Erik's web

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

Two months of MFD, looking back

E-mail Print PDF

I had my MFD-equipment for almost two months. During that time I have done a lot of testing and made some nice but not truly outstanding images. I have a lengthy blog about the experience here: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/76-my-medium-format-digital-journey but I felt it would be nice to present a snapshot of my findings.

Samples with raw files


All the directories here held raw files, full size JPEGs etc.



The positives

  • I am impressed by the image quality, especially regarding sharpness and resolution. I would say that resolution is the main technical advantage I have seen.
  • The old Hasselblad lenses I have perform much better than I would have expected.
  • The Hasselblad V camera I have is within good tolerances.
  • Everything works decently well together. None of the used stuff I have bought has any significant issue as far I know.

The negatives (none of which was unexpected)

  • I expected focusing being a problem and to a certain extent it is. The P45+ I have lacks live view in the field, and I regard live view at actual pixels as the only reliable means of accurate focus.
  • Color aliasing is a significant problem, it seldom shows up in landscape shots, fortunately.
  • Moiré filters in Capture 1 and Lightroom are not helpful with Bayer demosaic artifacts.

Observations on color

Initial tests indicated problems with color rendition in the problematic area of red, violet. In my case a flower which (at least to me) was red visually came out violet with LRs AdobeStandard profile. Creating a DNG profile overcame that problem, see below:



What I found using that profile was that it resulted in unnatural yellows/green and caused generally oversaturated colors. It is well possible that Capture One (which I also own) offers better colors, but
I sort of never made friends with that program. I start knowing it better, though, more on that later.

So I found that more and more I go back to Adobe Standard rendition.

Tweaking colors in Adobe DNG Profile Editor


Adobe DNG Profile Editor builds a DNG profile using one or two DNG images. One of the DNG images would have an incadescent illuminator and the other one daylight. The DNG profile can be tuned. I built a tweaked DNG profile where I tried to reduce saturation in the offending areas. The links below show the following images:

  • Adobe standard profile
  • New DNG Profile Editor generated from two ColorChecker cards (dual illiminant)
  • The above DNG profile tweaked
  • A DNG profile generated by Color Checker Passprt software
  • An image converted by Phase One's Capture One software
Color balance was on grey card on the ColorCheckers back side.

The lenses

I have the following lenses:
  • Zeiss Distagon 40/4 CF FLE (not the IF version)
  • Zeiss Distagon 50/4 CF FLE
  • Zeiss Planar 80/2.8 CFE
  • Zeiss Macro Planar 120/4 CF
  • Zeiss Sonnar 150/4

Of the five lenses, I would say that the Sonnar 150/4 is the best one. The weakest lens is the 120/4 Macro Planar when used outside it's designated macro range. The main issue at infinity is field curvature.

What I have found is that Macro Planar is one of my mostly used lenses. Yes, it needs stopping down to f/11 or f/16, but in real world shoots I have to do it anyway, to get depth of field. 

When I bought the 40/4 and made initial tests I found it to be an awful lens, but in real world conditions I feel it performs OK even at f/8.

Do I have to many lenses? Yes! Am I getting rid of any? No! Why?

I would say that the 50/4, the 80/2.8 and the 150/4 are first rate. But the 120/4 is perhaps my most used lens. The Distagon 40/4 FLE is the only really wide lens, as the P45+ sensor is really a crop factor lens on 6x6.

P45+ diensions are 49.1x36.8 mm while 6x6 film used to be 55x55, this would give a crop factor of 1.27, diagonally.

Diffraction

Diffraction is sort of laws of physics. A lens designer can do little to reduce diffraction as it is essentially a function of aperture and little else. I made an aperture series on my Sonnar 150/4 and it peaked around f/56.




On the other hand, diffraction is benign to deconvolution type of sharpening. Smart sharpen with "Gaussian" "PSF" and well cosen "amount" and "radius" is pretty decent on restoring image detail.

Some practical examples  are given here

Capture One vs. LR5


Doing comparisons, I much prefer using "common ground" processing. On the other hand, it has often been suggested that Capture One gives much better results on P-series backs than other converters.

Regarding C1 vs LR5 I certainly have a bias, being an LR user since first public beta. One issue with C1 is that it defaults to "Film curve" which looks like 1.5 stops overexposure. Results using linear curve is similar to LR5. The issue with 'film curve' is that it may make you underexpose images by 1.5 stops, looking good but wasting a lot of image qulity. Don't use film curve, or at least don't trust it! Use 'linear curve' which shows the real image!

Other than film curve, I simply don't know. Lightroom has probably better algorithms for tone mapping and has better sharpening options, but more options give us more rope to hang ourseves. 

The way I use Lightroom, I much depend on parametric editing and on a workflow essentially avoiding raw conversion, unless needed. The codeword is: "Tiffs are bad". Switch to another tool breaks that workflow.

Why TIFFs are bad?
  • Doing conversion to TIFF throws away information
  • Doing TIFF includes demosaic
  • TIFFs take like five-six times more disk space and still have less information than raw files, example: IIQ file 45.6 MB, TIFF file 236.1 MB. You can alway create a TIFF from the IIQ but not an IIQ from TIFF.
But C1 has some advantages and may have some others ( I don't have the experience to tell)

I have to little experience to tell, but it may well possible that C1 has some advantage in some areas. What I may have seen is that:

  • C1 has somewhat less color aliasing artifacts
  • C1 one may have somewhat better sharpening (but LR  has more options, so it is hard to tell)
  • C1 implements locald adjustments as layers, good idea, but I am not sure it is better than the corresponding LR solution
  • C1 has it's own workflow and database facilities

What I do right now?


I process all my images in Lightroom but sometimes check them out in C1.

Aliasing artifacts

Aliasing occours when the lens transmits significant contrast beyond the resolution of the sensor. Aliasing may show up as artifacts:
  • Color moiré is the most obvious one
  • Christmas tree artifacts (color demosaic artifacts)
  • Fake detail 
What I have seen is that artifacts in general show up in test images, but much less in real work. I guess that in real world artificail detail may matter little, also, real world images may be less perfect than 'lab' work. For instance we often stop down to gain DoF. 


The pair of images below show a subject with identical focus but different aperture. The medium aperture shot shows color artifacts that are gone in the f/22 shot, where diffraction acts as an OLP (Optical Low Pass) filter.


The second image (shown below) was taken with small aperture, diffraction reduces resolution and the color aliasing is gone:



The image below is a studio shot of a feather with a Sonnar 150/4 at 3.5m. The image projected on the sensor has detail the sensor cannot resolve, so it invents detail that does not exist in the subject. Some areas are marked below. Unnaturally bent strains, cross hatch and other effects. This link shows the same feather shot from the same position with a similar lens with three different cameras. 


Dynamic range

Regarding dynamic range, it was not a problem for me with any of my cameras of recent. I tried to shoot some comparisons, and I am pretty confident that my Sony Alpha 99 SLT has a wider dynamic range than the P45+.

In real life I would suggest that there are other factors playing a larger role. My guess is that DR is limited by lens flare in many shots. 


Should I have bought a D800E instead?

A very good question!

Unfortunately, I have no answer, as I never had a D800E or Nikon stuff (well except an FE and a 28/4 shift lens). For half the money I spent on a 10-20 years old Hasselblad, P45+ back and lenses one could buy a D800E with quite a few lenses. But, I wanted to find about MF.

Looking at MTF data for the lenses I have and MTF data for Nikon and Zeiss lenses I would say that a well focused Nikon D800/D800E with the best lenses they have can probably match my 555ELD, Zeiss lenses and P45+ back in most aspects.

Last Updated on Tuesday, 27 August 2013 21:51  

Statistics

OS : Linux u
PHP : 5.2.9
MySQL : 5.0.67
Time : 14:37
Caching : Enabled
GZIP : Disabled
Members : 1510
Content : 72
Web Links : 1
Content View Hits : 325566

Login

Banner1

J!Analytics