Michael's review of H5d60

[jerome_m]

The other thing to keep in mind is this subject had rich olive colored skin, which is the easiest to photograph.   The hardest is very light caucasian skin with that light epidermal layer that allows red to bounce back through, (think pasty white politicians).

I certainly agree on that. Fair skin, especially on young ladies or even babies/children, is the kind of colour where MF shines out of the box. I am still not sure whether another camera can be tweaked to produce the same colour or not (*), but I certainly agree that MF does a very convincing job by default.

(*) I am not sure means I do not know. I am not saying it can, I am not saying it cannot.

[jerome_m]

Of course, a camera that has weird CFAs is going to go crazy if used to image a box of cosmetics ....

I think I could easily make up an anti-target designed to break the renderings
And one could use the original stuff as a hard reference.

Now, that is a pretty good idea.

[eronald]

J,

 Hope you caught up with sleep and are feeling better to some degree.

 This set of images reminds me of Leaf's tests which I could never really get to terms with, precisely because of the olive skin of the models. I find it very hard to infer the real world color of the models from their images.

 Canon always have at least one woman with fair skin in their official camera test shots. And one can usually see how much "plastic skin" one will get from that image.

 I think you're right, one learns to be more discriminating on these images after doing a lot of them, partly because the clients themselves show you what they are looking for, and when it is hair or lipstick colors the women are very very picky.

Edmund

I see a huge difference, maybe because I've done so much of this type of hair and beauty photography.

The p65 then the hasselblad's show real color detail within the hair.  Not as pretty out of camera because they pick up every tone and color, but when your working in fine detail, it's much easier to distinguish and match color if the information is there vs. the cameras like the d3x and that has a global warm color.

The camera I find interesting is the D90.   I still own one, never really use it, but bought it as it was the first dslr that shot video.  I was working in Korea with mixed light and had fits with the color with all of my cameras.  For the heck of it I shot the d90 and the skintones  were beautiful golden natural in the scenes I was working.

The only issue was in full length horizontals that little d90 didn't hold detail that well.

But to me this test was excellent and I did see a great deal of difference in color response, though people should keep in mind that most digital capture, regardless of sensor size of pixel count produce a lot of detail this close up.  Once you pull back you see a difference.

The other thing to keep in mind is this subject had rich olive colored skin, which is the easiest to photograph.   The hardest is very light caucasian skin with that light epidermal layer that allows red to bounce back through, (think pasty white politicians).

IMO

BC

[synn]

Hi,

I have checked out my old A2 prints I made from the Hasselblad D50 and the D3X. What I see is that lady has dark grey hair on the D50 and dark brown with red tint on the D3X. I cannot really judge skin tones. The D3X seems to have slightly higher contrast.

The red sensivity curve on most CGA-s is very steep around 550 nm, so a colour spike at 530 nm would be rendered very differently from one at 570 nm.

Sensor sensivity curves are very different from human vision, see enclosed figure from wikipedia.

Best regards
Erik

This is getting into comically bad territory and someone has to break the news to the emperor that he has no clothes.

Dear Erik, has it ever occurred to you to just LOOK at the images for a second as a work of art rather than think of the next possible way you can micro-analyze them ?
Do you also rank Pink Floyd, Led Zeppelin and Deep Purple based on what frequency spectra they have recorded in?

You mentioned in the other thread that you haven't seen the magic in MF. Maybe; just maybe if you'd spend a fraction of the time you spend analyzing on careful compositions and understanding good light, you'd have seen it. Instead, you're taking test shot after test shot in unremarkable light with unremarkable compositions.

The majority of this guy's work is with the same back you have: http://bulbexposures.com/ . DO you see anything different there?

You're in a very picturesque part of the world. An unbelievably beautiful country. For once, please take this good natured advice; leave your test charts at home, go out with your gear and work hard for getting one killer shot in great light. You'll see the magic. Trust me. If not, please sell off your MF gear as your Sonys are everything you'd ever want from an imaging device.

Please don't take this as an insult or an attack. It just pains me to see such great gear going to waste.

[eronald]

This is getting into comically bad territory and someone has to break the news to the emperor that he has no clothes.

Dear Erik, has it ever occurred to you to just LOOK at the images for a second as a work of art rather than think of the next possible way you can micro-analyze them ?
Please don't take this as an insult or an attack. It just pains me to see such great gear going to waste.

Synn,

 We were all looking at this set of images from various cameras.
 Artistic license only takes you so far - the camera is supposed to show some decent approximation of reality.
 FOR SOME IT MATTTERS IF SHE IS A MOUSY BRUNETTE OR A PUNCHY REDHEAD
 Take a deep breath. Go there. Look. Come back with your thoughts, we'll be delighted to hear them.

Edmund

[ErikKaffehr]

Hi,

I didn't say REDHEAD, I saw dark brown with some red tint, that red tint was very subtle. I did look at a lot of distances and different light.

The technical part was actually addressed to you. I have noticed the steepness of the red curve on many CGA plots I have seen. I think it is pretty obvious that would cause color shifts with a spiky spectrum in that area. Someone was mentioning spiky spectrum, it thought it was you? I don't know if studio flash spectrum is spiky, I got the impression that greybody radiation dominates over gas spectral lines. I have not found any spectral plot of studio flash light, have you seen any?


I looked at prints I made back in 2009 when I downloaded the original images in full size, I enclose a screen dump below. My prints came out less colorful than the on screen images, need to check in that.

Best regards
Erik

Synn,

 We were all looking at this set of images from various cameras.
 Artistic license only takes you so far - the camera is supposed to show some decent approximation of reality.
 FOR SOME IT MATTTERS IF SHE IS A MOUSY BRUNETTE OR A PUNCHY REDHEAD
 Take a deep breath. Go there. Look. Come back with your thoughts, we'll be delighted to hear them.

Edmund

[synn]

Synn,

 We were all looking at this set of images from various cameras.
 Artistic license only takes you so far - the camera is supposed to show some decent approximation of reality.
 FOR SOME IT MATTTERS IF SHE IS A MOUSY BRUNETTE OR A PUNCHY REDHEAD
 Take a deep breath. Go there. Look. Come back with your thoughts, we'll be delighted to hear them.

Edmund

Dude, I posted my thoughts several posts ago and none of it had anything to do with graphs borrowed from Wikipedia.
Is it really possible to turn EVERY conversation into one revolving around numbers and graphs? Seriously...

You can get her hair to look like whatever in post, but the undertones captured in the MF files can't be magically recreated in the other files. Why does thy fly right past all the number crunchers?

[eronald]

I said redhead to be funny, but the lady next to me has informed me that the word *in french* is "Auburn"

I'm gonna try doing some studio flash tests this weekend, this set raises enough questions to make it worthwhile. I believe studio flash is pretty clean, but the hair dye may be fluorescing in a narrow band.

Edmund

Hi,

I didn't say REDHEAD, I saw dark brown with some red tint, that red tint was very subtle. I did look at a lot of distances and different light.

The technical part was actually addressed to you. I have noticed the steepness of the red curve on many CGA plots I have seen. I think it is pretty obvious that would cause color shifts with a spiky spectrum in that area. Someone was mentioning spiky spectrum, it thought it was you? I don't know if studio flash spectrum is spiky, I got the impression that greybody radiation dominates over gas spectral lines. I have not found any spectral plot of studio flash light, have you seen any?


I looked at prints I made back in 2009 when I downloaded the original images in full size, I enclose a screen dump below. My prints came out less colorful than the on screen images, need to check in that.

Best regards
Erik

[ErikKaffehr]

Hi,

I will look a bit more into this, as I see even more questions.

Best regards
Erik

I said redhead to be funny, but the lady next to me has informed me that the word *in french* is "Auburn"

I'm gonna try doing some studio flash tests this weekend, this set raises enough questions to make it worthwhile. I believe studio flash is pretty clean, but the hair dye may be fluorescing in a narrow band.

Edmund

[Aphoto]

Synn,

 We were all looking at this set of images from various cameras.
 Artistic license only takes you so far - the camera is supposed to show some decent approximation of reality.
 FOR SOME IT MATTTERS IF SHE IS A MOUSY BRUNETTE OR A PUNCHY REDHEAD
 Take a deep breath. Go there. Look. Come back with your thoughts, we'll be delighted to hear them.

Edmund

Ha, funny, the only picture I was absolutely sure it was shot with a MF-Back (and not 35mm) was the fat pixel CF22-Picture.

[jerome_m]

Since I am a bit confused about the different white balances of the portraits, I tried to download them and let the computer correct the white balance automatically (using the same patch of skin as a reference). These are the results. Indeed hair colour changes amongst cameras and that cannot be due to light alone. I find that the 3 Hasselblad pictures are the closest, but we should know that manufacturers try to get consistency in their range of cameras. Phase One is close to Canon. I only tried the MF cameras, Canon 5D and Nikon D3X.

[ErikKaffehr]

Hi,

Just to tell. I made a somewhat sobering experience today. I looked at A2 print I had of the Hasselblad D50 and the Nikon D3X images, and found that the lady had greyish hair on the Hassy and brownish with some red tint on the Nikon. Than I checked both the images in Lightroom, and found that the hair on the Hassy image was a lot warmer in the JPEG than in my print. After that I verified with soft prof that colours were shifted with quite a few profiles, including Durst Lambda.

Anyway I decided to make a different test, I asked a lady at office about the colour. She felt that the Nikon image was awful. She felt the Hassy image had credible skin tones and the hair looked as natural hair, while she felt the Nikon D3X had reddsh skin tones and unnatural red tone in the hair, like has it been artificially coloured. "Synn" won that round on KO!

Best regards
Erik

Since I am a bit confused about the different white balances of the portraits, I tried to download them and let the computer correct the white balance automatically (using the same patch of skin as a reference). These are the results. Indeed hair colour changes amongst cameras and that cannot be due to light alone. I find that the 3 Hasselblad pictures are the closest, but we should know that manufacturers try to get consistency in their range of cameras. Phase One is close to Canon. I only tried the MF cameras, Canon 5D and Nikon D3X.

[Chris Livsey]

I need a better monitor 

[jerome_m]

She felt the Hassy image had credible skin tones and the hair looked as natural hair, while she felt the Nikon D3X had reddsh skin tones and unnatural red tone in the hair, like has it been artificially coloured.

But if that lady has dyed her hair, which seems probable, the D3X is more accurate, isn't it?

The white balance on the Nikon and Phase One were particularly bad, BTW. I don't think it is reasonable to judge the capacities of MF on that series without correcting the pictures.

[ErikKaffehr]

Yes,

That could be the case. But I asked the lady for her opinion and I got it. She really did not like the Nikon image but that also had to do with the posing of the lady.

I am a landscape shooter, so I know little about skin tones.

Best regards
Erik

But if that lady has dyed her hair, which seems probable, the D3X is more accurate, isn't it?

The white balance on the Nikon and Phase One were particularly bad, BTW. I don't think it is reasonable to judge the capacities of MF on that series without correcting the pictures.

[ErikKaffehr]

Thanks for sharing both thoughts and images!

Erik

Since I am a bit confused about the different white balances of the portraits, I tried to download them and let the computer correct the white balance automatically (using the same patch of skin as a reference). These are the results. Indeed hair colour changes amongst cameras and that cannot be due to light alone. I find that the 3 Hasselblad pictures are the closest, but we should know that manufacturers try to get consistency in their range of cameras. Phase One is close to Canon. I only tried the MF cameras, Canon 5D and Nikon D3X.

[jerome_m]

I am a landscape shooter, so I know little about skin tones.

I suspect that you need to watch young ladies more often, preferably naked.

[ErikKaffehr]

Yes, thanks for good advice ;-)

Best regards
Erik

I suspect that you need to watch young ladies more often, preferably naked.