Recently, I attended a Hasselblad workshop which focused on their H-system of digital medium format cameras. During the workshop we were given access to the newest H4D cameras and a comprehensive set of lenses. Two of the workshop organizers, Oliver Klink and Stephen Johnson, were nature photographers who personally shot with the Hasselblad. We also had a very knowledgable staff member from Hasselblad, Eric Peterson, helping us as well.
My main motivation was to see if using medium format digital is something that I’d like to consider for my own photography which has primarily been with 35mm film and digital cameras. As you can see from my website, I typically shoot travel images, although I have also done a little portrait work. Generally during the day I shoot handheld when there is enough light (I consider IS/VR a godsend) and with tripods for surise and sunset.
The workshop occured over two weeks with a weekend at Point Reyes National Seashore photographing and a review where we looked at images printed in size from 21″x28″ to 30″x40″. I also tested out the camera with a few portrait pictures separately.
Hasselblad has three variations of the H4D camera which are refered to as the 40, 50, and 60 depending on the number of megapixels in the back. At the workshop, I used the H4D-40 which has a 40MP sensor made by Kodak yielding a humongous digital file 7304×5478 pixels in size. This is enough to make a 40″x30″ print at 180dpi without any uprezzing. A few other people in the workshop were using the 50MP back where the sensor is made by Dalsa.
Cameras in the H-system are modular which means that you can take it competely apart into separate components for the body, digital back, and viewfinder. Note that although the system is “modular”, the backs are not easily interchangeable as they need to be calibrated for a specific body.
One advantage of the digital back is that the entire sensor is accessible making them very easy to clean. However, by the same token, since the sensors are completely exposed and are vulnerable to scratching. In our workshop, we found that two of the backs had huge vertical scratches in the sensor ruining the images (luckily the folks found out at the beginning). Since the scratches were located in roughly the same spot, we speculated that they occurred during the mounting of the back.
We had access to almost every Hasselblad lens available for the H-system including primes ranging from 35mm to 300mm, two zooms (35-90 and 50-110), and the HTS 1.5 tilt-shift adapter. The first thing you notice about the Hasselblad lenses is their physical size: they are enormous in comparison to 35mm both in terms of size and weight. For example, the Hasselblad 35-90mm zoom is over double the weight of my canon 24-105mm f/4 zoom and is 50% longer. See the table below for some lens size comparisons.
|Hasselblad||80mm f/2.8||3.3″||2.8″||1.0 lbs||$2100|
|Canon||50mm f/1.4||2.9″||2″||0.64 lbs||$370|
|Hasselblad||35-90mm f/4-5.6||4.0″||6.6″||3.1 lbs||$7200|
|Canon||24-105mm f/4||3.3″||4.2″||1.5 lbs||$1060|
|Hasselblad||300mm f/4.5||n/a||n/a||4.7 lbs||$4160|
|Canon||300mm f/4||3.5″||8.7″||2.6 lbs||$1259|
In general usage, I found the Hasselblad lenses very pleasant to use. They were very sharp (none of that mush in the corners that I see with some Canon wide angles), had little distortion, and gave a very warm color rendition.
Hasselblad lenses also have built into them leaf shutters which allow for flash sync at any shutter speed. This is useful if you want to use flash outdoors and overpower sunlight but it is not really a concern for my photography.
For static subjects, I thought the autofocus on the Hasselblad camera was comparable to my Canon 5D mark II. The Hasselblad has only one autofocus spot located in the center of the image so you have to focus and then recompose. However, I use my Canon in the same manner with the center AF point (I don’t find the additional AF points as reliable).
Speedwise, I did not notice any differences from the 5D in terms of autofocus but it’s possible that the Canon could be significantly quicker. I did not perform extensive side-by-side testing, and without that it would be difficult to make a definitive statement about their relative autofocus strengths (other than they seem similar at first glance).
With the H4D, Hasselblad instituted a focus adjustment called “True Focus” whose purpose is to adjust for the slight differences in subject to sensor distances when you focus and recompose. I used true focus for all my shooting simply because the true-focus button doubled as an AF-on button on the back of the camera. For the most part, I don’t think it makes a difference unless you are shooting portraits wide-open.
The only time I had difficulty with the autofocus was late a night after the sun had gone down. While I was shooting the docks and water in the image below, I could only achieve focus on the window and lights in the middle of the image. The AF would not lock on the dock in the foreground. In fairness, at this time it was very dim and the exposure was over a minute at ISO 100.
Using the Hasselblad meant shooting almost everything from a tripod. Handheld shots were possible but it was difficult to keep the camera steady enough for sharp images unless there was a lot of light. In the workshop, I only have a few shots that were handheld that I kept such as the picture of the cow below.
I think this may be a problem for general shooting, especially for available light portraits that I like to do handheld. For example, in the picture below I needed to use the tripod to keep everything sharp. Normally with my Canon’s I wouldn’t hesitate to shoot this handheld with IS to ensure sharpness (I would probably increase the ISO and open up the lens somewhat).
I also discovered that my tripod setup using an Acratech Ultimate Ballhead, and a Velbon el Carmagne 640 carbon fiber tripod was borderline for the H4D. I would get creep if I didn’t tighten the ballhead as much as possible. Walking around with the camera and tripod over my shoulders was a big no-no and the ballhead would slip to point where the camera would be pointed straight down.
One big benefit of the H4D is the larger viewfinder which made it easier to compose. When I shoot with my DSLR I often have problems with the horizon coming out at an angle (even when shooting off a tripod). but with the H4D almost everything came out level.
As mentioned previously, the camera is very heavy especially if you are used to holding lighter DSLRs. However, with the exception of weight the camera fits very nicely in my hands and I found the controls easily accessible.
Navigating the camera menus is a different matter and I spent a significant amount of time setting up the camera even with the help of the Hasselblad representative. For example, it probably took me about half an hour to figure out how to set the self timer with a 3 second delay between mirror-up and the shutter firing (because I didn’t have a cable release). Perhaps the most confusing aspect was that I was never sure when a setting would be saved permanently, reset after a single shot, or reset when the camera was powered down.
The H4D has a profile system where you can quickly switch between a group of settings. For example, a landscape profile might have low ISO, aperture priority and mirror-up whereas an action setting might have higher ISO and no mirror-up. Although I set up profiles initially, I did not use them and found it more convenient to simply change the settings manually.
The LCD display on the back was acceptable but not great — for the price of the camera you would think that they would put in a higher quality display. On the LCD, you can zoom in to check sharpness and see most of the information that you would expect. The only thing that I felt was missing was a color histogram which I find useful when photographing highly saturated subjects such as flowers to ensure that none of the color channels are clipped.
The H4D-40 produces raw files (with a .3FR extension) that are 50-80MB in size depending on the amount of compression. However, once you convert them to 16bit tiffs you get a massive 240MB file. These files are considerably slower to work with on my computer setup (MacPro with 10 GB of RAM) then my 5DII raws.
Hasselblad also has their own software to process the images called Phocus. One of the advantages of Phocus is that it uses parameters specific to each lens, not just the model as a whole, to fix issues such as distortion, vignetting, and chromatic aberration. However, I did not have much experience with Phocus and I found it painfully slow to work with especially for editing. In the end, I decided to use Lightroom to edit and process my images with the idea that I would reprocess my selects with Phocus.
Image quality is the main reason for getting the Hasselblad and it did not dissapoint. With the 40MP back you get a file which supports huge prints at high dpi settings. My initial impressions of the files were that (1) they were very sharp on a per pixel basis, possibly because there is no anti-aliasing filter over the sensor, and (2) the colors were pleasing to my eye and didn’t need much processing or alterations.
I think the lenses also contributed significantly to the image quality. With my Canon DSLRs the first thing I do with a lens is test it to check to see how it performs in terms of sharpness. I have sometimes been dissapointed with the quality and even bought multiple copies of one lens (17-40L) trying to get a better version. In contrast, the Hasselblad lenses all seemed very good even in the extreme corners.
Normally when I test image quality, I begin with static architectural subjects for reproducibility and consistency. However, I didn’t have this option at Point Reyes so instead I will show below a few crops at 100% magnification to give you an idea of what the camera system can produce. For the crops, the images were processed in Lightroom with default sharpening settings (amount = 25, radius = 1.0, detail = 25, masking = 0) to produce to TIFF files. These were then converted to sRGB and saved as jpegs in photoshop with quality level 9.
Figure 5 shows a crop taken from the center of the frame from my portrait picture of Victor Stangenberg (Figure 4). If you click on the image to see it at 100% magnification). Looking at the unresized pixels, I am impressed by how sharp they appear.
Figures 6 through 8 shows a picture taken at Limantour beach for a flowering bush with 100% crops in the center (Figure 7) and bottom left corner (Figure 8). The image is still very sharp in the extreme corner.
Figure 9 shows an image of a cow taken Historic E. Ranch. Figure 10 is a 100% crop of the eye area which is very sharp and you can see individual eyelashes. There are also specular highlights just beneath the eye. Figure 11 shows a 100% crop of the hair on the cows head just to the right of the eye. This image is the only one where I can see some undesireable artifacts in the hair. However, I doubt these would show up in a print and it may be that processing with Phocus would eliminate them.
Thoughts on Image Resolution
In terms of the raw number of pixels, the Hasselblad’s greater resolution will let one to make prints one size larger than with my Canon DSLR system. For example,
- The Hasselblad produces a 7304 x5478 pixel image which at 180 DPI yields a 40″x30″ print
- The Canon 5D II produces a 5616 x 3744 pixel image which at 187 DPI yields a 30″x20″ print
Although this doesn’t sound like much with a linear increase of only 33% in the long dimension, after looking at the larger prints I can say that visual impact is much greater (at least for myself).
With 35mm digital systems, I feel like I am coming up against a limit and further increases in sensor MP will only result in small increases in print detail and quality. When I moved from the Canon 5D with 12MP to the 5DII with 21MP the detail improved but not by as much as you would expect based on the number of megapixels alone. I think this is because with a fixed size sensor size, as we increase the number of MP the lenses and shooting technique become the limiting factor.
Another alternative to get this resolution with DSLRs is to make digital composite pictures by stitching images together. For example, with a tilt-shift lens I can stitch three Canon shots together and get a 40MP final image which is basically the same size as the H4D capture. With a panorama head I can make even bigger captures. While this works wonderfully, there are many drawbacks and it is not appropriate for all shooting situations such as when there is subject movement or rapidily changing lighting conditions. Furthermore, stitching requires more post-processing work which I’d like to avoid if possible.
For the most part, I didn’t do much testing of the high ISO capability of the back. Generally the workshop organizers discouraged us from using the higher ISO settings on the camera and I didn’t see much reason to disagree given that we were shooting from tripods. However, I did make a few images at 800 ISO using the 50MP back as shown below. As you can see, the Lightroom processed image shows considerable noise, much more than my Canon 5DII would produce. Processing the same image in Phocus, Hasselblad’s own RAW processor, yields a much cleaner image although there appears to be a slight loss of detail.
I have no kind words for the battery life on the H4D as it was simply pathetic. For example, one morning during the workshop we shot the sunrise of Drakes Estero and after an hour and half people were already running out of battery power. In fairness, there may be a couple mitigating reasons why the batteries dropped so low so quickly:
- The batteries we had were new and I was told that they need a few charges to reach full capacity.
- The power meter on the camera may not have been calibrated properly against the battery. For example after fully charging a battery and checking capacity in the camera I would get varying readings that were as low as 85%.
- People were not fully comfortable with the camera and may have been shooting and reviewing more than they typically would with their personal camera systems.
Using the Hasselblad has definitely peaked my interested in medium format digital cameras although I am not sure that the H4D is the right choice for me. In terms of image quality and resolution the H4D delivers without question, but the camera would introduce significant logistical problems related to its weight (portability) and cost (an H4D-40 kit with one lens is about $20k). I also haven’t yet explored other MF options such as the newly released Pentax 645D which may be better suited for my photography.
You can see more images taken with the Hasselblad in my Point Reyes gallery. The H4D pictures are those with a 4 to 3 aspect ratio.