UPDATED: So, as Johnny Danger pointed out in the comments below, I didn’t upsample the 5D to match the native resolution of H3Dii in the original test article. I have added two new slides to illustrate this and it really shows how much better the H3Dii is compared to the Canon 5D. I’m still amazed that this little blog post gets about 100 unique visitors a day and is the #3 link when you do a Google search for h3dii.
As you may have followed in my Daily Photoblog, I’ve been testing out the new Hasselblad H3Dii 39.1MP Back. The camera is just awesome and really fun to shoot with after being in 35mm format DSLR for the past 4 years. And the files I pulled off it were amazing. The image quality, sharpness, and dynamic range just blew my mind. But is it worth $30,000? Better yet, is it worth a $500/day rental versus a $150/day rental for Canon 5D?
Why I did this test:
My goal is to provide the best photographic image for my client and I made this to show the difference. Feel free to reference to it, but please link back to it with full credit line, ©Lincoln Barbour - www.lincolnbarbour.com
Control:
In order to get results that made sense to me and my workflow, I converted the H3Dii raw file in FlexColor to DNG. My Canon files are always converted to DNG on import. I processed both images in Adobe Camera Raw 4.2 using the same default exposure settings, sharpness set to zero, noise reduction set to 10. I did change the color temperature so they would match better and be a little more neutral. Most of my color issues are dealt with in post, so I wasn’t too concerned about the native color of each camera.
On the Canon 5D, I used the 50mm 1.4 at f/5.6. On the Hasselblad HD3ii, I used the 80mm 2.8 at f/8. These lenses are equivalent in focal length and both are good quality lenses. The reason for a smaller aperture on the H3Dii was to get same amount of sharpness on a larger image area. I’m not sure if one stop is mathematically correct, but it’s close enough. Both cameras were on 100 ISO. My shutter speeds varied due to the natural light conditions (it was a partly cloudy day, so light levels were constantly changing).
Test 1: Image Size
As you can see, the bigger sensor creates a bigger image. Kind of a “no duh!”, but I wanted to show it visually. The 5D image size is 9.707″x14.56″ @ 300dpi and the H3Dii is 18.04″x24.053″ at 300dpi. This example is a scaled down version of the two images at the same resolution.
Test 2: Resolution
Cropping in at full resolution, this is what we get. It’s really hard to compare the two, other than “Holy crap! That spout is huge!” But, you can tell how much more detail the H3Dii gets pixel per pixel. The more information you have, the better your final output will be.
Test 3: Sharpness & Detail
To compare apples to apples, I reprocessed the H3Dii file as a 12.6MP to match the 5D’s native output. I then also processed the 5D files at 25.2 MP and upsampled it (PS bicubic smoother) to 39.1 MP to match the H3Dii’s native resolution. As you can see, the H3D has a lot more detail and also appears sharper. Towards the bottom of the image, you can really see the difference by examining the fine powder specs in the metal. The specs are much more clearly defined and you see more of them. The 5D starts to mush things together.
Test 4: Shadow Detail
This is a tougher test to see at first, but the 5D is clipping to black faster than the H3Dii. In other words, there’s more shadow information in the H3Dii file. Shadow detail is so important and though the 5D does do a good job, the H3Dii gives you just a little more push and pull. Plus it’s a smoother transition to pure black, which makes a huge difference when it’s printed.
Conclusions:
I’m definitely sold on the H3Dii. For me, it comes down to three things: Size, Detail, and Information. Cost isn’t really a factor, because my clients deserve the best and I want to give them the best. Though it’s hefty price tag, the H3Dii is really worth the extra money. Especially if you look at it from a rental fee point of view. You get such better quality images for an amount of money that is relatively small to entire project budget.
The 5D is awesome camera and I will still use it for some editorial assignments and personal work. You really don’t need 39MP when you photo is going to run 1/4 page. Plus, walking around decaying urban areas with a $34,000 camera and $10,000 in lenses can be a little nerve racking. 
I hope you found this informative. I’m happy to answer any questions you may have about the camera or my test set up. I’ve also created a web forum for Hasselblad Professional Photographers called HASSELBLADusers.com. Check it out!
-lincoln
Entries (RSS)
October 26th, 2007 at 11:43 pm
I’ve used a digital back on my old Hasselblad CM and it has produced equally impressive results at a more moderate cost (£4,000 for a refurbished back). Wrote a blog about it a while back. The H3D is an awsome machine and I will hire it for the right jobs.
Thanks for all the time and effort in putting this article together.
Cheers,
Paul
October 27th, 2007 at 12:10 am
Very informative, it would interesting to se how 1DS Mark III would do the job. Thank you.
Sincerely,
Jannik
October 31st, 2007 at 1:39 am
The one thing this test demosntarted to me more than anything, is how good a camera the 5D really is. For a $3500 AUD camera body, the 5D is a superb piece of gear, as you mentioned but really needs to be compared to likes of a 1DSMkII or III.
I think the assumption that this camera is worth the money (hire or buy) is one we can make without hesitation. In order to really prove it’s worth, we should be comparing it to the likes of a Phase One’s backs. At $45,000 AUD for a Hassleblad H3DII body kit (no lens) V $50,000+ AUD for a P45+ back only, you would then be comparing apples with apples and then possibly proving the worth of the Hassleblad. Although I think the reverse may ensue would be a demonstartion as to why one would consider buying the Phase One technology as opposed to Hassleblad (Imacon) :0)
October 31st, 2007 at 2:10 am
Hi Peter,
I agree with you, the 5D is awesome camera. I shot with it today, in fact.
This comparison test I did was for my own purposes and I thought I’d share it with the world. It’s been a pretty big hit as far as I can tell. I’m getting 30 to 40 visitors a day reading it. It wasn’t meant to be a guide, just my own impressions of the two cameras.
It wasn’t the sensor that impressed me with the H3Dii. It was the lenses and the firmware of the back working together to fix problems like moire, chromatic aberration, and so forth. In other tests I made (not pictured), I shot the H3Dii with the Hasselblad 28mm. It was incredibly sharp with no obvious chromatic aberration and beautiful highlight tones. Way better than anything I’ve shot with the 5D or the 1Ds Mark II.
I shoot a lot of architecture and H3Dii will do a really good job for my needs. I’m hoping Hasselblad will develop a tilt-shift version of the 28mm and maybe a tilt-shift 35mm. That would be awesome.
I’m sure the Phase One is a great back and you would know since your company is a dealer. I would love to see some wide angle shots taken with it to compare to the H3Dii. I’m not loyal to any brand, just to the best image.
I’m happy to test it for you if you want to send me one. Cheers.
-lincoln
October 31st, 2007 at 2:14 am
Also, I will be doing another side by side in the future. I’m going to shoot an architecture job with the H3Dii and the 1Ds Mark III and see what all the fuss is about.
November 5th, 2007 at 9:34 pm
very interresting , did you read the test on luminous-lanscape P 45+ & 1Ds Mark III ?
thanks
erick
November 6th, 2007 at 7:11 am
“Also, I will be doing another side by side in the future. I’m going to shoot an architecture job with the H3Dii and the 1Ds Mark III and see what all the fuss is about”
Architecture ? Perspective control ?
while you can use Tilt and shift lens with the canon (24+45+90+ x1,4M), there’s no substitute in the hassy arsenal… then the Arca R m3d or Alpa XY make sense…
November 6th, 2007 at 5:44 pm
Yeah, that’s nice thing about the H3Dii back. You can mount it to a view camera to get perspective control back when you need it. I have a Horseman 450 for this very reason.
I’m betting Hasselblad will develop a series of tilt-shift lenses soon.
November 6th, 2007 at 11:25 pm
I couldn’t understand some parts of this article Canon 5D vs. Hasselblad H3Dii - A Real World Test, but I guess I just need to check some more resources regarding this, because it sounds interesting.
November 10th, 2007 at 6:20 pm
Hi. I’m very curious about how well the H3DII would perform in the studio against the 1Ds Mark III shooting tabletop while using hot lights.
I’m currently using a 1Ds MK II.
Thanks,
Jesse
February 3rd, 2008 at 1:59 pm
Hi,
I have a couple of Hasselblad 500c’s and have been wondering
weather I can use the backs for the V system on them and if
not then I would appreciate if you can suggest some other cost effective digital back alternative?
Another thing is that I too have been looking into getting a digital SLR and have been contemplating between a Nikon D200 and Canon 30D. Which one would you recommend?
Appreciate your help and the time you put into helping the likes of us out.
Thanking you sincerely,
Abbas
February 5th, 2008 at 4:36 pm
Hey Abbas,
You’re probably better off with a CF or CFV back than the H3 back. Not sure on what your budget is though. A CFV back will set you back $9000 USD
You should read/join my Hasselblad Users forum http://www.hasselbladusers.com Great info is available there.
As for the D200 vs. the 30D, I would say they’re about the same. Both are great and both have their drawbacks. If you have any lenses for either camera, I would stick with that brand though.
Thanks for commenting on my blog.
All the best,
Lincoln
March 14th, 2008 at 4:29 am
Hi,
thank you so much for sharing this. The 5d really is worth its money.
March 21st, 2008 at 3:56 pm
Thank you for this excellent comparison. I am amazed how well the Canon 5D compared, bearing in mind the massive price differential. I’ve got the money for the Canon 5D Mk2 waiting in my bank account, can’t wait till this camera comes out!
March 22nd, 2008 at 1:05 am
Is it just me or does this test lead readers to believe (incorrectly) that the 5D resolution is nearly as good as an HD3ii? Sizing the HD3ii DOWN to 12.6 megapixels is… kinda bassackwards IMO. Try sizing the 5D UP to 39.1 megapixel size and then the detail differences will become painfully obvious. Granted the 5D is a great camera, but 12.8 megapixels just isn’t that close to 39.1 megapixels no matter which way you cut it.
Why would people shoot the HD3ii? Most likely because they need 39.1 megapixels for unmatched detail in large prints, not because they want to make a nice image and size it down to 12.6 megapixels.
HD3ii looks great obviously. 5D is great too. Side by side… apples and oranges.
March 22nd, 2008 at 1:13 am
After re-reading my post, I realize that it may come across a bit trite. I’d like to thank Lincoln for producing the test as I’m always interested in pixel-peeping, but I guess I just don’t see the value of the comparison in the way it was presented. When portability and smaller files are required obviously the 5D will make do just fine, but when you need a 39.1 megapixel smoking camera… well, we all know where to go. Let’s see both results at 39.1 megapixel crops to get a gauge of the large printing abilities of both.
March 22nd, 2008 at 7:51 am
Re: Johnny Danger,
You know, I totally agree with you. Pixel to pixel, the two cameras compete. But with the H3Dii, you get 27 more MP of info. That’s what I was trying to ge across originally. When I have the time, I’ll post upsampled 5D crops of the comparrisons.
Thanks for your feedback!
-lincoln
April 29th, 2008 at 4:22 pm
Thanks for putting the time in doing this test, I my self are in the same situation for different reason, I just want the best image quality money can buy.
But when it comes to quality we all know that beside the obvious personal talent as a photographer is a must as no great gear makes a great photographer, but in deed great gear makes a great photographer, well … greater!
I’m familiar only with Canon, and have owned and used extensively the 5D, 1D IIn and the new 1D III, I only have “L” series lenses, and use mostly Primes.
Now in the market for a new camera, to be the 1Ds III or the H3D II.
I believe that each camera should work better with the lenses which were created for it, but there are obviously exceptions, they always are.
I think that for me there will be no other alternative before I make this next Photo gear purchase, but to rent out both the Canon 1Ds III and the Hassie H3D II with comparable lenses, shoot away on the same subjects and exact lighting, and see the results at full res on each camera, and each file processed first on the same application, and secondly process the Hasselblad image on the proprietary Fhocus, which claims wonders.
Thanks again for the post.
Ciao,
KETCH ROSSI
May 1st, 2008 at 9:04 am
Good article. I’m also waiting with money burning a hole in my pocket for the 5D Mark II.
I wonder what the results would be (shadow detail, resolution, sharpness) if, rather than upsampling the 5D image, several photos were taken with the 5D so that the resulting file was the same size as a Hasselblad file (a bit like stitching a panorama). A different lens would have to be used, or subject-to-camera distance adjusted, but wouldn’t this would be a truer comparison of Canon pixels versus Hasselblad pixels?
September 30th, 2008 at 12:05 am
Of course the Hasselblad has sharper photos. It is a simple matter of the area of the sensor. The Hasselbad H3Dii has a sensor measuring 48mm x 36mm =1728mmsq area compared to the Canon 5D 36mm x 24mm = 864mmsq, that is half the area for the canon. It means that the same quality lens can create a large image on the larger sensor giving the effect of a considerably sharper image. This is recognised by photographers who have moved up from APC-s sensor to full frame sensors. When 35mm was introduced it was introduced for its convenience and quality took back seat, a bit like reel to reel tape to cassette audio tape recorder. On my Canon 5D using EF50 f1.4, the same as used in the test, can’t produce fine detail that hasn’t lost considerable amount of contrast. It will be only a matter of time before Canon and Nikon will be forced to manufacture a medium or large format digital camera.