Foo - Video Peeps: Fun Aspect Ratio Question

Okay, so I was Googling something, and it created a question, rather than answering one (weird, huh?) and Google couldn't seem to answer it.

I have been playing around with 4:3 and 16:9 footage (standard def) since my camcorder does both.

How the heck does 4:3 and 16:9 have the same pixel count, just with different pixel sizes, and result in more coverage, rather than stretched coverage?

My 4:3 footage is 720x480x0.9 and 16:9 is 720x480x1.2121 (0.9 and 1.2121 are pixel ratios).

Isn't a pixel a pixel? So if the pixel is sqaure, almost square, or a rectagle, how can it have more stuff in it?

If the 4:3 screen is 720 pixels across (x480) and the pixel is almost square (0.9), then you change it to 16:9 footage, also 720 pixels across (x480) but the pixel is a rectangle (1.2121) how does that get you more footage, rather than just strecthing the footage across the larger screen space?

If none of this makes sense, it sheds light on why I am so confused about the whole thing.

My camera shoots 16:9 at 848x480... not sure if that's meaningful at all... just thought I'd mention it.

you're confusing me, cite sources.

on a regular CRT TV screen, the pixel is not square.
it squishes the 720x480 image to fit into a 4:3 format of 640x480 pixels

DVDs and cable TV come through at 720x480, but your TV will squish the image and your PC video player converts it into 4:3

http://lipas.uwasa.fi/~f76998/video/conversion/

My source which got me started is my Project Wizzard in Sony Vegas Movie Studio 9. If I set up a 4:3 project, the format is 720x480 with pixel ratio of 0.9. If I set up a widescreen project, the format is still 720x480 with the pixel ratio of 1.2121.

The only difference is the pixel ratio, but there is clearly more material on the sides, rather than stretching it.

(Anyway, this all started because I want to make use of YouTube's new widescreen format with my SD 16:9 footage. I have since figured it out, but I was trying to use the application's wizzards, which made it much more complicated than it needed to be, and in my Google search for an answer, I came across this.)

Look up anamorphic stretching. This was used in the old days to squeeze widescreen format onto normal 720x480 DVDs. Your software isn't set up for true HD 16:9 formatting using either 1280x720 or 1920x1080 size.

Also look at the raw-DV footage downloaded from the camera. If it's still 720x480 for widescreen, the camera doesn't really do widescreen. What a lot of them do is crop the top & bottom to give you simulated widescreen image of 720x405. Then they stretch that vertically into a 720x480 image. But actual angular field-of-view is still the same. A true widescreen camera will pick up extra stuff on the sides that wasn't there in the 4:3 view without cutting off the tops & bottoms.

Here's how anamorphic stretching works. The widescreen-source is squeezed horizontally to fit into a 720x480 space:
http://i42.photobucket.com/albums/e346/DannoXYZ/Photography/sopranos-native.jpg

When displayed on 4:3 TVs, the player will compress the image vertically and insert letterboxing:
http://i42.photobucket.com/albums/e346/DannoXYZ/Photography/sopranos_lbx.jpg

When displayed on 16:9 TVs, the image is unstretched horizontally and will fill the screen fully without any letterboxing needed:
http://i42.photobucket.com/albums/e346/DannoXYZ/Photography/sopranos-wide.jpg

However, this is still stretching the same 0.35mp (720x480) across a 16:9 screen and won't be as crisp as a higher resolution image.

My 4:3 footage is 720x480x0.9 and 16:9 is 720x480x1.2121 (0.9 and 1.2121 are pixel ratios).

Isn't a pixel a pixel? So if the pixel is sqaure, almost square, or a rectagle, how can it have more stuff in it?

SD Video pixels are non-square. If you're shooting 16:9 on an SD camcorder, you're probably shooting an anamorphic, or squeezed, image. If you edit that image in a project set up for 16:9 video, your computer stretches it out for you. Make sure your export settings are also 16:9, and you should be able to make wide-screen movies for youtube.

I think you're overthinking the problem to try to understand what happens to the pixels, but yes, the pixels of video info that you shot get stretched horizontally. The wizard seems to be trying to tell you that with the pixel ratio number.

You may find it helpful to shoot some tests - get a big sheet of paper and draw a perfect circle on it, and make some numbered marks on a horizontal line with a ruler. Shoot it at 4x3 and 16x9 settings on your camera, making note of what you see. Then import it, and change the settings in your video editing project and watch how the image changes. Do the same thing with the output to make sure you're not squeezing or letterboxing it when you're done.

Anamorphic footage on SD tapes is used all the time in television post-production and transmission nowadays, because we're in the middle of a huge conversion from 4x3 to 16x9 standard aspect ratio tv. Some parts of the world also use a 14:9 standard, just to make it more fun.

I have noticed a few things with my camcorder: The raw footage for 16:9 mode shows the same top to bottom, but a lot more on the sides, than 4:3 footage.

In fact, the raw footage on the computer of 16:9 footage has even more side to side than what I can see in the LCD screen on the camcorder. I framed a doorway edge to edge in the camcorder in 16:9 mode, but the raw footage on the computer showed stuff beyond the doorway!

I haven't figured out the actual resolution of my raw footage yet, though.

Thanks for the input!

What model camera? Does the 16:9 show the extra-width footage compared to 4:3 when you're zoomed out to the widest angle?

It's a JVC, only a few months old, and just before the HD level. I will look for the model number when I get home.

Regarding 4:3 and 16:9, I have done the same test--frame a door in 4:3, but then when I switch to 16:9, much more stuff shows up beyond the door on the sides, but is stays the same top to bottom.

And as I said, when I import the footage into the computer, there is more footage still side to side compared to when I am using the camera. If I have time tonight, I will even include screenshots.

Heh, heh... reminds me of old times. I bought the GR-DVM5u and GR-DVM90u when they first came out. There's a physical-limit to the zoom-lens and sensor-chip. What the camera's doing is zooming out a little when you switch to 16:9 to generate that extra coverage on the sides. But when fully zoomed out at max wide-angle, there's no more it can back off. So 16:9 at this point is the same as 4:3 in angular coverage. Try measuring the height of the door-frame relative to the frame-height when zoomed fully out in both modes. ;)

I have the GR-D850U. What you say is probably true, but it works for me. Anyway, here are some screencaps. I realize now that my scenery is terrible, but you get the idea:

Here is footage of 4:3 mode. Sony VMS 9 calls this 720x480 pixel 0.0909. The actual screencap is 655x480, though:

http://img.photobucket.com/albums/v220/sonata1/For%20Posts/Image1.jpg

Here is 16:9, without moving the tripod. Vegas MS still calls this 720x480, but pixels 1.2121. There is a lot more on the sides, even though the room was dark, they are not bars. The actual screencap size is 873x480:

http://img.photobucket.com/albums/v220/sonata1/For%20Posts/Image2.jpg

So, even though the application says they are both 720x480, they aren't because the screen cap sizes are way different???

Haha, I played with this again last night, and to make things more complicated, I have to manually crop all of my "16:9" footage for a 16:9 screen or else there are tiny bars on the top and bottom, as apparently my 16:9 footage is a little wider than a 16:9 TV ratio.

Haha, I played with this again last night, and to make things more complicated, I have to manually crop all of my "16:9" footage for a 16:9 screen or else there are tiny bars on the top and bottom, as apparently my 16:9 footage is a little wider than a 16:9 TV ratio.

time to get an HD video camera.