I created an HDR image, why can't I save it as JPEG?
An HDR image stores color information in 32-bit precision and as such can't be directly saved in a Low Dynamic Range format such as JPEG.
An HDR image in itself is not very interesting for photographers since it can't be correctly displayed on monitors and even less printed. This is why the HDR image you created from several exposures needs to be further processed for photographic use. It needs to be tone mapped so that the values of the image data fit into the limited range that your monitor can display.
Once you have applied the Tone Mapping tool to your HDR image, the details in highlight and shadows captured with your multiple exposures will then become visible on screen and you will be able to save the result as JPEG (or as 16-bit TIFF image if you choose this option).
My camera does not shoot RAWs. Can I still use your software with JPEG pictures?
Definitely. Photomatix does NOT require images in RAW format. Photomatix works with images taken under different exposure settings, and works great when those images are JPEGs as well.
Most results shown on our example page have been produced from differently exposed jpegs, by the way.
So, the important feature to look at in a camera is Auto Exposure Bracketing (AEB): the ability to automatically take three or more exposures, preferably in two EV steps as detailed here. The higher the number of auto-bracketed frames, EV step and frame rate speed, the better it is for HDR processing.
When I load Raw files in Photomatix, why does it ignore my Camera Raw edits?
The settings added by Adobe Camera Raw in XMP sidecar files are parameters of the Raw conversion engine of Adobe Camera Raw (ACR). Photomatix does not have access to ACR's Raw conversion engine, which is a proprietary technology, so Photomatix cannot use information set by ACR in the XMP files.
The same applies to Adobe Camera Raw settings embedded in DNG files.
In order to preserve the edits you made in Adobe Camera Raw, you will just have to convert your Raw or DNG files to TIFF, and then process the TIFF files in Photomatix. See above for more information on this workflow.
Why does the size of the Photomatix output slightly differ from the size of the original Raw file in Photoshop?
When you load Raw files directly in Photomatix, the Raw conversion is done by Photomatix, and the Raw conversion process differs depending on the software doing it.
The Raw conversion used by Photomatix tries to extract as many pixels as possible from the original Raw data captured by your camera's sensors. For some camera models, this results in a width and height that are a few pixels more than the width and height you get when the Raw file is converted by other software.
To avoid this, you would just need to convert your original Raw files first in Photoshop (or other Raw conversion software you are viewing the Raw file in), then load the converted TIFF or JPEG files in Photomatix. The pixel size of the images produced by Photomatix will then be same as the size of the source images.
Note that if you use the alignment option and need to overlay the Photomatix image with the original photos, you will have to leave the 'crop aligned images' box unchecked to ensure the size is not reduced.
What is the maximum file size that I can work with?
The file size to consider for Photomatix is the size expressed in number of pixels, i.e. width X height. Since Photomatix has to decompress the images for processing, the compression factor of the input images does not make any difference in the ability to process large files.
The maximum file size (in number of pixels) that you can process with Photomatix depends on the following:
- the RAM your computer has
- whether your OS is 32 or 64-bit
- the pixel depth of your images
- the number of bracketed photos you are combining
- the number of other memory-hungry applications opened on your computer
- the free space available on your hard drive
Additionally, it is important to note that there is an upper limit to the amount of RAM that is made available to applications like Photomatix. On Windows XP 32-bit, this limit is as low as 2 GB, regardless of your RAM. That is, it will still be 2 GB via virtual memory if you computer has less than 2 GB RAM, but it won't be more than 2 GB if your computer has 4 GB RAM (unless you enable the /3GB switch).
Then, external memory fragmentation further limits the memory available to Photomatix, by making it impossible to allocate a contiguous block of memory large enough to contain the whole images data. External memory fragmentation is a problem on Windows OS, and particularly on Windows XP. It means that the system does not organize the available memory efficiently, making it unable to re-use the memory that Photomatix has released.
When you create in another application a large 32-bit HDR image file that you want to tone map in Photomatix Pro, you have the option to open the file in so-called "Preview" mode. This will load a low resolution of the file, thereby avoiding saturating the memory available for the user interface when adjusting the tone mapping settings. The final process will then run in the background from the file on disk. Note that the "Preview" mode is only available when the 32-HDR file is saved in the Radiance format (.hdr extension).
For an idea of the memory necessary to process your images, the following formulae give a rough estimate of the amount of memory (RAM) needed (in bytes)
Merging bracketed photos to HDR:
width * height * (3 * (bit-depth/8) * numberOfImages + 6)
Tone Mapping with Details Enhancer a 32-bit HDR image file opened in "Preview" mode with the Lighting Effects Mode box
checked (the memory needed is significantly higher otherwise):
width * height * 18
This means that merging three 100 MegaPixels 16-bit images to HDR requires around:
100,000,000 * (3 * 2 * 3 + 6) = 2.4 GB
And tone mapping a 100 MegaPixels HDR image with Details Enhancer (Lighting Effects Mode box checked) requires around:
100,000,000 * 18 = 1.8 GB
The amount of memory needed to merge photos to HDR can be quite high when the source images are in 16 bits/channel mode and/or when there are many of them. However, there is an option in Batch Processing that allows creating the 32-bit HDR image file one strip at a time, provided the source images are TIFF files. When this option is checked, Photomatix Pro will only load, process and write a given number of rows from the source images instead of the entire images. This way, you can create a 32-bit HDR file stored in Radiance (.hdr) format from any number of large source images, even on a computer with limited RAM.
How does your software handle color profiles?
Photomatix processes the RGB values of your source images directly, without converting them to another color space. This means that the resulting images produced by Photomatix will be in the same color space as the one specified by the ICC color profile of your source images.
When a color profile for the source images is available, Photomatix embeds it into the resulting tone mapped or combined image.
Please note, though, that in the case of a tone mapped image, the color profile can only be passed through if the Generate HDR and Tone Mapping steps are done in the same session. If you have first saved the HDR image file, then the ICC profile information will be lost and the Tone Mapping step won't know the color profile anymore. This means that you will have to assign the color profile of the original source images to the tone mapped image yourself.
However, since version 3.0, Photomatix Pro saves the name of the color profile in the header of the HDR image file when this is saved as Radiance (.hdr extension). This means you will not have to re-assign the color profile, provided the HDR image file has been saved as Radiance and the color profile is either sRGB, Adobe RGB or ProPhoto RGB.
Additionally, Photomatix Pro is color managed for the display as well. This means it will show the correct color values based on the ICC profile of the image and the color profile set as display profile for your monitor.
On Mac OS X, the display profile is set under System Preferences->Displays->Color.
On Windows, it is under Settings->Control Panel->Display->Settings->Advanced->Color Management.
I can't get sharp results with your software!
Lack of sharpness on tone mapped HDR images is often due to the use of Shutter priority or Program mode instead of the recommended Aperture priority mode when bracketing shots.
It it important to set your camera to Aperture priority when shooting with Automatic Exposure Bracketing. This way, only the shutter time will vary, and the Aperture will remain the same. If you set it to Shutter priority instead, the depth-of-field will change between the shots, and this will lead to inferior results on the combined image. Additionally, there may be vignetting issues too.
Are there ordering rules for loading the images?
No. You can load the images in any order, regardless of their Exposure Value.
Creating an HDR image requires to assign an exposure to each source image. Photomatix automatically retrieves the exposure information from EXIF data. When the images do not have EXIF data, Photomatix will let you enter the exposure values manually, or automatically estimate them when running in batch mode.
What are the differences between Exposure Fusion and HDR/Tonemapping?
Both processes start from the same source files: differently exposed Low Dynamic Range (LDR) images. And both attempt to produce as final result an LDR image that shows tonal details of the entire dynamic range captured by the different exposures.
The differences are in the process itself. Exposure Fusion consists in combining the differently exposed images in such a way that highlight details are taken from the underexposed photos and shadows details from the overexposed ones. Since the bit-depth does not change throughout this process, the basis of Exposure Fusion algorithms is a type of weighted average of the source images.
One of the advantage or Exposure Fusion is that it is easy to understand and you can see what you are doing. Also, it is rather familiar to photographers who are used to doing this process manually in image editing applications. Another advantage of Exposure Fusion is that it reduces noise.
HDR Tone Mapping is composed of two steps. The first step creates an HDR image from differently exposed photos. This HDR image can not be displayed correctly on a Low Dynamic Range monitor, which is why a second step called Tone Mapping is necessary. Tone Mapping consists in scaling each pixel of the HDR image, so that details in highlights and shadows show correctly on monitors and prints (those details are available in the HDR image but not directly visible in both highlights and shadows because of the low dynamic range of the display).
Tone Mapping algorithms vary from a simple gamma curve (which is often what cameras are doing when converting 12-bit RAW data to 8-bit JPEGs) to more complex operators commonly divided into two categories:
- Global operators: mapping depends on the pixels' intensity and global image characteristics, but not on spatial location
- Local operators: mapping takes into account the pixels' surroundings (in addition to intensity and image characteristics).
The main advantage of global operators is fast processing. Local operators require longer processing times but they are better at producing a "good-looking" photograph (the human eye adapts to contrast locally). In Photomatix Pro, the Tone Mapping method "Details Enhancer" belongs to the category of local operators and the method "Tone Compressor" to the category of global operators.
The pros and cons of Exposure Fusion vs Tone Mapping in Photomatix Pro are detailed under the section below.
Does Photomatix make use of dual and quad processors?
The current version of Photomatix Pro supports multiple processors for the Tone Compressor tone mapping method and noise reduction on source images, as well as parts of the Details Enhancer tone mapping method, Exposure Fusion/Natural method, alignment by matching features and RAW conversion. Further versions will add more multi-threading support.
However, it is important to note that most processes of Photomatix are memory-intensive, which means that multi-processor support will not speed up processing times as much as one may expect. For a memory-intensive process, the bottleneck regarding processing times comes from memory accesses rather than a high number of operations. This means that the processor has to stay idle for many cycles, waiting for data to be fetched in memory. Adding more processing power in this case will just waste even more cycles.
So, multi-core support will not improve much the processing speed of the most memory-intensive functions of Photomatix, and may even increase processing times because of the overhead incurred. However, there are still processes in Photomatix that are not particularly memory-intensive and can thus benefit from multi-threading -- the Tone Compressor tone mapping method is probably the best example, but some parts of the Details Enhancer method can benefit from it as well.
How many images can I merge?
With Photomatix Essentials, the number of images you can merge is limited to 5.
With Photomatix Pro, the number of images you can merge is unlimited. However, The fusion method "Highlights & Shadows - 2 images" merges only two images. If you have loaded more than two images, Photomatix Pro will let you select the images to process.
For HDR generation and all other fusion methods, you can merge as many images as you want.
When you have 3 or more bracketed shots and are using the "Exposure Fusion" functions of Photomatix Pro or Photomatix Essentials, we recommend that you also try to merge fewer images, eliminating for instance images that do not add much values (e.g. an underexposed image where even the highlights are too dark to provide interesting details).
Is it true that Photomatix converts RAWs to JPEGs for internal processing?
No. Photomatix does not convert RAWs files to JPEG for internal processing, and never did it. It would not make sense to do this anyway, given that converting to JPEG would result in quality loss and moreover would add processing time.
When you load RAW files in Photomatix, the files are converted in linear space into an uncompressed image with 16 bits per color channel, i.e. 48 bits per pixel.
The only moment Photomatix converts to JPEG is when you want to save the image created by Photomatix and choose to save it as JPEG. This applies to a tonemapped or fused image created by Photomatix, and not to the original image you loaded.
Can I use your product for combining multiple scans?
Yes, Photomatix can be used to combine two or more scans from the same film scanned under different exposure settings.
There is an example produced with 3 scans from a film's negative here. You may also try with scanned slides, but it is better to do it with negatives, as the dynamic range for film's negatives is higher than for slides.
We would recommend combining the differently exposed 16-bit outputs from your scanner with the Exposure Fusion method named 'Fusion/Natural'. Since this fusion method is mostly automatic, it is recommended to use the Batch Processing function to process the 16-bit scans to avoid out-of-memory issues.
The scans will need to have the same size before processing them. Also, the option "Align images" will have to be checked in order to correct for possible mis-registration of the scans.