1440x1080 formats generally make a 16:9 picture. Theoretically you can make a 1440x1080 format that makes 4:3 pictures if you wanted (and you could do this in Vegas).
I don't know of anyone who "considers" 1440x1080 (which format?) to be 4:3... because no format being used does that.
mainconcept reference encoder and sorenson squeeze seem to think that 1440 x 1080 is a 4x3 aspect ratio because if you frameserve out of vegas into them for encoding the file is identified as 4x3. they do not seem to recognise a 1.333 pixel aspect ratio.
Even converting to cineform hd intermediate file and importing this into the encoders results in the file being recognised as 4x3.
Yeah, but those number represent standard definition
1920x1080 is certainly considered 16:9.... and does in fact have a mathematical 16:9 ratio.
What I'm getting at here is that it has been hinted that 1440x1080 was set up so that high definition could live in the 4:3 world but work in the 16:9 world. When HD was first devolved, the NTSC signal recognized 4:3 only, as a legal size and therefore 1440x1080 with a 1.333 pixel aspect was used to kind of "cheat" the system. Trying to find out if there is any truth to this.
I've never heard this but that doesn't mean much. I'd just assumed that this was a reasonable way to save bandwidth for some HD formats that needed the savings to boost performance.
The fact that 1440x1080 is (unmodified) a 4:3 image might have some advantage if it was expected that early adopters would be using SD TVs to view it. Maybe. I don't think this really works out since this format would just look squeezed if displayed at 4:3
Keep in mind that SD is 4:3 in name only. The entire frame, even in an analog signal, is a little wider than 4:3. That 4:3 ratio just represents the portion of the entire SD image that televisions were supposed to show and given how variable TVs were at the time this standard was set that 4:3 window could be anywhere on the overall image. To me, combined with the anamorphic nature of 1440x1080 video, this suggests that there's really no strong relation of 1440x1080 video to SD. But I could be wrong.
In any case, people usually refer to a frame size by whatever it's final output will be. I don't know of any cameras that shoot 1440x1080 as a 4:3 image.
No it's not "4:3" because the 1440x1080 formats (HDV, HDCAM) use non-square pixels and therefore the picture aspect ratio is 16:9. The SD formats also use funky non-square pixels.
When you downconvert HD-->SD, you handle the difference in pixel aspect ratio by letterboxing, cropping, or pan and scan.
The subsampling is used to reduce bandwidth... though IMO it's not really a good idea.
The 1400x1080 has no relation to "cheating" the system.
Standard NTSC DV is 720x480, which is 1.5:1, but it uses non-square pixels for a display aspect ratio of 4:3. (This has confused a lot of people when they have created content for computer display and not compensated for the 1.5:1 to 1.333:1 difference.)
Storage Aspect Ratio (SAR) is the width to height ratio of the frame in actual pixels.
Pixel Aspect Ratio (PAR) is the width to height ratio of the pixels themselves.
Display Aspect Ratio (DAR) is the width to height ratio of the display screen measure in inches or centimeters, not pixels.
You multiply the SAR by the PAR to get the DAR. Clear as mud. :)
There was a plan it seems a long time ago for a HD broadcast standard that would have been 4:3. I don't know the details as it originated before I got involved in video but at a guess the idea could well have been to use 1440x1080 as 4:3 at a PAR of 1:1 and as 16:9 at 1.333:1 PAR much the same as is done for SD. Possibly one reason they were thinking of going down that path was the high cost of manufacturing 16:9 CRTs. The advent of afforable LCD and plasma displays would have changed the thinking.
Don't get hung up on the 1440x1080 pixel depth of HDV, XDCAM HD, and HDCAM. The fact that your HDV camera spits out 1440x1080 does not mean that the tiny CMOS inside of it is actually picking up that much data necessarily either.
There is an extensive lecture at Panavision's website which explains this phenomenon.