Aspect Ratio Vs Resolution

[matteob50581971]

Hello,

I have a problem that I can't quite understand and it might be a stupid question but I need to ask anyway.

I'm currently working on a indie movie.

The movie was shot in 4K (3840x2160) in 16:9 ratio.

At first, to make a cinematic look, I created a sequence on 16:9 and then I added the cinematic bars (Top and Bottom) to simulate the 2.39 ratio.

Then I looked a little more into it and it seems like the best solution is actually create the 2.39 ratio sequence and so I did.

I created a sequence that is 3840x1606 with the intention to export later in 3840x2160 which will automatically create the black bars.

First of all does anybody think that this is really the best solution?

Second: am I losing any quality by doing that? somehow, the footage looks a little less sharp when the sequence is in 2.39 aspect ratio. I tried rendering it and it still not 100% as good as before.  (please keep in mind that my video preview setting are: 1920x1080 or in the case of the 2.39 sequence: 1920x803. But so it was before, and still when I use the 16:9 ration, it looks sharper) Is it all in my Mind?

Thank you so much to anybody willing to share an opinion.

[Mike Dziennik]

Try making your sequence 2.4:1 rather than 2.39:1 giving 'rounder' raster dimensions which will probably render sharper.

Ie.

1920x800

or

3840x1600

and yes...if you don't have to output to an external monitor then cutting in a 2.39:1 (or 2.4:1 in this case) sequence is perfectly acceptable and it will be easy to add a letterbox on export for a 16x9 delivery.

[matteob50581971]

Hey Mike,

Thank you for your answer.

I will def try.

Can I ask you why though? How could it be that even less pixels make it sharper?

Sorry to bother you, but I'm trying to really understand this.

[R Neil Haugen]

Different formats and codecs are built with different mathematical formulas as to what size of blocks they use to compute image when compressing or decoding.

Some are four pixel blocks, some eight I've been told. So the safest choice is to make frame-sizes divisible by four. Making a frame-size outside the codec's natural math requires every bit of the frame be modified. Not nearly as accurate and therefore not nearly as detailed or sharp.

This exhausts my knowledge of the subject but there are others here who know the full details better.

Neil

[Mike Dziennik · Accepted solution]

It will appear sharper because Premiere doesn't have to blend pixels together as much when the maths involved in the transform is simpler. Imagine a vertical bar of 4 pixels:

Resize it by half (divide by 2 - simple math)

Nice result.

However, resize it to make it 3 pixels tall (more complex)

Smooth perhaps - but not sharp.

[Ann Bens]

I would edit in 1920x1080 or 3840x2160 add bars for editing sake and crop in the export settings to what ever ratio.

[matteob50581971]

That actually makes it clear.

Thank you!.

So, if different cameras have different codecs and compression calculation, how is it possible that 2.39 or 2.35 is a universal ratio and more people use those ?

Also, Would you think it's the best way to use the cinematic bars?

Is there a better way I'm not aware of?

Thank you so much,

I really appreciate your help!

[Mike Dziennik]

Of course it is more complex than this - and as Neil says, different codecs handle it differently and it's about blocks if you want 'perfect' scaling.

[matteob50581971]

Ok, I will use 2.4:1 and see what happens.

But, Am I actually losing quality by doing that?

What's the best way of exporting with black bars?

(let's say for a theater projector, not youtube or social media)

Thanks

[MyerPj]

matteob50581971  wrote

Ok, I will use 2.4:1 and see what happens.

But, Am I actually losing quality by doing that?

Give it try on a clip or two put together in the timeline... then you'll look at it and know... in fact you could even try other resolutions and project sizes... uh, give it a try, no?