Re: Best software for increasing resolution?

On Oct 23, 3:39 am, IggyZiggy <bloc...@xxxxxxxxxxxx> wrote:
On Mon, 22 Oct 2007 22:49:42 GMT, "SS" <xsx20...@xxxxxxxxx> wrote:
I have seen that Genuine Fractals and Photozoom seem to be 'intelligent'
software but have also seen a review that puts the fred Miranda 'stair
interpolation' Photoshop action better than these. I wonder what
experiences/advice the subscribers to this newsgroup can give. Is there
anything even better on the market now.

Thanks

Any of the methods that are using PhotoShop's native bicubic interpolation and
its 16-bit math foundation (like Miranda's stair interpolation or that
Richardson-Lucy iteration) will be abysmal compared to other more evolved
interpolation methods. PhotoShop is still only using 16-bit math (think Windows
3.1), it has to throw away so much data required for the detail. If you are
using 16-bit images you're already at the limit of PhotoShop's math-bandwidth.

It might surprise you to know that many of the modern scientific
deconvolution codes had their roots in PDP-11 with FPS-120 arithmetic
array processors strapped on the side. Images were usually held as
16bit data because almost no techniques could originate source data
with that dynamic range (and disk space was scarce).

Trying to do anything complex with 16-bit data in a 16-bit math environment is
pushing it beyond what it was designed for and inherently limited to. No

Great care is needed to avoid overflows, and the main calculations are
done in floating point, but there is nothing inherently wrong with
16bit source data for storing images. Very few imaging systems use all
16 bits.

There's no one-size-fits-all solution when it comes to upsampling.

That at least is a true statement. You have to decide what you mean by
"best" before the question is answerable.

Do you want a plausible pretty looking image?
or
An image which shows the highest resolution detail and optimised
signal to noise consistent with the measurement process? (the latter
can be important in forensic work, but is almost never pretty to look
at)

Regards,
Martin Brown

.

Relevant Pages

  • Re: Positive image interpolation
    ... so any interpolation methods returning negative values does not ... properly model the image formation process." ... natural images are more accurately modelled by convolution followed ... Fourier coeffcients for frequencies close to the Nyquist rate. ...
    (sci.image.processing)
  • Re: Blackman interpolation instead of linear or cubic
    ... interpolation with Blackman interpolation, ... what we would like to see when filtering or reconstructing images. ... Fourier sense by taking the discrete fourier transform, ... The Blackman function is not meant as a filter in itself. ...
    (comp.graphics.algorithms)
  • Re: Scientifically Based Presharpening for Enlargement
    ... Your starting images should be uncompressed and unsharpened. ... interpolation, which is familiar to many people, and doesn't ... You might address why the sharpening should be a preprocessing step. ... the sharpening is a linear filter, ...
    (sci.image.processing)
  • Re: Convolution-based image interpolation
    ... It is quite common to interpolate images by FIRs. ... interpolation by non-simple integer ratios is needed, ... even be told the name of a single bell-shaped self-overlapping function ...
    (comp.dsp)
  • Re: Best software for increasing resolution?
    ... software but have also seen a review that puts the fred Miranda 'stair ... interpolation' Photoshop action better than these. ... photoshop's math, which is actually 15-bit, not 16-bit. ...
    (rec.photo.digital)