Well, at last we did it. We received and installed our first scanner (Minolta Dimage Scan Elite). Installation went smoothly and here is our first scan. What are the steps (and in which order) you go through. We used Photoshop 5 LE to reduce the size of the file, otherwise we did no adjustment at all. Seeing such a picture, what are the adjustments that come to your minds. Do you know a simple reference that could be useful for beginners? Thanks

By the way, this picture was shot by my husband using an old Zeiss COntaflex 50 mm ... and our Leica.

-- Angelique (abischop@earthlink.net), October 04, 2001

Answers

It wouldd have been too beautiful if we had succeeded after ourr first trial... What s wrong?

-- Angelique (abischop@earthlink.net), October 04, 2001.

Another try ...

-- Ang. (abischop@earthlink.net), October 04, 2001.

HELP !

-- Ang. (abischop@earthlink.net), October 04, 2001.

go to administration at the bottom of the question page there is a complet explanation.

Good luck to both.

-- r watson (al1231234@hotmail.com), October 04, 2001.

Angelique:

You missed one step: When you are at Photonet at your picture and want to copy the web address, FIRST click and hold on the picture until you get a little drop-down menu and then select "Open this image".

THEN: use THAT web address to create your link. It will be very similar to the Web address that you used that didn't work, but will contain the phrase /photo-display/ somewhere in the middle. This is the critical piece to make the picture come up.

(I forgot how to do it, too, for a few months. Then I rechecked Tony's instructions and rediscovered this critical trick.)

Here's the link to Tony's instructions: SPECIFIC TO PHOTONET IMAGES

-- Andy Piper (apidens@denver.infi.net), October 04, 2001.

Ooops: the critical phrase will be /image-display/

(For other folks: you can get to Angelique's picture in the meantime by clicking and holding on her broken image icon and selecting "open this image" from the drop-down menu yourself).

-- Andy Piper (apidens@denver.infi.net), October 04, 2001.

Thanks for your helps! Hope this worrk now...

-- Angelique (abischop@earthlink.net), October 04, 2001.

Let s try bigger ... sorry for the mess...

-- Angelique (abischop@eartthlink.net), October 04, 2001.

Now, getting back to your questions: there are tons of Photoshop books available at bookstores and computer stores. I learned most of my Photoshop hands-on, anyway, so my knowledge and opinion on the books is limited. (I supervised the installation of a Macintosh-based image pre-press (scanning) system at a small newspaper 8 years ago and have done my own personal scans since 1994).

My steps (for B&W) are as follows: (and if any of this seems like gobble-de-gook, please feel free to ask for clarifications)

FIRST: save the image so that you'll have a backup in case something horrible happens (like a power outage or a really bad mistake).

THEN: I do most exposure/contrast manipulations BEFORE reducing the file size - that way if the histogram (the graph of light-to-dark pixel distribution) starts to get spikey, the resampling process cleans it up substantially by interpolating the missing grays.

My first step is to set black and white using either the Levels or Curves dialogue boxes. You want the range of tones to just BARELY reach pure black and pure white, at least to begin with - if too many pixels go white or black you are losing detail, and if none go black or white you aren't using the full dynamic range available and the image will look flat.

My second step is to adjust highlight and shadow separation using the Curves dialogue box, by twisting the straight-line curve up or down (or both). I often end up with an S-shaped curve to open up shadows and separate the highlights.

Then, in order, I:

Do any other tonal range adjustments (localized dodging and burning)

Use the Despeckle Filter (ONLY ON SOME IMAGES) if there is too much grain. Despeckle will sometimes soften some fine detail, so check the image afterwards and UNDO if too much detail fuzzes away.

Reduce the file size (if needed/wanted)

Use the UNSHARP MASK filter to restore sharpness lost in the scanning/ resampling process. For high-resolution (150 dpi or above) my standard setting for Unsharp mask are 200%, .5 pixels, and a 4 levels threshold. For lower resolution images (e.g. for the web - 72 ppi) I use 135%, .3 pixels, 4 levels.

Use the DUST & SCRATCHES filter on areas of the pictures that DON'T contain sharp detail (e.g. skies, out-of-focus areas) at a setting of 5 pixels and about 20 levels (+/-) to eliminate the bulk of dust specks and scratches. (With a picture as detailed as your fire escape I'd skip this step...Photoshop tends to interpret all those fine lines and intersections as defects, and tries to 'spot' them all away.)

Use the CLONING tool (looks like a rubber stamp) to retouch dust and defects in detailed parts of the picture. Magnify the picture up to 1:1 OR LARGER to do this and use the soft and hard "brushes" as needed depending on how much fine detail you have to work between/around.

Done. and Save!

Side note: There are at least two ways to dodge and burn areas. Photoshop has 'dodge' and 'burn' brushes, which work for small/medium areas. For larger areas, I select the area and feather it quite strongly (60 to 200 pixels) and then use the CURVES dialogue to lighten or darken selectively. If, for example, you are really trying to lighten a very dark face, be aware that the tones may have lost a lot of contrast by being compressed into the shadows. In this case I also use the brightness/contrast dialogue box to restore some macro- contrast as I lighten the image.

Finally. remember that, as with traditional printing, you have to experiment and undo and start over until you get a feel for the medium - it may take as long to learn Photoshop as it did to learn traditional printing; in my case 1-2 years to be competent - and I'm still developing skills (pun) 30 years later.

Regarding this particular picture - I would open the curves dialogue and hit the AUTO button to set black and white (and then look at the image to see if I agreed with the auto-set - you don't have to close the dialogue box, just move it around). The I would probably pull down on the center of the curves graph to darken the mid-tones and highlights - it seems washed-out like an underexposed silver print. Then I would probably go into the masking layer and drop a gradient from white to black from the top of the image to the bottom (ask if you didn't get that) to select the top half of the picture only with a strong feathering to the bottom, and use curve again to 'burn in' the top of the frame.

If you get conflicting advice, it ALL may be right - there are often two or three ways to do the same thing, more or less, in PhotoShop.

-- Andy Piper (apidens@denver.infi.net), October 04, 2001.

thank´s Andy, I think I need much of your advice, but you know if I seat back the pictures in the screen seem fine. Just kidding, wish I could what I know in the lab here in my computer, I shall give me time to learn the craft.

Angelique, congratulations, now we can go on seeing some more of your work, my first impresion on this was that something falling down was going to hit me. do you do you´r own printing?

-- r watson (al1231234@hotmail.com), October 05, 2001.

Andy

Wow, this is a thorough yet concise crash course which I would almost fully subscribe. Great post.

I would add just a few tips. In tune with your recommendation to do basic tonal tweeking before resizing and retouching I would extend this and counsel tonal corrections *before* Photoshopping - while scanning, that is. The better the scan the more options are left for tweeking it. I second you to absolutely save a raw scan before starting to work on it if you can afford it spacewise - not only for blackout safety but also just in case you wish to use a different size, resolution, cropping(!) or even tonal approach of the same scan. Third addendum: Get a dedicated JPEG plug-in for Photoshop such as BoxTopPro JPEG to dramatically decrease file sizes. It is so comfortable to use, allows you to preview in detail the effects of compression and in general cuts down file sizes to about a third of what Photoshop can do by itself. For further hints on scanning I recommend http:// www.scantips.com/ which has been pointed out in another thread on scanning tips.

As for the photo, I really like it, Angelique. Especially the fifties' touch of that apperently less than super-sharp and multicoated lens. So, I personally wouldn't try to cover that upper edge flare - but that's a matter of taste, in the end.

And it's nice that you, too, start getting a "face" - at least a first shade of one...;o) Cheers.

-- Lutz Konermann (lutz@konermann.net), October 05, 2001.

...tweeking is tweaking of course... and BoxTop Pro is available here.

-- Lutz Konermann (lutz@konermann.net), October 05, 2001.

Wow, this really is a crash course, alright -- like Roberto said, I was afraid the whole thing was going to land on my desk here.

-- Michael Kastner (kastner@zedat.fu-berlin.de), October 05, 2001.

Great post. Most sources say sharpening should be done last, and I do it just before I reduce size, and then find that after the size has been resampled another round of sharpening helps a lot.

-- Michael Darnton (mdarnton@hotmail.com), October 05, 2001.

Thanks Andy, you gave us a lot to work on but at least it is a very useful way to start. More questions come to me when I read all the posts here. For now I have one more question:

When scanning, I use the Minolta software. You have a pre scan after which you can do some adjustments and then scan the final picture. Is it preferable to scan at full resolution WITHOUT any adjustment and do the rest of the work with Photoshop? Or is it better to control the adjustment BEFORE the final scan?

-- Angelique (abischop@earthlink.net), October 05, 2001.

For what it is worth I adjust the histogram before scanning, and with B/W this is quite easy, I too have a minolta scanner so I guess you will have this option on your model. when you look at the histogram you will effectively see the tonal range of the negative across the full tonal range of the scanner. You have three traingular pointers that indicate the maximum minimum and centre or average position. Look at where your information is and capture the span of this only, and then adjust the central position for the best brightness/contrast balance as you see it. You can always make finer adjustments in photoshop later, but in this way you will maximise the dynamic range of the negative for want of a better descriptive term.

-- Richard (richard@designblue.co.uk), October 05, 2001.

All: I do the sharpening ALMOST last. But sharpening often reveals additional dust and other artifacts that need 'spotting', so I do my cloning/retouching AFTER sharpening. Otherwise I often have to do it twice.

Richard:'Dynamic range' is EXACTLY the right term.

I use Nikon scanner/software. It also has some pre-scan tone controls, and I use them to some extent (opening up deep shadows, e.g.).

I didn't go into them because there is probably some variation between Nikon and Minolta; e.g. Nikon, in addition to a tone graph (like the Photoshop CURVES control) also has a GAMMA control under one of the menus, which I only recently discovered, and which is very useful in preventing blocked highlights or shadows while getting the best possible contrast rendition in the mid-tones.

I ALSO have another trick, for film scanning.

Problem: If you use the scanner's AUTO exposure function (at least with Nikon) the scanner looks for the thinnest and densest parts of the film and pegs those to black or white (or with slides, white and black). This can end up 'clipping' off some of the detail or at least compressing it towards the ends of the 'dynamic range'.

Solution: I took some pieces of processed film leader, which are half totally fogged and half unexposed, and mounted them in slide mounts (a sample for each film I use).

I let the scanner "autoexpose' these pieces of film - in effect telling the scanner "This is what pure black and pure white really look like." Then I TURN OFF the autoexposure (with the Nikon it's a little 'lens aperture' button) and scan the real film.

In this way the scanner exposes for film-base-plus-fog as the lowest density and opaque film as the highest density and ignors the actual minimum/maximum, so that if there is ANY detail available it registers as something other than pure black or white, instead of dropping to level zero or punching up to 255 (white).

THEN I use the scanning tone control/graph/whatever to carefully set the highlights/shadows myself, usually JUST SHORT of pure B&W, so that I get the best possible dynamic range in the final scan WITHOUT losing any detail.

This is especially useful for Velvia (with its very high maximum density) and for slightly over/underexposed negatives.

It is an adaptation/analogue of the Zone system processes for determining basic print exposure time and (N)ormal developing time - you want the absolute minimum exposure or development that just barely gets you to a pure black (or white).

-- Andy Piper (apidens@denver.infi.net), October 05, 2001.

Great input here once more, Andy! Thanks.

-- Lutz Konermann (lutz@konermann.net), October 06, 2001.