H39MKII vs Canon 5DMKII image crop comparison

[ziocan]

The Canon shot is framed too close on order to be cropped as the Hassy one. You shot it about 25% too close. The canon shot should have been cropped below the knee for a proper comparison. If you put the two faces nect to each other you can seee the the Hassy's one is at least 25% smaller.
In any case shadows transitions are smoother on the Hassy shot, though the Hassy image is strangely dull for a 39mp shot.
Something is wrong.

[lisa_r]

Who knows when you are looking at a web jpeg. If I had to choose one to make a print from based on what is presented here, it would be the Canon. Perhaps Neil would make the RAWs available via yousentit.com so everyone can peep their brains out ;-)

By the way Neil, which lens on the Canon?

Thanks.

[neil snape]

The Canon shot is framed too close on order to be cropped as the Hassy one. You shot it about 25% too close. The canon shot should have been cropped below the knee for a proper comparison. If you put the two faces nect to each other you can seee the the Hassy's one is at least 25% smaller.
In any case shadows transitions are smoother on the Hassy shot, though the Hassy image is strangely dull for a 39mp shot.
Something is wrong.

No I don't think so . I frame as I see fit to shoot. I didn't have the cameras on tripods and do a sync shoot. That is not my style. I want to see how a camera works for me. This is how they work so it it is a very good test . Is the H39 flat? The histogram should tell. IF you look at the screen capture of the info on each you see the frame on each image. So yes as I said before there is a difference in cropping , but that should be obvious. I do have a test on a tripod of a still life black on black. Unfortunately it was the day before I bought the Canon MKII so it is a much lower res. Come to think of it that was also the H39 MKI.
The transitions are definitely smoother on the H39. To be expected. Actually I'm a bit disappointed with the new Canon. I thought by going to 14 bit it would have been a lot better but in fact it's not much better than at 12 bits.

[neil snape]

Who knows when you are looking at a web jpeg. If I had to choose one to make a print from based on what is presented here, it would be the Canon. Perhaps Neil would make the RAWs available via yousentit.com so everyone can peep their brains out ;-)

By the way Neil, which lens on the Canon?

Thanks.

I have the exported psd's on my server, but at 120 and 220 MB it's a hefty download. The raw's no they are here. Too bad I didn''t keep Flexcolor on the edit Mac.

Canon was shot on the famous lens I use 100mm 2.8 macro. It is the only lens capable of rendering everything for close ups at around 100mm which I shoot. The Hassy was the 120 macro which in theory should be the equivalent.
For most shooting I see the Canon as the faster way, easier way, as you can shoot into LR or CO or whatever else you like.
Quality wise I'd say if Phocus worked out the bugs, would be the better choice. The differences though are a lot less than some want to believe.

PS if the Hassy shot is a bit flat, it is because I didn't have time to check the images until after (Phocus is so slow it's unusable while shooting) and realised that the Canon underexposes compared to Hasselblad for the same ISO and F stop. So I pulled the H39 expo down 0.27 or something. IF you leave the exposure high with the default settings unfortunately there is too much funky stuff going on with oversaturation in the 1/4 tones. Again is that because Phocus needs work?

[ziocan]

No I don't think so . I frame as I see fit to shoot. I didn't have the cameras on tripods and do a sync shoot. That is not my style. I want to see how a camera works for me. This is how they work so it it is a very good test . Is the H39 flat? The histogram should tell. IF you look at the screen capture of the info on each you see the frame on each image. So yes as I said before there is a difference in cropping , but that should be obvious. I do have a test on a tripod of a still life black on black. Unfortunately it was the day before I bought the Canon MKII so it is a much lower res. Come to think of it that was also the H39 MKI.
The transitions are definitely smoother on the H39. To be expected. Actually I'm a bit disappointed with the new Canon. I thought by going to 14 bit it would have been a lot better but in fact it's not much better than at 12 bits.

I do not see it as a matter of shooting style, unless the main purpose of using those cameras is recreational or just inkjet printing.
If you need of printing those two images on a magazine, which normally have a ratio of 4 by 5, a portfolio that let's say is 11 by 14, or web for a customer web site, which 99% of them use a 4 by 5 ratio, on the canon shot you need of discarding 25% of the image unless, you want to have a "nice" white band on one side. while the hasselblad image is already pretty much proportional to the final media.
As it is framed the Canon images can be printed only on 11 by 17 format, which does not have many commercial use applications that I can recall of, or it is suitable for inkjet printing and hanging it on the wall. Not even model's portfolio have an 24 by 36 ratio.

What many people forget, it is that the for "24 by 36" format to be used on commercial media, we needs to discard between 20 to 25% percent of the image (top and bottom). Any time I work with the Canon or Sony, I keep that in mind, otherwise i'm going to deliver images that are not suitable for the client. we can make an exception if we are working in studio with a seamless background and we plan to extend the background on the sides.
As is the test is not valid for let's say commercial/professional use of the images.

[neil snape]

The posterization that I was mentioning is also clearly visible whether in my browser (Safari, which does handle colour spaces) or CS4. I am talking about the facial shadow area on the left side of the image, especially around her right eye and the corner of her mouth. The shadows are so bad as to be useless here. My original 5D didn't give me results like this, and I don't understand why it is happening at all. I cannot imagine the results being accepted for critical use by any magazine? Is it always like this? The rest of the image looks fine, but that area is horrible.

Safari is not going to be the place to look at images even if they take into account embedded profiles.
Hmmm. Well Canon in skin tones has always done that horrible transition if you don't fill light it. Would it show in CMYK on press? Maybe depends on the press op if they run light for shadows. On the proof is it there? Yes. Can you do anything about it? Not much , perhaps CO is better, but I think it is the DR of the captor.

I did the shadow on purpose BTW. Black velvet covering a Styro board. I wanted to see if the 16 bit MF could dig into the shadows or not. It does, but the best thing is transitions are as they should be. Phocus is creating points though that definitely should not be there. I tried setting the moire to 3 and above but it blends the jeans onto her skin so it is unusable at 3.

Maybe I should send the RAW to Dustbank.

[neil snape]

I do not see it as a matter of shooting style, unless the main purpose of using those cameras is recreational or just inkjet printing.
If you need of printing those two images on a magazine, which normally have a ratio of 4 by 5, a portfolio that let's say is 11 by 14, or web for a customer web site, which 99% of them use a 4 by 5 ratio, on the canon shot you need of discarding 25% of the image unless, you want to have a "nice" white band on one side. while the hasselblad image is already pretty much proportional to the final media.
As it is framed the Canon images can be printed only on 11 by 17 format, which does not have many commercial use applications that I can recall of, or it is suitable for inkjet printing and hanging it on the wall. Not even model's portfolio have an 24 by 36 ratio.

What many people forget, it is that the for "24 by 36" format to be used on commercial media, we needs to discard between 20 to 25% percent of the image (top and bottom). Any time I work with the Canon or Sony, I keep that in mind, otherwise i'm going to deliver images that are not suitable for the client. we can make an exception if we are working in studio with a seamless background and we plan to extend the background on the sides.
As is the test is not valid for let's say commercial/professional use of the images.

I actually just add background as I need to. Not the purest way I agree. I used to do this with drum scanned 6x6 too as it's my recreational style of shooting.

Yet I did say the test is up to you to see  which means it is subjective. Not sure what shooting commercial is about, I just shoot what I want, as I can produce what I need for the magazines and ads I've done.

[ziocan]

I actually just add background as I need to. Not the purest way I agree. I used to do this with drum scanned 6x6 too as it's my recreational style of shooting.

Yet I did say the test is up to you to see  which means it is subjective. Not sure what shooting commercial is about, I just shoot what I want, as I can produce what I need for the magazines and ads I've done.

Let's say Franca Sozzani  decide to print that image on Italian Vogue, you will need to crop the canon shot 25% off, or they will print it with a white band on one side(very few magazines do that anymore). You hardly see ads printed on 24 by 36 ratio unless are spread of a magazine. If you shoot on location and you cannot stretch the background to your liking on post, you will need of framing the picture differently, unless you do not mind chopping whatever is on the edges.
I have yet to meet an art director on the entire world that love to deal with chopping hands, shoes or foreheads, because the photographers were shooting 35mm without taking care of the basic rule of leaving enough space for cropping.
To each his own.
cheers.

[Nick-T]

Hi Niel
Thanks for doing the test. I'm not a canon user but I was struck by how bad the individual channels looked and the overall redness in the skin, I'd be interested to hear from a 5D2 user in the light of my following comments. To repeat, I'm not a Canon user, I am however a Hasselblad user, and at the risk of being rude must ask you Neil how well you know the (Hasselblad) system. The H3D image is in no way representative of the sort of quality I get from my back every day and as such I think the test is a bit flawed.
Nick-T

BTW I wanted to add that I'm familiar with Niel's beautiful work and in no way doubt his skills as a photographer.

[neil snape]

Hi Niel
Thanks for doing the test. I'm not a canon user but I was struck by how bad the individual channels looked and the overall redness in the skin, I'd be interested to hear from a 5D2 user in the light of my following comments. To repeat, I'm not a Canon user, I am however a Hasselblad user, and at the risk of being rude must ask you Neil how well you know the (Hasselblad) system. The H3D image is in no way representative of the sort of quality I get from my back every day and as such I think the test is a bit flawed.
Nick-T

BTW I wanted to add that I'm familiar with Niel's beautiful work and in no way doubt his skills as a photographer.

As I said in the first post I just received both the Canon and Hasselblads both MKI + II. Phocus is the only software I have for RAW dev on the FFF files. There are not any LR 5D II profiles as they have for the 5D, nor did I make a camera calibration as I had previously. The problems I see in Phocus might not be present in Flexcolor, that I do not know. I can't say the quality is unexpected as the Canon and the Hasselblad look similar so I think this is just the way it is.

I did change the add a curve to the developed H39 file to bring it into line with the Canon.
The girl has this beautiful orange red skin which is quite unusual. A bit of a mix but a nice one at that.

[neil snape]

Let's say Franca Sozzani  decide to print that image on Italian Vogue, you will need to crop the canon shot 25% off, or they will print it with a white band on one side(very few magazines do that anymore). You hardly see ads printed on 24 by 36 ratio unless are spread of a magazine. If you shoot on location and you cannot stretch the background to your liking on post, you will need of framing the picture differently, unless you do not mind chopping whatever is on the edges.
I have yet to meet an art director on the entire world that love to deal with chopping hands, shoes or foreheads, because the photographers were shooting 35mm without taking care of the basic rule of leaving enough space for cropping.
To each his own.
cheers.

You know in French we'd say sans blague.
When you're shooting something it's going to have to fit into your camera. If your goal is to have the whatever space as headroom then that's what you do. It doesn't take one long to figure out that a crop on the traditional 355 mm frame is always long for the width for magazines. I thought that was something you learned when you picked up a camera, 35mm for me was many years ago.
So I suppose you are dwelling on something that is obvious , a 35mm frame is cropped for magazine or practically any other use.  So a 21.1 Mpx becomes something like 18 , and yes the ratio of a MF is better adapted just as the Fuji 6x8 was. But I'm going to tell you a secret, the covers I've done, all the covers I know are highly retouched, and if you have to add bleed you just do so. From what I've learned from reviewing pages coming back, you'd be challenged to see a print quality difference from MF vs drum scanned film, or the recent 20mpx Dslrs.
I have said , and will say it again, there is more detail in the 39mpx images period. This also gives you room to crop, absolutely right. Freedom at last, but at a price of liberty of hand holding and shooting quickly.

[Nick-T]

You know in French we'd say sans blague.

I think that's French for "No Sh#t!"

Niel's test is exactly how photographers test cameras.

The pixel peeper will pick up a camera and talk about aspect ratios, lens equivalents, relative apertures, choice of glass, and goodness knows what else. The pixel pitch of a certain sensor for example, although relevant, has very little to do with photography IMO.

The photographer will pick up a camera and shoot some frames to see which works best in his (or her) hands.
My previous comment was just to point out that I don't think Niel got any where near the quality out of the Hasselblad that he could have.
Nick-T

[Dustbak]

Exactly my thoughts as well though I am not sure where it went less than optimal.

Neill you can download Flexcolor from the (or any) Hasselblad website. You have to register and login.

[James R Russell]

The pixel peeper will pick up a camera and talk about aspect ratios, lens equivalents, relative apertures, choice of glass, and goodness knows what else. The pixel pitch of a certain sensor for example, although relevant, has very little to do with photography IMO.

The photographer will pick up a camera and shoot some frames to see which works best in his (or her) hands.

Nick-T

I somewhat agree with Nick that it doesn't have to be an either/or situation.  Sure I'd love my digital backs to work as fast, easily and have the lcd and iso of a dslr but the flip side to this is I'd love my dslrs to have a different cropping ratio for vertical or sometimes like a removable aa filter, but they don't.

For me the real answer is I need more than one system and today I seem to use the dslrs more than the medium format cameras, but depending on the gig, maybe even my mood it might go the other way.

The upside to all of this is the dslrs have improved a great deal and medium format has come down in price so it's really possible to get more cameras, more use and have more options than ever before.

[Nick-T]

The upside to all of this is the dslrs have improved a great deal and medium format has come down in price so it's really possible to get more cameras, more use and have more options than ever before.

As Cooter says maybe we'll all end up like it used to back in the day where most pro's (in my world) had 35mm, medium and large format cameras and used whatever camera suited the job...

Nick-T

[ziocan]

I think that's French for "No Sh#t!"

Niel's test is exactly how photographers test cameras.

The pixel peeper will pick up a camera and talk about aspect ratios, lens equivalents, relative apertures, choice of glass, and goodness knows what else. The pixel pitch of a certain sensor for example, although relevant, has very little to do with photography IMO.

The photographer will pick up a camera and shoot some frames to see which works best in his (or her) hands.
My previous comment was just to point out that I don't think Niel got any where near the quality out of the Hasselblad that he could have.
Nick-T

Mais bien sure Nik.
And I'm a photographer too.
Than again, if we make a comparison to asses quality and detail on two different cameras, we need to at least take on consideration the size of the subject in the frame, and the lens, making otherwise it is just a waste of time. or at least try of having the main subject of the same size on both photos. that does not require a too much of work and not even the camera to be on tripod.  
But what is the purpose of comparing two photos from two different cameras and posting crops (especially jeans crops) if not for looking at details and quality, which like it or not involve pixel peeping.

Nik, I was talking about aspect ratio, simply for making a practical example of real life applications for those kind of cameras which could be fashion editorial publishing, catalogue or commercial printing. Sure we can take photos completely disregarding the aspects ratio relative of the media where the photos are supposed to be printed and following our deepest primordial artistic instincts, but that will not instantly turn anybody in the hottest artist of the year, but rather lead to have to reject a lot of supposed to be "great" shots, simply because once fitted on the page they will not work.



By the way, "Sans blague" means: "No kidding..." Nik.

Do you really need to pixel peep on order to see that the chick on the photos is much larger in size on one photo rather than the other?

[Nick-T]

By the way, "Sans blague" means: "No kidding..." Nik.

And "No kidding" means "No sh#t"... but anyway

I take your point on the cropping of the two files I guess it would be nice to see tests that were half way between Niel's and a real pixel peeper one of a newspaper. Also it would be nice if there was a way to have someone really familiar with each system to advise. For example it would be a waste of time me doing a test with a Canon as I know very little about them.
Nick-T

[ziocan]

I somewhat agree with Nick that it doesn't have to be an either/or situation.  Sure I'd love my digital backs to work as fast, easily and have the lcd and iso of a dslr but the flip side to this is I'd love my dslrs to have a different cropping ratio for vertical or sometimes like a removable aa filter, but they don't.

For me the real answer is I need more than one system and today I seem to use the dslrs more than the medium format cameras, but depending on the gig, maybe even my mood it might go the other way.

The upside to all of this is the dslrs have improved a great deal and medium format has come down in price so it's really possible to get more cameras, more use and have more options than ever before.

Well, using the right camera for the job is the most sensitive thing to do.
It was like that on film days and it became somewhat affordable to do it on digital as well. that is why most of us, use 2 or 3 different formats depending on the job. at least i do.

for some people there is "the camera do it all" and for others "there is the camera for the job".
there are photos than cannot possibly be taken with a digital back, but can be taken with the dslr, yet the opposite is becoming less the case.
there are more situations where a 20+dslr can be used instead of a DB, rather than the opposite. Some situation that were DB domain, until not long ago, they can be done with a DSLR.
But this should be another topic.

[Dustbak]

As Cooter says maybe we'll all end up like it used to back in the day where most pro's (in my world) had 35mm, medium and large format cameras and used whatever camera suited the job...

Nick-T

Pretty sure we are going there. At least I am and most people I know that have MF also have LF & 35mm.

[neil snape]

Pretty sure we are going there. At least I am and most people I know that have MF also have LF & 35mm.

With Hasselblad being the first group to cut the price in half, they opened the door for the current users of film , and Dslr to move up and have the possibility of using both.

I am one of those potential clients, and there are a LOT more around. Phase one followed, and it is logical that Leaf should do the same.


as I said the camera, the viewfinder, the system is night and day above the 5D. It has more detail, transitions are better. I think you can pull more from them too.

For the person stuck on the framing and size between the two , well there will not be a 39 mpx Dslr in the near future to compare same framing to anyway. Yet the repro from the Canon is not only good enough for pixel peeping, but good enough for shooting anything if you forgive it's bit depth and less margin for error.
Personally I don't care , as it certainly never will be me that would say I wouldn't want both Dslr and a MF for what they do best. For what this test shows is the quality of the Dslr is not that far off the H39. I don't shoot MTF charts, never will. I shoot what I want how I want, and know enough to shoot according to the format , with lenses that meet the requirements for what I do. Hasselblad have a lame comparison in Victor mag with a Canon on a 50 1.2 lens shot at f2, the H at 80mm 2.8 wide open. Neither is ideal but the Canon at 1.2 , now that is a crazy thing to do considering it is studio. This quick test is exactly what happens in a studio, with a contrasty light both at around the ideal f , exposure close enough to the same. I said the settings were as close as default as possible. Believe me , if I want to improve the default settings I know how. Setting grey balance for example has and is still very important for MF, much less so for Dslr.

I will try to dl Flex but I don't have much time between a shooting and retouching like crazy this week.