[Dave]

OK, I mentioned elsewhere that I had read somewhere that reducing image sizes in a stage approach provides better PQ than 1 immediate reduction.

So I thought I had better test it on PS CS2 with default settings. Here is the results from a resize of original image from 2974 x 2211 to 800 x 594, in 2 manners :-

1. Image size reduced by 200 pixels each time until final result.
2. Image size reduced immediately to final result.

Below are crops from final photos for comparison :-



Correct me if I'm wrong but the 1st photo has darker tones, especially on left eye, plus definition round nostrils is sharper ? It's the kind of subtle difference others have found between medium & expensive glass......except this costs you nothing but a few seconds of your time

What do others think ?

[fingerz]

That's surprising actually. The top (gradually-reduced) is indeed slightly better.

It's weird because all the 'laws' of digitisation usually say that whatever you're doing, you should do it from the highest-quality original you can. And any time you make a digital manipulation (like resizing) you're reducing the image quality. So the gradual process ought to get progressively worse, not better.

There must be something odd happening in photoshop's algorithims. Hope they sort it out.

Edit - maybe the resizing process uses a tiny amount of unsharp masking? So by doing it several times you're applying a few sharpening passes, or something.

[SammyC]

I agree with your finding there Dave, I can't explain it but it does look sharper. I'm wondering if the resampling caused by each reduction is actually causing a kinda of USM side effect?

From a purists POV I would have to say that a single resize would lose less information than a progressive resize. But then that has nothing to do with what it looks like.

Veeeery strange!

[fingerz]

I'd be interested to see if the gradually-reduced image still looks better after some USM is applied (same amount to both). Especially when using a large amount of USM since (if it did look worse) that would suggest it already had some sharpening used on it.

[CT]

I'm not convinced yet.

I've noticed on my flat screen monitor that the image gamma changes slightly as you scroll the image up and down on the screen. This also seems to be dependant on the angle I'm viewing the monitor screen at, but it's a possible explanation for the very slight differences between those pics. I might add it's an expensive Proview monitor.

If you could post the images side by side I think it would be a more plausible test, but fair does for having the enquiring mind to pursue it. I hope you're wrong - I spend enough time in front of the monitor as it is!

[Dave]

As requested....gradual to left :-



Think you have a point there CT re side by side vs vertical stack.....I'm struggling to see see difference now, although highlights in left eye look more defined on gradual shot ? .....or have I been looking at them for too long

[SammyC]

They are still different but it's not a strong now. Right hand one does look a little softer but looks like it has more detail in it though.