From: "pixel_gnome", ray.landis@techbooks.com
Date: Mon, Mar 18, 2002, 2:21 AM
RE: [colortheory] Slides enlarged over 800%
We often receive slides that our customers want enlarged 800 to 1500%. Needless to say they then complain when they print grainy and out-of-focus. We have had some success by raising the resolution from 300 to 6 or 800 but these results still reflect a soft image. Does anyone have any other solutions to this type of work?Thanks,
Ray Landis
Techbooks
ray.landis@techbooks.com
From: jharmon@levi.com, jharmon@levi.com
Date: Mon, Mar 18, 2002, 4:58 PM
RE: RE: [colortheory] Slides enlarged over 800%
Have you had any luck upsampling with Lizard Tech's Genuine Fractals Pro? I've heard good things about its interpolation method, and many people use it to make billboards and such...http://www.genuinefractals.com/products/genuinefractals.php
(No, I don't work for them and have no financial or personal reason for endorsing their products.)
-- Jeff Harmon
Colorhythm
From: Jan Steinman, Jan@Bytesmiths.com
Date: Mon, Mar 18, 2002, 4:57 PM
RE: [colortheory] Re: Slides enlarged over 800%
>We often receive slides that our customers want enlarged 800 to
>1500%. Needless to say they then complain when they print grainy and
>out-of-focus. We have had some success by raising the resolution
>from 300 to 6 or 800 but these results still reflect a soft image.
>Does anyone have any other solutions to this type of work?You sound like a printer -- talking in percentages. That's just an observation, not meant as any sort of insult!
What are you scanning on? If you're limited to "300 to 6 or 800," it sounds like you may be using an inexpensive flatbed scanner.
Even a cheap film scanner should have no problem scanning at 2400 spi, which would be 800% with a 150 line screen. A prosumer film scanner, drum scanner, or high-end flatbed should do 4,000 spi, which is pretty close to 1500%. So you might consider sending those scans out, if your scanner doesn't have at least 4,000 spi.
A lot depends on the slide. If it's shot at ASA 800 its likely to be sharp, but grainy. If its shot at ASA 100 without a tripod, its likely to be blurry, but relatively grain-free. I have yet to find a tactful way of telling a customer, "Garbage in, garbage out!"
All that aside, if I have to go beyond my scanner's capabilities, I turn to Genuine Fractals first. It will often (not always) do a better job than Photoshop's bicubic interpolation, and will occasionally do a spectacularly better job, especially on scenes that are naturally fractal-like, like foliage. I have a mossy tree, printed 24"x36" from 35mm Kodachrome 64, that experienced photographers examine closely, then ask, "So, what sort of large format equipment do you use?" :-)
-- : Jan Steinman
: Bytesmiths
: 19280 Rydman Court, West Linn, OR 97068, 503.635.3229
From: "merckp", pmerck@cinci.rr.com
Date: Mon, Mar 18, 2002, 7:51 PM
RE: [colortheory] Re: Slides enlarged over 800%
> We often receive slides that our customers want enlarged 800 to
> 1500%. Needless to say they then complain when they print grainy and
> out-of-focus. We have had some success by raising the resolution
> from 300 to 6 or 800 but these results still reflect a soft image.
> Does anyone have any other solutions to this type of work?
Dear RayWhat kind of scanner, drum or flatbed ? You should try to oil mount the trannies. If you're seeing alot of grain and you can can manually focus the scanner (some drum can auto focus) on the grain of the trans, then guess what, the photographer can't focus his camera. I can't tell you the number of times I've scanned a 35mm trans up to several thousand %, focused on the grain and been told my scanner screwed up. Blurry is blurry, look at the trans under a loop. Also some detail is too small to be resolved by the film. As for grain adjust your unsharp masking settings instead of dpi.
Pete Merck
From: "pixel_gnome", ray.landis@techbooks.com
Date: Tue, Mar 19, 2002, 7:58 AM
RE: [colortheory] Re: Slides enlarged over 800%
We have attempted these enlargements on an Eversmart Supreme. The film grain is noticeable under a loop and of course when enlarged becomes even worse. We also have a Screen 608 drum which we also have tried (although hesitantly - due to costs) which seems to help only marginally. We also have used Genuine Fractals, which on these subjects has done no better than interpolation through Photoshop 6. The complaint from our client is "too grainy and background is too blurry". Our attempt to improve by increasing the scanner resolution on the Eversmart seems to have worked the best, even allowing a small degree of unsharp masking to be applied. But overall, the images are less than stellar. The one comment which seems most applicable is "garbage in, garbage out" but I was hoping that someone here may have found a more effective solution.Thanks,
Ray Landis
Techbooks
From: "Stephen Marsh", samarsh@ozemail.com.au
Date: Tue, Mar 19, 2002, 12:58 PM
RE: [colortheory] Re: Slides enlarged over 800%
--- In colortheory@y..., "pixel_gnome" wrote:
> We have attempted these enlargements on an Eversmart Supreme. The
> film grain is noticeable under a loop and of course when enlarged
> becomes even worse. We also have a Screen 608 drum which we also
> have tried (although hesitantly - due to costs) which seems to help
> only marginally.That's a shock about the Drum, I have some minor experience with the ES Supreme/oXYgen combo and was never impressed when it came to slides, negs or prints. Good for small or MF tranny or batching many slides for lesser size/quality reproduction though. Grain and noise always seem an issue, as was a lot of dust. The workflow seemed the biggest sell, as well as those great proprietary separation table results that blow away profiles in most cases. There is acceptable grain removal found under the sharpen section (perhaps turn off USM but use the grain softening). Although Photoshop smart blur/edge masks and/or Dans duped darken/lighten curve blends are good too (check at foot of this message). For serious work, plugs like GrainSurgery, Quantum Mechanic Pro or Neat Image (PC). Perhaps a slide scanner with GEM or other hardware grain reduction.
> We also have used Genuine Fractals, which on these
> subjects has done no better than interpolation through Photoshop 6.
> The complaint from our client is "too grainy and background is too
> blurry". Our attempt to improve by increasing the scanner resolution
> on the Eversmart seems to have worked the best, even allowing a small
> degree of unsharp masking to be applied. But overall, the images are
> less than stellar. The one comment which seems most applicable
> is "garbage in, garbage out" but I was hoping that someone here may
> have found a more effective solution.There was a method mentioned by Katrin Eisman on another list, if I recall correctly - dupe layer, apply emboss filter, blend layer in overlay/hard-softlight. Adjust opacity to taste. Do not expect much, but if you can reduce the grain and define the edges better via this and USM tricks or selective minor blurring on non critical areas to make the other areas appear sharper etc.
Or accept the flaws and turn them into design elements...
Regards,
Stephen Marsh.
======
[Dan's quick grain/noise/spot/artifact removal steps]
Message archive #1551
1) Make four identical layers. Layer 1 remains untouched.
2) On Layer 2, Gaussian blur, Radius 2.0 (for starters, may adjust).
3) Having blurred Layer 2, apply a curve that lightens the midtone very slightly.
4) Change layering mode on Layer 2 to Darken. This will disallow most of the layer because of Step 3, and will only fill in small white blotches.
5) On Layer 3, apply Gaussian blur at double the radius of Step 2. (Reason: darker spots on the skin tend to be larger than white spots).
6) Having blurred Layer 3, apply a curve that darkens the midtone very slightly.
7) Change layering mode on Layer three to Lighten. This will disallow most of the layer because of Step 6, and will only correct age spots and other dark blemishes.
8) Compare Layer 3 to Layer 4 (an untouched original) and merge the two as desired with a layer mask.
This, incidentally, is how one "unsharpens" a picture, when a client has been so helpful as to supply something that's been oversharpened, without giving us access to the original.
Dan Margulis
From: ks@mountain-mall.com, ks@mountain-mall.com
Date: Tue, Mar 19, 2002, 12:09 PM
RE: Re: [colortheory] Re: Slides enlarged over 800%
On Tue, 19 Mar 2002 12:50:47 -0000, your pet rodent escaped it's cage, climbed the computer table, ran across the keyboard and accidently hit keys that wrote:>We have attempted these enlargements on an Eversmart Supreme. The
>film grain is noticeable under a loop and of course when enlarged
>becomes even worse. We also have a Screen 608 drum which we also
>have tried (although hesitantly - due to costs) which seems to help
>only marginally. We also have used Genuine Fractals, which on these
>subjects has done no better than interpolation through Photoshop 6.
>The complaint from our client is "too grainy and background is too
>blurry". Our attempt to improve by increasing the scanner resolution
>on the Eversmart seems to have worked the best, even allowing a small
>degree of unsharp masking to be applied. But overall, the images are
>less than stellar. The one comment which seems most applicable
>is "garbage in, garbage out" but I was hoping that someone here may
>have found a more effective solution.
Could it be that the slides were shot on an older 400 or higher speed film that is known to have grain the size of baseballs? Now if this were Velvia or Provia 100F then there is a processing problem, but if it is a source problem, then nothing will help. Is the slide really sharp and in focus? Was the exposure correct? Is the slide's color properly saturated?What I'm trying to say is you could have a source problem, not a processing problem. Have you been able to use this process on other slide or is this your first attempt?
Karl Snyder
Boulder, Colorado
http://www.RockyMountainNationalPark.Info
http://www.EstesParkOnLine.Com/
http://www.MtEvans.Com/
Moderator of the Yahoo OutdoorPhotographers group
To join:
OutdoorPhotographers-subscribe@yahoogroups.com
From: RJay Hansen, rjhansen@myrealbox.com
Date: Tue, Mar 19, 2002, 11:44 AM
RE: Re: [colortheory] Re: Slides enlarged over 800%
This may be too obvious, but have you tried scanning at a higher-than-needed resolution and then downsampling? (i.e. if the scan needs to be made at 1600 ppi, scan at about 3500 ppi then downsample to 1600). This softens the focus, but I think tends to remain truer to the original than something that is upsampled. I would also agree with the earlier statement that playing with the unsharp mask settings, during the scan may be helpful.RJay
From: Gary Roushkolb, garyr@donlevylitho.com
Date: Tue, Mar 19, 2002, 4:47 PM
RE: Re: [colortheory] Re: Slides enlarged over 800%
Try just Blurring the channels with the excessive grain like the yellow. the channels with less grain can be sharpened. I do this with my eversmart scans. I love the comment "we want less grain but make it sharper" I've had that written on proofs.Good Luck,
Gary Roushkolb
Scanner operator, Donlevy Lithograph
From: "Dave King", kingphoto@mindspring.com
Date: Tue, Mar 19, 2002, 10:10 PM
RE: Re: [colortheory] Re: Slides enlarged over 800%
Make sure there is no default sharpening when scanning, then sharpen, then use history brush to revert the smooth tone areas that are most "grain offensive" back to unsharpened.Dave
From: James Carambat, jcarambat@kwinc.com
Date: Wed, Mar 20, 2002, 7:48 PM
RE: Re: [colortheory] Re: Slides enlarged over 800%
> I've been lurking in the background for awhile now and this is my first
> posting. What I do on my drum scanner is to drop the Sharp down equal to the
> Smooth then crank up the Threshold, maybe twice what I normally run.The
> results are super smooth, but still holding "good detail" Good Luck!James Carambat
Scanner Operator
K&W Inc.
From: Dan Margulis
Date: Fri, Mar 22, 2002, 1:04 AM
RE: RE: [colortheory] Slides enlarged over 800%
Jeff writes,>>Have you had any luck upsampling with Lizard Tech's Genuine Fractals Pro? I've heard good things about its interpolation method, and many people use it to make billboards and such...>>
I reviewed it extensively and my findings were supported by other members of the group who have made the comparison. We did not find it was significantly better than just upsampling in Photoshop. However, in certain cases, chiefly where there were fine lines in the image, GF did a better job of presenting them than Photoshop did. OTOH, GF tended to posterize fleshtones and we preferred Photoshop's rendition of them.
Dan Margulis
From: Andrew Rodney, andrew@digitaldog.net
Date: Sun, Mar 24, 2002, 7:35 AM
RE: Re: [colortheory] Slides enlarged over 800%
on 3/23/02 6:49 PM, Dave King at kingphoto@mindspring.com wrote:> Fine lines in the sense of simple contrast juxtapositions, yes I agree, but
> in areas of complex texture (in my test image, foreground grass) GF creates a weird sort of "computer generated" look.Yes, I saw exactly the same issues. Kind of a grid like pattern which doesn1t look all that great. Given the choice between a soft looking resize or this odd grid pattern you mention, I1ll stick with Photoshop.
Andrew Rodney
From: "Dave King", kingphoto@mindspring.com
Date: Sat, Mar 23, 2002, 9:12 PM
RE: Re: [colortheory] Slides enlarged over 800%
Fine lines in the sense of simple contrast juxtapositions, yes I agree, but in areas of complex texture (in my test image, foreground grass) GF creates a weird sort of "computer generated" look. Bi-cubic in PS at 100% on the monitor seemed considerably more "natural" looking to my eye, but those differences in a large format print were minimized to the point there is no significant difference between bi-cubic and GF, except that GF is considerably slower.Perhaps for rezing up very small files to extremely large sizes GF comes into it's own?
Dave King
> I reviewed it extensively and my findings were supported by other members
> of the group who have made the comparison. We did not find it was
> significantly better than just upsampling in Photoshop. However, in certain
> cases, chiefly where there were fine lines in the image, GF did a better
> job of presenting them than Photoshop did. OTOH, GF tended to posterize
> fleshtones and we preferred Photoshop's rendition of them.
>
> Dan Margulis
Adobe Photoshop training classes are taught in the US by Sterling Ledet & Associates, Inc.