Progressive Segmented Frames (PSF)
The end result of many years of interlacing meant that most broadcast systems, especially hardware, were built to read interlaced footage. These systems are still widely used today because SD still is interlaced.
However, HD standards support progressive footage in addition to interlaced video. The younger generation of engineers had to find a way to make progressive and interlaced video work together on systems that are meant to show both SD and HD.
We are so used to software running things today that we forget that, traditionally, hardware runs on fixed circuit boards that can’t have firmware upgrades (firmware is software). This meant that an only-software solution was not practical. The solution had to marry both interlaced and progressive technologies.
The way engineers tackled this problem was to cut up a progressive frame and transmit them like interlaced video. A cut-up progressive frame is called a Progressive Segmented Frame (or PSF).
Whereas in interlaced video each subsequent field is shot at a different time and don’t match when juxtaposed with one another, PSF fields match. They match because they are not shot at twice the speed.
E.g. let’s say a progressive video is shot at 30 fps:
In order to make it look like interlaced footage, this is split up like this:
Then it is broadcast or transmitted as interlaced footage, in 60i at 29.97 fps.
The key difference between PSF and interlaced footage is that interlaced footage has information corresponding to 60 fields per second (1, 2, 3, 4, 5 and 6 counters); however the PSF footage repeats each field twice (no data is lost since they are cut up in even and odd fields that when added together give back the full progressive frame) and even though it appears to be 60i it only has the counters: 1, 3 and 5.
PSF is mainly used in data transmission, where progressive footage has to be displayed or read by a system that is expecting interlaced footage.
Many prosumer cameras outputting uncompressed progressive HD signals via HD-SDI uses the PSF method. If the recorder interpreting this signal makes a mistake, then the progressive signal is in danger of being transcoded and rewritten as interlaced video, or vice versa. Unless a manufacturer has clearly written documentation in support of a certain system (how many times does this happen?), be careful!
Or better yet, just be careful no matter what!