The public – including filmmakers – associates the photographic image with high quality. And that amorphous complex that we know as "the customers", wants their speeches and products to enjoy high quality images... obtained, of course, with barely enough budgets to work in video.
The intention of this article is to try to define the differences between the look of cinema and the discredited look of video and television, and thus provide the reader with some tools that allow him to take control of his video production processes with a cinema look.
Two ideas to get you started
Well, the market asks us to generate video images that look like cinema, a difficult but not impossible mission. To address this problem, we must start from the following two premises.
The first points out that careful work allows to obtain high-quality video images. The aesthetic distance between film and video can be significantly reduced if the working systems of film production are adopted, designed so that each shot is recorded in the best possible way. The "news look", hard, flat and pale, comes hand in hand with hasty work.
In fact, any ENG-type video camera can deliver images that can eventually lead our customers to say "it's great... it looks like cinema!" ..., as long as we have enough time, adequate lighting, good optics and perhaps some of the photographer tricks we talked about above.
Obviously, better results will be obtained with higher quality equipment and technical personnel: someone who knows how to "move the buttons" correctly can "paint" the video signal of a camera to achieve what they want. And in many cases you also have to make a special effort to choose the appropriate equipment for each situation.
The second premise states that video and film cannot be directly compared. We compare video and telecine, which is film-transferred-to-video using a complex digital process that usually includes rigorous color correction, shot by shot. When we receive a telecine tape, we have images that have benefited from a significant effort to get the best out of the raw material available. And, in addition, this effort allows to correct or mask serious photography defects.
With the above we want to emphasize that the post-production of the cinematographic image begins in the development and is perfected in the telecine. In fact, experienced cinematographers often use chemical and electronic tricks to obtain some particular effect. But when post-production of a video-based project includes a color correction session with dedicated high-quality equipment, things change significantly.
The Four Small Differences
Despite the above, high-quality video is still different from cinema. We can all detect the origin of the material with a single glance. But, exactly, what makes film and video so easily differentiated?
Taking as a reference the latest generation of 35 mm film stocks, we could reduce the perceptual differences between cinema and video to the following factors:
1. The movie offers a higher contrast ratio than the video. The most recent film stocks can handle contrast ratios above 100:1, while the best video cameras available today do not exceed a 30:1 ratio.
This translates into the possibility of preserving detail, both in dark areas and in large lights, while it is still feasible to capture a good amount of intermediate tones and obtain greater color saturation. The film offers excellent results when it comes to capturing images with large variations of light within the field, situations in which video cameras – even the best quality ones – tend to generate saturated signals, without detail in the darkest parts and with a very abrupt staggering of the intermediate values.
Keep in mind that the inability of video to reproduce the richness of tones captured by the film, is not only given by the performance of electronic cameras, but by the limitations of current standards of video recording. CCIR 601 signals carry digital video with 4:2:2 sampling and a color depth of 10 bits, offering bandwidth that may be insufficient to preserve all the subtleties recorded by the film. The use of analog formats or lower quality digital standards (such as the ubiquitous DV), will obviously increase the visible effect of these limitations.
A good quality telecine work allows to make a very efficient translation of the film image in the color space of the video, making the most of the possibilities of the video signal to transport information. Most modern telecine equipment records the photographic image as a signal with a spatial resolution between 2 and 8 times higher than that of HD video, with 4:4:4 sampling or higher and with a color depth that can exceed 24 bits for each channel. From these images with oversampling, the processes of color correction and transfer to recording media with 4:2:2 sampling are handled. And it's really not very difficult to get top quality results in those conditions.
Thanks to oversampling recording and DSP processing, the most modern video cameras allow you to adjust a gamma control to optimize the response of the camera within a certain luminance range, but until today it has not been possible to match the performance of the film in this aspect.
This procedure improves the distribution of the luminance values of the original image in the limited color space of the CCIR 601 video, but even the most sophisticated electronic cinematography equipment has problems handling lighting situations that could be solved on film with a small iris adjustment.
One of the most important restrictions for DSP processing in video cameras is that they have to manipulate the signals in real time, while in the off-line work of modern telecine systems there is no pressure to process signals with high speed, nor practical limits for the computing capacity that you want to invest in performing these tasks.
2. The movie has a much higher resolution than the SD video. In fact, the 35mm film far exceeds the effective resolution of the HD cameras available today.
The public's reactions to variations in effective resolution are puzzling, to say the least. When doing subjective tests, it has been found that the typical viewer associates definition with image values such as contrast, color saturation, or use of filters.
More modern video cameras emulate the behavior of the film and offer images with softer edges, which do not give the effect of "line-that-separates-colors", usually associated with the "video look". This is achieved through the selective diffusion of luminance or some colors, DSP technology that has been applied for several years in functions such as skin-detail filters used to reduce detail in skin colors. Unfortunately, the result of this processing does not always offer adequate results, since its success depends, directly, on the degree of detail of the image.
3. While the video cameras capture the image using still arrays of sensor elements, the film records the image over an irregular collection of photosensitive particles.
The presence of visible grain in the image is one of the determining factors of the filmlook. Paradoxically, while film makers strive to reduce the visibility of grain in their film formulations, filmlook fans are concentrating on recreating it in their video-originating images.
In practice, the visible grain does not survive the telecine. The grain only appears in the interlaced video in cases where its appearance has been intentionally sought. However, the presence of grain in the photosensitive emulsion is responsible for most of the peculiarities of the photographic image.
Thanks to the high density of the grain present in the film it is possible to obtain an excellent record of details, and thanks to its irregularity it is impossible for the cinema to present moire effects. Precise control of grain distribution in the emulsion makes it possible to improve the emulsion's speed, its exposure latency, and therefore its contrast ratio. The spatial non-correspondence of the granules that correspond to each area of the image, their change between frame and frame, seems to be responsible for the high-definition sensation generated by the film, even after it is transferred to video.
The most recent DSP processes are relatively successful when it comes to grain reproduction. Manufacturers do not insist much on this point because the results are more like the noise characteristic of analog video than the effects of grain in the photographic image.
4. Cinema-to-video transfer, both in PAL and NTSC, produces characteristic effects on the image.
This is another paradoxical case: the defects that are produced by the periodic interleaving of fields to reach 30 fps are associated with good quality images. The slight loss of fluidity in movement – if it is NTSC – or the slight strobe effect that is perceived in PAL, seem to have an effect similar to that of exposure to a 24 fps projection in a dark room, and this reinforces in the viewer the feeling of facing a cinematic experience.
Subjective tests show that material recorded on DV camcorders at 24 fps and converted to 30 fps by a simple electronic pulldown is accepted by the unsuspecting viewer as film-originating material. Apparently, the entertainment industry has capitalized on this phenomenon with DVD-Video players, which have the ability to apply the same conversion procedure from 24 to 30 fps "hot", thus reinforcing in the viewer the feeling that he enjoys a cinematic experience.
Real-world options
A review of the previous points allows us to affirm that it is not yet possible to achieve the ideal filmlook in the output of the video camera. DSP technology, in its current state, does not allow to generate "live" material that has a quality equivalent to that of a good telecine. So, in the same way that images originating in cinema require the intervention of a colorist to reach their full potential, "raw" video requires very careful handling during filming and some kind of subsequent treatment to be able to resemble the photographic image.
The industry offers filmmakers several routes to producing video-looking film. The most attractive is electronic cinematography, which allows to obtain HD video recordings similar to distribution copies in 35 mm, although to obtain the correct look they usually require a special effort of lighting and filtering, and in many cases a subsequent processing of the material is resorted to to simulate the visible grain.
The second option is the ex post processing of the material recorded in SD, either through plugins, pieces of software that can be "inserted" into the editing systems,
or through business processes that offer very interesting results. These systems simulate the effects of grain, pulldown and compression of whites, which is the most visible effect of the highest contrast ratio of cinema. Although this type of procedure offers very satisfactory results, in most cases its effectiveness is limited by the characteristics of the camera that has originated the material. And, as always, by the expectations of users.
This type of process offers first-rate results when applied to material originated with high-end cameras that include sophisticated DSP functions, which allow to emulate the behavior of the most frequently used film stocks. Despite the high cost of operation of this type of camera, the cost/benefit ratio of a production flow based on this scheme is still much more attractive than "native" production in 35 mm.
Another alternative is production based on HD or SD cameras with 24 fps logging. However, low-cost cameras that offer 24fps operation may not achieve image quality that allows for satisfactory results with software-based processes.
Finally, there are companies that offer rental and sale services of modified SD cameras to get a satisfactory filmlook without much effort. The intervention in the cameras, generally, is based on the manipulation of the encoders and the firmware that controls the DSP modules, complemented almost always with the addition of high quality optics.
Either way, the choice that is made must go through the entire process of realization, from the production design to the distribution of the finished product. The most important thing to remember is that there are no magic formulas, and that it is virtually impossible to achieve the look of cinema in a single step.
Any of the options you take will require judicious photography work, which will probably need to include some sort of optical filtering, special camera adjustments, and very careful exposure control, all of this to pave the way for you to get the best results with post-processing of the material.
Leave your comment