View Full Version : Two questions
03-08-2004, 09:56 PM
<B>First</B>: If I scan a print at 300dpi, adjust brightness/contrast/color, resize, sharpen and save for Web. Where in this process should I run the image through NeatImage for the best result? And is the same appliable for a digital image?
<B>Second</B>: If I have a photo with a bright upper part and a dark lower part, so when processing it I'll end up having either an overexposed or an underexposed half. Is there - in photoShop 7 - a way of adjusting brightness and/or contrast in <I>one half</I> of the image, without the two halves looking too disparate?
Thanks in advance,
03-08-2004, 10:26 PM
First: I don't know.
Second: The best way is to use a Mask. In the main toolbox just below the area for setting foreground and background color, there is two buttons. When you hover the cursor over the left a little label pops up saying "Edit in Standard Mode", over the left it says "Edit in Quick Mask Mode". Choose "Edit in Quick Mask Mode". Now you can use any tool to paint a mask on the image visible as a red layer, you can also make a gradient. The "thickness" of the mask will later determine how much any kind of editing will affect the image later (as long as the mask is in use).
In your example I would use the Gradient tool to paint a mask that starts solid at the bottom and fades out towards the top. When the mask is as you would like, press the "Edit in Standard Mode" button and the well known "walking ants" edge of a selection appears.
Now you can apply the editing like eg. darkening and see how it affect the image according to the mask. You can toggle between mask and selection at anytim on the two aforementioned buttons.
03-08-2004, 10:38 PM
Thanks Torben, I'll try that. I have an image which is extreme and might be very good for trial and error. If I succeed to my satisfaction, I'll upload it. My latest posting, which is a re-post due to this problem, had to be cropped as I knew of no other way doing it.
03-09-2004, 02:07 AM
I will take a stab at your first Kaj. For me, when I use Neat Image, I run it first thing. What I do to allow the most flexibility is to run Neat Image over a photo and save the result as a different file name. Neat Image does this as a default. Then, I will open the original photo up in Photoshop, as well as the Neat Imaged version. I then copy the Neat Imaged version and paste it as a layer over the original. From there, I will use the eraser tool to erase away any areas of the NeatImaged layer where too many details are lost. For me, I will usually erase the Neat Imaged layer completely from faces and hands, as these are the areas I want maximum sharpness. When using the eraser tool, I will often use a lowish opacity (maybe 35%), so that I will have to make a number of passes over an area to erase it completely. This way, I can somewhat dial in the effect I want. Also, I can vary the opacity of the entire Neat Imaged layer to control the effect as well. After I am satisfied with what I have created in Neat Image, I flatten the image and go on with the other adjustments (curves, color balance, etc). It is important I think to do this at the beginnning so that you will get consistent color reactions throughout your photo.
Alternatively. you could add a layer mask (hide all) to the Neat Imaged layer and then use a white brush to paint in the effect you want, but for me this is no more effective and takes longer, so I don't go this route.
Hope this makes sense.
03-09-2004, 02:34 AM
Thanks very much, Darren. Your note was highly informative, as you not only told me when to apply NeatImage, but also how to do it. It sounds like a very effective way to use it in the post-processing, still being in full control of how the picture develops. I like to do one thing at the time and be able to go back one or more steps when I'm not satisfied with the result. I will take your instructions into great consideration next time I'm scanning a print for upload.
And I have a lot of them.
Next step will be learning how to scan color slides, of which I have even more... ;o)
03-09-2004, 03:13 AM
First: If you run it through NeatImage at all I'd either use it on the original scan, or after you've adjusted the scan for brightness etc. The only time I use NeatImage is if I feel there's an excessive amount of noise in the image, or if I want to smooth out skintones somewhat - none of my current TE pictures use NeatImage.
Second: Consider using Levels or Curves, both are very powerful - if you're careful you might not need to use Layers or Masks at all! Which you need depends on what the problem is - Levels is most useful for correcting colour casts or increasing contrast (or brightening midtones), Curves is more useful for brightening (or darkening) a specific range of intensities. Generally I'll do Levels first, and then follow through with Curves if required.
03-09-2004, 03:46 AM
Thanks Adrian. As I understand it, you haven't used NeatImage on any of your scanned photos uploaded to TE. I thought this was done as a rule, but I haven't seen the expected result. And by that I mean prints - not color slides, as I saw somewhere that you have an awsome slide scanner.
The photo I'm mostly thinking of, when it comes to brightness/darkness, is taken in the transit hall of Frankfurt airport, which had a metallic roof thus reflecting all light from upward lamps and windows. While all people are in darkness among lots of dark furniture and clothings. This makes the contrast large and I have uptil now no way of dealing with that.
I'll take all your advises and play around with them, until I find a way that suits me. If the improvements are considerable I'll present the result on TE!
03-09-2004, 04:08 AM
You're right, most of my shots are either digital originated or slide scans. I used to scan prints though. I've never needed NeatImage for Fuji Frontier produced prints, I have on an older Kodak digital print though.
You might find that to get the most dynamic range you'd be best off scanning the image twice - once for highlight detail and once again for shadow, and then combining them.
Feel free to send me a message and I'll have a go at the picture if you wish :-)
03-09-2004, 04:50 AM
Gee thanks, Adrian. Let me have a (couple of) go at it first, to see if I'm able to convert your tips into real action. I'm very much into that learning-by-doing and trial-and-error stuff, and so far I think I've done pretty good. But there are always these cul-du-sacs and usually I'm able to turn to my son, but when it come to photography and post-processing - other than plain images - he's lost. So I'll be happy, if I may get back to you in case of this happening.
And also thanks for your casual scanning tip! That might do it!
03-09-2004, 08:06 PM
As useful as PhotoShop is, in answer to your second question there's usually no substitute for a bit of additional work before you take the shot. I find that using a neutral density graduated filter allows you to compensate for a wide range of brightness in a scene. If you search the equipment section in the gallery for the Cokin P121 filter you'll see numerous examples of where I've used such a filter -
<A HREF="http://www.trekearth.com/photos.php?cat=filter&id=8">Cokin P121 shots</A>
In many of these shots it's not obvious that a filter has been used at all, and that's what you should be aiming for in general.
The thing to note is that you <b>can</b> retreive underexposed areas in a digital image, but you can rarely get detail back from overexposed areas. Therefore, post-processing isn't always a solution as a bright white area will only become drab gray. A filter is a much better solution, because you're controlling the range of detail that's captured.
03-10-2004, 04:38 AM
Thanks Ron, for your advise. I'll really look into the issue of Cokin filters. As you write, it's better to do all of the basic work before or when taking the shot. But I also needed advise on how to deal with photos already taken. I am a beginner when it comes to digital photography and some of my pictures have a contrast bright/dark that I didn't foresee. I'll try to save what is possible and fully understand that overexposure equals "blown out".
03-12-2004, 11:51 PM
Another technique (if your camera has this facility) is to check the histogram on the LCD display after you take a shot. If there's a spike at the right hand side that suggests that you may have a number of overexposed areas - which you won't be able to recover detail in during post-processing. My camera (Nikon D100) also has a feature to highlight blown out areas on the shot directly. If you have the histogram feature you might be able to make some adjustments while shooting.
03-13-2004, 02:50 AM
Thanks for the tip, Ron. I have an Olympus C-750UZ and yes, it has the histogram feature. Now I understand the function of this and also how to use it while shooting - I'm beginning to grasp the importance of adjusting while taking photos and being able to do so at no extra cost. The advantages with digital photography!
vBulletin® v3.8.4, Copyright ©2000-2013, Jelsoft Enterprises Ltd.