Showing posts with label dng. Show all posts
Showing posts with label dng. Show all posts

Thursday, September 5, 2013

The InfraWeek...

This week been somehow a busy week with my infrared filter. After some trials that were done last week with the infrared filter, and after learning some new methods to process infrared RAW files, I've been planning for new shots and even visiting old shots for re-processing.
Before I delve deeper into this matter of the infrared venue, I do like to thank all visitors for peaking up my Flickr profile to more than 2000 hits in a single day! This is a first time to be that way. I think my own personal record was something a bit more than 1000 hits in a single day before!

Click to Enlarge

Infra
One of the main problems in processing infrared images (beside working with unconverted camera) is the fact that the white balance is limited when working in ACR, or even in Canon's DPP. Canon's DPP though was somehow a good tool before but currently and with my new acquired shots from last week, it didn't make much difference. Anyway, I've gone through some articles online and I've discovered a free software provided by Adobe, namely the DNG Profile Editor.
I won't get into the details of this tool and how to use it, but in a brief sentence, all I can say is, it makes a profile to be used in ACR, and it would maximize the range of the white balance in some way. This step alone, made me re-think about previous shots and the possibility of shedding them under a new light.

Frozen Coal
Cropped from original.
One of these old shots was a shot taken in the beginning of this year, in winter specifically, when we were still using coal trays or something like that to enjoy some traditional tea! Anyway, the shot that was taken and processed in the traditional way and even back then, swapping the channels (of Red and Blue to give an alien feeling to the image) did not yield or give much sense. Right now, though, the story is completely different and swapping the channels after fixing the white balance in the RAW file can yield a promising result. I say "can" here, so don't take it for granted!

InfraCoal

However, the coal shot was an "indoor" shot and not outside, thus I had to dig in my archive and see if there is anything like that taken with infrared. I've found one shot, for which I needed DxO to deal with its distortion and other criteria and this one, too, did a good job in swapping the channels.

Infraarbeitplatz


Despite this promising result from a previous shot, but it didn't really work out for much of the shots taken back when I just had the filter. Some of them didn't work out as before and some of them were better rather in red shade like the Ciudad en Rojo.

Ciudad en Rojo
Canon EF 18-55mm @35mm, f/8, 3.2sec, ISO100.

Anyway, after looking back at my stock of infrared shots, I've decided to increase the stock a bit more just to work with processing more images of this kind. This is my preferred method of learning (specially in Photoshop), and reading alone won't do me any good!
For this reason I've decided to take my camera to work with me and try to find something just to shoot with my infrared filter.

One of the first trials at work is a little plant which looked like a palm tree (but it's not). I leave the classification of this plant for the professionals! Anyway, it was a shot in a hurry as the background and the distance were not suitable really.

Paulo Palma
Canon EF 50mm, f/2, 1.3sec, ISO100.

Naturally, some infrared shots would turn completely or partially to black and white. I think this is the magic of infrared photography. Somehow, the result can be unpredictable. Anyway, there is a major difficulty with unconverted camera (i.e. camera with modified sensor for IR photography). As I will show later, this was a problem in some shots which required a fast shutter speed in a breezy weather. However, in Paulo Palma, the weather was relatively stable and not much of a breeze in the air, but surely had a problem framing the image while the sun striking on my neck and on the LCD! I've used a wide aperture relatively (and hence the shallow depth) just to ensure a faster shutter. Otherwise, it could have taken up to 3 or 6, or even 10 seconds for this shot.

One of the major hassles while working with IR images is the noise level even when ISO is set to the lowest. Next, comes the soft focus. These two matters makes it uneasy for me to upload such images on the web, and specifically into stock agencies. Thus, imagine when you are forced to increase the ISO under such conditions. The noise are simply "ridiculous". However, the noise can be less pronounced and just get more attention when channel swapping is formed, that is, replacing the places of the Red and the Blue channels in the RGB mode. At this point, the typical noise reduction plugins in Photoshop won't do much good, but the major trick is to play around with Hue/Saturation to twist the colors and make the noise virtually negligible. Then comes the next trick of adjusting the contrast to suppress some grain, and then we have the change in the color space by using the Assign command in Photoshop (not Convert). The Assign command, for some reason, does reduce (or maybe re-distribute?) the noise level, sometimes to a great extent. This is some of what I've learned doing long exposures last summer on the beach!

Infra Columnae
Canon EF 50mm, f/1.4, 6-1sec, ISO200.
InfraDoor
Canon EF 18-55mm @22mm,
f/16, 8sec, ISO200.
Despite the general look of the infrared shots, which tends to be in black and white, but the fact is lot of shots I've done were low saturated and just close to the black and white level. This is done by virtue within the processing of such shots, specially via Hue/Saturation when I try to reduce the noise by such an adjustment layer. The fact is, IR photography got a potential for various aesthetic settings or processes just like, or almost like, the HDR technique. The hard part, might be, to predict the final look of the shot. With HDR and its processes, you can imagine what's the final image is on location (specially if it is some metallic surface). However, this is still not so evident for me when I work with IR. Shots like Infra Columnae and InfraDoor were so various and I didn't know whether they would be colored slightly, or I'll be needing channel swap, or simply turn everything into black and white. The main puzzle is how to settle down the white balance. Usually, I do like to set the white balance from any green surface in the image if there is any but that doesn't mean I would like it. I suppose the green surface is the first choice because the green color is in the middle of the spectrum, not sure about this theory though! InfraDoor specifically is a subject for an HDR shot and it got me back to my old habit of still life and abstracts by geometrical means, like I used to do before around my work place - anyway this might be a subject for another post some other day!

Then one day came a challenge that took me 2 days to perform. It's a small bloom that was barely 1cm (~0.394in) in height. Not only I had to get so low with the tripod, but approaching the bloom and getting closer to it was critical. The first day, I tried to get to it by the end of the day before leaving my work place but the sun was so harsh that, beside my personal troubles, the LCD was barely visible and I had nothing to cover it (things to put on my purchase list as it seems). Thus, I had to come the next day and try to get to it in the early morning time just before the sun strikes that spot specifically.

Truncartis Parvis
Canon EF 100 macro + ET, f/8, 30-1sec, ISO200.

I'm putting Truncartis Parvis here for comparison with shots that were taken with infrared filter later. In Truncartis Parvis there were two main problems: first, the low light level as the sun wasn't shining on the spot (ironic), and secondly, the breeze which required faster shutter speed to stop the motion of the tiny bloom. I didn't bring my speedlite with me that day thus I had to depend on the simple flash on-camera. Because I couldn't get so close to the bloom, I had to use the macro 100mm lens with extension tubes added gradually to it. By using 12, 20, and 36 tubes gradually I was able to make some estimate about the relative size of the bloom within the frame. Truncartis Parvis (which means the small jungle by the way) was taken with all tubes (i.e. 68mm) added to the macro lens. Thus, I was trying to hit some balance between the 3 variables: ISO, speed, and aperture. I didn't like the idea of raising the ISO too high for such a "normal" shot thus I had to put down the aperture to f/8 (after trying f/16). They say in macro photography, it's better not to go below f/8. However, such balance was even harder to achieve and I had to give up with the ISO when I started taking infrared shots with that setting (remember the sun was not striking the place, yet).

Glowing Bloom (Gold)
Canon EF 100mm macro + ET, f/4, 15-1sec, ISO1600.

As you can see above, in Glowing Boom (Gold), the shutter speed was barely acceptable amid the breezes in the weather, and this is only achieved with ISO1600 and  shallow depth of f/4 with maximum flash output. Probably it would have been an easier task if my speedlite was with me but I had to deal with the situation as it is and yet the shot was of a surrealistic approach somehow. The noise level, however, was a novel on its own right.
The shot above was done without channel swapping (and of course with some processing on the side to suppress the background noise). Hence the "(Gold)" part in the name. After Red and Blue channels swapping, it appeared as beautiful as the first as it seems...

Glowing Bloom (Blue)
Canon EF 100mm macro + ET, f/4, 15-1sec, ISO1600.

It is amazing how the infrared filter clears off the differences in tones and colors and makes the petals as if they are made with one color (compare with the one taken without filter above). At the end and just before I take off, I've decided to do one last trial but this time with one less extension tube just to include more background into the frame. Ironically though, I've decided to crop the final image into a square to reduce the background!

Gecenin Çiçek
Canon EF 100mm macro, f/5.6, 10-1sec, ISO1600.

That was "some" of the trials done with the infrared filter for this week, but not the end of it. I need to study the artistic chances and venues behind this filter. It is cumbersome to use it of course with an unconverted camera, yet it does help in long exposures anyway!

Conclusion
Potentially, there is so much to be done in the IR field. Something behind the typical vivid and surreal landscape shots that we usually see under this category of photography.
However, I do find myself debating myself about the aesthetic prospect of such shots. Usually, what I read in articles online about IR photography is that the grains (or noise) and the fuzzy focus is something natural and even characteristic of this type of photography. There are even some Photoshop tutorials that deal with photo manipulation; that is to edit a normal photo and make it look like it is taken under IR, and one critical step in the overall process is to add grain or noise to the whole image (beside reducing the saturation). However, do people accept that, or do agencies? I'm reluctant about uploading such images to stock agencies. I did already upload to some of them but it's just a test and I'm very doubtful about their acceptance for such shots with such high grain level. This, somehow, makes IR photography a highly fine-art branch I believe, and agencies wouldn't love this type of photography much, unless some of them do know what does it mean to take IR shots with all that noise and defocus.
The almost-random aspect of processing IR shots makes it as vivid and wild as HDRs but let's say with different tones. You can do crazy things with the HDR techniques and lot of people hate the halos produced in over-tone-mapping HDR images, and IRs also got the same potential. If you check Infra Columnae back up above again, you'd notice that the image is almost black and white except of a spot that shows clearly as green. I'm pretty sure some people won't like it that way; just a matter of how the human brain works I believe.
In the near future, however, I do need to work more with such filters specially to see how does it change the skin tones. I've done some simple trials on myself but these are not enough and I need more. First though, I have to know what kind of relation there is between studio (or let's say the speedlites) light and this kind of filters.
After all, it is just something to get my mind busy and away from my collapsible life for the time being. I do miss doing lot of things; panoramas specifically. Anyway, got to wait for winter to arrive. This summer I'm literally melting down my body, and mood...



Thursday, May 31, 2012

Cephalargia...

If it wasn't for the frequent headache that have been striking me now and then from the beginning of this week, this week should have been fair deal with me. I've started this week with a severe headache (or migraine, call it whatever you like) that almost kept me paralyzed, but even with that I went to work on Sunday (weekend: Friday, Saturday), and didn't feel like going to work on Monday. This break from work in fact allowed me to do many things (since I've been awake in the morning). I think now, I do need a vacation just to do my own business!

I've started now recording the 6th chapter of Alexander's story, and in hope that the engine won't stop. In average, I'm planning to vocalize 4 verses a day, if possible. The break I had on Monday made me do lot of things, including the recordings, and also some images and panorama preparations from my session on Saturday (will discuss below).
The vocalization is now available for verses starting from 121 to 432, here. Of course you can always refer to the English version here, in the same website.

I. Scientific Center Workflow:
On Saturday, I had to push myself out of bed and head to the Scientific Center; a plan that I've been delaying for about a month or so. This trial now is different, specially with my 2 relatively new lenses compared to the last time I've taken pictures from there (April, 2011). After some negotiations with the receptionists I was able to get in. The difference here, beside the lenses, I went in with my VT-head and tripod. Heavy sum of tools that I had to carry through the corridors, and actually beside the panoramas and few simple shots with the macro lens, I don't think I needed the tripod after all.
One of the mistakes that I've made here is to bring the ball-head for the tripod instead of the pan-head. Despite the theoretically-easy movement of the ball-head, its lock system and fine-tuning for the shots proved to be a failure. I guess I will always stick to my old pan-head; it is easier to control.

Velbon pan-head (owned)
Source: B&H
Inside the aquarium, I've been mostly using my new macro lens (100mm) which works well as a telephoto lens as well. There were few instances where I would have changed the lenses for some particular locations, but the main lens was the 100mm.
The new lens has f/2.8 which theoretically works better in the darkness rather than lenses of f/5.6 and f/3.5 (that is, the maximum aperture). However, that did not avoid me from using high ISO ranging from 800 to 1600 at some instances. In fact even at ISO1600, there were problems following some movements and most of the images were blurred, which is normal in my case.

At some points, certain points, I've used my 15mm fisheye lens, to take panoramas of course. The panoramas were hard to take, and at a certain point I had to wait for 45 minutes just to shoot a simple 360 panorama. At the end, and with my last supposed panorama, I had to stop in the middle of the process after the arrival of some people, who eventually did not have any intentions to budge. Some panoramas, like the last one, was done in single shots in order to finish as soon as possible.

Aquacurvatic
Rokinon 8mm fisheye f/3.5 t/15 ISO1600
 One of the earliest shots to be processed was the shot taken between water tanks at the end of the aquarium almost. The play of light was enough to catch my eyes, but I had to wait for a significant time just to get this shot done (I think 15 minutes). I've taken this shot in portrait and landscape orientations but at home, I've realized that the landscape orientation fits better for showing more curving, and also for the planned DxO corrections. Starting from this image in fact, I've made up my mind for a new approach and a new workflow, specially when it comes to pictures taken with high ISOs.

Pengo
Canon 100mm f/2.8 t/160 ISO800
This approach involves using DxO first as the main RAW editing tool, followed up by using ACR for final touches. I realized that I might do better if I reduce the noise levels gradually. First by DxO (beside editing other aspects like contrast and saturation), which has ready-made profiles for a wide range of lenses and helps with reducing the luminance and chromatic noise significantly. Starting from this point of automatic correction (if available) I would reduce the amount of correction leaving some noise left in the image still. Finalizing the edit, the file would be saved as DNG (Digital Negative).

Fish Profile
Canon 100mm f/2.8 t/25 ISO1600
ACR has some advantages over DxO, as I believe. On one side it has easier controls, and it has direct editing tools on the other hand. DxO is complex and good, but sometimes you need simple touches like fill light to just add the right amount of light in the shadows; something you can struggle with somehow with DxO. Also, in ACR editing, the color space can be assigned directly, while in DxO it involves loading profile files and save it to the file. For this reason, I never assigned color space while working in DxO, but only when I open the DNG file in ACR.
The workflow continues with reducing the luminance and chromatic noise a bit further in ACR too. The sharpness amount is not touched though. Finally, a final noise reduction is achieved using NeatImage or Noise Ninja. Noise Ninja is preferred though since it can make a profile from different spots in the image and also the sharpness can be controlled easier than in NeatImage.
This lengthy workflow might not be practical in sense of speed but it did yield some good results I think. Also, it cannot be done for HDR processing. Aquacurvatic itself was originally captured in brackets for some HDR processing but I realized one of the slides was enough to achieve the desired look.

II. Too Cool?
Ever had an image, that after the first step of processing, you just find out that the image is so good that you just don't know what to do about it or how to process it because of the many possibilities and the racing ideas in your mind?
At least I know this is what was happening with me when I reviewed and tried to correct an image of a horned viper from the Scientific Center. The image was taken with Fluorescent WB (and mounted on tripod). Turns out that the image came out bright and not as contrasty as I hoped. With simple correction to the WB, the image became another story and I've realized its potential for different settings. However, I limited my editing to two versions only.

Tarnung (disguise)
Canon 100mm f/22 1.6sec ISO400

Tarnung - BW

The black and white version specifically was delicate to achieve, with many contrast and level adjustment layers to achieve the look. I thought of adding a tint of light blue to reflect a silver-ish look, but that would be even another path, another story.

III. Pano Pano:
As I've stated before, I took the chance inside the Scientific Center to take some panoramas. Not all of them are good I have to admit, since lot of visitors were there and in some places I had to work as fast as I can, but I did have some peace in one particular location despite the visitors: The Greenhouse; as I call it. It is the place where the penguins are and some birds are left to fly freely beside fish tanks and snakes, and some green cover. It looks indeed like some tropical forest, but with controlled air circulation!

Greenhouse #3
When stitching this panorama, I've tried to get out of the conventional box of certain ideas, like doing a vertical panorama using a flat spherical projection. Instead, I've tried the Mercator projection and converted the image into a vertical upside down panorama.
The main reason for this is that the empty space in the middle (or the ground where I stood) pushed all the interesting features on the side and a flat spherical panorama done vertically won't give much attention to such details (while stretching this space in the middle and make it look like an ellipsoid). Tone-mapping, on the other hand, was done in a simple form and not depicting any dramatic emotions, but I've tried to convey a natural feeling to it as much as possible (by using the Compressor option in Photomatix instead of the DetailEnhancer).
I've stitched this panorama 4 times, each time in a different projection style. I'm planning to present my work to the people responsible for the gift shop inside the center, in hope that I might get a deal out of such images if possible. As for now, I have to prepare all my papers and pictures before calling the main principle.

I've made simple stitches for other simple panoramas but did not proceed with processing. Some of them would be hard to represent, specially those captured in single shots and not in HDR. Yet, there is at least one panorama that is not touched yet at the time this post is written.

IV. Trends of Business:
Till this very moment I didn't hear anything from the Marina Mall desk which was supposed to call me back (after an optimistic conversation). However, and because I've been busy with other tasks, I didn't try to recall and check again.
The latest for me now is that I've been trying to make a layout design for a request paper that includes some prices for some services. Still not finalized, but the major aspect of it are set. I think I do need some help in creating this form (and also a help to measure its legal consequences).
On the other hand, and away from this request paper that I'm trying to create, I'm preparing a set of images that I've taken from the Scientific Center in order to make some sort of a proposal (in a form of a wholesale) that they might be interested in. The idea is to offer posters of some sort as a souvenir item for visitors, but that depends on their views and whether they think visitors would be interested in such an item or not. One step at a time...

This is some of the work done this week, and yet the image processing is still on its way and more to come. I didn't even put everything I've done here.
I do want so much to write some poem but my feelings are swinging so much and the work load is over, making it hard to concentrate on proper words.
Yet, I do ask myself, what's the worth of all of this... ?