genuine fractals?

It works better than re-sampling in PS but it isn’t magic. The more data you start with the better it works in my opinion. for instance, it can turn 4000 pixels into 8000 pixels better than it can turn 500 into 1000… Admittedly, I have only used it a few times. There website boasts, "Up to 7X" enlarging. I imagine it all depends on the detail in the original image..

I wouldn’t worry about installing the trial version unless you have fussy PC.

Once installed you will find the interface under File>Automate

"Peter" wrote in message

I do a lot of mining of old shots. This consists of taking fractions of images and blowing them up to approximately 12 x 18. so far I have used Photoshop CS2, with Bicubic smoothing. I know that I cannot recover details that are not in the shot, there, but I would like to recover all detail that is present. I have seen some comment here that GF is like "snake oil." It seems to me that since GF converts the bit map image to a vector image, it should allow one to make the type of enlargements I like to do, without loss of detail.
Has anyone had any experience with this product? I would prefer not to install, (even a trial version,) unless there is a reasonable chance it will work.

TIA

—
Peter

Peter wrote:

I do a lot of mining of old shots. This consists of taking fractions of images and blowing them up to approximately 12 x 18. so far I have used Photoshop CS2, with Bicubic smoothing. I know that I cannot recover details that are not in the shot, there, but I would like to recover all detail that is present. I have seen some comment here that GF is like "snake oil." It seems to me that since GF converts the bit map image to a vector image, it should allow one to make the type of enlargements I like to do, without loss of detail.
Has anyone had any experience with this product? I would prefer not to install, (even a trial version,) unless there is a reasonable chance it will work.

TIA

I have an older version of it, but don’t use it any more. It does work, but only very slightly better than Photoshop’s Bicubic. In my own tests I see no difference between the two up to about double the size. In the really large jump ups, GF does seem to do a bit better. Luckily, I rarely need to enlarge that much.

Remember that none of these will add anything to the picture. i.e. You will never add detail that wasn’t there to begin with. So, the real question is how smooth is the up sizing? That is what my comparison was.

One of the reasons that I got GF was that it would make much smaller files. They were a lot smaller in "near lossless" mode. I never saw any quality loss in that mode. In the "lossless" mode they were about half the size of PSD. Years ago, I didn’t have the computer space and that was an issue. Now I have plenty of space. Besides, JPEG2000 is a wavelet compression method that does lossless compression even smaller than GF.

Clyde

"Clyde" wrote in message

Peter wrote:

I do a lot of mining of old shots. This consists of taking fractions of images and blowing them up to approximately 12 x 18. so far I have used Photoshop CS2, with Bicubic smoothing. I know that I cannot recover details that are not in the shot, there, but I would like to recover all detail that is present. I have seen some comment here that GF is like "snake oil."
It seems to me that since GF converts the bit map image to a vector image, it should allow one to make the type of enlargements I like to do, without loss of detail.
Has anyone had any experience with this product? I would prefer not to install, (even a trial version,) unless there is a reasonable chance it will work.

TIA

I have an older version of it, but don’t use it any more. It does work, but only very slightly better than Photoshop’s Bicubic. In my own tests I see no difference between the two up to about double the size. In the really large jump ups, GF does seem to do a bit better. Luckily, I rarely need to enlarge that much.

Remember that none of these will add anything to the picture. i.e. You will never add detail that wasn’t there to begin with. So, the real question is how smooth is the up sizing? That is what my comparison was.
One of the reasons that I got GF was that it would make much smaller files. They were a lot smaller in "near lossless" mode. I never saw any quality loss in that mode. In the "lossless" mode they were about half the size of PSD. Years ago, I didn’t have the computer space and that was an issue. Now I have plenty of space. Besides, JPEG2000 is a wavelet compression method that does lossless compression even smaller than GF.

Thanks to all who replied.
When I used PS 7.0 the images seemed to posterize at high magnifications, even though I used the 10% rule when increasing size. With CS2, I don’t get as much posterization using bicubic smooth.
For mining purposes I scan my old slides and negatives at 1,200 for those shot that will not be highly cropped and not much larger than 12 x 16. 3,000 dpi for others. I can go up to 7,200 but speed and my lack of patience becomes an issue. If I need a very small area, I can always rescan at the higher res. (provided I can find the original.

—
Peter

In article <436a7906$0$42518$>,
"Peter" wrote:

I have seen some comment here that GF is like "snake oil." It seems to me that since GF converts the bit map image to a vector image, it should allow one to make the type of enlargements I like to do, without loss of detail.

Genuine Fractals does not convert the image to vector; that is inaccurate. What it does is perform a fractal wavelet compression on the image. The image remains raster.

Genuine Fractals retains all the information in the image already. So does Photoshop’s interpolation. That’s the problem. An enlarged image looks bad because there is not any more information in the image; the information contained in a raster image is sharply bounded by the total number of pixels in the image. When you resample upward, you are spreading the same amount of information over a larger number of pixels. As the size of the image increases linearly, the number of pixels increases exponentially.

—
Art, photography, shareware, polyamory, literature, kink: all at http://www.xeromag.com/franklin.html

"tacit" wrote in message

In article <436a7906$0$42518$>,
"Peter" wrote:

I have seen some comment here that GF is like "snake oil." It seems to me that since GF converts the bit map image to a vector image,
it should allow one to make the type of enlargements I like to do, without
loss of detail.

Genuine Fractals does not convert the image to vector; that is inaccurate. What it does is perform a fractal wavelet compression on the image. The image remains raster.

Genuine Fractals retains all the information in the image already. So does Photoshop’s interpolation. That’s the problem. An enlarged image looks bad because there is not any more information in the image; the information contained in a raster image is sharply bounded by the total number of pixels in the image. When you resample upward, you are spreading the same amount of information over a larger number of pixels. As the size of the image increases linearly, the number of pixels increases exponentially.

Thanks for the clarification on how GF works. But, my qauestion is still open, for making extreme enlargements, original file size .12 x .2 to 12 x 20, does GF help. On screen I see little difference, but as many here must know, there is a vast difference between wht we see on screen and the final print.

—
Peter
All things being equal,
a fat person uses more soap
than a skinny person.

Peter wrote:

"tacit" wrote in message

In article <436a7906$0$42518$>,
"Peter" wrote:

I have seen some comment here that GF is like "snake oil." It seems to me that since GF converts the bit map image to a vector image,
it should allow one to make the type of enlargements I like to do, without
loss of detail.

Genuine Fractals does not convert the image to vector; that is inaccurate. What it does is perform a fractal wavelet compression on the image. The image remains raster.

Genuine Fractals retains all the information in the image already. So does Photoshop’s interpolation. That’s the problem. An enlarged image looks bad because there is not any more information in the image; the information contained in a raster image is sharply bounded by the total number of pixels in the image. When you resample upward, you are spreading the same amount of information over a larger number of pixels. As the size of the image increases linearly, the number of pixels increases exponentially.

Thanks for the clarification on how GF works. But, my qauestion is still open, for making extreme enlargements, original file size .12 x .2 to 12 x 20, does GF help. On screen I see little difference, but as many here must know, there is a vast difference between wht we see on screen and the final print.

Actually, you’ve been answered many times. If you look back in newsgroups there are many more answers. Then again, you’ve already started the testing, why not finish it? Print the different versions and see if you can see a difference. If you can’t see a difference that matters to you, it doesn’t matter.

Clyde

In article <436e69ef$0$39503$>,
"Peter" wrote:

Thanks for the clarification on how GF works. But, my qauestion is still open, for making extreme enlargements, original file size .12 x .2 to 12 x 20, does GF help. On screen I see little difference, but as many here must know, there is a vast difference between wht we see on screen and the final print.

No. What you see on screen is what you see in print.

For some images, Genuine Fractals sometimes produces results that are slightly better than Photoshop–emphasis on "some images," "sometimes," and "slightly." If it’s better on screen, it’s better in print, and vice versa.

For extreme enlargements, nothing can beat, or even come anywhere close to, just making the image at the correct resolution in the first place.

—
Art, photography, shareware, polyamory, literature, kink: all at http://www.xeromag.com/franklin.html

"Clyde" wrote in message

Peter wrote:

"tacit" wrote in message

In article <436a7906$0$42518$>,
"Peter" wrote:

I have seen some comment here that GF is like "snake oil." It seems to me that since GF converts the bit map image to a vector image,
it should allow one to make the type of enlargements I like to do, without
loss of detail.

Genuine Fractals does not convert the image to vector; that is inaccurate. What it does is perform a fractal wavelet compression on the image. The image remains raster.

Genuine Fractals retains all the information in the image already. So does Photoshop’s interpolation. That’s the problem. An enlarged image looks bad because there is not any more information in the image; the information contained in a raster image is sharply bounded by the total number of pixels in the image. When you resample upward, you are spreading the same amount of information over a larger number of pixels. As the size of the image increases linearly, the number of pixels increases exponentially.

Thanks for the clarification on how GF works. But, my qauestion is still open, for making extreme enlargements, original file size .12 x .2 to 12 x 20, does GF help. On screen I see little difference, but as many here must know, there is a vast difference between wht we see on screen and the final print.

Actually, you’ve been answered many times. If you look back in newsgroups there are many more answers. Then again, you’ve already started the testing, why not finish it? Print the different versions and see if you can see a difference. If you can’t see a difference that matters to you, it doesn’t matter.

I have to apologize for my blankout moment. You are right, of course.

—
Peter
All things being equal,
a fat person uses more soap
than a skinny person.

original file size .12 x .2

..12 x .2 What? Inches @ 300dpi?

If you can print it at 300 dpi and get anywhere close to the size of a 35mm negative, then you can scan it on a flatbed that can do slides or negatives at 3200 or large PPI you "Might" then be able to print it at a magnification of the magnitude you are after.

Genuine Fractal bost the ability to Enlarge an image, "Up TO" 7X, You are asking for 100X You can’t do that via any type of re-sampling and expect to have good results.

http://imagequest.netfirms.com

"Peter" wrote in message

"Clyde" wrote in message

Peter wrote:

"tacit" wrote in message

In article <436a7906$0$42518$>,
"Peter" wrote:

I have seen some comment here that GF is like "snake oil." It seems to me that since GF converts the bit map image to a vector image,
it should allow one to make the type of enlargements I like to do, without
loss of detail.

Genuine Fractals does not convert the image to vector; that is inaccurate. What it does is perform a fractal wavelet compression on the image. The image remains raster.

Genuine Fractals retains all the information in the image already. So does Photoshop’s interpolation. That’s the problem. An enlarged image looks bad because there is not any more information in the image; the information contained in a raster image is sharply bounded by the total number of pixels in the image. When you resample upward, you are spreading the same amount of information over a larger number of pixels. As the size of the image increases linearly, the number of pixels increases exponentially.

Thanks for the clarification on how GF works. But, my qauestion is still open, for making extreme enlargements, original file size .12 x .2 to 12 x 20, does GF help. On screen I see little difference, but as many here must know, there is a vast difference between wht we see on screen and the final print.

Actually, you’ve been answered many times. If you look back in newsgroups there are many more answers. Then again, you’ve already started the testing, why not finish it? Print the different versions and see if you can see a difference. If you can’t see a difference that matters to you, it doesn’t matter.

I have to apologize for my blankout moment. You are right, of course.
—
Peter
All things being equal,
a fat person uses more soap
than a skinny person.

tacit writes:

tacit> In article <436a7906$0$42518$>,
tacit> "Peter" wrote:

I have seen some comment here that GF is like "snake oil." It seems to me that since GF converts the bit map image to a vector image, it should allow one to make the type of enlargements I like to do, without loss of detail.

tacit> Genuine Fractals does not convert the image to vector; that is tacit> inaccurate. What it does is perform a fractal wavelet compression on the tacit> image. The image remains raster.

tacit> Genuine Fractals retains all the information in the image already. So tacit> does Photoshop’s interpolation. That’s the problem. An enlarged image tacit> looks bad because there is not any more information in the image; the tacit> information contained in a raster image is sharply bounded by the total tacit> number of pixels in the image. When you resample upward, you are tacit> spreading the same amount of information over a larger number of pixels. tacit> As the size of the image increases linearly, the number of pixels tacit> increases exponentially.

Quadratically, not exponentially.

—
Andrew Hall
(Now reading Usenet in alt.graphics.photoshop…)

In article <436a7906$0$42518$>,
Peter wrote:

I do a lot of mining of old shots. This consists of taking fractions of images and blowing them up to approximately 12 x 18. so far I have used Photoshop CS2, with Bicubic smoothing. I know that I cannot recover details that are not in the shot, there, but I would like to recover all detail that is present. I have seen some comment here that GF is like "snake oil." It seems to me that since GF converts the bit map image to a vector image, it should allow one to make the type of enlargements I like to do, without loss of detail.
Has anyone had any experience with this product? I would prefer not to install, (even a trial version,) unless there is a reasonable chance it will work.

TIA

It really depends IMHO on the quality of the image in question just as everyone said. I use GF for a magazine that I design and I’m often given small images that I need to enlarge. If the quality of the image is good (lighting, focus, etc ~ too many variables to list) then GF does a great job. But as an unknown genuis (atleast, I can’t remember who said it) once said ~ you can’t polish a turd. If the image sucks, no resampling will improve it. You’ll just have a
bigger-but-just-as-crappy image.

And note: back when LizardTech owned GF it was $178 for a license for the PrintPro version . . . now since it was sold to OnOne Software it’s $299 and I can’t see any visible improvements made in the plug-in. Go figure.

Good luck!!

—
Constance Pierce
principal/designer

"you can’t polish a turd."