Hyperfocal Design
Interface
CS2 makes HDRI creation very simple. There is no lingo to learn, no camera curve graphs to mess with, no need to look at or edit tables of numbers and no manual tweaking of any kind. Depending on what you prefer, you may either like the streamlined interface or wish for more control.
Creation
Creating a HDRI in CS2 is as simple as selecting the files to use, then selecting the shutter speed and f-stops used for each image. Then choose if you’d like CS2 to automatically align your images (but you used a tripod anyway, right?) And that is it. The camera curve creates itself with no user intervention, the stops are all figured out from your shutter speed/fstop inputs. Before you merge the HDRI you are presented with a preview window showing the HDRI in the center screen and LDR images used to construct it down the left hand side. Control the initial exposure level using a slider on the right, and you can even pick and choose which exposures you want to include in the finished HDRI. If you are getting some colour problems or something has moved in one image or you don’t need so much dynamic range you can remove those images. The HDRI will automatically recreate itself and display a preview without that image. Once you are happy with your settings, press ok and you have finished your HDRI.
The finished HDRI comes out perfect time and time again, and you should experience the same thing if your initial photographs were taken and setup correctly. While I can’t remember ever having a problem with any HDRI created with CS2, due to the limited amount of control over the image means you have no where to turn if the image doesn’t form correctly.
You will have to whitebalance your LDR images proir to merging them aswell, since there is no tool available to perform this task after the fact. This is definitely a shortcoming if you have gone to a lot of trouble editing or fixing an image then later wish to tweak the white balance.
As far as speed goes, testing was performed on a fairly low end PC (Athlon 1.7 gig with 768 megs of RAM) and still CS2 had no problems. Loading the files prior to merging takes quite a while for 6-7 images, but considering each file is 500megs, Photoshop remains stable and never just sits there as if it maybe just hung your machine. Performing the final merge to HDRI operation on an image at 10,000×5,000 took around 5 minutes, but again on a slow PC this is FAST for such a large file. Plus the program always tells you what is happening with progress bars so you know it is still alive. Also stability wise, HDRI in CS2 is faultless every step of the way.
CS2 supports the following file formats for saving out HDRIs: PSD, Tiff, PSB, PBM, HDR and OpenEXR
Editing
For its first version supporting HDRI, Adobe has included a fairly impressive number of tools for editing HDRIs. Most of the functions below work on a selected region, making them more useful again. However since layers are not supported in HDRI format you have to ask how useful some of these tools are.
Included are:
- Exposure adjustment
- Exposure preview (doesn’t affect base exposure level)
- Blur
- Lensflare
- Sharpen
- Offset
- Channel Mixer
- Photofilter
- Transformations
- Crop
- Resize
- Canvas size
- Fill and Stroke
- Cut/paste and selections with feathering
- History brush
- And cloning!
The following paragraphs outline some problems and highlights of editing in CS2:
32 bit Preview
First of all you’ll need to be careful with 32 bit preview exposure changes vs changing the actual exposure level. If you’re trying to edit/clone at one exposure level but you have the preview level set at another level, you can create problems for yourself. To edit dark areas effectively you’ll need to reduce the image exposure level, and to paint bright areas you’ll need to be painting with the image set to a lighter exposure level.
Resizing Problems
Resizing works fine most of the time, however in high brightness and/or contrast areas, resizing in BiCubic mode can leave black pixels surrounding these areas. You can either do some grunt work and start cloning pixels, or you can use the only real work around, which is using Bi-linear resizing. Chris Cox, the programmer behind the HDR code has confirmed this and suggested that there would be a fix made available in the future.
Lensflare
Lensflare is next to useless, since the lens reflections can’t be turned off.
Cloning
Cloning usually works as you would expect, however you have to take into account exactly how bright the cloned areas are. Your cloning can look fine at one exposure level but be off at other settings. You can also find it difficult to paint darker areas into lighter areas without full brush pressure/strength. Use this tool to clone out the photographer, fix up moving objects, remove unwanted elements and so on. Definitely the most useful editing tool for HDRI in CS2.
Selections
Since you can make changes to a selection, you can artificially boost exposure in certain parts of an image, or apply blurring, sharpening, colour changes and so forth. You may not get layers but using selections is the next best thing and with a bit of hard work you can get around this limitation.
Speed
Again CS2 excells in image editing, everything from cloning to photofilters are fast even on a low end machine with large files.
Compressing to a 16 or 8 bit image (Tonemapping)
If you are a photographer looking to tonemap your HDR to be viewed in 8 or 16 bit format, you can convert your image using 4 options:
Exposure and Gamma: Choose a range of exposure within the HDRI file and don’t compress any tonal value. This really isn’t even tonemapping or compression at all, and if you are using this option you may as well have just chosen one exposure from your original LDR images or adjusted the RAW files slightly.
Highlight compression: This function takes detail from the brightest parts of an image and squeezes them into a LDR image. It is next to useless, it generally creates poor images over which you have no control over (there are no settings to adjust).
Equalize Histogram: Similar to the above but this mode will compress all the ranges into a LDR image providing a higher contrast version of the above. No control is available in this method either, making it just as useless as highlight compression.
Local Adaptation: By far and away the best and really the only option you will want to use, local adaptation allows you to adjust the resulting image via a curves like window. Since I am not a highly technical person I have no idea what is going on behind the scenes here, but with a bit of tweaking of values and curves you can get precisely the effect you want.
Mysteries
Since CS2 offers cloning you would think it would only be too easy to draw solid colours with a standard brush, yet it’s not available. It also seems odd that since you can use selections and feathering with such control, why are there no layers or masking? You can’t even brighten or darken using a brush, but you can make a selection and adjust exposure. But I won’t spend too much time on this, since I am not a programmer and there may be perfectly reasonable explanations for all of this.
Conclusion
From extensive experience with CS2 I fully recommend this software for use in creating and editing HDRIs. The pros far outweigh the cons with extra easy HDR creation and some great editing tools. There are some shortcomings – no HDRI white balancing, resizing bugs and some lack of control, but these issues can be predominantly worked around. The fact that you are getting a fully operational, easy to use HDR package as well as a professional image editing program means you can kill two birds with one stone. At a recommended retail price of $599 for a perpetual license you can’t go wrong.
For more information or to purchase Adobe Photoshop CS2 visit www.adobe.com
View Hyperfocal’s range of HDRIs
