VIEWPOINT, THE NEW KODAK PROFESSIONAL PRO 14n DIGITAL CAMERA
One Photographer's Opinion
by Pat St. Clair
I recently had the opportunity to work with the new KODAK Professional Pro14n Digital Camera. The main purpose of the work was to shoot a camera comparison test series in the studio with the KODAK DCS Pro14n and seven other 35mm style professional digital cameras. But I was excited by the opportunity to do some tests with it as a VR camera as well.
The short story is, the Pro14n has a very strong feature set and puts out very high quality digital images. In terms of QTVR work, shooting digitally with a full frame sensor is wonderful, and is a long awaited feature, as is the ability to choose from three resolutions: 14MP, 6MP and 3.4MP. I find that the Pro14n captures an extended tonal range as well, rendering highlights and deep shadows (simultaneously) very nicely. The camera is considerably smaller than all previous Kodak Professional digital cameras, and it's light weight works nicely with existing VR rigs. It has full manual controls if that's what you need, and it accepts a wide range of existing Nikon lenses. Until the Pro 14n came along, my choice has been to shoot 35mm film and scan rather than use a digital camera without the focal length, feature set or image quality that I need in my work.
With 25 years of professional commercial photography behind me, and ten of them shooting professional VR work as well, my professional opinion is that the Kodak DCS Pro14n is a breakthrough camera for the VR world. But, decide for yourself; I've placed five QTVR panos in an FTP server for any who are interested in downloading them. (see FTP info at bottom)
THE PROJECT
Being able to shoot VR with the camera was actually an added extra - the project was really to test the camera against seven other professional digital cameras in a studio setting.
This was an internal project for Kodak Professional Worldwide Marketing people at Kodak headquarters in Rochester, NY using the Pro14n and seven other 35mm style professional digital cameras. Although Kodak funded the project, I was given total freedom in designing the scene and designing the image processing protocol, and was free to present the results in my own way. Kodak did not attempt to bias the test or my conclusions in any way. They voiced no opinions or expectations and did not suggest protocols or attempt to direct my work - just gave me the assignment and let me work it out to my own satisfaction. As a result, any comments expressed here are independent and strictly my own; I am not speaking for Kodak.
The project went like this:
Part I
Design a still life scene that would prove challenging to all professional digital cameras and includes lots of color and texture, have a full tonal range with deep shadow detail, a mid-range neutral, and delicate highlight detail, and have a reasonable representation of a flesh tone somewhere in the scene.
Part II
Shoot that scene at three ISO's and a fixed f-stop with the KODAK DCS Pro14n, the KODAK DCS ProBack, the KODAK DCS 760, the NIKON D1X, the NIKON D100, the Canon D60, the Canon 1DS and the FUJI S2 as well as on 35mm and medium format film.
Part III
Design a protocol for processing the images (working from RAW files, not JPEGs) that is fair and objective and produces images that will facilitate true "apples to apples" comparisons from camera to camera. Use each manufacturer's own software to work from the RAW files, i.e., KODAK DCS Photo Desk, NIKON Capture 3.5.1, Canon File Viewer Utility and FUJI FinePixViewer. There was to be no subjective actions taken in the image processing (no sharpening, color balance, hue or saturation adjustments in Photoshop, etc). I worked very hard to come up with a non-biased protocol. Then I put the blinders on and let the chips fall where they may.
Part IV
Analyze the resulting images and design comparison layouts for 16 x 20 300 dpi output.
Observations
Of the eight cameras I tested, by my eye the Nikon D1X was the lowest image quality with the Nikon D100 next lowest (personal opinion only, I am not speaking for Kodak in any of this). If you add software functionality on top of image quality observations, then I'd put the FUJI S2 lower on the list as well. I felt the image quality of the S2 is good, but I found the software very slow to use. (Again, this is strictly a personal observation, and please note, any of this could change with the next release of manufacturer's software or firmware).
One of the most telling layouts I created from the test series was a comparison of all eight cameras bringing files into Photoshop WITHOUT ANY SHARPENING (noise reduction turned off as well where I had that capability). In my mind, this sort of "unveiled" the RAW files from all eight cameras . . . removed software intervention (the smoke and mirrors, if you will), to see what we really have to work with coming straight out of the camera. The Nikon images appeared noticeably soft, the three Kodak's and the Canon 1DS looked pretty good before any intervention.
And by my eye, the two best cameras in the bunch were the Pro14n and the Canon 1DS. Judging on image quality alone, the Canon had a slight lead over the Pro14n (as judged by zooming in to a fairly extreme degree . . . at a reasonable viewing distance, they were fairly indistinguishable). But, the Canon is currently about twice the price and the software is a little clunkier (once again, personal opinion only . . . I'm sure strong Canon supporters would dispute that, and I'd respect their opinion). So, other than price, I'd rate the Pro14n and the Canon 1DS as fairly equal, but at the top of the list.
Color Rendition
Something I noticed from camera to camera was saturation level, especially if you experiment with the different color rendering intents offered by the various software. With the Pro14n and the current version of DCS Photo Desk (3.0), the saturation is a bit lower than most of the other cameras. (This is with the rendering intent set to the product look). On the flip side, some of the other cameras put out images that appear too saturated to my eye. It's certainly no big deal with any of them though, because that's something you "season to taste" in your workflow, either through custom profiling or judicious use of the hue/saturation command in Photoshop. However, I did not adjust saturation in PhotoShop for any of the test images even though most photographers would address saturation levels in a production environment.
Overall Impressions
My overall impressions of the KODAK DCS Pro14n are very, very strong.
One thing . . . Kodak is taking a lot of flak with people saying the Pro14n is noisy. Here's how I react to those comments (as a long time KODAK digital camera user, I know their software better than most):
1) The KODAK system (i.e., camera and DCS Photo Desk software together) is misunderstood. The Kodak software offers the highest degree of control in regard to noise reduction and sharpening of all the major manufacturers' dedicated software I used in this project (i.e., FUJI FinePixViewer, Canon File Viewer Utility, Nikon Capture, Nikon View 6 and Kodak DCS Photo Desk).
2) In the film days, did you ever have one film that did all things perfectly? No. What happened when you push processed a film to get a couple extra stops? Grain. Why do we have this unrealistic expectation that digital should be any different, that one camera needs to be best at everything? (Frankly, I use a different digital camera when I need to operate at extreme ISO's, just like I'd use a different film a few years ago).
3) In my opinion, the Pro14n at ISO 80 is extraordinary, and not just for VR work. The Pro-14n at ISO 400 is really, really good (especially, if you make the effort to learn how to properly use DCS Photo Desk's noise reduction tools).
4) Cheap Shots. Much of the "data" photographers have available to them in making this purchase decision is anecdotal, based on hearsay. Dealers have learned that they can bash the Kodak cameras as noisy and get away with it . . . and that's becoming the market perception. So all the competition takes pot shots, and word spreads that the Kodak camera is noisy. I'll bet none of the detractors can show you objective evidence of it, at least none with a strong working knowledge of the noise reduction features in DCS Photo Desk.
Kodak's system is terrific in my opinion, but they don't explain the software well. The great majority of the photographers I've talked to do not understand the key features of the Kodak software. And I blame KODAK for not rectifying that.
The test series I did for Kodak is highly informational, and my protocol was very impartial. I understand that some will question my objectivity because I'm a long time Kodak user, so I worked very hard to develop a protocol where no camera was favored, no subjective decisions were made, and I could let the resulting images speak for themselves. With that said, I feel the images and layouts I created are unbiased and provide much new visual data to help cut through the anecdotal information that unduly influences many photographer's knowledge of today's professional digital camera market.
I'd love to go on the road with this data to present comparison layouts and files to groups of professional photographers without trying to guide them toward any conclusions . . . sit back and let them arrive at their own decisions about the data. (Yet I'd be there to answer questions about the shoot and the image processing protocol).
TECHNICAL DATA
PANOS #1 - 4 (four resolutions of the same pan)
This pano was shot in a floral conservatory with a Nikon 20mm lens on the Pro14n. Here's technical data for the shot:
- PRO 14n set to six megapixel capture
- ISO 80
- Focus on manual
- Metering set to aperture priority at F16
- No exposure compensation,
- Available light without fill flash
- Product look rendering intent
The building was basically a greenhouse, and it was a sunny day, so the lighting ranged from direct sunlight to deep shadow.
I saved tiff files from the camera RAW files using Kodak DCS Photo Desk 3.0 set to low sharpening and medium noise reduction and no exposure compensation.
I did no retouching to any of the files with the exception of one image, which came in a little too dark because the in-camera meter read a lot of sky in that shot and reacted accordingly. So I brightened that one image a little in PhotoShop. Otherwise, I went with camera defaults and software defaults.
I stitched a master Pict file from the full res tiffs, and resized that master Pict file to obtain multiple resolutions of the final pano. I authored the panos in QuickTime VR Authoring Studio using JPEG medium compression.
There are four resolutions available on the FTP site. Here is the data for those four (my floating text increased the file size a bit, especially on "Floral(1)" . . . sorry):
NAME.............SIZE OF PICT FILE..........SIZE OF PANO
Floral(1).............11808 x 2972............. 6.7 megs
Floral(2).............3456 x 872.................944K
Floral(3).............2400 x 604.................488K
Floral(4).............1728 x 436.................276K
For those who bother to download the 6.7 meg version, you can zoom in forever.
PANO #5
This pano was shot in a DigiQuick digital portrait studio in Rochester, NY, again with a Nikon 20mm lens on the Pro14n. Here's the technical data:
- PRO 14n set to 3.4 megapixel capture
- ISO 80
- PRO 14n set to manual mode
- Focus on manual
- 1/60th sec at F11
- Product look rendering intent
- Lighting for the scene was the lighting the portrait photographer had set up for his shot. I did have one additional softbox on the photographer for the two shots in which he appeared. Otherwise, no extra lighting. Seeing deep into the shadows surprised me, as I did not light with that in mind.
I saved tiff files from the camera RAW files using Kodak DCS Photo Desk 3.0 set to low sharpening and medium noise reduction and no exposure compensation.
Stitching the 3.4 megapixel tiff files yielded a 52 MB stitched pict file. I resized that by 50% in Photoshop and ended up with a 12 MB file. I did some retouching on the 12 MB file, but not much. I authored the panos in QuickTime VR Authoring Studio using JPEG medium compression.
Part of what amazed me is the pano taken in the DigiQuick portrait studio. I shot it at 3.4 megapixels, and then reduced the pict file by 50% (which reduced the PICT file's size from 52 megs to 12 megs) and you can still zoom in to a significant degree. I think even more importantly, you can see well into the shadow detail (which I didn't really plan on with that lighting set-up). In the floral conservatory shot, I worked with default images right out of the camera. Other than the brightness of one of the twelve images, I did zero adjusting in Photoshop. Color, tonality and sharpness were great right out of the box (using Kodak DCS Photo Desk to work with the RAW files). I'm particularly impressed with the way it carries deep shadows to bright highlights as a matter of course.
FTP INFO
You can download the five panos, along with three layouts from the still life studio tests, via FTP. Regarding the three layouts, two of the files are screen res level 8 JPEGs, the third is a 16 x 20 at 300 dpi saved as a level 12 JPEG. All images in the still life tests were shot at F16. (The files I'm providing all exist as 16 x 20 Tiff files, but that's too much for my FTP service). One of the two screen res layouts shows the still life scene taken with the Pro14n at ISO80, the other taken with the Pro14n at ISO400. Even though they are screen res JPEGs, you'll get the idea. The 16 x 20 at 300 dpi is the Pro14n at ISO 80.
FTP Protocol using the following information:
HOST: ftp.stclairphoto-imaging.com
USER ID: anonymous
PASSWORD: guest
DIRECTORY: /pub/KODAK
OR via your web browser at the following address:
ftp://ftp.stclairphoto-imaging.com/pub/KODAK/
About the author:
Pat St. Clair has a bachelor?s degree in marketing from Miami University (Ohio) and a bachelor1s degree in professional photography from the Rochester Institute of Technology; he has been photographing commercially since 1978. St. Clair serves a corporate clientele that includes agencies of all sizes as well as direct corporate clients such as Eastman Kodak Company, Palm, Xerox, 3Com, DaimlerChrysler, ExxonMobil, Microwave Data Systems and more. He was an early adopter of digital technology and has worked with Eastman Kodak Company on digital capture projects and digital image quality issues for the last eight years. He has worked with QuickTime VR since 1994, is a charter member of the IQTVRA, and was a speaker at the IQTVRA Summit In the Rockies October 2002 in Boulder, Colorado.
He owns and operates St. Clair Photo-Imaging in Rochester, NY. More about Pat St. Clair and his work can be found at www.stclairphoto-imaging.com
Email: pat@stclairphoto-imaging.com
|  | |
| The purpose of this banner is to raise funds for a new VR community project VRMag will launch in a few months.
| |
