Erik's web

  • Increase font size
  • Default font size
  • Decrease font size

My medium format digital journey

E-mail Print PDF
Article Index
My medium format digital journey
raw images for download
Where do I stand now
The experience
Color Profiles
Resolution Tests
Real World Samples
Color rendition
Dynamic Range
Moiré & aliasing
DoF and diffraction
More real world images
Three format comparison
Diffractio series
The lenses
Fcousing
The Bad News
Capture One or LR 4.4
Some comments on Phase One business model
The Ultimate Platform?
All Pages

I recently acquired an MF Digital equipment, it arrived yesterday and now I write a blog about about my experience. I just started writing this, the stuff arrived yesterday (2013-06-14). Much more info will be added later on.


Updates:

2013-06-15: Initial posting
2013-06-16: Added sample images
2013-06-17: Added 1:1 crops and some discussion on detail
2013-06-18: Preliminary color evaluation using Imatest Colorcheck added
2013-06-19: Some artifact examples added
2013-06-19: DoF vs. Diffraction discussed
2013-06-19: Dynamic Range example added
2013-06-19: I go abroad for a few days - don't expect new postings for a week.
2013-06-24: Comment added on dynamic range (better dark noise with C1)
2013-06-24: Three format comparison
2013-06-24: Real world samples
2013-06-25: Added a Capture One processed version of the fountain image
2013-06-28: Added some more images 
2013-06-28: Capture One or LR 4.4?
2013-06-28: Some comments on Phase One business model
2013-06-29: Diffraction series, some results
2013-07-03: Initial lens samples added
2013-07-04: More lens samples added
2013-07-08: New section added, where do I stand now?
2013-07-16: Alternative tools for focusing added
2013-07-21: Fixed som errors on part regarding focusing tools
2013-07-24: Bought an Distagon 40/4 FLE (non IF) on Ebay. Corresponds to 28 mm on FX.


A short summary (updated 2013-07-04)


  • The Hasselblad 555ELD and PhaseOne P45+ combo is workable, if you can live with slow work.
  • The Hasselbald lenses are OK.
  • Manual focus on "Acutematt" screen seems to work, at least with Hartblei 4X focusing hood.
  • Split image works reasonable well.
  • Waist level shooting is not for me!
  • Good sharpness, but I don't see the cutting edge sharpness I have expected. My impression of sharpness improved much after recent shooting.
  • Moiré is much more frequent than suggested, but not as much a problem as I expected.
  • Reproduction of greens may be oversaturated and have a yellow shift. May be a problem for me.
  • Nice to work with a classical camera.
  • The aspect ratio is mostly helpful.
  • I acquired a PM5 viewfinder and use it with a Zeiss Victory 3x12 monocular as a focusing aid, this is the solution that works best for me.
In general, the Hasselblad 555ELD / P45+ is a nice experience, it is like a classic device. Regarding image quality I still don't know. It seems that the Hassy / P45+ has a real resolution (MTF) advantage over the Sony equipment I use. Regarding other aspects of image quality I simply don't know. 



Raw images

A set of raw images are here: http://echophoto.dnsalias.net/ekr/Articles/MFDJourney/RawImages/Samples/


More will be added later. Right click on image or download link for raw image.

Background

I have had a long interest for medium format photography, I was shooting 67 on Velvia for a long time. Like many other I converted to digital around 2005 and never looked back. MF digital was always around but has always been horribly expensive. 

Also, I have also for long been interested in lenses. I am the kind of person who is highly skeptical about myths. There are a lot of myths about Zeiss lenses so I decided to test. I had a couple of Sony zooms, labeled Zeiss and I compared them to older Minolta and Sony lenses. I failed to see the Zeiss effects.

After that I considered to buy a Hasselblad lens, just to find out. After looking at MTF charts I decided on the Sonnar 150/4, that is about the best of the old Hasselbald lenses according to Zeiss MTF curves. Well, the 180/4 is a tiny bit better, but I felt I had more utility for a 150/4, so I got one. Gosh, that lens was good! No, it does not outshine mine Sony zoom lenses when mounted on a Sony full frame camera, but you really see the potential of such a good lens on full frame MF. Actually, the Sonnar 150/4 is a tiny bit better than the othr lenses I tested, but the difference is small.

I had some interest in MF digital fo a long time, and I always felt that I would go for it if I could assemble an equipment for around 100 kSEK (Swedish Crowns).  Lenses and stuff for Hasselblad V series is now very affordable, and I recently found a P45+ for V-series at decent price, so I went for it.

The lenses were really the cause I went for the V-series, H-series is very nice, I guess, but the lenses are quite expensive, so camera back and lenses would probably not fit my 100 kSEK limit.

Folks involved

Originally, I was hooked on an offer from "Mr. Rib" for an Aptus back on a Hasselblad H2. The same seller also had a P45+ in V-mount. He is based in EU, making stuff easy. I was tempted by this, but I asked my friend (I think I may say) Stefan Steib of Hartblei fame, for his opinion, and he strongly suggested the P45+, especially if I intended to go with the Hartblei HCam later. 

So I started looking at V series bodies. I had very good experience with ClassicCamera in Gothenburg, where I acquired several of my Zeiss lenses, and he had a very nice Hasselblad 553EX. On the other hand Mr. Rib suggested that a 555ELD he could arrange was a better option, being made for digital. So I bought the body from Mr. Rib.

I also put an order on a Hasselblad 50/4 FLE and a an 80/2.8 CFE from Classic Camera. So I now have four lenses:

  • 50/4 FLE
  • 80/2.8 CF
  • 120/4 Macro Planar CF
  • 150/4 CF


Why do I need two very similar lenses, the 120/ and the 150/4? Well the 150/4 is a very good lens. The 120/4 is for close range, at long distances it has a lot of field curvature. That said it is a very nice lens, I love it! 

Stefan Steib suggested using the Hartblei focusing hood, a proposal I gladly expected as I had no pleasant experience of early Hasselblad viewfinders.



Where do I stand now?


Actually, I don't know where I stand. 

  • Working with the Hasselblad is fun, it is a classic.
  • Focusing is hard. It feels like the camera is in decent register. Working hard to learn better focusing. 
  • Hartblei loupe is good for focusing, but PM-5 prism is much more comfortable. 
  • With the DSLR I take one picture, because I know what I do, with the Hasselblad I make many exposures. Are the results better? Hmmm...
  • Working with fixed focals is a compromise, possibly a good one, not sure about that.
  • The P45+ can play it's advantage, if everything is OK, but the full frame DSLR I also carry makes a tremendous job. No walkover victory for MFDB.
  • I struggle a bit with colors on the P45+, oversaturated and yellowish greens? Sony much more accurate.
  • I shoot landscape and no studio portrait so I do know little about skin tones.
  • I had some negative experience with C1, and had seen a good demonstration of it's advantages.
Some things that keep me confused:

  • I built a DNG profile for the P45+ back, but it seems that I mostly prefer the Adobe Standard profile.
  • I have a Zeiss 120/4 Macro Planar, a lens everyone seems to love, but also a lens that has about the worst MTF curves seen, at infinity. Close up it is quite good. My experience is similar to the MTF curves. Still working on it.
A short comment on lenses:

  • Sonnar 150/4 is the champ, sharp and crisp over the field even at full aperture!
  • Planar 80/2.8 CF and Distagon 50/4 CF (FLE) are very good at center and also very good in corners at f/8 or so.
  • The Planar 120/4 Macro? I don't know. Right now I don't feel it does match my 70-400/4-5.6G lens used on the DSLR, but it may be very good at close up distances.Of course, it is possible that I have a bad sample. On the other hand the MTF curves indicate a clear weakness, and that is what I can see.
  • I need to stop down for depth of field! Easy to forget...




The experience

It is a very nice camera. In a way it is not that much different than using a DSLR on a tripod. I have no zooms for the Hasselblad and that means moving around the tripod a lot more than with the DSLR and the zooms. Other than that it is not that different. I focus both manually, the DSLR using live view at actual pixels and the Hasselblad using my Hartblei focusing hood, which is workable. 

Using essentially waist level viewing makes the tripod quite low and a bit top heavy, as I have both a leveling plate and a pretty heavy Arca Swiss D4 head. It seems I don't need a light meter, just guess the exposure and adjust until I get a decent histogram, pretty similar to the DSLR experience. Still, using the DSLR is less work, of course. The DSLR kit (10.5 kg)  seems to be more heavy than the Hasselblad kit (8.5 kg), but DSLR kit offers more flexibility.

Using classical equipment and the nice haptics of the Hasselblad is kind of a pleasure. 

Putting the stuff into use

I needed to sort out some things but I was up and shooting within a hour. 

One thing I missed was a viewfinder mask, but I made my own, you can download it here, print it on overhead film, cout out and insert as suggested in the P45+ manual. 



Color profiles

I normally use Lightroom 4 for handling my images. I have acquired Capture 1 from Phase One, too. Capture 1 is a fine program for sure, but I am not fond of it. That said, it seems that Capture 1 produces less aliasing artifacts on P45+ images than LR 4.4 does. Colors out of Lightroom 4.4 are not horrible but not really acceptable. Shooting a Color Checker Passport with dual illuminants and generating a DNG profile resolves that problem.

In the cases below the Adobe standard profiles are to the left and the profile I generated with Adobe DNG Profile Editor is to right. The profile used is the only difference between the two images.








Resolution test

Note! I corrected an error, images are OK but raw files not updated yet.


Right now I am somewhat confined by pollen allergi and a bad cold, so outdoor shooting is not on todays schedule, but I made some resolution tests.

The way I dis this was to assume a given print size, 50x70 cm and a printer resolution of 360 PPI. This is not entirely arbitrary chosen, as Epson printers normally have 360 PPI as native resolution. 

A complication is that  DSLRs and MFDB have different aspect ratio. The images here were produced this way:


The pair on the left uses the upper edge of the CC passport and the top of the flower as boundary. This image utilizes the height of the P45+ sensor fully. Also, in the latter case focus was on the flower.

The pair on the right was using the same horisontal crop, ignoring the wide dimension. The image on top is P45+ and bottom is Sony Alpha 99. The focus in this case is not on the flower itself, but on a resoltion taget outside the crop. Alignment error between focus target and flower is about two cm at 1.8 m.

Click on the image to see the full size crop! You may need to click once more to see actual pixels in the browser!



The resolution target crops below may be as informative:

Click on the image to see the full size crop! You may need to click once more to see actual pixels in the browser!

Short discussion

No doubt that P45+ outresolves the Sony Alpha 99. The advantage of the P45+ may be smaller than expected, however. The resolution difference is also depending on the crop ratio to be used. If we use all the 135 format the difference get smaller than if we can use all the vertical pixels of the P45+.

It is interesting to note that the P45+ image sharpens really well in LR4.4. I can use similar settings to the Alpha 99.

Also note that the resolution target has aliasing artifacts on the grey scale.


Real world sample images


Castle

Raw images are here (DNG only): chophoto.dnsalias.net/ekr/Articles/MFDJourney/RawImages/Castle/

I was shooting some real world samples with the Hasselblad and my Sony Alpha 99. Light was changing rapidly, so no comparison, but the P45+ images came out on top. Top P45+ image,  bottom Alpha 99. 





The image below is probably more compatible with the Alpha shot as lighting was more similar. 




Color seems to be better on P45+, cannot explain why, need to look more into it. Light was changing rapidly, I forgot to make white balance shots. Anyway, the P45+ images came out better, at least this time.

Looking for details, I made some side by crops from the latest two images, which both were made unde cloudy conditions. P45+ is on the left and Sony on the right.





At infinity we see that the P45+ image is not sharp at the pixel level. Image was probably shot at f/16 the Sony image was also at f/16. 


As a comparison of two systems this was very sloppy, shooting in the field with two different system doesn't really help careful and elaborate testing.


Color rendition

Latest shot I did I also shot a color checker, mostly for gray balance. Today I run Imatests Colorcheck utility on my color checker shots. I used defaults expect exposure correction and adjusting saturation. The images were adjusted for correct exposure and 100% saturation according to Imatest. On LR I used my own DNG profile.





It seems that for the color checker and under these conditions, that is bright midday sunlight, LR produces more correct color than C1, using my DNG profile. The Sony gives considerable more exact color.


Now, I am fully aware that correct color is not the same as pleasant color and also that color can be tweaked. 



Dynamic Range


I looked at dynamic range, comparing my P45+ to my Sony Alpha 99. Sony Alpha is much better, but I seldom find DR a limiting factor. Click on image below for full size. The Sony image has much cleaner shadows. Both images are exposed to the right, with the P45+ image having some clipping in the clouds. The Sony image is not clipped, see RawAnalyzer data below image.



This an actual pixels view, the P45+ has 62% more pixels, downsampling to 24MP resolution would improve DR, but very little.
NOTE 2013-06-24: The impulse noise visible in the LR-converted P45+ image is not visible in the Capture One converted image. Will look more into this.

Note 2013-06-25: It has been pointed out that image on the left is not sharp. The reason for this is that depth of field is limited, due to using a longer focal length on the Hasselblad P45+ combo. So this depends on focusing distance and aperture combination. 




Raw analyzer data for P45+ below:


Raw Analyzer data for Alpha 99 below:






Moiré and aliasing artifacts

The P45+ does not have an OLP (Optical Low Pass) filter like DSLRs normally have. The sensor has relatively large pixels, 6.8 micron pitch. The combination of large pixels, good lenses and no OLP filtering is expected to produce aliasing artifacts and color Moiré. Regarding aliasing it may be not to bad, but I have seen some color moiré on most of my test pictures. It seems that Capture 1 yields less moiré than Lightroom 4.4 and both have quite efficient tools to reduce color moiré.

Here the image on top is converted by LR4 and the bottom with C1.



The feather image


The image below was shot with the P45, Sonnar 150/4 about 3.5m distance. The image contains a lot of aliasing artifacts, except color moiré it has a lot of discontinous strain and downward bent strains on the right side. Note also a hatch pattern at the bottom. I compared the P45+ image with two images shot with my Sony, also with 150 mm and at the same distance and same light. The Sonys produce a much cleaner image, due to higher resolution (because denser pixel pitch) and OLP filtering. 

Comparison below shows P45+ to the left and the Alpha 77 to the right, would the Alpha 77 have the same sensor size as the P45+ it would have 114 MP.
P45+Alpha 77 SLT
P45 to the left, Alpha 77 to the right.

The image above essentially says that high sensor resolution with OLP-filtering is much preferable to lower resolution without OLP-filtering. Moiré reduction in LR4 or Capture 1 removes the color moiré, by local desaturation of color, but does little to the other artifacts.



DoF and diffraction


I made a series of images of a fountain at the castle with different apertures. Focus was in all cases on the head in the fountain statue. 

Here the ledt image is Sony Alpha 900 at f/8, center Hasselblad at f/32 and Hasselblad at f/11. Here we see decent sharpness in both left image from foreground to background, with the yellow flower being a bit odd in sharpness.



The second crop is Sony Alpha 99 on top, followed by Hasselblad at f/32 and Hasselblad at f/11. We see that the head, which is supposed to have focus is much sharper in the f/11 image than in the f/32 image. Stopping down increases diffraction and larger formats are no way immune to it. 


In images requiring a lot of DoF larger formats may not show a true advantage, as they use longer focal lengths to keep field of view and need more stopping down. Stopping down excessively increases diffraction.  


More real world images

I shot some real world images today, raw files are here.

Here are some examples:


The image below show some spider silk:


The above image processed in Capture 1:







The image below is taken with the 120/4, it actually shows the weakness of the lens, the edges are quite soft. This seems to depend on excessive field curvature at infinity. The lens is absolutely OK at close range. For infinity the 150/4 Sonnar is much better. The 120/4 is designed for close up range, at long distance it has a curved field. Also note that focus here is on the tree in the foreground, so maximum focus may not reach out to the stone wall. If shooting at a long range the area that has focus will be critically sharp, as the lens is well corrected, but other things at same distance may be slightly out of focus. Stopping down helps.






Todays shooting...



In this case I have seen som impressive sharpness, see 1:1 crops below:







Three format comparison


I have tried to compare three formats, P45+, Sony Alpha 99 (full frame) and Sony Alpha 77 (APS-C), the latter two are both 24MP and were used with the same lens. I composed with the Hassy and tried to make the other images have the same horisontal crop. The Hassy was probably at f/11 and the Sonys at f/8. The images are all scaled to P45+ width and the two crops shown are actual pixels.

Hassy on the left, Alpha 99 center and Alpha 77 to the right. The results are somewhat puzzling, I expected a greater advantage for the P45+ and I expected less difference between Alpha 99 and Alpha 77. Again this may illustrate that lab testing is more accurate than real world shooting. Note that although the Sony Alpha 99 image looks sharp, it does suffer from upsizing artifacts. This test was done using the Planar 80/2.8 and the Sony SAL 24-70/2.8 ZA on the Sonys. 








Diffraction series

I have shot a diffraction series today. 


P45+ Diffraction results

What this shows is the sharpness peaks at f/5.6. A significant drop occours at f/16.


The images below show difference in central sharpness a5 f/8 and f/22. Please not that a significant portion of the sharpness lost to diffraction can be regained by adequate sharpening, so stop down when needed, but for optimum results medium apertures are normally the best choice. See also: http://www.onlandscape.co.uk/2012/07/diffraction-limited/






The lenses

The Macro Planar 120/4 at infinity


A good way to test lenses is to shoot a subject with distant treetops. I shot some images yesterday, with all my lenses. All performed reasonably well or as expected according to MTF. The Macro Planar is not intended for distant treetops so it needed to be stopped down to achieve decent quality. The Macro Planar is intended for close up work. This fundings are in accordance with Hasselblad's MTF curves.


MTF at infinity, below:

Planar 120/4 MTF at infinity


MTF at close up range below:

Planar 120/4 MTF at close up

Here are some samples from a very good lens:


Sonnar 150/4 at full aperture, extreme corner.

Sonnar 150/4 at 4, corner


Sonnar 150/4 at 4 center


The Planar 120/4 needs to be stopped down a lot for acceptable corners, due to field curvature. At close distances it would have flat field.


f/16 below:

Planar 150/4 at f/16 corner'


At f/8 it is not sharp at all:

Planar 120/4 at f/8 corner


The Planae is outright unsharp in the corners at f/4:

Planar 120/4 at f/4 corner

Update 2013-07-12


I got some suggestions from Stefan Steib, who pointed out that the Planar is very sharp, but needs to be accurately focused and stopped down to f/11 for optimum sharpness. A page with comparison of all apertures is here.


It seems that the Planar 120/4 stopped down to f/11 works as well at infinity as a Sonnar 150/4 at any good aperture. 

The Sonnar 150/4 

According to published MTF data the Sonnar is an excellent lens, and it does indeed impress! The samples below are at full aperture!






Planar 80/2.8


Not very good fully open 


But pretty good stopped down


Distagon 50/4


Corners are awful at f/4


Corner quite OK at f/8




Focusing


I have three different focusing hoods. The standard focusing hood:



The PM5 - prism viewfinder

The PM5 can be combined with a Zeiss Victory 3x12 monocular, this combo yields a 9X magnification and the monocular can be adjusted to individual eyesight.



The Hartbley 4X loupe is very good and can be adjusted to individual eyesight.





Clearly, my favorite is to use the PM5 with the Zeiss monocular.

The image belows shows a crop of a large image with a small USAF-type test target. I looked at USAF-target in the viewfinder which group could be clearly resolved in the viewfinder.

PM5 - red frame
Hartbley - blue frame
PM5 + monocular yellow frame 



The standard viewfinder loupe actually worked very well, but I cannot give a figure for that one.

Clearly, lens/sensor has much higher resolution than the viewfinder, even with the 9X PM5/monocular combo. So focusing will not be based on resolution alone but on conceived contrast. 

My viewfinder has also a split edge. This is probably very accurate as it takes help of the vernier acutance of human vision. The split edge is very consistent with the PM5+monucular combo. The problem with split edge is the it needs very good and clear edge to focus on, like a flagpole. It landscape photography it is pretty difficult to find such good edges. This is an area where live view manual focus at actual pixels has a great advantage. On the other hand, most lenses may have residual aberrations and even focus shift making live view focus harder.

Discussion

To me it seems that my findings essentially confirm what could be expected from the MTF curves published by Zeiss. The Sonnar 150/4 is excellent at all aperture while the other lenses need to be stopped down for good corners. Mostly I shoot at f/11 - f/16, because I want good DoF, that will probably loose some sharpness at center but makes fo decent corners. Stopping down also reduces moiré (acts as an aliasing filter).





;-( The bad news ;-(


Yeah, there are some bad news. One is the crop factor. The P45+ has a crop factor of 1.1 and that is not too bad. Problem is that the crop factor applies to 6x4.5 format, on full rame 6x6 it is much longer. So my 50/4 is essentially a an 35 mm and the 40/4 is more like 28 mm. The 40/4 is the widest lens for the Hasselblad V, except the 35 mm fisheye.

The other bad news is that contrary to P45+ manual the P45+ cannot be mounted vertically on 5XX EL bodies, the battery/motor compartment blocks the release key.
Statement above wrong, there is an alternate release for vertikal but it only engages one of the locking hooks. Should be good enough and has no play. Great work Phase One!

Tripod head


I normally use an Arca Swiss D4 head. It works just fine. Working with a ball head causes nausea for me. I was surprised at the difference, it may vary between people, but side inverted viewing looking down the viewfinder hood on a ball head is not for me!


Capture One or LR 4.4?


I am still using LR 4.4. I have tried Capture One, but I simply cannot reproduce my way of processing images in Capture One. It is certainly a good product, and I have seen some benefits, but it seems that me and Capture One sort of don't make friends. 

With both programs I have problems with oversaturated greens. From time to time I find that the Adobe standard profile gives the best results. I have a license for Capture One Pro since a few months.


Some comments on Phase One business model

Phase One works trough a dealer chain. If you have an issue, you can contact Phase One support, but it seems that any issue you have needs to be resolved trough a dealer. That approach has probably both advantages and disadvantages.

I have little experience with dealer network, but the experience I had this far has been positive. I contacted Phase one for part number for view finder masks for V-series Hasselblads, and support gave me the part number within hours. Once I contacted an accredited dealer I got the parts I needed swiftly and at a reasonable price.

Phase One is selling both backs and their raw converter. It seems that Phase One does not make all data available for other makers of raw converters. I am strongly opposed to this strategy. In my view, the photographer needs to have full ownership to all data and have the option to use any software. 

I would suggest that Phase One needs to support DNG (which seems to be the only open format) without discrimination. 


The ultimate platform

Hopefully, I can buy a Hartblei HCam in a year and so. The Hartblei is a technical camera that can use almost any lens and back. The idea is to use the Hartblei with a Mirex T&S adapter, so I can utilize all the image circle of Hasselblad lenses on the relatively small P45+ sensor for tilt and shift. Extreme wide angle capability can be added using the excellent 24mm and 17 mm T&S lenses from Canon.  Hartblei integrates it all. 

OK, what is the Hartblei HCam?

Well it is a thin technical camera built around the Mamiya shutter. The camera is essentiall solid metal (I guess pericipation hardened aluminium), with an integrated computer handling things. The HCam is tremendously flexible, can take any back and almost any lens. For viewing there is a motor driven slider that shifts either a focusing hood or the MFDB into image position. Great stuff as explained below.

Last Updated on Thursday, 08 August 2013 12:38  

Statistics

OS : Linux u
PHP : 5.2.9
MySQL : 5.0.67
Time : 03:29
Caching : Enabled
GZIP : Disabled
Members : 1510
Content : 72
Web Links : 1
Content View Hits : 372611

Login

Banner1

J!Analytics