Recent from talks
Film look
Knowledge base stats:
Talk channels stats:
Members stats:
Film look
Film look (also known as filmizing or film-look) is a process in which video is altered in overall appearance to appear to have been shot on film stock. The process is usually electronic, although filmizing can sometimes occur as an unintentional by-product of some optical techniques, such as telerecording. The process has the opposite result to VidFIRE, used to restore a video look to telerecorded video.
Today, most digital cinematography video cameras in use feature 24p format (24 progressive frames per second).
When shooting with old cameras, 50 Hz interlaced video (usually used with most forms of PAL and SECAM) can be relatively easily processed to give 25 progressive frames per second, which is the framerate that the PAL/SECAM telecine process also uses. Every two video fields can be "blended" together, every other field can be decimated and the remaining fields can be shown for double the length (this noticeably reduces vertical resolution), or a motion estimation process can be applied to achieve one frame out of every two fields. This technique is sometimes called Field-removed video or FRV. Some modern PAL video cameras do offer the ability to produce 25 frame-per-second progressive video, negating the requirement of post-processing the video to get a temporal similarity to film.
On the other hand, it is much more complicated to convert 60 Hz interlaced video (used with NTSC and PAL-M) to a framerate resembling that of film. Doing the same as PAL/SECAM filmizing will yield 30 frame-per-second video, which is significantly faster than film. Two out of every five fields could be dropped (and 3:2 Pulldown can be applied to the remaining fields), but any motion after this process will look very uneven. Sophisticated computer motion estimation and field blending is usually used to convert NTSC video to 24 frames-per-second – something which could not have been done until recently, and still does not yield as realistic results as PAL filmizing conversion.
Many computer editing programs can de-interlace video to give it more of a film look. An interlaced frame is actually the combination of two fields, one providing the odd-numbered scan lines and the other the even-numbered. Interlacing results in a type of motion blur known as "combing", and also shows "interline twitter" where vertical details approach the resolution limit, neither of which occur in film. De-interlacing can remove or reduce these artifacts, resulting in an appearance closer to that of film.
Some inexpensive consumer editing programs achieve de-interlacing by deleting one of the fields.[citation needed] The result gives half the vertical resolution of the original frame, and sometimes adds a jagged effect to the picture.
For each frame, video cameras normally expose their sensor as long as they can, while film cameras only expose the negative up to half this time, so that they can transport the negative in the remaining time. Many video cameras now allow adjusting the shutter timing manually, though, so this is no longer a big concern.
Old video technology only had a 5-stop exposure dynamic range. Modern HD video cameras have up to 14 stops. The exposure range is therefore less of an issue than before, although there is still a popular belief that video is considerably worse than film in the shoulder of the gamma curve, where whites blow out in video, while film tends to overexpose more evenly and gracefully.
Hub AI
Film look AI simulator
(@Film look_simulator)
Film look
Film look (also known as filmizing or film-look) is a process in which video is altered in overall appearance to appear to have been shot on film stock. The process is usually electronic, although filmizing can sometimes occur as an unintentional by-product of some optical techniques, such as telerecording. The process has the opposite result to VidFIRE, used to restore a video look to telerecorded video.
Today, most digital cinematography video cameras in use feature 24p format (24 progressive frames per second).
When shooting with old cameras, 50 Hz interlaced video (usually used with most forms of PAL and SECAM) can be relatively easily processed to give 25 progressive frames per second, which is the framerate that the PAL/SECAM telecine process also uses. Every two video fields can be "blended" together, every other field can be decimated and the remaining fields can be shown for double the length (this noticeably reduces vertical resolution), or a motion estimation process can be applied to achieve one frame out of every two fields. This technique is sometimes called Field-removed video or FRV. Some modern PAL video cameras do offer the ability to produce 25 frame-per-second progressive video, negating the requirement of post-processing the video to get a temporal similarity to film.
On the other hand, it is much more complicated to convert 60 Hz interlaced video (used with NTSC and PAL-M) to a framerate resembling that of film. Doing the same as PAL/SECAM filmizing will yield 30 frame-per-second video, which is significantly faster than film. Two out of every five fields could be dropped (and 3:2 Pulldown can be applied to the remaining fields), but any motion after this process will look very uneven. Sophisticated computer motion estimation and field blending is usually used to convert NTSC video to 24 frames-per-second – something which could not have been done until recently, and still does not yield as realistic results as PAL filmizing conversion.
Many computer editing programs can de-interlace video to give it more of a film look. An interlaced frame is actually the combination of two fields, one providing the odd-numbered scan lines and the other the even-numbered. Interlacing results in a type of motion blur known as "combing", and also shows "interline twitter" where vertical details approach the resolution limit, neither of which occur in film. De-interlacing can remove or reduce these artifacts, resulting in an appearance closer to that of film.
Some inexpensive consumer editing programs achieve de-interlacing by deleting one of the fields.[citation needed] The result gives half the vertical resolution of the original frame, and sometimes adds a jagged effect to the picture.
For each frame, video cameras normally expose their sensor as long as they can, while film cameras only expose the negative up to half this time, so that they can transport the negative in the remaining time. Many video cameras now allow adjusting the shutter timing manually, though, so this is no longer a big concern.
Old video technology only had a 5-stop exposure dynamic range. Modern HD video cameras have up to 14 stops. The exposure range is therefore less of an issue than before, although there is still a popular belief that video is considerably worse than film in the shoulder of the gamma curve, where whites blow out in video, while film tends to overexpose more evenly and gracefully.