[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
  Web www.spinics.net

Re: Genuine Fractals - Yet Another Test/Opinion

Open Letter to Mr. Zax,

It's too bad that Altamira doesn't support your on your positions (below) on
this product. But, if they did most interested parties would be informed
that they don't need the product or can't use it.

on 8/12/01 9:34 AM, Jon Zax at lotus@ionix.net wrote:

> As some on this list might recall, I was on the beta team for the GF products,
> and submitted
> them to tests beyond any that have been reported or referenced here in recent
> posts.
> First, the source image is critical in evaluating the performance of the
> plug-in. In my tests
> I used scans from a polaroid ss plus, drum scans, a kodak consumer digital
> camera and
> a high end studio 3 pop digital back on a sinar 4x5.
> The old computer adage garbage in garbage out  really applies here.
> Frankly, while the polaroid and nikon scanners do allow one to actually do
> digital imaging,
> they are toys.

Your claim that Nikon and Canon film scanners are toys, and suggestion that
these are not appropriate for the Genuine Fractals, rings hollow. Genuine
Fractals is bundled with a variety of "toy" scanners. Are you suggesting
that Altamira Group and the scanner purveyors misunderstand the purpose of
the product and bundle it inappropriately?

> What I found was that there is a base level of image data that needs to be
> there for GF to
> really work.
> I made drum scans at various resolutions from 2 megs to a full size file that
> would print out directly at 20x24. All prints were made on a Iris printer. The
> sweet spot was between
> 20 and 30 megs. Prints from that file size when rezed up in GF were as good as
> the full
> sized file.
> The workflow that this suggests is that all one needs is a 30 meg file to
> print out to any size
> with GF, not to store all your files as .stn.

I would think folks who have access to high quality scans are least likely
of any potential buyer to need this tool because they already have access to
more than enough high quality image data for any typical print purpose.

I contend that what you call the sweet spot (about 25MB of image data)
happens to be significant for enlargements in a psycho-visual sense: roughly
25MB of solid image data (e.g., from a drum scanner), carefully handled, can
fill a normal person's field of vision under a wide range of viewing
conditions with a striking high quality image. 25 MB is the sweet spot for
most image applications, regardless of the use of Genuine Fractals.

> The files from the kodak camera were pretty bad no matter what I did, but the
> files from the
> studio digital back, while only 12 megs, did produce amazing results. I shot
> the same set up
> with film, then had a digital 4x5 imaged on a 16k solitaire film recorder and
> showed them
> around side by side. although there was a slight difference, no one could tell
> which one was
> digital without a loupe. The file size for the film recorder was 120 megs.

> Technically, GF does not interpolate pixels or resample. It converts the image
> to something
> more like a vector file using fractal algorithms. When asked to resize a brand
> new file is
> rendered at the new size. It is certainly true that no new detail can be
> added, but the fractal
> nature of the file keeps it from having to guess edge and color detail, as is
> done in
> resampling.

This "vector file" claim is so mind-numbingly empty and tedious.

I call you to demonstrate your claim that whatever it is that Genuine
Fractals does it not interpolation by providing a formal explanation of
vector image encoding and compare and contrast this with a similar
definition for raster image encoding.

For extra credability, explain what "guessing" means in the sense of the
behavior of a computer program, and how interpolation can be considered a
form of guessing whereas whatever it is that Genuine Fractals does should
not be considered guessing.

> This is a highend product and not for most users.
> J.Z.

Altamira Group claims basically nothing for this product, so it's not hard
to satisfy their claims. But their marketing spin includes a lot of innuendo
that seems (if it says anything) to contradict your assertion that this is a
"high end product," in the sense of "high end" meaning that drum scans are

Altamira Group recommends the product for all kinds of image files, and
promises the results will be "beautiful."

The way I see it, Genuine Fractals doesn't ruin images, so when my images
are beautiful, I can expect the same from the output of Genuine Fractals.


Turn off HTML mail features. Keep quoted material short. Use accurate
subject lines. http://www.leben.com/lists for list instructions.

[Photo]     [Yosemite News]    [Yosemite Photos]    [Scanner]     [Gimp]     [Gimp] Users

Powered by Linux