Ad
Ad
Ad
Pages: [1] 2 3 »   Bottom of Page
Print
Author Topic: Nikon D800E v.s Hasselblad H4D40: my in-studio test-review  (Read 24028 times)
AlexKoloskov
Newbie
*
Offline Offline

Posts: 29



WWW
« on: August 03, 2012, 03:06:25 PM »
ReplyReply

I've got this test-review posted, did it mostly for my own curiosity (I am in upgrade mode from an old H1 and P25+ DB), and i think it might be interesting here :-)
So, I've tested Nikon D800E v.s Hasselblad H4D40 in terms for details, shadows and highlight recovery. Selected H4D40 as the closest current MF camera to Nikon 36Mpx D800E

My "Rambo-style" shot with both cameras:


Part one: http://www.photigy.com/nikon-d800e-test-review-vs-hasselblad-h4d40-35mm-against-medium-format/

Part two:
http://www.photigy.com/nikon-d800e-v-s-hasselblad-h4d40-the-end-of-medium-format-superiority-round-two/

I know that F16 is not the sharpest aperture, but this  is what I use the most in a studio product photography, so I needed to see the difference at closed iris.
Enjoy the read and let me know what do you think.
Alex
Logged

Alex Koloskov,
The Lighting Magician at
www.photigy.com
PdF
Sr. Member
****
Offline Offline

Posts: 296



« Reply #1 on: August 03, 2012, 03:22:29 PM »
ReplyReply

Thank you for this confirmation: the difference between a digital MF and the Nikon D800E is very slight.

PdF
Logged

PdF
evgeny
Sr. Member
****
Offline Offline

Posts: 481



« Reply #2 on: August 03, 2012, 03:58:22 PM »
ReplyReply

Thank you for this confirmation.
Hasselblad images show reacher colors and better dynamic range, as expected.

Evgeny
Logged
Don Libby
Sr. Member
****
Offline Offline

Posts: 729


Iron Creek Photography


WWW
« Reply #3 on: August 03, 2012, 04:38:12 PM »
ReplyReply

Great write-up with good samples.  Thanks for the effort.

Don
Logged

bjanes
Sr. Member
****
Offline Offline

Posts: 2825



« Reply #4 on: August 03, 2012, 04:57:47 PM »
ReplyReply

I know that F16 is not the sharpest aperture, but this  is what I use the most in a studio product photography, so I needed to see the difference at closed iris.
Enjoy the read and let me know what do you think.
Alex

Alex,

Thank you for an excellent comparison, but you have not taken into account how sensors scale with respect to pixel size and sensor size. With smaller pixels one has to use a larger aperture (smaller f/number) to prevent loss of MTF from diffraction. This is well demonstrated in Figure 8 of Roger Clark's post. You should really use a larger aperture (perhaps one f/stop) with the D800e to offset the effects of diffraction. Since the smaller sensor has more depth of field for a given aperture, this would also tend to equalize differences in depth of field between the two sensor sizes. Roger also discusses this factor in his post.

Regards,

Bill
Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7655


WWW
« Reply #5 on: August 03, 2012, 06:04:22 PM »
ReplyReply

Hi,

Thank you for publishing the images. I made a quick comparison. Opened both images in Lightroom 4.1, white balanced both images on one of the buttons  on the lady's shirt and sharpened as I would on landscape for f/11. A bit excessive for portrait, I know. Then I uprezzed the Nikon image to in height using straight bicubic.

What I note:

1) The Nikon image shows some staircase and interpolation artifacts. That may come from the enlargement.
2) DoF is much shorter in the Hasselblad image. The Nikon image could use f/8 or even f/5.6 for equiavlent DoF. That would give Nikon an advantage regarding diffraction.
3) Hasselblad image is cleaner and has little more detail.
4) Note hair crossing the pupil of the eye on the Hassy image. It has very artificial look. May it be a missing column of pixels in the image? Does Phocus software show same behavior?

Which image has more natural colors? Hard to say...

Best regards
Erik
Logged

ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7655


WWW
« Reply #6 on: August 03, 2012, 06:42:57 PM »
ReplyReply

Update: Original posting had the wrong images, corrected!

Hi,

I tried to look at color accuracy. Note that this is about accuracy, not pleasantness.

I converted both images to 16 bits TIFF using Lightroom 4.1 (that is what I use). White balance was on the second grey patch from the left. Adjusted both image to similar exposure and analyzed the images using Imatest Color Check. According to Imatest the images were 0.7 step overexposed, so exposure was reduced for both images by 0.7 steps. The results are shown in the enclosed screen dumps.

First image shows the colors compared to color checker colors. The second one shows the position of the colors in lab space. The colors in the Nikon image, as processed, are significantly more accurate. In both cases the colors are probably more saturated than the reference colors, probably for better visual impression.

I would like to thank the OP for posting raw files of very adequate test images.

Best regards
Erik
« Last Edit: August 04, 2012, 07:56:23 AM by ErikKaffehr » Logged

ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7655


WWW
« Reply #7 on: August 03, 2012, 07:04:46 PM »
ReplyReply

Hi,

I tried to reduce saturation in LR 4.1 to get better match with reference colors, I arrived at:

-13 for Hassy
-4 for Nikon

Color still has a better match on Nikon.

The next step is to build a DNG color profile for each and repeat the experiment.

Best regards
Erik
Logged

EricWHiss
Sr. Member
****
Offline Offline

Posts: 2427



WWW
« Reply #8 on: August 03, 2012, 07:06:41 PM »
ReplyReply

Color accuracy vs pleasing colors is one thing which I won't get into, however to me the nikon has less color variation or tonality.  The tonality is what lends the image a more palpable feel or depth.  No question the nikon is capturing lots of detail, but not color detail, just detail in the luminosity sense. I've read different conjecture on why - color filter choice being one of them, but I don't know why for sure.  Anyhow another big difference between the DSLR and MF is the DOF which is shallower on the MF.  This is good and bad depending on what you are trying to shoot. You can see it in the portraits of the little girl.

I thank Alex for doing the test and sharing the results. 
Logged

Authorized Rolleiflex Dealer:
Find product information, download user manuals, or purchase online - Rolleiflex USA
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7655


WWW
« Reply #9 on: August 03, 2012, 07:23:51 PM »
ReplyReply

Hi,

According to Imatest Color Check it seems that LR 4.1 produces a bit more saturated colors on the Hassy. How do you think bumping up the saturation would affect the Nikon image?

The observation on DoF is absolutely correct. I actually think that would the Nikon lens be used at f/8 or f/5.6 would level the playing field. As pointed out by the OP, the Nikon lens is a "cheap one" a 105/2.8, but it would still probably work better at f/5.6 than at f/11. On the other hand, the only Nikon 105/2.8 I heard about is the Macro 105/2.8 and it is said to be very good.

In all a very good test, and a lot of thanks to Mr. Koloskov for sharing raw images. Obviously, my results may be different if Phocus was used.

Best regards
Erik

Color accuracy vs pleasing colors is one thing which I won't get into, however to me the nikon has less color variation or tonality.  The tonality is what lends the image a more palpable feel or depth.  No question the nikon is capturing lots of detail, but not color detail, just detail in the luminosity sense. I've read different conjecture on why - color filter choice being one of them, but I don't know why for sure.  Anyhow another big difference between the DSLR and MF is the DOF which is shallower on the MF.  This is good and bad depending on what you are trying to shoot. You can see it in the portraits of the little girl.

I thank Alex for doing the test and sharing the results.  
« Last Edit: August 03, 2012, 07:51:34 PM by ErikKaffehr » Logged

EricWHiss
Sr. Member
****
Offline Offline

Posts: 2427



WWW
« Reply #10 on: August 03, 2012, 08:16:38 PM »
ReplyReply

Hi,

According to Imatest Color Check it seems that LR 4.1 produces a bit more saturated colors on the Hassy. How do you think bumping up the saturation would affect the Nikon image?


Best regards
Erik


I compared some figurative images shot in my studio by my friend with his D800E and my AFi-ii 12 last month using the same model and lighting.   I observed the same points.  Shallower DOF on the AFi (and in that case I shot at f/14 or f/16 while my friend shot at f/10 using the same cropping.   Color differences were very pronounced but I only used C1 for both files.  Nikon files looked flat both at fit to screen and at 100% compared to the AFi-ii 12.   I was impressed with the detail the D800E could capture, but again it seems to be all in the luminosity channel, not in the individual color.    The AFi images looked to have more depth and feel.  Overall I was impressed with the D800E but its not superior to MFDB at base ISO.   

RE: Dynamic range.    I did run the Stouf001.NEF shot that I think Bernhard posted to the forums through Imatest and compared it to a shot I took of my own Stouffer transmission wedge.  It seemed that the Aptus 12 (similar to the IQ180) had about .4 stops of DR over the D800 at all the levels Imatest presents in the chart.   I can't say its a conclusive test since one shot was taken by someone else with a different step wedge and lighting conditions as mine.  At this level of DR I am also wondering how much differences are introduced by different lenses.   I'm not going to post the results since I'm not sure they are fair. 

My overall impression of the D800 is that its a very good camera, but still does not match the quality capable of the current MFDB's.  The real advantage to that camera is the better autofocus and higher ISO capability - and of course cost. 

Logged

Authorized Rolleiflex Dealer:
Find product information, download user manuals, or purchase online - Rolleiflex USA
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7655


WWW
« Reply #11 on: August 03, 2012, 09:27:35 PM »
ReplyReply

Hi,

I have a hunch that DSLRs have less orthogonal CGAs, meaning that there is more overlap between the color channels. That would be beneficial to high ISO performance. The new monochrome version of the Leica M9 lacks CGA and has very high base ISO, as an example.

Some knowledgeable person claim that Sony Alphas have better color than Nikon/Canon, but they are also noisier at high ISO. Sony has some film heritage from the Konica Minolta days and that may show. Although using Sony I have no clear view on this. I see color as mostly depending on processing. That said I have compared images from Pentax 645D and Nikon D3X that Miles Hecker has made, and I couldn't get them to match. Which was more correct? I may guess the Nikon image, but I preferred the Pentax image.

Regarding DR and the Stouffer Wedge it is my understanding that it is not particularly easy to use, as lens flare affects the results. Arri (Arriflex) has developed a special target for DR measurements with ver small gray patches. I got the impression that DxO uses the Arri target for their measurements.

I played around a bit with Alex's raw images, bumped up saturation on the Nikon image a bit and increased exposure about 1/3 stop. I enclose a side by side screen dump from LR4.1. The amount of saturation increase was based on my Imatest analysis.

Best regards
Erik


I compared some figurative images shot in my studio by my friend with his D800E and my AFi-ii 12 last month using the same model and lighting.   I observed the same points.  Shallower DOF on the AFi (and in that case I shot at f/14 or f/16 while my friend shot at f/10 using the same cropping.   Color differences were very pronounced but I only used C1 for both files.  Nikon files looked flat both at fit to screen and at 100% compared to the AFi-ii 12.   I was impressed with the detail the D800E could capture, but again it seems to be all in the luminosity channel, not in the individual color.    The AFi images looked to have more depth and feel.  Overall I was impressed with the D800E but its not superior to MFDB at base ISO.   

RE: Dynamic range.    I did run the Stouf001.NEF shot that I think Bernhard posted to the forums through Imatest and compared it to a shot I took of my own Stouffer transmission wedge.  It seemed that the Aptus 12 (similar to the IQ180) had about .4 stops of DR over the D800 at all the levels Imatest presents in the chart.   I can't say its a conclusive test since one shot was taken by someone else with a different step wedge and lighting conditions as mine.  At this level of DR I am also wondering how much differences are introduced by different lenses.   I'm not going to post the results since I'm not sure they are fair. 

My overall impression of the D800 is that its a very good camera, but still does not match the quality capable of the current MFDB's.  The real advantage to that camera is the better autofocus and higher ISO capability - and of course cost. 


Logged

EricWHiss
Sr. Member
****
Offline Offline

Posts: 2427



WWW
« Reply #12 on: August 03, 2012, 09:51:41 PM »
ReplyReply

Erik,
Look at your image set and compare the regions under the eyes near the nose, the eyes themselves, the top of the forehead where the hair is starting but you can still see the scalp, the lips, the stain on her shirt to the left and down from the button.  The spot on her skin near to the left of her mouth.  These all show you which camera has better tonality.  Now that you've equalized the saturation etc.  which image looks more real to you?
Logged

Authorized Rolleiflex Dealer:
Find product information, download user manuals, or purchase online - Rolleiflex USA
AlexKoloskov
Newbie
*
Offline Offline

Posts: 29



WWW
« Reply #13 on: August 03, 2012, 10:32:26 PM »
ReplyReply

Alex,

Thank you for an excellent comparison, but you have not taken into account how sensors scale with respect to pixel size and sensor size. With smaller pixels one has to use a larger aperture (smaller f/number) to prevent loss of MTF from diffraction. This is well demonstrated in Figure 8 of Roger Clark's post. You should really use a larger aperture (perhaps one f/stop) with the D800e to offset the effects of diffraction. Since the smaller sensor has more depth of field for a given aperture, this would also tend to equalize differences in depth of field between the two sensor sizes. Roger also discusses this factor in his post.

Regards,

Bill

Yep, agree. did not think about it. I use F16 on my full frame canon, so following your advise I'd have to close down Hasselblad to F18-20 to match DOF and diffraction levels.
However, following this route would require to replace Nikon's lens to something closer to Hassy's as well:-)
Thank you everyone, I dig your comments :-)
Logged

Alex Koloskov,
The Lighting Magician at
www.photigy.com
MichaelEzra
Sr. Member
****
Offline Offline

Posts: 659



WWW
« Reply #14 on: August 03, 2012, 11:13:37 PM »
ReplyReply

Alex, thank you for this review and its great presentation and providing the raw files. I'd like to add a few points.
I played with the portrait raw files using RawTherapee, without using custom DCP profiles, but using dcraw camera matrices, which seem to be the same as Adobe.
I'd like to point that due to composition differences D800E file is in inferior position as portrait is captured in a smaller scale (illustration 1)
Additionally, there are some differences in skintones due to movement in relation to the light.
Lenses on both cameras were set at F11 which leads to different rendering of the DOF and overall image look, also in favor for sharpness in critical focus for H4D.

Nevertheless, this illustration 2 (http://minus.com/m2aWArEH6) can be used for comparing skintones from both cameras as well as resolution. I added identical sharpening to both images.
As owner of D800E I am very happy with this result!:)
Attached in the zip file are the pp3 sidecar files that can be re-used in RawTherapee (version 4.9.50+)
Logged

K.C.
Sr. Member
****
Offline Offline

Posts: 653


« Reply #15 on: August 04, 2012, 12:22:32 AM »
ReplyReply

My overall impression of the D800 is that its a very good camera, but still does not match the quality capable of the current MFDB's.  The real advantage to that camera is the better autofocus and higher ISO capability - and of course cost. 

But the reality is in many markets the cheaper Nikon will allow new competition entry into your market, no matter what it is. There will be countless threads about how to give the Nikon images a look that comes close enough to move clients from studios that shoot MF to guys working cheap to gain clients. 

I love my Hasselblads and I'm buying the Nikon. My clients can decide which they want me to use,
Logged
EricWHiss
Sr. Member
****
Offline Offline

Posts: 2427



WWW
« Reply #16 on: August 04, 2012, 12:39:52 AM »
ReplyReply

Well really how much has changed?  Tons of work has been done with the 1Ds, 5D, 5D mk2, etc.     
Logged

Authorized Rolleiflex Dealer:
Find product information, download user manuals, or purchase online - Rolleiflex USA
FredBGG
Sr. Member
****
Offline Offline

Posts: 1651


« Reply #17 on: August 04, 2012, 01:47:42 AM »
ReplyReply

I think it is very important to point out the lenses used for this comparison and the aperture used.

A fair comparison would have been shooting at least on stop more open with the Nikon, because that is what would
give both images about the same depth of field.... thus comparing diffraction more fairly.

Second thing I want to point out is that the focal lengths are not equivalent either.
85mm vs 120mm would have been more equivalent.

Lastly the 105 2.8G macro is not the best Nikon lens.

Here is a comparison between the 85mm 1.4G and the 105mm 2.8G

85mm 1.4G


105mm 2.8G


Clearly the 85mm 1.4G would make up for the difference in the D800 vs Hasselblad test

There are quite a few Nikon lenses that are better than the 105 2.8G

50mm 1.4G


105mm f2D


even a zoom 70-200mm 2.8G
« Last Edit: August 04, 2012, 01:57:09 AM by FredBGG » Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7655


WWW
« Reply #18 on: August 04, 2012, 02:57:02 AM »
ReplyReply

Hi,

To begin with, Nikonians may rejoice now they have a 3000$ 36MP camera instead of a 8000$ 24MP one. I guess that is without doubt a good thing.

The other impression I have is that the test images are quite close. An assumption may be that they would be even closer in print.

In my view, this is a very valid test. There are a few factors favoring the Hassy but there may be other factors favoring the D8000. I wouldn't call this a scientific test. In a scientific test we try to eliminate as many variables as possible. A scientific test also needs to be reproducible, therefore dollar bills and color checkers are used instead of young ladies. The tests that DPReview and ImagingResource do are more scientific, they ought to be for they do testing for living.

To me it seems that the Nikon D800/E is a very interesting alternative to low end MF. Little doubt that better image quality can be achieved with high end digital, like the IQ180, if that equipment is put to perfect use. In addition, digital backs have a lot of flexibility.

Best regards
Erik




Well really how much has changed?  Tons of work has been done with the 1Ds, 5D, 5D mk2, etc.     
Logged

ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7655


WWW
« Reply #19 on: August 04, 2012, 03:17:53 AM »
ReplyReply

Eric,

Thanks for hints where to look, I see some of your points. I don't know which looks most real to me. I'm essentially shooting landscapes.

My guess is that if you made prints from these images and sold to a customer the customer would buy the one with better facial expression, regardless of tonality.

Best regards
Erik


Erik,
Look at your image set and compare the regions under the eyes near the nose, the eyes themselves, the top of the forehead where the hair is starting but you can still see the scalp, the lips, the stain on her shirt to the left and down from the button.  The spot on her skin near to the left of her mouth.  These all show you which camera has better tonality.  Now that you've equalized the saturation etc.  which image looks more real to you?

Logged

Pages: [1] 2 3 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad