I read an article (forwarded via mail) about how too much post processing is masking real talent and was thinking about it ever since. Recently, I am browsing through a lot of photo blogs (ofcourse, looking to learn and emulate) and saw that a lot of post-processing is indeed being done. Once I learned a few basic tricks, a lot of the pictures actually look 'too much sharpened' or 'too high contrast' or 'nothing much in it, other than post-processing'? Does, anybody out there get a similar feeling?

Ofcourse, there are lots of people who does very minimal post-processing and come up with astonishing results. Those are the people, I would like to emulate! As a rule of thumb, I dont think about post-processing until I see the pic in my comp (except may be when I am trying an HDR: see 'HDR image from Golden Temple, Bylakuppe' or a moon merge: see 'The Full Moon Glory' - a very old post). Once, I have the picture in the comp, I would play a little bit in GIMP: trying to improve sharpness, color, contrast and do some cropping to remove unwanted portions. Thats about it. Some of the pictures I posted here are almost untouched. But, I am sure that there will be some people who may still think that I do too much of GIMP. I am sure, my dear wife is one of them :) The fact is, everybody would have their own levels of 'too much'!

If u r wondering where this is going, I was just thinking aloud and there is no conclusion :) Infact, I am sharing a heavily post-processed picture today:

I have a fascination for rural scenes, involving a lot of colors and I grabbed the opportunity on seeing this lady on my way from K M Doddi to Mandya. I had good light and shot this picture with the Canon EF 180mm f/3.5L (macro) lens, handheld and set in Av mode, ISO-100, f/3.5 for 1/500s. But, the picture didnt turn out to be as smashing as I thought ... in spite of applying an Unsharp Mask, improving the constrast and turning up the color in GIMP!

Hence, I sat and played a lot more than usual with that image. I could have done more, like trying sepia toning instead of Black and White and coloring the entire foreground, but I was too lazy to sit and do that much :)

For those who may be interested in knowing the 'How-To' part, here is what I did:

1. Open the picture in GIMP and open the Layers Dialogue (Ctrl-L).
2. Create a new layer (lets name it the 'B W Layer') and select the background layer.
3. Go back to the image, select all (Ctrl-A) and copy (Ctrl-C).
4. Open the Layers Dialogue (Ctrl-L) again, select the 'B W Layer', go back to the image and paste the copied layer back to the Image (Ctrl-V).
5. Go back to the Layers Dialogue (Ctrl-L), right click on the pasted layer and 'Anchor Layer'. The pasted layer should merge with the 'B W Layer'. Select the 'B W Layer' again.
6. Go back to the image, right click, go to Colors -> Hue-Saturation, move the 'Saturation' slide bar to its extreme left position and click on 'OK'. The 'B W Layer' should now turn black and white.
7. Now, select the 'Eraser' tool from the toolbar and erase the portions you want colored in the final image. Double clicking on the 'Eraser' tool will give you options to control the size and shape of the eraser.
8. Once, erasing is done, go back to the Layers Dialogue (Ctrl-L) again, right click on the 'B W Layer' and 'Merge Down'. You should now have a selectively colored image.

One has an option of converting the image to Sepia and using it instead of the 'B W Layer' for a sepia toned background. Ofcourse, the most tedious process is the erasing part and there are lot of different ways to do it, like 'Select by Color' / 'Select Contiguous Region' and delete (Ctrl-X). It might get tricky to properly erase some portions and increasing the size (Ctrl-+) will help. Infact, this is the reason why I got lazy and chose to color only the clothing, instead of completely coloring the lady and the bundle of grass :)

Update: I am setting myself a personal limit for post processing. Anything, I can do in-camera, I would try to do it in-camera. But, there are always limits to existing technology (say, in terms of dynamic range), equipment (lens not fast / sharp enough) or photographer (unsteady hands, improper focus, non-optimal frame requiring crop) and post-processing can help in some of these cases. Ofcourse, it should not be thought of as a fallback option while shooting, unless its used to overcome a technology limitation (read HDR, super imposing moon). The aim is to post the pictures as it is :)

To get an idea of the lighting situation inside the temple complex, check out the non-HDR images in an earlier post :Golden Temple @ Bylakuppe, from a Dec 2007 visit to the same place. In that picture, the idols are too bright and the remaining portions are dark. This is indeed an ideal situation to try out the HDR technique.

This HDR is created from three pictures, taken using the AEB settings in Canon EOS 400D. I set the AEB to +/- 2EV in Av mode, f/8, ISO-100 for three images, 1/3s for the base image, 1/13s for the dark image and 1.3s for the bright image. The camera was mounted in a tripod with the Canon EF-s 18-55mm f/3.5-5.6 kit lens @ 31mm.

Dynamic Range is a measure of the camera's capability to collect details from dark as well as bright regions. When the camera's dynamic range is not sufficient enough for a scene, one may use more than one frames with different exposures and merge them using software for a higher dynamic range.

Usually three shots (mid-exposure, under-exposure and over-exposure) are taken but more than three is also possible. All shots should have exactly the same frame but different exposure values. Its advisable to use a tripod for this, since any non-aligment of frames will spoil the merge result. Also, this is possible only for frames with little or no movement. Canon cameras have a setting called Automatic Exposure Bracketing (AEB) which will help to take three different shots with different exposures.

The basic idea of merging is to use the intensity of the image itself as a mask for merging. We should use the mid-exposure image as base and add brighter portions from the under-exposed image as well as the darker portions of the over-exposed image. I have used GIMP to do this but the procedure should be similar in all softwares.

Let me outline the procedure of merging three different exposure pictures for a HDR result in GIMP:

1. Start with the normal exposure picture (lets call it the base image) and open the Layers Dialogue (Ctrl-L).
2. Create a new layer (lets name it the 'Dark Layer') and select that layer.
3. Open the under-exposed picture. Select all (Ctrl-A) and copy (Ctrl-C).
4. Go back to the base image and paste (Ctrl-V). Go to the layer's dialogue, right click on the pasted layer and 'Anchor Layer'. The pasted layer should merge with the 'Dark Layer'.
5. Right click on the 'Dark Layer' and 'Add Layer Mask' (default settings will do). Keep the selection on 'Dark Layer'.
6. Go back to under-exposed picture and convert it to gray scale (Image -> Mode -> Grayscale). Now select all (Ctrl-A) and copy (Ctrl-C).
7. Go back to the base image and paste (Ctrl-V). Go to the layer's dialogue, right click the pasted layer and 'Anchor Layer'. The pasted layer should now be the layer mask for the 'Dark Layer'.
8. Right click on the 'Dark Layer' and 'Apply Layer Mask'.
9. Right click on the 'Dark Layer' and 'Merge Down'.
10. Create another layer (lets call it 'Bright Layer') and selct that layer.
11. Open the over-exposed picture, select all (Ctrl-A) and copy (Ctrl-C).
12. Go back to the base image and paste (Ctrl-V).
13. Go to the layers dialogue, right click on the pasted layer and 'Anchor Layer'. The pasted layer should now be the 'Bright Layer'.
14. Right click on the 'Bright Layer' and 'Add Layer Mask' (default settings will do). Keep the selection on 'Bright Layer'.
15. Go back to over-exposed picture and convert it to gray scale (Image -> Mode -> Grayscale). We need to invert the image. For this open the curves dialogue (Colors -> Curves), move the bottom-left point to top-left and top-right point to botom-right. Save this, select all (Ctrl-A) and copy (Ctrl-C).
16. Go back to the base image and paste (Ctrl-V). Go to the layer's dialogue, right click the pasted layer and 'Anchor Layer'. The pasted layer should now be the layer mask for the 'Bright Layer'.
17. Right click on the 'Bright Layer' and 'Apply Layer Mask'.
18. Right click on the 'Bright Layer' and 'Merge Down'.

The HDR picture is ready. This procedure may be extended to merge more than three pictures as well. To control the layer mask, one can adjust the color curves or edit the gray scale image in step 6 and 15.

I'd been adviced to watermark the pictures I post in public domain for a while now. Hence, I was trying to atleast find out how to do it, using my favourite image editing tool: GIMP. The procedure turned out to be fairly simple:

1. Open the picture in GIMP and Cut paste ur watermark logo into the image or bring up the text tool to add the text u want
2. Bring up the Layers dialog (Ctrl-L) and select the background image layer
3. Right click on the image and do Filter => Map => Bump Map with the default settings (experiment with the parameters to obtain different settings but default should work just fine)
4. Go back to Layers dialog and delete the text (or the pasted logo) layer

The watermark effect is not very conspicuous here and hardly spoils the image, but is barely there to convey the source of the picture. Isnt this what the watermark is supposed to do? :) I used the "Script MT Bold, Bold" font, with size 144 and placed the watermark at the top left corner in the picture.

I still havent made up my mind whether to put the watermarks in to the pictures I post here. It hardly makes any sense, especially since I am posting only small sized (max 500 x 750 pixels) pictures here. But I may end up doing it for wallpaper sized pictures at other places in the site.

Yeah .... the picture is taken @ RT Nagar house (indoor, night, artificial lights) using the EF f/3.5L 180mm lens, mounted on a tripod. I used Av mode with a -1Ev exposure compensation @ f/13 and ISO-100 for a timer triggered 30s exposure.

I'd been trying to use GIMP for Sepia toning and stumbled across this article in gimp.org, which talks about doing Sepia toning using GIMP. Though, sepia can be easily done using the Old photo built-in script available in GIMP dialog : Script-Fu : Decor : Old Photo, the procedure outlined in the article clearly outlines how Sepia toning is done. But, the best part in this procedure is the fact that Sepia (Red=162, Green=138, Blue=101) is not the only tone / color which may be used here. Various shades of blue (especially for pics with water), green (especially for pics wth greenery) and a lot of other colors may be used too. In this picture, taken in August 2007, near Ponnani river, using my Canon Powershot S2IS, I used GIMP and a blueish toning (R=0, G=128, B=255) for this result.

The original and the sepia toned versions are below:

Nowadays, every morning, when seeing the light coming in to the house ... I think of taking a pic :-) This time, the light was a little brighter and I cud use shorter exposures and higher DOF (f/6.3 and 1s for the 1st pic, f/6.3 and 1/2s for the 2nd).

Obviously, as is evident from the bored expression of Achan in the 2nd pic, this was more of a staged photo session and lacked the impromptu feel that I would like in portraits.

One thing I just cudnt decide was whether to use sepia or not. The pics, especially the 1st one has so many colors in it and it sometimes made sense to retain them. But, after doing sepia conversion ... I was in two minds. Just wasnt sure what to put ... finally I put both!

Best thing I cud've done on a monday morning :-) I wanted to take the pic against the light coming from the background. Obviously, the pic didnt come as well as I visualized ... and it lost a lot of sharpness after resizing ... but any pic of Achan is always good :-)

Against the light, manual mode was the only option @ f/3.5 and 2.5s. Needed a few retakes to make sure that Achan stay still thro' that 2.5s :-) Also, am not sure if sepia is the way to go here ... mostly I'll revisit this frame sometime.

I always loved using Sepia Mode. Here is one more situation where I prefered the sepia toning over the normal color. The lake in the pic is somewhere near Siddapura, while coming back from a Kodachadri Trek

In between all this I did have a few decent pics in my collection and one fine morning I decided to put them all together in a Photo Album. While doing this I used one of my all time favourite software, GIMP and got hooked playing with the pictures.

The picture above, the original taken in the Bengaluru - Krishnagiri highway (NH7) was modified with a different sky and slightly darker road to become the picture on the right. I am still not sure, if the modifications improved the picture ... but one thing for sure ... this is probably were I started developing a real interest in photography.

Another picture I spent so much time on was a picture of my aquarium taken in Aug 2004. The background here was so bad and the glass had a few scratches on it ... May be this is the time I started wondering why I didnt take care of all these while taking the photo! It is another story that I may not have been able to do anything about it at that time ... but I should have atleast noticed that the pictures are not coming the way it should.

A 3rd interesting pic was that of a snake, taken from Silent Valley in March 2005. A slight rotation of the snake to fill the frame better (at the same time removing some distractions), a little higher contrast and color and the picture looked very different. I shud've taken the picture like this only. Isnt it?

All this time, a good picture for me was nothing but a good frame and right colors. It changed slightly when I tried to blur the surroundings of a flower pic. There were too many distractions in most of the pics that I have taken, especially of the flower pics. Sometimes I have not even focused the flower properly or there were hardly any details, especially when the flowers are of brighter colors like yellow or white.

Had I paid a little more attention to the basics, things should have been a little different. But, I still didnt know what basics I was missing! And it took a few more months for me to understand some of the basic things I didnt know about taking photographs. First of all ... I had to try things other than Auto / P and Stitch Assist ... It took a few months patient efforts to improve the way I was taking photos ... and a lot more to go before I can proudly show off my pictures to the best people at it!