I've read about aspect ratios, anamorphic video and how video is stored on DVDs but I still do not understand a few things.
1. If I'm not wrong, then anamorhpic video is 16:9 video that is squished horizontally in order to fit 4:3 screens. Is anamorphic video simply called "anamorphic video" or is it called "4:3 anamorphic" or "16:9 anamorphic" as I've come across confusing terms while reading forums?
2. If video was originally 16:9 but was stored on DVD as 720x480, wouldn't that destroy some horizontal resolution since you would later be converting it to 848x480 or 852x480?
3. Can you tell me if this is right?
- PAR (Pixel Aspect Ratio): How the pixels are shaped (ratio of width to height)
DAR (Display Aspect Ratio): The ratio (width to height) at which the video is actually displayed on your screen
SAR (Sample/Storage Aspect Ratio): The ratio by which the PAR is stretched or compressed to match the DAR
4. Why is 1440x1080 considered 16:9 and not 4:3?
5. Do 1080p/720p Blurays store native 16:9 content or do they store the video with letterboxing to make it 16:9? What resolution (excluding letterboxing) are the movies on Blurays stored in? What resolution are 4K Blurays stored in?
+ Reply to Thread
Results 1 to 12 of 12
while i don't know it all
it is not squished to fit 4:3 screens, at least not when displayed, it is squished for 'Storage'
think of it like a fun house mirror, that distorts images
the image is stored in a 'distorted' compressed format to fit in the DVD standard frame, if you looked at it as it is stored, circles would be tall oval eggs
people would look skinny
but the dvd player expands the image horizontally stretching things out making wheels round and people normal
In general terms, anamorphic could be squished horizontally or vertically. In real life situations, due to historic optical habits with film and due to complications of vertical with interlacing, it always operates horizontally.
All anamorphic footage is non-square PAR, but not all non-square is anamorphic. Though most here do not see the distinction, it has to do with intent & application. So for example all dvd footage is stored non-square regardless of resolution or AR, but really only 16:9 material may be stored anamorphically. And remember, 16:9 can and has been stored on dvd on hardcoded letterbox mode as well (but still non-square as is all dvd material).
#2 Yes, resolution is less. But it retains more resolution than the letterboxed version. Remember, especially these days you are seeing dvd rez being blown up to hd screen's rez and it is much better when provided with more data to start with.
#3. Par is right as is dar. Sar could be sample (refering to pixel), but in truth there is no such figure as storage ar. It only exists as shorthand in the minds of hobbyists who avoid working with direct resolutions in math calculations. Par, dar, and resolutions are actually specified and written into the file - storage ar NEVER is.
#4 1440x1080 in some contexts is 4:3 (with pillarboxing added later) and in other contexts is anamorphic 16:9 (early hd cams).
#4 if bd material is sd rez, it is stored just like as on dvd - 4:3 or 16:9. If hd or uhd rez, it is stored square pixel & non-anamorphic only (which may or may not include built-in letterbox padding, depending on the intended image AR), but the overall contained AR is always 16:9.
#5 dvd can store as 720x480/576, 704x480/576, 352x480/576, or 352x240/288 for ntsc/pal.
Hd bd can store the 1st 2 of dvd's (but not the latter) for sd rez. For hd, it is 960x720 or 1440x1080 in anamorphic or 1280x720 or 1920x1080 in standard square pixels (the latter 2 being by far the most common). Uhd bd can do all that hd bd can, but for uhd material can store 3840x2160 in square pixel format. Only.
It would be natural to just design something like --DAR 16:9 for that command line, to just specify DAR all the time (PAR is obvious - it is resolution ratio coming into encoding, in our case 1440/1080) because that'd be much more simple in general or particularly in DVD case for example to avoid that SAR calculating twist while encoding.
DAR = PAR x SAR
In my experience, PAR used to mean pixel aspect ratio, the shape of individual pixels. And SAR used to be the storage aspect ratio, the ratio of the frame dimensions. So lots of old documents use:
Display Aspect Ratio = Storage Aspect Ratio * Pixel Aspect Ratio DAR = SAR * PAR
Display Aspect Ratio = Frame Aspect Ratio * Sampling Aspect Ratio DAR = FAR * SAR
Last edited by jagabo; 23rd Aug 2016 at 17:50.
I'm not sure it this was explicitly mentioned above, but the anamorphic transforms done for film, by using camera and projector lenses, is not uniform across the width of the picture. More stretching is done at the edges of the picture, where you won't notice as much the resulting increase in artifacts. By contrast, anamorphic stretching that is done for digital formats is generally the same at the edge of the frame as it is in the center.
Thank you, everyone, for the answers. I understand what FAR is too now. Could not find it anywhere before.
I don't understand the concept of square pixels though. Does 1280x720 have square pixels? Is a monitor that is 16:9 have square pixels or are square pixels subjective? If I'm not wrong, then square pixels will have a PAR of 1. So if a 1280x720 is square pixeled, then it wouldn't be so on a 21:9 or 4:3 monitor?
Thank you again.
You might also encounter while viewing some 4:3 content on 16:9 TV that video is stretched out and therefore not proportionally correct (or the other way around seeing letterbox for true 16:9 content). Some TV channels for example, then you'd need to grab remote control and fix that, TV's have functions for that. Or you don't, there is many, many people watching 4:3 content stretched out and even feeling somehow happy that they have picture somehow "bigger" or something. Like watching MeTV 4:3 content only on 16:9 TV etc.
Avid Media Composer, which is still the tool of choice for most high-end productions has the following commonly used aspect ratio masks and presets. So the answer is, it depends. 804-822 are common but often rounded to accommodate mod4 or mod8 encoder requirements.
If that's a 1280x720 monitor receiving a 1280x720 image then a single white pixel at that location would light up those RGB subpixels, and only those subpixels.
Got my answers. Thanks, everyone.