Grex Photography Conference

Item 5: Digital manipulation of images

Entered by raven on Thu Apr 6 05:17:45 1995:

        This is the item to discuss digital manipulation of phtographic images.
Examples might include, what is your favorite photoshop filter, what projects
are you working on, what refernces do you find helpful?
119 responses total.

#1 of 119 by mcpoz on Sat Apr 8 01:43:24 1995:

This is really interesting to me, but I know ZERO about it.  I am considering
spending some hard earned cash to play around with digital photos in general.
I hope we get some informed people to start a healthy discussion here.
Thanks.


#2 of 119 by zagman on Thu Apr 27 14:27:42 1995:

I am interested in purchasing one of the software programs to change/modify
the photo image from disc.  Could this technology make the darkroom obsolete?
If the poducts are half as good as the adds claim, they just might.  Is 
anyone familiar with such software?


#3 of 119 by rcurl on Thu Apr 27 20:21:42 1995:

Popular Photography (magazine) has had a number of articles about this
recently, and many examples. 


#4 of 119 by mcpoz on Fri Apr 28 00:20:12 1995:

I have no experience with digital manipulation, but Adobe Photoshop seems to
be most frequently mentioned.  Another thing to consider is the RAM level
of your computer.  Again, I am not experienced, but I think you need 16 M and
perhaps as much as 32 M.  Anyone else?


#5 of 119 by mwarner on Tue Jun 27 21:29:11 1995:

I'm considering buying a Microtek 35T slide scanner, which comes bundled
with Photoshop LE (limited edition).  It offers 8 bit times 3 passes
scanning for color.  The difference between this and a typical flatbed
scanner is a higher optical resolution: more dots per inch.  BTW, this
requires 8 Meg to run (which I have :) ).  This unit seems about the same
in optical quality as a Polaroid model costing around 2X as much which
features faster scanning.  I wonder if anyone in this cf has any
experience with these types of scanners.  The 8 bit models run from $950
to $2,000, while fancy 10 or 12 bit models now being advertised go for
upwards of $2,500 to $6,000.  As I am mostly interested in an archive of
of files I can use via a browser like Netscape, and not for a Nth degree
quality presentation, I won't be considering the top end models...


  Maybe this item should be linked to one of the computer cfs so we can
learn something from some of the users of this sort of hardware that
might not be on the photo cf.




#6 of 119 by mcpoz on Wed Jun 28 00:50:09 1995:

I have droped an e-mail note to the fw's of the micro cf - (Omni & Jshafer)
asking them if they think it is appropriate to link, and if so to go ahead.
We'll see!


#7 of 119 by ajax on Thu Jun 29 16:38:28 1995:

  I use Photoshop on my Mac...it's a great program, though for basic
functions (contrast/brightness/hue/saturation adjustment) there are
cheaper/free programs.  It's also an incredible memory hog...I run it
with just 8 megs of RAM (half of which is used for the operating
system), which makes things darned slow, but even with 16 megs it's
pokey.  Depends on image size too...Adobe offers some rule of thumb,
like you ought to have 8 megs plus two or three times as much RAM as
the image size, or something like that.
 
  I'm not too familiar with darkrooms, but from what I've seen, if you
have an okayish print, you can scan it and run software to do anything
you'd do in a darkroom.  Though the resolution on the computer is
usually lower, and converting the image to hard copy, with screen &
printer colors synched, is an art unto itself.  On the other hand, it
wouldn't surprise me to see film pretty obsolete in a couple decades.


#8 of 119 by mcpoz on Fri Jun 30 01:16:41 1995:

Can you burn and dodge on a computer?  (Selectively over or under expose?)


#9 of 119 by mwarner on Fri Jun 30 03:03:30 1995:

one pixel at a time....


#10 of 119 by ajax on Fri Jun 30 15:17:15 1995:

  I'm not sure of the exact effect you want, but you can select
a subsection of an image, and apply a transformation on it
(brightness adjustment would be similar to under/over-expose).
Normal retouching of part of a picture this way is quite easy.
You could also easily make a section redder, sharper, or blurrier.
If you're trying to make neon-like glows around parts of the
image, it's a bit trickier, but you can do that too.


#11 of 119 by helmke on Fri Jun 30 21:08:21 1995:

The big problem is reproducing in the digital realm the effect known as "the
magic of film".  If you have ever seen a "making of" about a movie and all
the explosions look super-cheesy, it is because the special is on video tape
vs. film.


#12 of 119 by srw on Sun Jul 2 04:56:10 1995:

Yes you can burn and dodge in photoshop (3.0). Select the "toning tool"
it looks like a lollipop. Double click it to set options (and brush
size/shape). It also offers a "sponge" option. I'll have to go get the
manual to figure out what that does.

Photoshop is demanding on your hardware, but if you feed it enough iron,
and ram, it works really well.


#13 of 119 by mcpoz on Sun Jul 2 11:02:27 1995:

I currently don't have enough power to run Photoshop.  If I get 8 meg ram to
add to my 486 dx would that do it?  Do I need anything else like a Graphics
accelerator card (whatever that does)?  Finally, if I don't invest in a super
snazzy color printer, can I go to any place in AA (Kinko's) and have the files
printed from a disk?


#14 of 119 by srw on Sun Jul 2 15:53:00 1995:

I am only familiar with the Mac ports of Photoshop. THe original port,
requires a 68020 or better MC68k processor and requires 6MB of
application memory. For a 68K Mac running System 7, 8MB would be a squeeze,
but you could do it.

For the PowerMac version, it requires a PPC601 and 11MB application
memory. A 16MB system would be required.

For the 386/486 PC version, I don't have any literature. My guess is
that 8MB Ram would be inadequate, although Rob Argy says he can do it.
I am not familiar with Photoshop on this architecture, as I said, 
so I'm only guessing. 16MB ought to be enough I would think, unless
you were running Windows NT.

Photoshop includes its own virtual memory implementation, so if you
are willing to wait while it swaps images in memory, it can do a lot
more than you might think with less memory. I have gotten fed up
with waiting and added more RAM to my system, though. I don't
recommend challenging Photoshop's VM by running it in a small space unless
you just can't afford more.


#15 of 119 by ajax on Sun Jul 2 20:45:01 1995:

  Actually, I run PS on my Mac with 8MB of RAM, though I'm not sure if
I need virtual memory for that.  I prefer PS 2.5.2 to PS 3.0 for most
things because it's quite a bit faster.  A friend of mine uses a PowerPC
Mac with 8MB RAM with PS 3.0 a lot, but it is excruciatingly slow, and
I know it uses virtual memory.
 
  The downtown AA Kinko's has the best publicly usable computer equipment
I've seen.  They have a nice new Apple color postscript printer, though
it's $2/page, and an HP black & white printer, and perhaps some others.
Photoshop is available at least on their Macs, and they can read PC disks
if they don't have it installed on their PCs.  They also have an Apple
OneScanner on one Mac ($20/hr to use it), or they can scan images for you
on either platform for $10/scan.


#16 of 119 by mcpoz on Mon Jul 3 00:50:36 1995:

Thanks.  I guess if I decide to go for it, it'l have to be 16meg (ouch).


#17 of 119 by srw on Mon Jul 3 05:37:38 1995:

A PowerPC Mac with 8MB is going to be very slow. I have my PPC Mac
set to give Photoshop 3.0 it's 11MB suggested. It works great!


#18 of 119 by mwarner on Mon Jul 3 05:43:53 1995:

There is also something called "Ram Doubler" which is a background
application to double your ram via various economies, with disk swapping
as a last alternative.  It is supposed to work much better than virtual
ram in terms of not slowing down a system.  I'm not familiar with how well
it would work with a real hog like Photoshop.  I also understand that it
works best when you are starting with 8 or more meg, but will also work
with a minimum of 4 meg.  I have it installed on my 8 meg Mac, giving me
an apparent 16 meg of ram.  I haven't put it to a serious test yet,
though.  IT is available for both Mac and Windows for about $55.



#19 of 119 by ajax on Mon Jul 3 23:07:09 1995:

  Another alternative is to get something other than Photoshop!  HSC
(publishers of Kai's Power Tools) recently released a competitor, with
one of its main claims to fame being *speed*...it allegedly edits half
gig images faster than PS can edit half meg images.  Quark is also
coming out with a competitive product, which is supposed to be more
speed-oriented than PS.  I'm not totally clear on the details, but
both sound like they're more into storing images and modifications to
the images, giving a speed boost over PS's approach of changing each
affected pixel of an image as its modified, even if it's off-screen or
too small to see on screen.  Apple also publishes an intro-level
image-editing package for around $100 that handles the basics.
 
  I'm thinking of popping $100 for a new 3d graphics package that
models the human body.  It looks like a fun tool to play with.  Not
sure if I'll need something else to render the wireframes though.
Can anyone recommend a cheap-but-decent ($100 or less) general-purpose
3d package for the Mac?  Seems like the good ones are $500 or more.


#20 of 119 by srw on Tue Jul 4 05:30:07 1995:

I strongly recommend Ray Dream Designer. It is not under $100, but it is
within reach at about $250. I like how it does solid and surface textures,
and text, especially (bevelled 3-d letters, for example). Relatively
fast rendering, too. No animation (I can live without it).
Modeling is pretty good too, but it doesn't allow you to edit a
surface point-by-point once you have constructed it, like the really 
expensive packages can. I think it's a great bargain, anyway.

Or get POV (shareware) for PC. It's much more limited with textures
and text, but less money. Not for my taste, though.


#21 of 119 by mwarner on Mon Jul 10 15:12:53 1995:

I bought a Microtek ScanMaker II, which works very nicely.  I'm not really
interested in creating 1G images, but that's what it would take to make a
decent size, high resolution color image.  Luckily, images scanned at
fairly low resolution look great, too.  (a newspaper image is only about
80 dpi), and besides who needs millions of colors when your hardware is
blind to all that resolution.  For compiling a fairly good collection of
images from photos and documents this seems to work well.  You *can*
create nice looking color images for <100K, with trial and error.  I
usually make a "rich" scan (2-4 meg) and then knock it down, paring away a
little quality and a lot of memory.  Some types of pictures look great in
16 colors, 16 grey scale or even straight B or W.

  One buggy problem I've been getting seems to be related to memory
allocation on my Mac.  I have enough memory to run the software (Color It!
with a Photoshop plug-in for scanning)  but I have started to get "not enough
memory" errors *on other programs* after using the scanner.  This has me
puzzled because a check of available memory shows I am using only a
fraction and should not be running into any walls.   I have virtual memory
turned of, but am using Ram Doubler.  The problem persists whether RD is
switched on or off.  Eventually the problem clears up, but I have not been
able to keep it from returning or decide exactly what the cause is.


#22 of 119 by ajax on Mon Jul 10 15:41:57 1995:

  Have you tried removing RAM Doubler from your system folder, as
opposed to clicking the "off button?"  Weird problem...I think every
Mac has a couple unique oddities like that :).
 
  I heard that the human eye can only consciously discern about 4,000
different colors.  I'm not sure I believe that, but if true, it seems
the "millions of colors" used in high quality images is either wasted
on us or maybe works at a subconscious level.  If my computer could
display millions of colors, I'd be curious to compare an image using
that with an image converted to 4,000 colors, to see if I could detect
any difference.  I can definitely see the loss when I go to 256 colors.


#23 of 119 by mwarner on Mon Jul 10 17:37:06 1995:

I don't think it's related to the doubler, but I might try physically
removing it.  I can debug the bug by using the "get info" menu and
tweaking up the preferred memory on the balking program.  The strange
thing is that it's a one time solution and not a fix.  The program runs
fine even if I turn around and lower the memory allocation... until the
next "blockage" occurs.  And leaving an increased memory allocation won't
prevent the bug from its return. 




#24 of 119 by rcurl on Mon Jul 10 21:50:16 1995:

Does the problem persist after rebooting (as we discussed f2f)? RAM
fragmentation could persist after running several applications
simultaneously.


#25 of 119 by mwarner on Tue Jul 11 14:58:31 1995:

Yes, and, as I share use of this computer, see:  Hot Water.


#26 of 119 by rcurl on Tue Jul 11 16:59:41 1995:

Does Color It! leave an INIT (extension) active upon reboot? If so,
are you have an extension (or control panel) conflict? If there is
one associated with Color It!, stick it in the disabled extensions
folder, and check for the problem after reboot. 


#27 of 119 by mju on Sat Jul 15 03:11:27 1995:

Re #22: The human eye actually has two color thresholds; we are
much more sensitive to differences in color when the two colors
are next to each other than we are when they're separated.
So it may be that 24-bit color is useful for providing smooth
color gradients, even though we might not be able to individually
discern all of the colors used.


#28 of 119 by rcurl on Sun Jul 23 22:32:04 1995:

I haven't studied this in any depth but.... the human eye can only
respond to three colors as it has only three visual pigments. All the
other colors are admixtures of responses of the three pigments. 256 -
or 4000 - colors are therefore differences in levels of intensities of
the admixtures, and in fact would depend on light intensity as well
as the source intensities (since the pigment responses are nonlinear). 
In view of all this, I am not sure that one can categorically state
that the eye is "more sensitive to difference...when (they) are next to
each other". First, I am not sure what is meant by "more sensitive", and
secondly, one has to define "next to each other" and "separated", and
thirdly, I think it would depend on the placement of the pairs in the
tristimulus space.


#29 of 119 by mju on Mon Jul 24 03:54:06 1995:

It is probably not correct to say that "the eye" is more sensitive
to such combinations, since all the eye does (as you stated) is
respond to stimulus as a combination of red, blue, and green.  However,
the perceptual centers of the brain may well be better at discerning
that color A is "different" from color B when they are next to each other,
versus on opposite sides of the visual field and separated by a third color.
This is similar, I would think, to our perception being more sensitive to any
kind of difference when we can compare the cases side-by-side.


#30 of 119 by mcpoz on Sat Oct 7 16:44:19 1995:

Well, soon I will be launched into the computer age in photography.  My wife
is getting a super mac (#7500) with all the bells and whistles.  She will also
get Adobe Photoshop.  I was starting to save for a PC capable of doing photo
work, but now, I can divert these savings toward the other stuff, like
printers, negative scanners, & possibly some sort of digital camera.  Anyone
have any experience/suggestions where they would start?

I have done black & white darkroom photo printing for 30 years plus, and I
thought I would start out with b&w digital if it is significantly lower in
initial cost. I would also appreciate any advice you may have here.


#31 of 119 by srw on Sun Oct 8 00:43:03 1995:

I just used my new APple Quicktake 150 Digital Camera (and some help from
photoshop) to produce the images on the GrexWalk photo homepage.
http://www.hvcn.org/info/grexwalk/grexwalk.html

I am a little disappointed that the jpeg compression it uses is so harsh.
It compresses 640x480x24 images to 68k in high res mode, and 34k in standard
mode. I have found standard mode just about useless. HIghres mode gives you
16 marginally usable pictures per camera load (1MB Flash EEProm).

It did a fine job for the Grexwalk, because I shrank the pictures to fit the
web site. If I left them at 640x480, I would have remained disappointed.
I wish it had a third resolution which was 132k and could only fit 8
pictures. It might be very useful.

If I wanted larger pix, i would use 35mm and photo CD. It's much higher
quality. I just hate waiting 3 weeks for developing. That's what it took.


#32 of 119 by mcpoz on Sun Oct 8 00:58:45 1995:

I haven't looked into it yet, but I thought I remembered ads for negative
scanners which were in the $500 range.  Do you know anything about these? 
As I recall, the scanner would do negatives or slides and they were supposedly
for high resolution scans.


#33 of 119 by srw on Sun Oct 8 04:13:11 1995:

I have never used a negative scanner. I would expect the quality of the
resulting scan toobe excellent in comparison to the low-end digital camera,
but the price you quoted seems way low to me.

The only transparency scanner in the mail order catalog I happen to be 
looking at is the Nikon LS-1000 (2700 dpi, slide feeder optional) at $2000

THere are a lot of flatbed scanners near that price range, though.
Some have transparency adapters, but because they are flatbed scanners,
they don't offer the resolution you really want for scanning negatives or 
slides.

Relisys has a REL2412/T Mac single pass 24 bit scanner with 1200x300 optical
resolution, It's a SCS devices which includes the transparency adapter
and runs $770, but 1200x300 is unacceptable resolution for negatives or slides.

That's the low end. At the high end of the Relisys line is the
RELI 9624/P Mac 2400x600 optical - $1500. Much better resolution, but still
not up to par for slides, and already 3/4 the price of that Nikon.

Admittedly these do come with bundled software, some of which is valuable.


#34 of 119 by mcpoz on Sun Oct 8 12:22:21 1995:

Well, it may be later rather than sooner when I buy the scanner, but I think
I'll keep it on my shoping list.  Meanwhile, I will try to locate that $500
scanner I "think" I saw.  If I locate it, I may post another note about it.


#35 of 119 by ajax on Mon Oct 9 00:06:53 1995:

Sounds way low to me too, but ya never know.  Choice of output device 
depends a lot on what you want to do with the output.  For proofs or
newsletters or casual use, most postscript laser printers are good
(especially some of the 600dpi ones with "resolution enhancement
technology" or whatever the particular manufacturer calls it).  For
high quality output (things you'd frame, or want a poster of), I'd
do some reading on the options available, and get the same picture
printed using a number of methods through service bureaus.  I wouldn't
pick any particular high-end device until you've compared the output
and understand the tradeoffs.  My dad did a number of color prints
using Iris prints (I think that was the one) a couple years ago, and
already the colors are changing.  Not desirable unless you're Andy
Warhol!


#36 of 119 by mcpoz on Mon Oct 9 01:27:12 1995:

Thanks for the insight & the precaution.  It'll be a while, but I will post
when I get the funds close and the research to my satisfaction.


#37 of 119 by mcpoz on Wed Oct 16 02:29:36 1996:

Does anyone have any knowledge about the "replacement" of standard emulsion
photography by digital photography?  Is there a future for standard
photography in any of the following fields:     
        (1) amateur     
        (2) Custom (weddings, etc)      
        (3) Special Corporate

I am interested if anyone knows if continued growth is predicted for the
standard wet chemistry/emulsion photography industry.

Thanks


#38 of 119 by olddraco on Fri Jun 19 21:03:43 1998:

Hrm...looks like the conference is dead? At least this post is...


#39 of 119 by arthurp on Sun Jun 21 19:56:32 1998:

Well, this item...
My experience is that when viewing a test pattern 32K colors leaves
significant steppings between colors.  64K colors is the same but with better
smoothness in the green range.  16M colors is smooth.  This would make 256
colors fine for 'office' work, but 16M would be needed for any kind of
graphics work at a fairly serious level.  Thousands of colors is probably
enough for the graphics that most regular people use.
YMMV.


#40 of 119 by vrondi on Fri Jan 1 21:55:18 1999:

About digital photography software:
Photoshop isn't the only thing out there by a long shot.
  My personal favourite is Paint SHop Pro by Jasc software.
www.jasc.com   You can download it from the website, and it does many many
of the things Photoshop does, and can use Adobe Photoshop plug-in filters.
IF you are a beginner with this, you can download the shareware version, then
by the time it runs out, you have some idea of a few of the things that can
be done with digital photography.  Paint Shop PRo will run TWAIN
devices(scanners, cameras, etc...) 


#41 of 119 by gregb on Sat Feb 27 22:26:31 1999:

I assume your referring to PSP 5.0.  I haven't tried that one yet.  I'm 
still using 4.0, which does pretty much what I need to do.  For more 
complex tasks, I use Painter 5, which does layering like Photoshop, 
plus some really nifty effects and filters.


#42 of 119 by jshafer on Mon Jun 21 07:16:14 1999:

And then of course there is the GIMP...  Haven't played with it yet,
but toking said he was impressed by it...  (And I guess there's a 
Windows port...)


#43 of 119 by gregb on Fri Jun 25 04:36:37 1999:

Re. 42:  I've played with Gimp.  Not bad.  I'm sure it'll be even more 
cool after I read the docs for it. 8-)


#44 of 119 by gull on Thu Jul 8 01:01:19 1999:

So far I haven't found anything Photoshop can do that GIMP can't.  Well,
except talk to my scanner, but that's a lost cause.  It's a parallel-port
one.


#45 of 119 by dang on Sun Jul 25 23:39:54 1999:

Linux will talk to parallel-port hardware as of recent releases.  I have
my parallel port zip disk running, and I saw when I was setting it up
that there are drivers for some parallel port scanners.  I'm not sure if
yours is supported, but you should check a recent 2.3 kernel.


#46 of 119 by jshafer on Mon Jul 26 06:23:03 1999:

This response has been erased.



#47 of 119 by jshafer on Mon Jul 26 06:37:03 1999:

photography 5 <==> micro 133 <==> graphics 5


#48 of 119 by rickyb on Wed Jul 28 23:45:13 1999:

Well I goofed... the 1st roll of photos I had done on a photo-CD I had done
as high-res scans (about 1MB each).  I'm happy with the results of the camera,
but can't send them to friends via e-mail (Yahoo, Juno, HotMail have size
restrictions on attachments).  Any suggestions on shrinking them digitally,
or do I have to have them re-scanned to do that?  (I don't want to simply
compress them with pkzip, etc, as I will eventually use that to group several
shots into a small enough file to send via e-mail).



#49 of 119 by gull on Thu Jul 29 00:28:29 1999:

Any image editing program will let you resize the image to a smaller
resolution, and save it.  (JPG is usually the best for distributing
photos...it's lossy compression, though, so you trade off quality for size.)
Adobe Photoshop will work if you have it.  If you don't have anything like
that, there are free tools you can use.  I'd suggest LView Pro.  Or Paint
Shop Pro, if you want something also capable of doing retouching.


#50 of 119 by drewmike on Thu Jul 29 02:12:30 1999:

Fireworks gives you a preview box, so you can see exactly what your graphic
will look like at a given compression, and adjust so you have a decent balance
between quality and image size.
 
Been a while since I've thought of a one-meg photo as "high-res"...


#51 of 119 by dang on Fri Jul 30 03:26:24 1999:

What do you think of as "high-res"?  I very rarely run across a photo
that is larger than 1 Meg.  Most are between 50 and 300 K.


#52 of 119 by gull on Fri Jul 30 04:06:27 1999:

It'd depend on the file format.  300K is a pretty big JPEG, but small for,
say, an uncompressed TIFF, which can easily be many megabytes.


#53 of 119 by drewmike on Fri Jul 30 04:39:32 1999:

I've done 17"x22" posters that had a 200 dpi full bleed CMYK TIFF in the
background. THAT's high res.
 
And compared to video, even that's nothing. The MacOS has a file size
limitation of 1.9 gigs, and usually at least once a week I have to work around
that.


#54 of 119 by dang on Fri Jul 30 16:34:15 1999:

True, I usually deal in Jpegs.  Hi-res jpegs are pleanty high for
anything I would need.  I agree that video is huge.  I'm wondering when
we'll get decent video compression, ala mp3 for audio?


#55 of 119 by gull on Fri Jul 30 17:38:49 1999:

MPEG2 is a decent video compression, though like MPEG Layer 3 (the MP3 audio
standard) there are noticable artifacts.  DVDs use the MPEG2 standard, I
think.

I use TIFF for anything I plan on editing in the future, since JPEG
compression worsens the picture quality every time you load and resave it. 
LZH-compressed TIFFs are large but not awkwardly so.


#56 of 119 by drewmike on Fri Jul 30 17:48:21 1999:

Note that MP3s are actually MPEG 1s (layer 3 audio). While there is such a
standard as MPEG 3, there's much more interest in the developing standard of
MPEG 4.


#57 of 119 by rickyb on Thu Aug 5 16:56:27 1999:

Thanks for the advice on the free utilities (I don't have photoshop).  I do
have one I got a long time ago called PhotoMorph which lets me re-touch and
make slide-shows, etc.  I'll have to go back and see if it'll let me re-size
to smaller resolution.

Question:  If I were to want a frame or two done at poster size (anywhere from
14"x18", 18"x24" or even a bit larger) would you recommend using the 35mm neg
or the 1MB jpeg?  Many shops take the jpegs over the phone (or on disk) and
blow them up very large.  _But_, is this either cheaper, or adequate
resolution for an "artsy" photo (ie; not commercial art, but to frame and
display on the wall).



#58 of 119 by drewmike on Thu Aug 5 18:27:00 1999:

If you're going to print a JPEG at that size, and it's only one meg, it's
probably going to be at a very low resolution, or compressed to the point of
severe visible artifacting. Last time I had photo-quality prints made, it was
an 8 x 10 and it ended up being at least 25 megs.


#59 of 119 by rickyb on Sat Aug 14 18:04:00 1999:

I'll be back to this item soon...just returned from a camping trip at
Ludington and had 2 rolls of 35mm and one (35mm) roll from an "underwater"
camera.  The river which flows into Lake Michigan was very clear, and I used
snorkle/fins to view the fish and underwater vegitation.  actually, i even
did some fishing with a 3' line tied to my finger and a small minnow on a
hook.  the fish came right up to me and took the bait...it was cool!

I took several 'close-up' photos with the waterproof camera of sunfish, trout,
bass, crawfish and underwater landscapes.  I'm having them scanned ("low-res")
and they should be ready later today.  I'll let you know how this turn
out...I'm really excited to see if this worked.



#60 of 119 by eprom on Mon Oct 25 02:55:07 1999:

re#44 have you tried using a program called sane (or xsane)?

anywho...Im using GIMP 1.0 and it works fine for what I do.
my favorite filter is "old photo" it makes your picture look old.
another good filter is the "glass lens" it gives a wierd halo look.

I usually use GIMP to crop and change the contrast/brightness of pics.


#61 of 119 by gull on Mon Oct 25 03:10:30 1999:

Yeah.  My old scanner wasn't supported.  My new one *is* supported by SANE,
but I can't get it to work; attempting to scan locks the machine solid.


#62 of 119 by raven on Tue Nov 2 20:55:00 1999:

I'm also impressed by the GIMP.  It has nice layering and so far it seems
to have just about everything phtoshop has except a smaller selection of
filters (and I assume Linux hackers are working on those).  I haven't
been able to find drivers for my Artec parallel port scanner, though. :-(


#63 of 119 by rcurl on Thu Jun 1 01:47:59 2000:

How would one remove the shadow around one side of an image from a
digital camera with on-camera flash? Say, in photoshop? It would
be OK to replace it with the adjacent background. 


#64 of 119 by eprom on Sun Jun 4 16:45:24 2000:

edit it pixel-by-pixel :)


#65 of 119 by rcurl on Mon Jun 5 04:39:02 2000:

I thought there was a way to select an area of particular characteristic.
Since the shadow is quite distinct from either the background or
objects, isn't it selectable in Photoshop? 

editing it pix-by-pix is 8^{, not :)


#66 of 119 by n8nxf on Mon Jun 5 11:35:26 2000:

You can select areas in Paint and copy and paste them.  It would be nice
if you could paste with a spray gun so that you can feather the paste in
and avoit the crisp, patchwork, lines.  I'm sure you can do that with a
better paint package.



#67 of 119 by rcurl on Mon Jun 5 14:55:10 2000:

Paint? More info, please.


#68 of 119 by n8nxf on Tue Jun 6 12:23:51 2000:

Microsoft Paint.  It comes with Windows 3.1, 95, 98 and others for all I
know.  I don't have access to a Mac at the moment, so I don't know if you
can do it under MacPaint, though I'm pretty sure you can.


#69 of 119 by rcurl on Tue Jun 6 17:25:02 2000:

It's not bundled with the Mac OS now - wasn't it once? Is there a
free/share-ware version of something similar?


#70 of 119 by n8nxf on Thu Jun 8 11:52:03 2000:

No, it never was.  I've got lots of paint type applications that came on
Macs the I got at PD for $1, bought at auctions or were given to me.  I
don't know if any of them would run on your iMac.  I would not be
surprised if there was a freeware package out there.


#71 of 119 by eprom on Sat Jul 8 23:42:50 2000:

hmmm..ever notice that the good graphic art magizines
such as "Digit" and "Computer arts" are from the UK?

[drift on]
I hate how the magizines from the UK are not the
standard american size.
[drift off]



#72 of 119 by scott on Sun Jul 30 21:54:23 2000:

I'm now using the "GIMP" (Gnu Image Manipulation Program) that comes with most
Linux distributions.  It's pretty cool, loaded with features I'll never get
around to figuring out.  And the price is certainly right, being essentially
free.  I've read that it's not quite Photoshop, but for what I do (mostly
cropping, resizing, and color adjustment) it's quite complete.


#73 of 119 by gull on Mon Jul 31 20:05:24 2000:

It does beat Photoshop in a few areas.  One of my favorites is that it
remembers the last directory you opened from and saved to -- SEPERATELY. 
Photoshop assumes you're reading and saving from the same directory, which
isn't usually the case if I'm working on a batch of images.  The result is a
lot of extra pointing and clicking, in Photoshop.

GIMP does have some glaring user interface flaws.  For example, it has two
File menus.  How to get to the more useful one is not immediately obvious. 
(You have to right-click in the image you're editing.)


#74 of 119 by rcurl on Thu Oct 19 00:52:04 2000:

I'd like to "overlay" one map image upon another, make the top one
semi-transparent, and then adjust the top one's orientation and size so
that they register correctly.  How do I do this in Photoshop 4? 

(The photoshop built-in help is nearly useless, assuming you already
know the meaning of terms like overlay, merge, channel, etc, while not
defining them anywhere. 8^P


#75 of 119 by scott on Thu Oct 19 01:08:20 2000:

Buy a cheap book.  There's plenty of them out there, and for the staggering
array of features in these sorts of programs you pretty much have to have some
kind of friendly guide.


#76 of 119 by eprom on Fri Oct 20 00:01:13 2000:

Its all about layering.

what i'd do (im using Photoshop 5.5 so it may be differant) is open one of
the images; then create a new layer, so you'll have the original on bottom
and a new blank layer on top.

now open your second image; re-orient and adjust the size of it. now copy that
to the blank layer and adjust the opacity slider scale.

I used my digital camera to take a picture of a chair; then I set the camera
to take a delayed photo, so I could hurry up and sit in the chair. I used
photoshop to aligned the photos and changed the opacity so it looks like a
ghost of me is sitting in the chair.

another cool layering thing is, taking a photo and make identical layers;
de-saturate one of the layers (it will be B&W now) and move it to the bottom.
now take the color layer on top and cut out what you don't want...so you'll
end up having a picture that looks like that girl with the red coat in the
movie "shindler's list".

(see http://members.fcc.net/eprom/radiocity.jpg)


#77 of 119 by rcurl on Fri Oct 20 04:52:00 2000:

I figured it out, and came up with something very close to what you
describe. I opened both images (each was layer 0), selected an area
of one and did a COPY, and then pasted that on top of the other. Viola -
two layers. Then it was just a matter of adjusting opacity. Those magic
words of select, copy, and paste, are not in the HELP directory, though
what they do say makes sense once you know how to do it. (I produced
a township map with an semi-transparent overlay of the bedrock geology.)


#78 of 119 by rcurl on Sun Oct 29 06:27:17 2000:

A PhotoShop question. I take digital pictures, which download in jpeg
format. I can open these in Photoshop (4.0.1), and there are a variety of
format options for saving them (over a dozen). However if I select a
portion of the image and paste it to a new image file, "save as"  offers
only one option - Photoshop format - for saving. I'd like to save the
selection as another jpeg file. How do I do that? 



#79 of 119 by scott on Sun Oct 29 13:17:07 2000:

As a generic programming thing, I'd guess that it's still saving undo info
or something.  Maybe there's an option to finalize or collapse layers or
something?


#80 of 119 by rcurl on Sun Oct 29 18:14:20 2000:

In PhotoShop format it is a 1+ MB image; in JPEG, about 70 KB. The
"save as" menu always shows the full list of options, but some are
"shadowed" and unavailable. All are, except, PhotoShop after I copy
and paste. I did try saving the original image as a TIFF file, then
selecting from that, copying and pasting to a new TIFF file, but then
could not save that as JPEG. So, how do I do what scott says in #79
("finalize" or "collapse"), before I collapse?


#81 of 119 by gull on Mon Oct 30 00:08:48 2000:

The problem is that when you paste something in, it's temporarily put in a
different layer, and none of the formats except Photoshop support multiple
layers.  I think clicking outside the selected area with the selection tool
will flatten the image.  You can also choose 'Flatten Image' from the
'Layer' menu.

You are aware that every time you load and resave an image in JPEG format,
you loose a little more sharpness and gain more artifacts, right?


#82 of 119 by rcurl on Mon Oct 30 04:06:34 2000:

I found it! (i.e., Eureka!). If you use the "Save a Copy" option instead
of "Save As", the selection may be saved in most of the available
formats. (The problem in discovering these things is you have to try
a lot of stuff until one works.). Of course, the online help doesn't
offer any help even for "save". I suppose I could run the CD-ROM tutorial...


#83 of 119 by scott on Mon Oct 30 14:38:20 2000:

I'll say it again:  For programs like these, even I'll buy a book on how to
use it.  The $20-$30 is easily justified if it explains only a couple of the
things that you are assumed to know (but don't).


#84 of 119 by rcurl on Mon Oct 30 18:41:26 2000:

OK OK - I'll look for "Photoshop for Dummies". I admit that I have found
the manuals for WORD and EXCEL useful on occasions. But, jeezle-peezle, I
would think that someone here knows the simple answers to simple questions
about these very widely used applications. People here often (usually)
ask questions about computers and their use (even about Grex), and they
are not usually told to go read a book... 8^}





#85 of 119 by scott on Mon Oct 30 20:40:55 2000:

Well, yes, we often do know the answer and aren't shy about sharing.

But still, it's pretty nice to have stuff explained without having to wait
for an answer.  

(I'm *not* trying to discourage questions, just trying to give a more useful
long-term answer)


#86 of 119 by rcurl on Mon Oct 30 22:31:09 2000:

But I wanted the answer *today* - like, right now....  8^} So I go to my
more knowledgeable colleagues and ask. But they don't know. I have
suspected for some time, though, that the real problem in getting answers
here is a decrease in participation. I initially found it hard to imagine
that I was the only person reading this conference that was working with
PhotoShop while not being the most knowledgable, but I have since become
to suspect that that is the case. 



#87 of 119 by scott on Tue Oct 31 13:51:54 2000:

This sort of thing happened in the Music conference not too long ago. 
Somebody came in talking about what a great fan they were of xxx, and wanting
to hear from the experts.  Well, that person *was* the expert on xxx, at least
on Grex.  

If you wanted the answer *now*, why did you put it on Grex?  ;)


#88 of 119 by rcurl on Tue Oct 31 20:04:06 2000:

Discouraging question... Where can one put that and similar questions
and get answers *now* (or, very soon...)? The vendor sure is not a source
of *now* answers.


#89 of 119 by scott on Tue Oct 31 22:58:14 2000:

Some are good, some suck in that regard.  Powersoft (before it got sucked into
Sybase) used to send out a CD with all their phone help and online
(Compuserve, in those days) Q&A in a searchable format.  That saved a *lot*
of trouble for me.


#90 of 119 by n8nxf on Wed Nov 1 13:23:23 2000:

Is usenet still around?  I have found good information there as well as by
just searching the internet in general.  After trying to diagnose a minor
problem with my car, one day I came across the answer on Joe's Garage Q & A
web page.  However, searching for information on the web is often time
consuming and frustrating.  It would be nice if there were a better way to
find the information that one is looking for.


#91 of 119 by scott on Wed Nov 1 14:52:04 2000:

Usenet is still there, but it's covered with trolls and scams.  But if you
go deep enough you can still find some useful info.


#92 of 119 by rcurl on Wed Nov 1 16:55:31 2000:

The usenet lode is mostly panned out - hard to find nuggets among all
the gangue. 


#93 of 119 by gull on Wed Nov 1 20:37:02 2000:

I've had good luck doing web searches for specific problems, though it
depends on whether the question is one that you can easily form a query for. 
If there's a USENET group that seems appropriate, their FAQ can be a gold
mine.  And while reading a group regularly means digging through all the
trolls and spam (not as bad on some groups as on others), if you pose a
question you only have to read your own thread.  then it's not so bad.


#94 of 119 by flem on Wed Nov 1 21:26:41 2000:

I imagine if you look hard enough, you can probably find a mailing list on
the subject, which will probably be somewhat better than usenet, at least.


#95 of 119 by cmcgee on Fri Jun 3 13:43:52 2005:

I'm trying to understand the difference between the resolution and compression
settings on my new Canon A95.  Why would you want to take a high-resolution
photo and then use high compression? 


#96 of 119 by eprom on Fri Jun 3 15:51:22 2005:

oooh...thats a really nice camera.

Compression is the compactness of the file size, the downside is 
that your images won't look as sharp, especially if you want to 
print them.

Resolution is the actual pixel count, higher resolution, means the 
images will be physically larger. Generally the higher the resolution,
the bigger the file size will be.

from the highest quality to lowest would probably go in this order:

1) High resolution and low compression
2) High resolution and high compression
3) Low resolution and low compression
4) Low resolution and high compression


#97 of 119 by scott on Fri Jun 3 16:56:40 2005:

Would you rather have a programmer deciding you shouldn't have a choice that
doesn't make sense?  :)


#98 of 119 by rcurl on Fri Jun 3 17:04:45 2005:

Compression does not necessarily compromise resolution. Compression per-se
just makes a file smaller. Uncompress it, and you recover the resolution.  
Resolution refers to the display; compression to the file size.  Some
compression techniques do compromise the recovery of resolution, so one
has to know the details of the compression routine. PDF applied to
word-processing documents is, for example, a file compression routine,
which can greatly reduce the file size without sacrificing "resolution"
when the file is gain expanded.



#99 of 119 by gull on Fri Jun 3 17:52:34 2005:

Every digital camera I've used has used JPEG compression, which is
lossy, so you lose sharpness at higher compression settings.  Some
cameras offer an uncompressed image format as an option.  Canon cameras,
for example, usually can store images in their proprietary RAW format,
which is not lossy.


#100 of 119 by eprom on Fri Jun 3 20:19:53 2005:

Ok...I was assuming it was jpg compression (since that what my Canon G2 
does) which does cause some noticable artifacting. 

There is LZW compression for tiff, which is loseless, but I don't know
any cameras currently on the market which do tiff anymore.


#101 of 119 by cmcgee on Fri Jun 3 20:56:25 2005:

re 98:  Thanks, that I understand.
re 99:  As far as I can tell, this camera will not let me download anything
in RAW format.
re 100:  Yes, I think it automatically uses jpeg compression when downloading
to my computer.


#102 of 119 by keesan on Sat Jun 4 03:41:30 2005:

It uses the jpeg compression to create the files that it later downloads to
your computer.


#103 of 119 by rcurl on Sat Jun 4 05:56:49 2005:

There is more than you want to know about JPEG at
http://www.faqs.org/faqs/jpeg-faq/part1/


#104 of 119 by rcurl on Mon Aug 21 16:59:44 2006:

I have two images of a group of people sitting on a low stone wall. In 
one, A took the image and B sat at the left end of the group. In the 
other, B took the image with a different camera and A sat at the left end 
of the group. I would like to cut B out of the first image and paste her 
onto the left end of the group in the other image, putting "everyone in 
the picture".

The main problem I have is that one image is 478 pixels (ps) by 640 ps, 
and the other is 2304 ps by 1728 ps, with both at 72 ps/inch but the 
scenes are essentially identical.

How do I scale (in Photoshop) the cut out image of a person to agree with 
the scale of the other image?

I also need to select the image to cut out, for which I presume the lasso 
tool is needed. Advice on the use of this tool in its three forms 
(ordinary or polygonal or magnet) would be useful.


#105 of 119 by cmcgee on Mon Aug 21 20:12:21 2006:

Actually, for something like this, I use the brush tool, and go down to the
pixel level to select.  You can toggle the mask back and forth easily if you
make a mistake, and use the same brush tool to deselect the error.

When I'm pasting a selection in, I create a transparent area at the
desitnation, and then resize the pasted pixels using a separate layer until
I get the right look.  

You can also use the Edit, Paste Into Selection command.

When making the selection, use some feathering to help make the seam between
the two pieces less visible.  


#106 of 119 by rcurl on Mon Aug 21 22:38:24 2006:

Since I've never attempted this and am unfamiliar with the tools you
suggest....I will study what you suggest and see how it goes. 


#107 of 119 by rcurl on Mon Sep 4 20:08:33 2006:

I looked at my image at the pixel level, and the brush tool, and fiddled 
with it a bit, but don't see how I can use that to cut out the part of the 
image I want to paste into another picture. Could you enlarge upon how to 
do this Colleen, "for dummiees"?


#108 of 119 by cmcgee on Tue Sep 5 15:34:04 2006:

Rane, before you can "cut out" anything in Photoshop you have to "select" it.
Working at the pixel level to select the parts of the image you want to
transfer helps get the shadows and other subtle visual cues transfered to the
receiving picture.  

I use the mask and brush tools to refine my selections after using the lasso
tool to make a rough selection.  

The premise of the selection brush is simple:  you paint in the image to
create a selection.  By toggling the brush you can add or subtract areas of
the selection.  I usually work back and forth from pixel level to full frame,
just to make sure what I'm choosing make sense in the total picture.

I use the same lasso-brush tools combination to erase the spot on the
receiving picture.  Then I use the Edit-->Past Into Selection  command to move
the sending picture selection into the receiving picture selection.  


#109 of 119 by rcurl on Tue Sep 5 16:31:29 2006:

I have a lot to learn....


#110 of 119 by gull on Wed Sep 6 20:46:38 2006:

Another way to do it, which is easier in some versions of Photoshop, is 
to paint the parts you don't want some unusual, uniform color, then do 
a select by color and invert the selection.  This is analogous to doing 
a "green screen" effect in TV.

In either case the tedious part is masking out what you want from what 
you don't.  Unless you're really lucky about the color of your subject 
and the color of your background, you're going to have to do it by 
hand.


#111 of 119 by rcurl on Wed Sep 6 21:47:50 2006:

I don't yet understand how to use a method to mask out what I don't want.
Then there is the matter of copying and pasting the remainder over what is
in the space it will go, so only the image I want covers its new location.
(I don't yet understand "layers" and a bunch of the stuff that has been
mentioned here - maybe I should take a course.....)


#112 of 119 by gull on Thu Sep 7 00:40:04 2006:

There are some Photoshop tutorials on line, but I don't have any good 
ones bookmarked right now.


#113 of 119 by arthurp on Sun Oct 1 05:49:59 2006:

In the making of Star Wars the foundations for the modern techniques
were laid.

Film an object in front of a uniform green color.
Film a background scene.
Use (now) computer help to eliminate the green to leave just the object
on the film where the rest of the frame will be clear.
Use also the computer to invert that step to make a clear object shaped
area in the frame for the background scene.
Then stack these films on top of each other to build the composite
image.

You have composite images already so you have to use the tool to cut
around the person to move, or paint the rest of that image to simulate
the filming in front of green screen.  Then the software should allow
you to paste on top of the other image to composite.  But perhaps you
will need also to use some method equivalent to makeing the clear object
shaped hole in the destination background.


#114 of 119 by denise on Tue Sep 4 00:31:27 2007:

I just started using iPhoto and am using it to organize and touch up my
photos...  I used  to think that digital photography and touching up the
photos was kind of cheating--the end  result not being what you actually
took.  And to some extent, I still feel that way.  However, I'm having
fun playing around with the program and seeing what I can come  with.

Is anyone else using iPhoto?  And what kind of cool stuff are you doing
with it?  


#115 of 119 by gull on Fri Sep 7 20:15:59 2007:

If it's cheating, then all the darkroom techniques that photographers
use (dodging, burning, etc.) are also cheating...


#116 of 119 by denise on Sun Sep 9 15:10:34 2007:

Hmm, with having had limited access to darkrooms, I never really did
much of that. But I  see your point.  :-)


#117 of 119 by cmcgee on Thu Oct 11 15:34:56 2007:

Here's a nice website with a shortcourse on using your digital camera.  

Going through the pages will take you from helpless to decent amateur.  


#118 of 119 by cmcgee on Thu Oct 11 15:35:05 2007:

http://www.shortcourses.com/use/

Duh!


#119 of 119 by denise on Fri Oct 12 11:38:29 2007:

Thanks, Colleen. I just scanned the headings/content and will go back to
 read it soon.


You have several choices: