With my recent acquisition of a Hasselblad H3D, I found myself in search of a new workflow. I've read the recent discussion about the best workflow when starting with the 3FR Hasselblad files. I am another user who is not at all satisfied with Phocus after having used Lightroom, especially the new revision of Lightroom. This is partially due to the fact that I'm using the PC version of Phocus, which still has many bugs, and only in the last revision (1.01) is it able to produce DNG files.
My previous workflow, before getting the H3D, was to start with film scans or Canon RAW files, go through Lightroom for some initial developing and overall adjustment, and then transition to Photoshop for extensive burning and dodging and adjustment. Very, very few of my images fit my vision without using Photoshop to a greater or lesser extent. My work is mainly B&W abstract landscapes.
My impression is that this separates me from many or most H3D users, whom I imagine to be studio photographers. If everything is well controlled and directed, I imagine, then there is much less need for extensive burning and dodging. I don't know if this impression is correct or not; it's just an impression. So I'm assuming that studio photographers have much less of a need for the capabilities of Lighroom and Photoshop, and are therefore satisfied with the capabilities of Phocus. Given what I do, I am definitely not. For now, and given my experience, I am going to use Lightroom, and the only question what type of file to start with.
To review for those who aren't familiar with it, Phocus offers 4 things that are not presently integrated into Lightroom:
- Lens distortion correction
- Noise Reduction in shadows
- Chromatic abberation correction (tuned for each lens)
- Vignetting correction (tuned for each lens)
A TIFF file generated by Phocus can contain these corrections, but it is a non-RAW file. A DNG file generated by Phocus cannot contain these corrections, but IS a RAW file.
I have read and heard that "DNG files generated by Phocus are lower quality than TIFF files generated by Phocus". I wanted to observe this, and to investigate in what ways Phocus generated DNG files might be worse than TIFF files.
I don't pretend that my testing was exhaustive or is universally conclusive. I'm reporting here to share what I've found, to see what others may have experienced, and to get comments. I don't claim to have an understanding of what is going on behind the curtains of Phocus or Lightroom.
Conditions:
Using Phocus 1.0.1b2 on a PC running Windows XP Pro, I generated four files for comparison:
1) DNG file
2) TIFF file generated with only lens corrections turned on (no other modifications)
3) TIFF file generated with all of the above corrections turned on, but no other modifications.
3) TIFF file generated with all Phocus corrections and modifications turned off
I imported the files into Lightroom 2.2 for comparison. I moved all of the Lightroom controls to neutral so that a direct comparison could be made without modifications applied to either TIFF or DNG. The image I used was a high contrast picture taken in bright sunlight. The scene included deep shadows and blown-out highlights.
Here are my conclusions separated into catagories.
Sharpness/Resolution
The TIFF files looked clearly sharper than the DNG when first imported. The question was whether the TIFF files were higher resolution, or just sharper. I modified the DNG file using the Lightroom sharpening tools, and found that I could sharpen the DNG to look indistinguishable from the TIFF file in terms of acuity. Looking at different parts of the image, shadow, highlight background, foreground, I could discern no difference between the TIFF and the slightly sharpened DNG. Conclusion: the basic "quality" of the images was equal - there was no difference in resolution.
Contrast
The TIFF files were clearly higher contrast than the DNG file, even with all contrast controls in Phocus set to neutral. Conclusion: I see this as a negative for TIFF files, since automatic changes give less flexibility.
Color
None of the files seemed to have exactly "correct" colors, although the TIFF files were much closer to the "real" colors than the DNG. With some modification in Lightroom, the DNG file could of course be corrected and brought closer to the "real" colors. Conclusion: The TIFF files are better as a starting point, but this is just a calibration issue.
Highlights
Highlight recovery in the blown-out highlights using the Lightroom "Recovery" slider was much more effective with the DNG file. The TIFF files also had a bad color cast in the highlights when the highlight recovery was used strongly. Conclusion: The DNG file is much superior for highlight correction.
Shadows
The TIFF files showed lower color noise in the deep shadows than the DNG file. However, the DNG file shadow color noise seemed easily corrected using Lightroom color noise controls.
Chromatic Aberration
The TIFF files were clearly superior for correction of chromatic aberration. Lightroom has chromatic aberration correction capabilities, but correcting the aberration in the DNG file in one part of the image means worsening aberration in another part (I am assuming this is due to lens distortion?). This is typical of working with raw images from Canon digital cameras.
Lens Distortion
Of course the TIFF files are superior for distortion, and there is no lens distortion correction available (yet) in Lightroom.
File Size
The DNG file was 1/2 the size of the TIFF file before modifications.
Conclusions
The TIFF files were superior in these catagories:
-Lens Distortion
-Chromatic Aberration
-Shadow Color Noise (but correctable in DNG)
-Color (but calibratable in DNG)
The DNG files were superior in these catagories:
-Highlight recovery
-Contrast
-File Size
Using DNG files as a default starting point is the best choice for me. There is no discernable difference in basic image quality, and the lower contrast and recoverable highlights have huge advantages for me. The smaller file size is not an overwhelming factor, but it is a factor.
The lens distortion correction would be nice, but it's something every photographer has accepted until now, and the differences shown in the files, at least for the one lens, are not all that significant. The chromatic aberration is more significant, but most of the work I do is in B&W, where the abberation doesn't result in a significant problem. Similarly, the shadow noise and initial color are less important because of my immediate conversion to B&W. For specific images or color images where these corrections might be more important, there is always the choice of starting with a TIFF file.
I look forward to the day when the Hasselblad corrections get incorporated into Lightroom, and then there will be no need for making this choice.
Chris