Short Films Page 9
RESOLUTION
Resolution is a term used to describe video image quality. It refers to how much detail an image contains and therefore its sharpness. It is widely used when describing the specifications of any technology that deals with image acquisition, storage, transfer or display.
Resolution is normally expressed in either numbers of lines or numbers of pixels and used to describe digital and analogue images alike.
Having a basic understanding of the resolution of different formats will be extremely useful to you because the image quality of your film will depend on which format you choose for acquisition and how you maintain that resolution through your post-production workflow. Although the technical specifications of formats can seem rather daunting to the novice, gaining understanding of how resolution and quality is described will be essential to you as a filmmaker.
Lines
Resolution within most filmmaking processes will be referred to in lines of resolution, generally in lines per image size, rather than per mm or inch. This is because the image size of a format can vary; television screens vary in size and so do projected images, whilst still containing the same image specifications. The amount of lines an image consists of is also not dependent on its aspect ratio.
In the simplest terms video images of any format are constructed from a tight weave of black and white lines that respond to input signals to create images. The tighter the weave of lines the sharper the image will be. A tighter weave needs more lines running both horizontally and vertically to create it.
An easy way to look at resolution lines is that the more of them there are the higher the image quality will be. This means that there are going to be two figures that affect an image’s resolution, the number of lines running horizontally and the number of lines running vertically.
However, the amount of vertical lines that an image contains is fixed according to format and cannot change; it is only the amount of horizontal lines that can fluctuate.
So for example, two digital video cameras of the same format will have the same amount of vertical lines per image, but one may offer higher quality because it has more horizontal lines, therefore more pixels and so a higher resolution.
In television broadcast the lines involved are often a greater number than the ones that actually make up the image, so for instance a standard-definition NTSC TV image would contain 525 lines yet only 480 would make up the picture, as the rest of the lines are used to store information and so are not active.
Pixels
If you imagine that the lines of an image are forming a weave of horizontal and vertical lines, then the tiny squares that this creates are referred to as pixels.
Some image resolutions are given just as pixels or in pixels per image size. These figures are often expressed in Mega pixels, which is a figure calculated by multiplying the number of vertical lines by the horizontal lines and dividing by one million.
Why use a higher resolution?
Higher resolutions give sharper, clearer pictures that lend themselves better to being blown up in size via projection. Because filmmaking involves so many different stages, it is often not possible to record, edit and screen video images at the same resolution.
It’s common for video cameras to offer capture resolutions greater than the resolution of the format they will be recorded on, or for television series that will only be displayed at standard TV resolution to be shot on 35mm film. This opens up the obvious question of why use a higher-resolution format if the final display resolution will be much lower? When a higher resolution is down converted to a lower one, much of the initial visual quality is captured, even though the resolution will be no different; the quality of an image shot on 35mm and then put onto MiniDV will look much higher in quality than an image shot on a DV camera although they will still be the same resolution.
PAL/NTSC
PAL and NTSC are two types of television broadcast systems. Although technically a type of colour encoding, they are broadly used to describe the line resolutions that each of them use: PAL images being 720 × 625 and NTSC 720 × 525, with PAL running at 25 frames, 50 fields a second and NTSC at 30 frames, 60 fields a second. NTSC has been the North American format and PAL the European. Although this technology is being phased out by HD it’s important to bear in mind that the two don’t really mix. So if you’re sending a copy of your film to a festival abroad or using a camera from another country, it’s important to check whether you are dealing with PAL or NTSC.
STANDARD-DEFINITION (SD)/SDTV
Standard-definition refers to a standard resolution for broadcasting and video capture formats, regardless of aspect ratio. This is an important point to bear in mind; just because an image has a 16:9 widescreen aspect ratio it doesn’t mean that it is necessarily high-definition.
Conventionally, standard-definition has been the standard for all digital video formats and for digital television broadcast, normally with a 4:3 aspect ratio and a resolution of PAL 720 × 625 (576 active) or NTSC 720 × 525 (480 active).
HIGH-DEFINITION (HD)/HDTV
High-definition is a higher-resolution format that has begun to supersede standard-definition, both as a digital broadcast standard and as a digital video format in its own right. It offers a huge leap in resolution quality, with almost four times the resolution of SD. Although only a few channels offer HD broadcast at the present time the entire film and television industry is on the verge of switching over to purely high-definition environments with the majority of productions now being shot and post-produced on an HD format even if they are still broadcast at SD. This means that everything from cameras to tapes to televisions are changing from the old SD format to the new HD format and once this has happened SD will be obsolete. Because this is a relatively new technology there are currently several versions of HD resolution and a fair amount of confusion over which is true HD. Currently HD is split between the options of 1280 × 720 or 1920 × 1080. Both are HD formats and many television channels will broadcast both. Digital cameras and tapes also offer one or both of these variations. To complicate matters further another factor is how these resolutions are scanned.
Figure 21. common resolutions, cour tesy of www.wikipedia.org
Interlaced scanning
Interlaced scanning is a process that was devised for broadcast of television images. It works by splitting a single frame of an image into two fields, with odd and even lines making up each. In a sense each one is only really half the image and each half image is interlaced with the next to create the whole until it is refreshed with the next two half fields. With NTSC the picture is refreshed 30 times a second giving 60 half fields, and with PAL 25 times a second giving 50 half fields. These are referred to as 60i and 50i. Although interlaced scanning has provided a valuable method for video and television capture and broadcast for a long time, the disadvantages such as the jagged edges and poor image quality have not made it ideal for filmmaking.
Progressive scanning
Progressive scanning is a newer concept that offers greater image quality by scanning complete fields from top to bottom instead of the half fields generated with interlaced. With progressive scanning in NTSC, 30 whole frames are generated a second and with PAL, 25 frames a second, also referred to as 30p and 25p.
24/25P
Apart from the potential image quality problems associated with interlaced scanning, the reason why progressive scan technology is so appealing to filmmakers is that 24 or 25 frames a second is exactly the same rate that film cameras and projectors capture and project films. 24/25 frame rates have a very specific visual quality that we associate with film, and watching video at this scan rate gives a film feel to the images. By adding only five or six more frames to take it up to an NTSC frame rate, the image gains a very different quality that we associate with television and video rather than film.
The first cameras to start capturing 24p images were developed by George Lucas, in an attempt to develop a video image that beha
ved like film. This technology has spread and has been available in prosumer cameras. There is also a version of 24p known as cineframe. This is not true 24p, it merely gives a similar visual effect and is found on many prosumer and domestic video camera ranges. HDTV broadcast also gives various options on receiving interlaced or progressive images.
1080p/1080i/720p
These are all HD resolutions with interlaced or progressive scans. At present these are the main three options that are offered in most HD technology, whether cameras or monitors. One of the problems with progressive scanning is that because it is generated from full frames rather than half fields, it requires larger files to be generated much faster. This is particularly a problem for video camera technology where the tape formats that have been viable for SD video are not really capable of handling HD progressive resolutions. As a result many current video cameras offer a variety of HD solutions, so 1080 interlaced or 720 progressive, often with a choice between them in a single camera. 1080 progressive is regarded as true HD and is the ideal for most filmmakers. It is also becoming available as an option in prosumer cameras that record straight to disc or storage cards as well as professional cameras that record to disc or very high-resolution tape formats.
As disconcerting as so many options can be, the bottom line for the moment is that they are all HD formats and look much, much better than standard-definition video. Ultimately it’s a trade-off between quality and practicality: what you lose in quality you might gain in ease and speed of use.
HDV
HDV is a technology offered by various video manufacturers that records HD resolution images onto DV or MiniDV tapes using MPEG2 Gop (group of images) compression. Because HD images are so high in data compared to DV, more compression is used to encode the images onto the small tapes. So although HDV technology technically offers much higher-resolution images, the images are far more compressed than standard-definition video captured on DV. However, the overall amount of data with HDV is far higher that SD video. Gop compression works by not recording every frame completely, just key frames, and then recreating the frames in between those from highly compressed data. This recreation process often occurs when capturing the footage onto a computer-based editing system, for example.
COLOUR SPACE/SAMPLING
The colour information or colour space of most video images is recorded by using varying intensities of red, green and blue (RGB) information. Three separate channels record intensities of the individual RGB information and then they are combined, making it possible to approximately recreate any colour that the camera picks up. The problem with RGB colour space is that the amount of information generated for the three different channels of colour is huge, so again, as with most digital video formats, this needs to be compressed to fit onto a tape or easily transferred. For this, most digital formats use chroma subsampling. With subsampling the RGB data is transferred into three components, but instead of them being three colour channels like RGB, the first of these components is a luminance (luma) and the other two are chrominance (chroma). Luma is the light and dark information and chroma is the colour information. Subsampling reduces the RGB data by preferencing the way the human eye actually sees; the eye is much more sensitive to luminance than it is to colour, so with subsampling there is always the full amount of luma data and then often much smaller amounts of the two chroma datas. In digital video, the three amounts of data of the three components are expressed with figures such as 4:2:0 or 4:1:1. The first figure is the luma, with 4 being the full data and 1 being a quarter of the full data. So for instance with 4:1:1, the two chroma figures after the luma have only a quarter of the amount of data as the luma. This is not necessarily as drastic as it may sound, with the image still looking convincing to the eye, but consisting of far less data than an RGB image. Without needing to understand the intricacies of colour space data, you can use these numbers as a guide to see how much colour information a format offers. Colour space is extremely important for high-end filmmaking, with processes like grading or chroma key effects relying heavily on the need for as much colour info as possible and, as you will see, different formats offer vastly different amounts.
VIDEO CAMERAS
Video cameras work on the same principle as film cameras, but instead of the light absorbed through the lens being channelled onto light-sensitive film it is channelled onto a light-sensitive chip or chips. These chips then convert the light information into electrical signals. Cameras designed to record a certain format of digital video such as DV or HDV often vary in quality through the features they offer. The type of lens, light sensitivity and amount of chips all affect the quality of the images they capture regardless of what format it is in.
Monitoring
One of the great advantages of shooting on video is that you can actually see the images you are getting as you record them. Although most video cameras offer LCD view finders and additional flip screens, both of these, due to their size, often don’t allow the user to see more than enough detail to judge composition and framing. To judge image information such as brightness and contrast more accurately, the video camera is normally hooked up to a field monitor, which is large enough to give sufficient information, but still feasible to move around on location.
CCD chips
CCD is the general type of chip used in video cameras; their size and shape is directly linked to the quality and shape of the image they produce. Domestic video cameras tend to just use a quarter inch CCD chip, while prosumer cameras will use 1/3 inch chips and professional cameras use 2/3 inch chips.
The higher-quality cameras will also use three chips rather than just a single chip, recording more information by splitting the light beam into red, green and blue and sending that information to an individual chip. The actual shape of the chip will also determine the aspect ratio of the image with standard-definition chips recording a 4:3 aspect ratio and high-definition cameras recording a 16:9 aspect ratio. New video cameras are also beginning to use CMOS chips, which are larger and more sensitive than CCD chips.
Tapeless
The first generation of cameras to bypass tapes completely has arrived. Storage cards that could only store small standard-definition video images have now been developed to capture large high-definition video. Portable hard drives are also being used instead of tape formats, allowing pure digital file workflows at very high resolutions. This will only continue in the future with videotapes becoming obsolete.
Figure 22. Panasonic HVX200 records straight to P2 memory cards or fire store hard drive, eliminating the need for tapes, courtesy of Panasonic.
VIDEO THAT LOOKS LIKE FILM
Video has always had a certain look associated with it that is different to film, both in quality and feel. As discussed, the main factors that have differentiated film from video have been frame rate, resolution, colour space and shallow depth of field. Despite numerous attempts at filmising video throughout the years, one or more of these factors has always meant that the results have been unconvincing.
In the last few years, however, HD has meant that video footage has obtained nearly all the qualities of film. Though still a long way from the resolution of 35mm, HD resolution cameras using progressive scanning technology have closed the gap considerably. Until recently the technology involved had been reserved for cameras that, due to their cost and demand, remained purely in the professional domain. Now ranges of prosumer cameras are starting to appear in the market that offer the same quality; while still expensive, they are nowhere near the price of the pro cameras. The HD capability of these cameras, coupled with developments such as the 35mm lens adapters (see lenses), means it is now possible for short filmmakers to shoot video that potentially rivals the look and feel of film for a fraction of the cost.
The final hurdle for independent filmmakers to compete with the technical quality of big-budget features really remains in the screening of these films. Films screened from 35mm prints have always been far better quality than
any type of video projection. This has been a stumbling block for most filmmakers who don’t have the amount of money needed to create a 35mm print. Standard-definition video, because of its low resolution, might look fairly good on a small screen, but when blown up to cinema screen size typically looks quite poor in comparison with film. But with the advent of digital film workflows and technology, cinemas are beginning to take the first steps into transferring to high-resolution digital projections. Films can then play from hard discs and digital tapes with outstanding results. Inevitably this will become the standard screening method for new films, offering independent filmmakers the possibility of producing incredible quality films on tiny budgets.
DIGITAL VIDEO FORMATS AND TAPES
Video formats are often confusing, with manufacturers introducing variations of tape for a single format. Cameras for those formats in turn offer their own unique strong and weak points, all of which affect the quality of the footage you can record with them.
DV
DV is a standard-definition format recorded onto DV tapes either mini or standard size. There are three main types of DV tape available within the format.
DV
DVCAM
DVCPRO 25/50
DVCAM and DVCPRO are professional versions of DV and offer faster tape speeds resulting in more reliable capture and playback, as well as improved colour space compared to standard DV. They are, however, all DV, so they use the same codec to capture and store the data and all have the same resolution. This format is widely available; cameras can be bought or hired comparatively cheaply.