| Notices | Welcome to RetouchPRO . You are currently viewing our boards as a guest which gives you limited access to view most discussions and access our other features. By joining our free community you will have access to post topics, communicate privately with other members (PM), respond to polls, upload images and access many other special features. Registration is fast, simple and absolutely free so please, join our community today! If you have any problems with the registration process or your account login, please contact contact us. | | Photo Restoration Repairing damaged photos | 
05-18-2005, 08:19 PM
|  | Member | | Join Date: Mar 2003 Location: Cambridge, MA
Posts: 37
| | Yipes! Stripes! (If you get that reference you may not be as old as me, but you're creeping up there!)
OK, to business:
I am attempting to clean up this photo in order to make enlargements of it. What's attached is just part of the picture, for reasons that will become clear, perhaps ...
The original is a bit over 2 by 2 inches. This would normally not be a huge challenge for me -- I've made creditable blowups from a scan of a single frame of a 35mm contact sheet, and this is much larger than that.
However ...
(and I think this is going to be an increasing problem for the likes of us as time goes on)
this print was made on an inkjet printer, something I didn't actually realize till I zoomed in on the scan. It looks OK to the naked eye (save for the blemishes I plan to remove), but scanned -- UGH!
You won't even see the worst of it in the piece I'm attaching, since it's only at 50%. If you zoom in to 200%, not only are the stripes all too visible, they are multicolored.
This sample is from a raw scan I made at 1200ppi. Other scans, with various corrections (including moire) appear a bit better at first, but are at least as bad to work on.
I'd prefer not to deliver my client something that looks as though it was photographed (indifferently) off an old B&W TV set!
Any suggestions? I'm nearing the end of my wits with this ...
Last edited by OlProfBear; 05-18-2005 at 08:23 PM.
Reason: attachment didn't show up
| 
05-18-2005, 08:34 PM
|  | Senior Member | | Join Date: Oct 2003 Location: Nanaimo, British Columbia
Posts: 1,213
| | | Using an FFT filter might give you the results you need. You can download it off the web.
Dave | 
05-18-2005, 11:34 PM
| | Senior Member | | Join Date: Oct 2002 Location: Los Angeles
Posts: 370
| | Quote: |
Originally Posted by Duv Using an FFT filter might give you the results you need. You can download it off the web. | This looked helpful, so I went looking for an FFT filter. I found a free program called RegiStax that is supposed to include one, but for the life of me, I can't find it in the interface, even after reading the manual.
Could you tell us which FFT filter you used, or suggest one. And maybe help get us started in understanding how to use it. | 
05-18-2005, 11:51 PM
|  | Senior Member | | Join Date: Oct 2003 Location: Nanaimo, British Columbia
Posts: 1,213
| | Hi Andrew. Here you go. http://www.pages.drexel.edu/~avc25/
There's quite a few posts on this forum that will help you how to use the filter including a thorough one by byRo. Try entering FFT in the Search field. It's not the easiest filter to use so if you need help just scream.
Cheers
Dave | 
05-19-2005, 01:01 AM
| | Senior Member | | Join Date: Oct 2002 Location: Los Angeles
Posts: 370
| | Quote: |
Originally Posted by Duv Try entering FFT in the Search field. | I don't know why I didn't think of this. I've searched here many times before. Quote: |
It's not the easiest filter to use so if you need help just scream.
| Thanks. | 
05-19-2005, 01:31 AM
|  | Member | | Join Date: Mar 2003 Location: Cambridge, MA
Posts: 37
| | | FFTery and such ... Well, I reread everything here on FFT, then went to see if I had the latest version of the filters. Turns out there is a new one available -- actually two new ones. One works like the old one, but the other eliminates the need to mess with HSB (I started to type HBS -- Harvard Business School -- which is right across the river here) ... it generates the two pseudo images in the R and G channels (which makes all the instructions regarding the G and B channels rather confusing).
Anyway ... it remained and still remains unclear just what constitutes a "star". I tried three different approaches to this. In the first, only roundish dots with or without halos were considered "stars". In the third, anything except an obvious line was. (My image created some cute arcs of dashes, and I called them stars for this try.) The second try was somewhere in between.
Of course, I refrained from messing with the central "star".
It was also not that obvious what to lighten on the other channel. For the first two tries I didn't touch it. For the third, I cloned out a couple of arcs of dashes, similar to the ones on the first channel.
The second try seemed to be the least bad. What that means, I do not know. And I can't tell whether the remaining graininess and possible artifacts are inherent in the image or the result of less than perfect FFTing. How critical are the steps modifying the pseudo images? The differences in the three tests I made were fairly subtle, so it looks like maybe not so critical?
This is still being frustrating, but at least it's a start, thanks!
Last edited by OlProfBear; 05-19-2005 at 01:34 AM.
Reason: left out a sentence! and made a silly typo
| 
05-19-2005, 06:28 AM
|  | Moderator | | Join Date: May 2004 Location: Goiânia, Brazil
Posts: 1,549
| | Quote: |
Originally Posted by OlProfBear ...Turns out there is a new one available -- actually two new ones. | Where? I'd be real pleased to find something easier to use, understand ( and explain) than the Alex Chirakov filter.
Rô | 
05-19-2005, 08:30 AM
|  | Moderator Patron | | Join Date: Mar 2002 Location: Milan, Italy
Posts: 2,058
| | | ...Sorry to disappoint you Roland ....It's the new 'Alex Chirokov' version ..... in other words .... same thing in pink! ..... | 
05-19-2005, 07:35 PM
|  | Senior Member | | Join Date: Oct 2003 Location: Nanaimo, British Columbia
Posts: 1,213
| | Not trying to ignore you Bill. The problem is that some photos show very clear stars and others small blotches. In fact there was one recently that no one thought they could see anything but I found two small dots on the y axis, removed them and the picture improved markedly. On your photo, I could only see the obvious stars on the vertical axis. I removed both the stars and the line that eminates out from them. It didn't get rid of all the texture but this tutorial might help you to improve it further. http://retouchpro.com/tutorials/?m=show&id=140
Dave | 
05-19-2005, 07:43 PM
|  | Member | | Join Date: Mar 2003 Location: Cambridge, MA
Posts: 37
| | Quote: |
Originally Posted by Flora ...Sorry to disappoint you Roland ....It's the new 'Alex Chirokov' version ..... in other words .... same thing in pink! ..... | Well said ... though this particular photo came out in a lovely shade of green with a glowing pink center.
So ... this time I split the channels, converted each back to RGB, applied the filter, used the healing brush, and converted each back to greyscale, then recombined them.
Got a relatively usable picture out of the deal (compared to what I had been getting), so I cleaned it up for a while and came up with this (which still has a couple of annoying artifacts near the head) ... | 
05-20-2005, 12:44 AM
|  | Member | | Join Date: Mar 2003 Location: Cambridge, MA
Posts: 37
| | | About as good as it's gonna get ... I can't spend the rest of my life fixing this picture, and anyway I seem to have reached the limits of my tools and present skills.
So I decided to sacrifice a bit of detail for a bit of softness, and came up with this. Will see what my client prefers ... |
Posting Rules
| You may not post new threads You may not post replies You may not post attachments You may not edit your posts HTML code is Off | | | All times are GMT -6. The time now is 09:57 AM. | |
|