In motion video, images are presented typically at 30 frames per second to create the illusion of continuous motion.
Field:
- In interlaced television systems, a field is one vertical scan period, typically scanning alternate lines of the complete image (frame). Thus, it takes two fields to scan the complete frame.
High-Definition Television (HDTV):
- A new broadcast format designed to replace NTSC/PAL. HDTV production requires the use of cameras and recording equipment designed for the format. An HDTV television set is required to view the improved resolution.
- The aspect ratio of HDTV is 16 x 9, versus 4 x 3 for NTSC TV.