Sunday, February 1, 2009

A couple of pictures I restored recently




This is a family photo ca. 1900. It was preserved in poor condition and got torn and stained by cockroaches or other insects. The photo was about 10x16", too big for my scanner. I had to scan it in parts and then stitch in Photoshop. The Digital ICE did good job in removing cracks, but screwed up some other areas – it doesn't work properly with opaque and monochrome images. I had to scan it again without ICE and use parts of that image to cover ICE's errors.


Here is the restored picture. Apart from some corner unsharpness, those large glass negatives captured very good detail (check out the grandpa's beard!). It shows that sensor size does really matter! 


The second picture was a pencil drawing cracked so badly that ICE and usual small brush cloning technique were useless. I took a large soft cloning brush and actually "painted" the whole image trying to preserve pencil texture wherever possible. 



All pictures clickable.



The new AGO is a usability nightmare. Part Deux.


Last week I visited the new AGO myself and found out that everything written here was true. And more!

The miniatures were amazing but the lighting was so dim that my eyes immediately started to hurt. When I finished the exhibition, I felt like there was sand in my eyes. Is ivory afraid of light? 

Another "surprise" was in the photo room. There was a wall with quite interesting photos of 1930s Germany, placed in 4 long rows, as 20x4 matrix. Needless to say, there were neither captions nor numbers near the photos, just three shabby booklets, one in English and two in French – perfect demographic decision! To find info about an interesting photo, I counted and memorized its position in the matrix and opened the booklet. "Row S... Row T..." What the hell?! Right! What they called "rows" were, in fact, columns. Probably, the unknown "genius" who created this layout had never opened Excel...

I saw just  two exhibitions that evening, but I don't feel like going there again.