Read this twice.
https://www.vegascreativesoftware.info/us/forum/render-settings-for-vhs-and-other--133700/?page=1

If you can make it through, you might have more insight. Then search for PAR, SAR, DAR for a treasure trove of 20th century video minutiae.

@Colin-Anderton Hi,

704 x 576 isn't 4:3 it's 11:9 ?

There's a couple of free Aspect ratio apps I use.

https://andrew.hedges.name/experiments/aspect_ratio/

https://calculateaspectratio.com/

If you haven't already have a look at the event - Right click - Properties (Pic below), does changing the Aspect Ratio correct your video?

If the black bars are burnt into the video, ie they're rendered into the actual video, you'll have to 'crop' the video &/or stretch it to the project/preferred/correct size,

Picture in Picture will stretch a video but also Pan/Crop will, Stretch to fill frame & Lock/Maintain Aspect ratio will give differing results.

will that return the video to its original shape?

That would depend on what shape it was before or should be, is the pixel aspect (AR) correct for that video clip & or prob more importantly was it rendered at the correct size & AR in the first place.

The answers to your question can differ depending on each individual clip.

Sorry, I should have mentioned that I want to render them out at 1.0926; I do understand about aspect ratio.

Gid's last comment, "The answers to your question can differ depending on each individual clip" is what I was afraid of!

So am I correct in thinking that, because some videos have larger black borders than others, that sometimes not all the picture area was utilised?

So am I correct in thinking that, because some videos have larger black borders than others, that sometimes not all the picture area was utilised?

@Colin-Anderton I can't think of the right words to describe that question but lets say it's a bit vague, so I don't think anyone can answer that, not without seeing the video or having a sample, & the answer prob wouldn't be simple. You'd need to see the true size of the video & if it incorporates the black bars.

Here is a video that is not 16:9, in the preview window it has black bars because it's on a 16:9 project. But if you look in the Pan/Crop window it shows that it doesn't have black bars.

If the orig had black bars it would look like this in Pan/Crop (I created a png of the orig clip to use as the example in this pic)

I have a number of videos that - somewhere along the line - had their width/height ratio altered. Is there a way I can put them back to their correct shape?

For example, if I set the editing programme to 704x576, and then stretch out the actual image pixels so they just touch the left and right edges of the picture, and set the height so those pixels are just touching the upper and lower borders (in other words, so the actual picture part of the video is 704x576), will that return the video to its original shape?

I ask this because many old 4x3 videos seem to have various size black borders?

I have a feeling I'm missing something here, so I would be grateful for any assistance.

Colin.

 

"I have a number of videos that - somewhere along the line - had their width/height ratio altered. Is there a way I can put them back to their correct shape?"

#1. You can change the pixel aspect ratio of clips by right clicking the media in VP project media, select properties and then change the Pixel aspect ratio: item.

"I ask this because many old 4x3 videos seem to have various size black borders?"

Just because they have different black bar sizes doesn't necessarily mean that there is anything wrong with the clips. You need to eyeball the clips and establish if the look right. If they look ok, items don't look too wide or too tall then VP will pick up on the correct PAR anyway. So in this case the video clips will simply display different size black bars but the "picture" looks ok.

If they don't look ok do #1.

Thank you all for your responses. I don't think I'm explaining myself too well, so please bear with me, I'll have another go.

I completely understand how to adjust the aspect ratio; what I need to discover (and this may well be different for different video clips) is how to calculate what the original shape was.

In those olden days of 4x3 TV, there was the area known as overscan. What I'm trying to find out is, was the actual picture area a standard 704x576 pixels, with the pixels having a ratio of 1.0926? Or was the black area in overscan non-standard?

In other words, is it going to be possible, by using the active picture area, to calculate the original picture geometry, or were the black areas off the side treated casually, with no absolute settings?

In those olden days of 4x3 TV, there was the area known as overscan. What I'm trying to find out is, was the actual picture area a standard 704x576 pixels, with the pixels having a ratio of 1.0926? Or was the black area in overscan non-standard?

I Googled that paragraph -

AI Overview

The 704x576 pixel area was indeed a standard for analog 4:3 video, but the black overscan area was considered non-standard. While the full 704x576 area represented the total possible picture information, CRTs (cathode ray tubes) varied in how much of that area was actually displayed. The overscan region was essentially a buffer area around the visible picture, and different TVs would cut off different amounts of this area. 

Here's a more detailed explanation:

Standard 4:3 Picture Area:

The standard 4:3 aspect ratio for analog television was 704x480 pixels for NTSC (North American standard) and 704x576 pixels for PAL (European standard). 

Pixel Aspect Ratio:

The pixels in these images were not square. They had a non-square aspect ratio, typically around 1.0926 to 1.21, to account for the 4:3 aspect ratio on the screen. 

Overscan:

Overscan refers to the practice of displaying a video signal with a larger area than what is visible on the screen. 

Why Overscan?

Early CRT televisions had inherent inconsistencies in how they displayed the image, with slight variations in positioning and size. To compensate for this, video signals were designed with extra picture information (overscan) beyond the visible area. 

Non-Standard Black Area:

The black area within the overscan was not a standard part of the visible picture, but rather a buffer zone designed to be discarded by the television. This meant that different TVs might show slightly different amounts of the picture depending on how much of the overscan area they cut off. 

Action Safe Area:

Because of overscan, a "safe area" within the picture was defined where important action or text should be placed to ensure it would be visible on all TVs

The 1.0926 PAR was applied to 720x576 PAL 4:3. It actually gives an image ratio of 1.36575 instead of 1.3333. Likewise the 16:9 PAR also gives an image which is slightly wide and I believe NTSC ratios are the same. That's why PAL SD 16:9 doesn't exactly fit a WS HD project of 1929x1080 which is exactly correct.

If your source clips do not have any black bars then you can alter their PARs to give a correct fit. The PAR value will depend on your output. Are you exporting with square pixels 768x576 or with 720x576 with stretched pixels?

If your sources have black bars in their image then you will need to crop that out and then and then adjust by eyeball as johnny-s suggests.

For info - 704x576 is/was an accepted size, in addition to 720x576, under the DVD specs. I've come across it in old PAL videos given to me. I forget what PAR was applied.

@Colin-Anderton  ... I'm not sure if the following might be of assistance, but Sony has a free video inspection app - Catalyst Browse - that includes in its report the Pixel Aspect Ratio of the video:

Thank you so much for these replies. This is all becoming clearer to me now. You've all been extremely helpful, and I'm sorry I didn't make a good job of explaining myself before.

So as I understand, if I set up my editing programme with a picture size of 704x576, with PAR 1.0926, the image will be correct if I take the width of the actual video image to 704, and the height to 576? I then render to 720x576 (PAR 1.0926), and the video will be correct, including eight blank pixels each side. Is this correct?

And, of course, because nothing is simple in this world, what about the noise at the bottom of the image, when the original is from video tape? Does that noisy area at the bottom have to be included in the height of 576?

Again, thank you all; you've been so helpful.

EricLNZ wrote: The 1.0926 PAR was applied to 720x576 PAL 4:3. It actually gives an image ratio of 1.36575 instead of 1.3333.

So I have to make allowance for the PAR being based on 720, not 704? OK, I can calculate from there what the image should be.

Thank you again.

what about the noise at the bottom of the image, when the original is from video tape?

What is the videotape source? 2" Quad, 1" C format, U-Matic, Mini-DV, Beta, VHS or other? It would be great if you could upload to this thread a still pic of the VT image so that we can see what the noise looks like.

Hi Dexcon,

I'm referring to the noise which most (maybe all) video tapes seem to have along the bottom of the picture. The tapes are S-VHS, recorded in the 90s.

I'm referring to the noise which most (maybe all) video tapes seem to have along the bottom of the picture. The tapes are S-VHS, recorded in the 90s.

IMO, once you've got the ideal aspect ratio sorted out via the pixel AR, I wouldn't mess around with that. If you need to remove noise at the bottom of the image, the best way to do that is to use pan/crop to zoom in slightly and then raise the pan/crop image to avoid the bottom noise area.

Dexcon - I'm afraid I can't use that inspection app, as I have just migrated to Linux, and they don't appear to have a Linux version.

But i do have Media Info, which is excellent.

OK thanks, I'll take your advice.

Thank you all for your help. I've got an archive of about 8,000 video clips to do, so you can imagine why I wanted to sort this before I got started!

Thank you again.

But i do have Media Info, which is excellent.

Yes it is. But the closest that I could find in MediaInfo to the Pixel AR item in Catalyst Browse is the "Bits(Pixel*Frame):" item under MediaInfo's Text or Tree view which - to me - is not all that helpful.

I've got an archive of about 8,000 video clips to do,

Do you have another computer with Windows so as to install Catalyst Browse ... or a friend, relative, workmate with a Windows computer who could install Catalyst Browse so as to check out a sample of your S-VHS video in order to get a Catalyst report of the S-Video's video's Pixel AR?

@Colin-Anderton I suppose you want to export to progressive, square pixel PAR 1 and AR of 4:3.

I would setup project settings as follows:

use double PAL framerate (or double NTSC framerate depending on your original framerate)

also in editing preferences:

and use following customized AVC rendertemplate:

RE: Render template: If you don't want to enter specific w/h resolution values, rather accept VP's detection (VP matching source media) then it's essential to check the box .. allow source to adjust frame size. If u don't do that then output WILL have black bars.

Worse case, letterboxing. Black bars on all 4 sides. Been there, done that.

I found that viewing on modern TV will tell if letterboxing has taken place. I have tried various settings in vlc and mpc for playback but couldn't always be sure if clip wasn't letterboxed.

I've got an archive of about 8,000 video clips to do,

Do you have another computer with Windows so as to install Catalyst Browse ... or a friend, relative, workmate with a Windows computer who could install Catalyst Browse so as to check out a sample of your S-VHS video in order to get a Catalyst report of the S-Video's video's Pixel AR?

What was I thinking? My internet computer has gone to Linux, but my editing computer is Windows (Vegas Pro doesn't work with Linux, does it?) so yes I'll try it out tomorrow. Time's running a bit short today. Thanks for your help.