HDV
HDV is an entry-level format for high definition video recording. HDV uses DV tape providing a cost-effective HD production solution, compared to previously developed HD formats.
The HDV format was developed by JVC and Sony to offer an affordable upgrade path from standard-definition DV video to high-definition recording. The format was initially supported by Canon and Sharp, and in September 2003 the four companies formed the HDV consortium.
Most HDV camcorders use MiniDV/DVC cassettes that are also known as "S-size" tapes. Some shoulder-mounted camcorders are also capable of recording onto larger DV cassettes known as "L-size" tapes. The recording time is the same as DV Standard Play. Unlike DV, HDV does not support Long Play speed.
Two major versions of HDV are HDV 720p and HDV 1080i. The former is used by JVC and is informally known as HDV1. The latter is preferred by Sony and Canon and is often referred to as HDV2. The HDV 1080i defines optional progressive recording modes, and in recent publications is often called HDV 1080 or 1080-line HDV as progressive 1080-line recording becomes commonplace.
HDV 720p
JVC ProHD camcorder
HDV 720p closely matches broadcast 720p video standard in terms of scanning type, frame rate, frame size, aspect ratio and data rate. Presently, JVC is the only manufacturer of HDV 720p camcorders. JVC was the first to release an HDV camcorder, the handheld GR-HD1. Later JVC shifted its HDV development to shoulder-mounted cameras.
JVC did not leave the market of consumer high definition camcorders, but instead switched to tapeless recording technologies. For example, its Everio GZ-HD7 camcorder, introduced in January 2007, can record MPEG-2 video to either built-in hard disk drive or to a Secure Digital memory card. JVC chose to use 1080i format instead of 720p on its consumer camcorders, with data rates, frame rate and frame size matching or exceeding parameters of HDV 1080i video. This format is informally known as TOD.
A common misconception is that JVC developed a proprietary extension to HDV called ProHD, featuring film-like 24-fps progressive recording mode and LPCM audio, for professional use. JVC states that ProHD is not a separate format, but "an approach for delivering affordable HD products," and stresses that 24-fps progressive recording and LPCM record/playback has always been part of the HDV format specification.
HDV 1080i
Canon XL-H1
HDV 1080i is an interlaced video standard. Interlaced video has been a useful compromise for decades due to its ability to display motion smoothly while reducing recording and transmission bandwidth. Interlaced video is still being used in acquisition and broadcast, but interlaced display devices are being phased out.
Modern flat-panel television sets that utilize plasma and LCD technology are inherently progressive. All modern computer monitors use progressive scanning as well. Interlaced video must be converted to progressive before it is displayed on a progressive-scan device. The process of converting interlaced video into progressive is known as deinterlacing. Progressive-scan television sets employ built-in deinterlacing circuits to cope with interlaced broadcast signal, but computers rarely have this capability. As such, native interlaced video must be deinterlaced if it is intended for viewing on a computer.
Some HDV 1080i camcorders are capable of recording progressive video within an interlaced stream, provided that the frame rate does not exceed half of the field rate. The first HDV 1080i camcorder to implement such Progressive Scanning was the Sony HVR-V1. To preserve compatibility with interlaced equipment the HVR-V1 records and outputs video in interlaced form. 25-fps and 30-fps progressive video is recorded on tape using progressive segmented frame technique, while 24-fps recording employs 2-3 pulldown. The camcorder offers two variations of 24-fps recording: "24" and "24A", the latter is also known as "24pSCNA". With the 24P mode the camera ensures that there are no cadence breaks for a whole tape, this mode works better for watching video directly from the camera and for adding "film look" to interlaced video. In the 24A/24pSCNA mode the camera starts every clip on an A frame and with timecode set to an even second margin.[7][8]
Prior to the HVR-V1, Sony was offering Cineframe, essentially an interlaced-to-progressive converter, to simulate film-like motion. The conversion process involved blending and discarding fields, so vertical resolution of the resulting video suffered. Because the camera was still shooting in interlaced mode, slower frame rate did not improve light sensitivity. In addition, motion, produced in the 24-fps variant of Cineframe was too uneven for professional use.[9] The same or better film look effect could be achieved by converting regular interlaced video into progressive format using computer software.[10]
In 2007 Canon commoditized progressive scanning, releasing the HV20 camcorder. The version for 50Hz market featured PF25 mode with PsF-like recording, while the version for 60Hz market had PF24 mode, which utilized 2-3 pulldown scheme. The HV30, released in 2008, implemented additional PsF-like PF30 mode for 60 Hz markets. Output is performed via component, HDMI and FireWire in interlaced form.[11]
To achieve full vertical resolution without introducing interlace artifacts the progressive scan video must be properly deinterlaced. 25P and 30P video must be deinterlaced with "weave" or "no deinterlacing" algorithm, which means joining two fields of each frame together into one progressive frame. 24P video must go through film-mode deinterlacing also known as inverse telecine, which throws out judder frames and restores original 24-fps progressive video.
HDV 1080p
Native Progressive Recording compatibility mark
There is no separate standard for HDV 1080p recording. Instead, the HDV 1080i specification defines optional progressive recording modes, which are getting popular as the users become more interested in shooting digital movies and creating content for the Web. All HDV camcorders that can record in 1080p format can also record native 1080i video.
HDV camcorders, capable of 1080-line progressive video, record it at rates of 24 fps (actually 23.98 fps) and 30 fps (actually 29.97 fps) for 60 Hz markets, and at 25 fps rate for 50 Hz markets. Video is output as true progressive video via an i.LINK/Firewire port. Output through other ports is performed in interlaced mode to preserve compatibility with existing interlaced equipment.[12][13]
The first 1080-line HDV camcorder to offer recording in native progressive format was the Canon XL H1, introduced in 2006. It was followed by the the XH-G1 and XH-A1. The camcorders generate and record progressive video from interlaced CCD sensors, which precluded Canon from calling these modes truly progressive. Instead, Canon reused the term Frame mode and a shorthand suffix "F" instead of commonly accepted "P" from its standard definition model lineup.[14] The camcorders can be made 50Hz/60Hz switchable for additional fee.[15]
In 2008 Sony announced HVR-S270, HVR-Z7 and HVR-Z5 models, capable of recording in what Sony have called Native Progressive Recording mode. The new Sony camcorders use true progressive sensors, which allowed Sony to claim superiority over Canon models by saying that Native Progressive Recording has been called 24F/25F/30F in some camcorders, while actually using interlace CCD imagers.
In January 2009 Canon announced the HV40 as a replacement for the HV30 consumer camcorder. Aside of minor cosmetic changes, the 60 Hz model sports 24-fps native progressive recording and uses the same progressive-scan CMOS sensor as the models it replaces.[16]
Sony designed a "Progressive" logo for the camcorders that are capable of native progressive recording. Canon has no special logo to identify cameras that can record in frame modes. Despite of the differences in branding, 24F/25F/30F modes offered by Canon and Native Progressive Recording offered by Sony are compatible. In particular, the XH-A1/XH-G1 Operation Manual states that 24F/25F/30F video is recorded to tape according to HDV native 1080p specifications.
Aside of aforementioned camcorders, other products that are capable of reading and recording in native progressive format are the Sony HVR-M15AU, HVR25AU,[17] HVR-M15AE, HVR25AE[18] and HVR-M35 HDV videocassette recorders, and the Canon HV20/HV30 camcorders when used in tape recorder mode.
HDV compression
HDV is based on MPEG-2 video, which compresses data both within each frame (intraframe/spatial compression) and between frames (interframe/temporal compression). This is the same type of compression used for DVD video and many network TV broadcasts, and allows HDV to achieve high spatial resolution at low data rates compared to other HD recording formats. HDV 1080i uses a recording data rate of 25 Mbit/s (3.125 MB/s) while HDV 720p records at 19.7 Mbit/s (2.46 MB/s). In either case the data rate is constant because the recording media — tape — is transported with constant speed.
Using MPEG-2 video enables HDV to achieve a higher compression ratio than recording formats without interframe compression, but constant data rate limits the video quality in scenes with lots of detail, rapid movement or other complex activity like flashing lights. Such scenes may exhibit visible artifacts such as blockiness or blurring, depending on the amount of movement and on the algorithm employed in the encoder. On contrary, when using a tapeless recording standard such as XDCAM or AVCHD it is possible to increase or decrease the data rate depending on complexity of a scene, thereby preserving quality of a busy scene while saving recording space on a low-detail scene.
Interframe compression also means that a recording dropout in HDV can affect several frames of video rather than just one, since the compression introduces dependency between frames. Hence it is best to use high-quality tapes for HDV recording such as "master quality" Mini-DV or specially formulated HDV tapes, and limit the re-use of tapes. Some users have gotten good results with HDV using generic Mini-DV tapes, but this is not generally recommended.
For audio, HDV uses MPEG-1 Layer 2 compression to reduce the audio bitrate to 384 kbit/s, compared to 1536 kbit/s for DV video and 1411 kbit/s for audio CDs. This makes HDV audio less desirable for situations where sound quality is critical, but MPEG-1 audio at 384 kbit/s is considered 'perceptually lossless.' For general video recording with an on-camera microphone, HDV audio is not a significant limiting factor.
It is important to view HDV's compression limitations in the proper context. Other HD codecs using lower compression ratios need more bandwidth and storage capacity for a given amount of video, requiring significantly more complex and expensive recording solutions. For example, the Panasonic DVCPRO P2 cameras use memory cards which can cost over $50 per minute of recording capacity, compared to a few dollars per hour for Mini-DV tapes. This cost differential has helped make HDV a popular HD recording format for consumers, independent videographers and low-budget TV programs. Another cost-efficient HD recording option became available in October 2008 with a release of the Panasonic HMC150 camera, which records onto readily available SDHC memory cards using the AVCHD encoding.