Aspact Ratio

[biferi]

I use Movie Studio 8.0 and it works good.
So far I was useing the 4:3 video projects.

Now I am going to be making a lot of JPEG images that I will be importing.
And I want them to be Widescreen.

So I know I pick NTSC Widscreen from my Movie Studio Projects Menu.

But my Graphic I make to import and have it show as Widescreen and stay the Right shape am I right I make my JPEGs 720X540?
Thanks.

[Cornucopia]

NTSC (actually the digital equivalent) Widescreen video is 720x480. So is NTSC 4:3 (Full) screen. Both use non-square pixels. Wide pixels are ~1.2121:1 and the narrow ones are ~0.9090:1.

If you have something like newer versions of Photoshop, you can create those exact dimensions with those PARs since non-square PARs are supported. Then exporting to JPG, which normally doesn't support non-square, if you preview the output it will look stretched/squished. But if you tell your NLE to interpret them as non-square, all will be fine in the timeline & on output. No scaling necessary.

If you generate JPGs from sources & apps that don't support non-square PARs (most, esp. most cameras), you will HAVE to somewhat scale your sources upon addition to the NLE timeline.
For widescreen, I would recommend ~854x480 square-pixelled sources.
For narrowscreen (I hate saying "fullscreen" anymore when it rarely is), I would recommend ~640x480 or 720x540.
Either way, one is scaling in only one dimension (it is slightly less accurate, and possibly more algorithmically wasteful, to need to scale in both dimensions). Personally, while I favor downscaling whenever possible to retain the most detail, I favor locking in the vertical dimension: the ntsc/pal scanlines are discreet so it is best to match that constraint and method.
Thus in the timeline & on export they should be conversely adjusted in one dimension to match the 720x480 frame.

Scott

[johnmeyer]

Right click on the event and select "match aspect ratio."

[biferi]

Thanks for all the help.

Now let me see if I have this right.

I am in the USA so I think this is how things work.

NTSC TV was Analog and was a 4:3 image.
Then when Digital came out they had to convert the Analog 4:3 image to Digital 4:3 with Pixels.
So we had a 640X480 Digital image.

so what is the 720X480 or what am I missing?

[Cornucopia]

You never had a 640x480 image (rare exception: pre-601 computer capture cards in 1980s that had different sampling rate, and scientific experimental cards) - you ALWAYS had a 720 or 704x480 (sometimes 486 if pro card). And it was usually 4:3 with non-square pixels, or occasionally 16:9 with non-square pixels.
You had a 640x480 EQUIVALENT image (when evaluated with square pixels) when referring to 4:3 and a 854x480 equivalent image when referring to 16:9. Either way, you have ~720x480s-worth of image detail (not counting pre-scaling or compression artifacts).
Only way anybody got a 640x480 image was subsequent scaling to square pixels and re-encoding. Or CGI synthetic image generation.

Clear as mud?

Scott

[biferi]

Ok this is what I thought.

I am not talking about Display Aspect Ratio when an video is Displayed on a TV.

I thought that when a Digital Camera takes a 4:3 video it uses a RES. of 720x480 but uses Square Pixels.

And if a Digital Camera take 16:9 Widescreen it uses a RES. of 720x480 but will use Rectangular Pixels?????????

[Bernix]

In my case info says (PAL 720x576) 4:3 ratio 16:15 and for 16:9 ratio 64:45

Bernix

[biferi]

Tell me if I have this right for NTSC in the USA?

If you record a 4:3 standard video with a Digital Camera it will take it with a RES. of 720x480?

Now when the Digital camera stores the video to it's Disc it will use Square Pixels to keep the video as a Square shape?

And if the Digal camera take a Widescreen video 16:9 it will also use 720x480 RES. but will store it with None Square Pixels to keep the Widescreen image as Wide?

Do I have it yet?
Thanks to everyoone on this site for the help.

[Cornucopia]

No, I just told you. If you have digicam and shoot SD (ntsc type) video, it either shoots AND saves 720x480 using non-square pixels in 4:3 (narrower than square) mode, or shoots and saves 720x480 using non-square pixels in 16:9 (wider than square) mode.

Both forms of SD are non-square in capture and in storage. No where does SD video normally originate with square pixels. Square pixels do exist in HD and UHD video origination.

When shown on modern HD displays it will of course scale to square pixels at full height (usu. 1080) and either (1440) pixels wide with pillarbars or (1920) pixels wide and full width.

Here's a secret: we say square & non-square pixels as if they have dimension. They themselves don't. They are single units, like point sources. It is just that they ENCOMPASS a square or non-square region of interest.
If you'd rather, you may understand it better thinking of the pixels as dimensionless points with either normal or more or less blank space between.

Scott

[jagabo]

If you record a 4:3 standard video with a Digital Camera it will take it with a RES. of 720x480?

Often, but not always. It depends on the camera and the settings.

Now when the Digital camera stores the video to it's Disc it will use Square Pixels to keep the video as a Square shape?

No. 720x480 with 4:3 DAR is not square pixel. Square pixel would result in 3:2 DAR. With square pixels the frame aspect ratio (often called the sampling aspect ratio) is the same as the DAR (Display Aspect Ratio is the final shape of the picture you see). 720/480 = 3/2 = 1.5. 4/3=1.333...

And if the Digal camera take a Widescreen video 16:9 it will also use 720x480 RES.

Often, but not always.

but will store it with None Square Pixels to keep the Widescreen image as Wide?

Since 720x480 is 3:2, obviously, the pixels are non-square.

Do I have it yet?

Apparently not.

The general algebraic equation is (on computers "multiply" is often represented as "*"):

DAR = FAR * PAR

where:

DAR = Display Aspect Ratio, the shape of the picture you see
FAR = Frame Aspect Ratio, the relative width:height of the stored frame
PAR = Pixel Aspect Ratio, the width:height of individual pixels (or when called SAR it's the relative horizontal:vertical distance between samples)

With Square pixels (PAR = 1:1):

DAR = FAR * PAR
DAR = 720:480 * 1:1
DAR = 720/480 * 1/1
DAR = 1.5 * 1
DAR = 1.5

with a 720x480 frame and 4:3 DAR:

DAR = FAR * PAR
DAR = 720:480  * 8:9
DAR = 720/480 * 8/9
DAR = 1.5 * 0.8888...
DAR = 1.333... = 4:3

with a 720x480 frame and 16:9 DAR

DAR = FAR * PAR
DAR = 720:480 * 32:27
DAR = 720/480 * 32/27
DAR = 1.5 * 1.185185...
DAR = 1.777... = 16:9

Then this is further complicated by the fact that with the ITU standard the 4:3 or 16:9 DAR is contained in a 704x480 portion of the frame, not the full 720x480 frame. So you'll often see PAR values of 10:11 and 40:33 for 4:3 and 16:9 DAR, respectively. And then it's even more complicated because 704 is a convenient approximation because it's a mod 16 value (many old codecs required mod 16 frame sizes). The real value is closer to 703. But most people ignore that.

[biferi]

OK

Pixels are not Square they are Diamond Shape I get this.
And if I take Digital Video NTSC SD 4:3 the RES will be 720x480 I get this.

Also when the Digital Camera stores the video it will use Pixels Higher then Wide I get this.

So do I have just this down?

[LMotlow]

Pixels are not Square they are Diamond Shape I get this.

Nope. Anamorphic pixels are rectangles, not diamonds.

[Cornucopia]

I think at this point someone is being purposefully obtuse.

Scott

[jagabo]

Pixels are not Square they are Diamond Shape I get this.

No. Theoretical pixels have no shape at all. They are points with no width and no height. Their aspect ratio refers to the distance between those points on a rectangular grid. That is, if you think of pixels on a piece of graph paper, the pixels are the intersections of the lines, not the boxes bounded by the lines. Of course, when you draw a pixel (on the monitor, paper, etc.) it may take on a shape. It's up to the renderer how those spaces are filled.

And if I take Digital Video NTSC SD 4:3 the RES will be 720x480 I get this.

Sometimes. Not always.

Also when the Digital Camera stores the video it will use Pixels Higher then Wide I get this.

When a 720x480 frame is used for 4:3 DAR, yes. Although, again, theoretically the pixels themselves have no dimensions. It's the spaces between them that are larger vertically than horizontally.

[LMotlow]

Pixels are not Square they are Diamond Shape I get this.

No. Theoretical pixels have no shape at all. They are points with no width and no height. Their aspect ratio refers to the distance between those points on a rectangular grid.

Physically speaking, that is the case.

[biferi]

OK so Pixels never ever change Shapre.
But how far apart they are from another Pixel is what they talk about when they say Square or Non Square.

So if we say Square Pixel we mean the space between the Pixels is Square.
And if we say Non Square Pixels we mean the space between the Pixels less wide.

Right?

[jagabo]

OK so Pixels never ever change Shapre.

They have no shape to change.

But how far apart they are from another Pixel is what they talk about when they say Square or Non Square.

So if we say Square Pixel we mean the space between the Pixels is Square.
And if we say Non Square Pixels we mean the space between the Pixels less wide.

Right?

The horizontal spacing can be less wide or more wide than the vertical spacing. With 720x480 4:3 DAR they are taller than they are wide. With 720x480 16:9 DAR they are wider than they are tall.

[_Al_]

When you start to think about pixel shape it could be misleading. Better to think of it like that resolution 720x480 will never be shown on screen. It is just a video put in the box to be delivered for someone. No rectangular or diamond shapes objects.

But magic starts always in the moment of playback. Player just reads DAR and calculates horizontal ad vertical resolution. Every time. For example, you have your player set to full screen 1920x1080 and aspect ratio is 8:9 , (your video is 720x480) then, 1080/480=2.25; 720x8/9=640; 640x2.25=1440; so it puts on screen 1440x1080 and fills black bars left and right

So for any resolution of your videoplayer, that is calculated, 640x480 is never depicted. Only if you happen to have your player selected to height 480, for example, when you select always show 1:1 scale or something.

That video only lives in the moment of playback , otherwise it is just boxed, waiting. That box has one law only, respect DAR when looking at it. Just like our universe interacting only on need bases, but of course many more laws has to be followed.

[biferi]

OK
I did some more Reading and from what I Read and what you told me I think I have it.

NTSC DV 4:3 image that is taken with a Digital Camera will take it at 720x480.

But as it captures the image the every Pixel Left to Right is closer to the next Pixel.
And this is what makes the image stay as 4:3 for if it is played back on an Old 4:3 CRT TV.

I get this now thanks to all.
But one thing why does an NTSC DV 4:3 have a Pixel Aspect Ratio of 0.9091 is this Cenimeters?

[_Al_]

ratio is a ratio, no lenght,
simply it is a ratio of stored lenght to viewed lenght, so units cancel each other out

whatever helps you to understand anamorphic video, like narrower pixel, as you say, but it could also be the other way, for example HDV video, it stores 1440x1080 but footage is 16:9, for example 1920x1080, so pixels could suddenly be further away somehow, or the same about widescreen DV

think of it like real horizontal video dimension as being stored or being digitized in different ratio as oppose its height is

since we started digital revolution one might do such a thing, they did it for whatever reason (keeping storing resolution or not having too much data to process etc.)

[hello_hello]

I get this now thanks to all.
But one thing why does an NTSC DV 4:3 have a Pixel Aspect Ratio of 0.9091 is this Cenimeters?

Pixel aspect ratio is much like display aspect ratio, it's just a shape. A picture with a 16:9 display aspect ratio can be 1920x1080 or 1280x720 or 960x540 etc. They're all 16:9.
Similarly, the pixel aspect ratio is the shape of the pixels, not the size.

The width in resolution multiplied by the pixel aspect ratio gives you the display width (before the whole image may or may not be resized for playback). If you have a DV that's 720x480 and the pixel aspect ratio is 10:11 ( 0.9091).
720 x (10/11) = 654.54
The height is 480 so 654.54/480 = 1.363637 which is the display aspect ratio.

If you use the "generic" pixel aspect ratio of 8:9 it gives you an exact 4:3 display aspect ratio.

720 x (8/9) = 640
640 / 480 = 1.33333 (4:3)

List or pixel aspect ratios. Generally you only need to choose either an appropriate mpeg4 or generic PAR.
The mpeg4 PARs result in a slightly wider display aspect ratio than the generic PARs, which give you exactly 4:3 or 16:9.

[jagabo]

But one thing why does an NTSC DV 4:3 have a Pixel Aspect Ratio of 0.9091 is this Cenimeters?

Since pixels have no physical dimensions centimeters has nothing to do with them

The 10:11 sampling aspect ratio comes from the CCIR rec.601 specification for digitizing analog video. Analog video has no pixels, it is a continuous waveform. The SAR comes from the rate at which the waveform is sampled when digitized at 13.5 million samples per second, the active picture region vertically and horizontally, and the timing of the 525 line NTSC signal. The math is complex and 10:11 isn't even the exact value, it's just a good approximation.

[biferi]

I know that Graphic JPGs are not the same Pixel aspect Ratio as Video.

I am making a lot of Graphic JPGs that I want to be Widescreen and seen on a Widescreen TV.

So because they are not NTSC DV Video they are Graphic JPGs I am not making them 720x480 I am making them 720x540.

And before I Import them into my Movie Studio I set the project setting to Widescreen DV and it sets the Pixel Aspect Ratio to 1.2121 Widescreen DV.

So I Import all my JPGs and Render As Widescreen DV.

After this I open my DVD Archatect Program.

I bring the Video I just made into it and set it to Burn NTSC DV Widescreen.

It Burns the Disc and I put it into my DVD Player in my Living Room.

I have a 72 Inch Projection TV. and it played with the Video going to the very top of the TV just Great and looked good.

The Left and Right sides of the Video did not come out to fill the screen.

It Looked 4:3 does anyone know what I did wrong?

Thanks.

[Cornucopia]

As I said before, you need to pre-adjust your graphic to either match square PAR or non-square PAR, and match 4:3 or 16:9 or some other oddball DAR.
Then have the editing app/NLE "interpret" the graphics properly upon import so that it knows what basis to start making adjustments from. If you don't do this, it will just try make assumptions based on the filetype/codec/container headers - which is square pixels for JPG.
Then, if the app doesn't automatically do a stretch/squeeze when dropping them onto the timeline and their DAR doesn't already match the project's, you will have to manually do the stretch/squeeze yourself, making sure the borders actually snap to the exact frame edge.

But your immediate question just now had a faulty assumption: you created jpgs (square pixelled) that are 720x540. 720 / 540 is a 4:3 DAR, not a 16:9. So the app is doing exactly what it should based on what you gave it, not what you intended to give it.
You should be creating jpgs -square pixelled- that are 854x480 for 16:9 DAR. (Some might say 853, or 852)
Then when you import and drop into the timeline, VERIFY that it is matching the project's output screensize.

Scott

[biferi]

Now I know a Pixel Aspect Ratio is how Far Apart Pixels are.

Now in my Video Editor I can have my NTSC DV set to a Pixel Ratio of 0.9091 DV or 1,0000 Square.
So if 0.9091 DV will give a 4:3 image then what is the 1.0000 Square??

[jagabo]

Square pixel has been discussed several times in this thread.

[biferi]

OK I know what you told me.

They say Square Pixels but they meen one Pixel is close to the next pixel.

But it has the words
NTSC DV 1.0000 Square.

But under it there is one other Ratio
NTSC DV 0.9091

So out of the two of them if I pick 1.0000 Square I think this is more 4:3?

Am I right or is the 0.9091 DV this is for CRT TVs or is the other one?

[_Al_]

NTSC DV is NOT square pixel. This is not matter of choice. That choice 1.000 is perhaps there but it was never meant to be selected as a default.
Select NTSC DV preset and default 0.9091 will be selected. If you select NTSC DV widescreen it will select 1.2121. But also field order will be selected and exact frame rate. None of these values should be changed. Follow default presets.

As for CRT TV, you can watch square pixel content on it. That has nothing to do with it.

But, is you intention to make a DVD? Then yes , same aspect ratio is used in DVD specs , same as for NTSC DV.

[Cornucopia]

1.0 PAR for DV is never, never, never a proper usage. It will never be seen coming from any camera or machine or broadcast encoding.
If you want to use it is up to you. But I wouldn't. Look at all the confusion you've already had over this one stupid option.

I will re-iterate what has been said: every attempt to figure out these variables ought to rely on basic princples of science & math. So use the universal formula (which I've corrected because FAR isn't really an accepted thing in the industry - it doesn't predict or describe anything about the image or the structure, plus this way is more versatile)

DAR (Display Aspect Ratio, or shape of the overall image = PAR (Pixel Aspect Ratio, aka SAR sample AR, or the shape/spacing of the inidividual pixels/samples) * Width / Height

Go back and look at what you were asking and what we were answering & plug in that formula each time, and you will see that you were misunderstanding these basics and our answers to you have been correct.

Scott

[Cornucopia]

(Duplicate)

...
Another rule of thumb to help you remember: digital formats that are/were based on Standard Def consumer or broadcast formats (especially if they are meant to be equivalents to analog NTSC/PAL/SECAM) are *ALL* non-square.
Digital formats that totally originated (and meant to remain) in the computer realm are primarily square.
High Def and UltraHD formats are also primarily square (there are a few exceptions).

So,
DV, DVcam, DVCPro, VCD/SVCD/DVD-specific MPEG1, DVD & DVB-specific MPEG2 are common non-square pixel SD formats.
BD, AVCHD, etc are square pixel HD formats.
Most older webcams, & those goofy VGA- or sub-VGA-based pocketcams are square pixel SD.
DVCProHD and HDV are some of those non-square HD exceptions, primarily because they are HD offshoots/extensions of those earlier non-square SD formats.
Generic MPEG1, MPEG2, MPEG4/DIVX/XVID AVC, HEVC can be many things (because they were meant to be versatile), but are most commonly square pixel. Similarly with your generic AVI & MOV.

If you are wanting make something for one of those non-square consumer formats, stay in the non-square realm.
Only convert square sources once to non-square and that's it.
Vice-versa, if your end result target devices use only square pixel formats, convert non-square sources once to square and that's all.

Scott