Enlarger AI 4 x 3 to 16 x 9 confused

datacaster

I run the Enlarger AI and a 720 x 480 video in 4 x 3 aspect ratio does not go to 16 x 9 aspect ratio where the video fills the screen without any sidebars. It shows in file details that video has gone from resolution of 720 x 480 to a file resolution of 1440 x 1080 but it still plays as a 4 x 3 aspect ratio, so I am confused to what program should do for me. I wanted to change old video to a nice 16 x 9 aspect ration without making it grainier like blowing it up.

Thank you for any help here!

signals

Enlarger AI is a powerful product but it cant create pixels where they do not exist in the source file. In most cases the black bars that appear on a widescreen display are actually encoded in the video and they would still be there on playback regardless. A non-letterboxed (anamorphic transfer) DVD will produce a widescreen (1920x1080) output with Enlarger AI. In either case, the aspect ratio of the Source is not changed (1440x1080=1.33:1=4:3 for a 640x480 DVD).

ord1117

Signals just said "It can't create Pixels where they do not exist in the Source file"

I don't understand that.

ANY video or Pic has Pixels in their Source File, or they would not exist.

When I heard about this Program I thought it could be BIG news.

Now I don't know what to think.

PLEASE Explain Signals etc. ????????

signals

It can’t make pixels where they do not exist. There are no active video pixels in a 4:3 DVD image outside of the left and right edges, so there is no way it can ever be a widescreen image after processing with Enlarger AI or anything else. You can take a 4:3 image and crop off parts of the top and/or bottom to make it appear to be widescreen but it’s really not and you wind up with parts of heads chopped off and other missing content. When time permits I will make some images to illustrate this.

ord1117

Thank you for responding Signals.

I still don't understand HOW you can have NO Active Video Pixels in a 4:3 DVD Image Video, except around the edge.

ord1117

There are different types of Pixels, maybe thats the Problem.

I believe a 4:3 DVD video is Probably Analog.

If you changed the Analog to Digital would that Help.

Seems like it would change the type of Pixels in the Video.

signals

It's because the active video is only 640 pixels wide. If you use Enlarger AI to upscale that to Blu-ray it will be 1440 pixels wide and have black bars on the sides when played on a widescreen (1920 px wide) display. The active video pixels between 1440 and 1920 do not exist in the original DVD, so they can't be there when you upscale it.

ord1117

Who says you have to Upscale it to a Wide Screen.

Just make the Box BIGGER.

Make the Ai so 'I' can choose the Resolution.

Blu-Ray & 4K already have Good Picture Quality (and Resolution).

Do something for DVD and other videos.

signals

?? It will upscale DVDs to Blu-ray resolution but it is not magic. There is no way to make a widescreen image out of a square image without cropping it, adding black bars on the sides or making it look squashed. I regret I can’t explain this in a better way.

ord1117

Your NEW Update, Upscales Blu-Ray.

I have previously sent Feedback and Logs.

About 4:3 Box DVD Shows.

Trying to get the 4:3 Box Bigger.

Still waiting and hoping that they will be able to do that one day.

I know it's a work in Progress, so we will see what happens.

datacaster

Signals,
I see what you are saying as the pillar area of 4x3 video displayed on 16 x 9 will just be black created by AI to achieve new resolution when viewed on a 16 x 9 display. The 640 pixels remains just 640 and balance of 1920 pixels - 640 pixels = 1280 pixels are just black filler so you have 1280/2= 640 pixels of black on left side and 640 pixels of black on the right side.

Now if I take a program like video redo and tell it to change my videos aspect ratio to 16 x 9 it fills the whole 16 x 9 screen, but might be grainier than 4 x 3 and might crop some of top and bottom but not sure 100%. I am assuming it is simply changing to anamorphic

If I take video redo program and change videos aspect ratio from 4 x 3 to 16 x 9 and run the converted aspect ratio video in AI afterwards, will that make it better or is it waste of my time?
It takes a couple hours to do one hour video at Highest quality using a Turbo RTX2080TI-11G

I am talking off my head here quickly and not really analyzing if valid thoughts or really possible, but if you take a 4 x 3 and intelligently expand video (Not zoom which chops top/bottom area off) and somehow interpolate that out to a 16 x 9 by some formula that takes original pixel width and divide that into new pixel width to determine needed pixels to increase/stretch width. With that number you duplicate a pixel every x pixels (formula logarithm) or stretch each pixel to achieve the new pixel width goal (Additional pixels necessary to make new width) . By doing this with width and height you maintain aspect ratio as closely as possible in formula and keep the duplicated pixels close to original and not chop off heads . Granted it most likely causes some artifacts that could cause less sharpness or image distortion as stretch takes more pixels or more pixel stretching as to what is required for height and this is just quick thought from my head . If you think of old analog world and TV you could adjust the height and width. Each video line got taller for height increase. If you double lines in given area and then stretch to fill screen each line is height of original line like taking 525 lines 6 inches high and making it fit 12 inch high screen. A 1050 line video 6 inches high stretched to fit that 12 inch high screen will have a the same line height of the 525 line image. First line starts at top of screen and is not chopped off and last line is set to hit bottom of screen.

Thanks and my mind is wandering around so sorry for probably bad thinking!

datacaster

I am trying Virtual Dub right now and using resize no cropping or zoom. Changing resolution to 1280 x 720 and I see top of video and bottom and start left side and end right side includes original video content so nothing is missing that I can see in preview. Takes over an hour to process as I am using Neat video plugin to clean up old video noise from analog 3/4" video tape that is 30 years old. Later I will run Enlarger AI and see what happens as I have a 16 x 9 format I hope using a lower resolution but it is correct aspect ratio and i am guessing this will work. later I can change pixel resolution to match Enlarger AI and try again to compare. I do not think there will be much improvement since this video is not common and one of kind and not in AI learning database if this is how it works to improve known motion pictures. Maybe Enlarger AI is known DVD database with ongoing analysis to improve those type videos? I know not how program truly works to improve a known video and if it can improve a home video or not.

ord1117

Just saw your Post DataCaster,

You are using Programs that I have not used before.

But I think you are on to something.

I hope DVDfab is Listening.

signals

This forum is not for discussing other products, this thread is closed.

If must say for the final time that it is not possible to create a 16:9 image from a 4:3 image without distorting the image in some way, either adding bars, stretching/squashing or adding graininess (zooming/cropping).

Note that properly authored anamorphic DVD transfers will be correctly decoded by Ripper and Enlarger AI to a widescreen image.