Pages: 1 [2] 3 4 5   Go Down

Author Topic: 5D3 vs. D800 - great real world head-to-head tests  (Read 57774 times)

John Camp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2171
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #20 on: April 20, 2012, 02:40:41 am »

I'm sure both cameras are excellent, but when you look at Michael's D800 comparison with the D800e, it makes me think that under the same conditions, the 5D3 would come in a distant third. That's because this is a landscape site, with a bias toward landscape work -- out of doors, with a premium on fine resolution. For journalism work, almost any current camera would work, well down into the amateur DSLRs. I note that the new Nikon D3200 will have a 24mp sensor, will sell for $700 *with* an 18-55mm (27-~82mm effective) lens.

On the other hand, *no* famous landscape photo relies for its fame on resolution.
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #21 on: April 20, 2012, 09:54:48 am »

I note that the new Nikon D3200 will have a 24mp sensor, will sell for $700 *with* an 18-55mm (27-~82mm effective) lens.
Yes: along with the Sony A77 and NEX7, the sensor resolution gap from very affordable hobbyist level to the D800 or even the Leica S2 or a HD4-40 is rather modest: at 300ppi, it is the difference between print widths of 20" vs 25" ... with Canon's 22MP sensors "languishing" at a mere 19" and the D4 and 1D X another inch or two behind.

And Erik K. has shown that a variety of lenses are capable of making good use of the A77 sensor's resolution, so hopefully no one will repeat claims about these "tiny pixels" being hopelessly diffraction limited, or absolutely needing "Zeiss glass" rather than crappy Nikon lenses.
Logged

johnkiv

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 52
    • JohnKishIV photography
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #22 on: April 20, 2012, 05:19:36 pm »

Indeed! I have been in NY for 20 years, and seen Ansel Adams prints a million times. Not one of them was 60" wide :P


It is a can of worms to even bring him up,  but Adams did not have the dynamic range in the toe that people seem to consider so critical now.  Adams prints had big swaths of deep blacks (dmax?) that seemed to swallow all the light in the room.  He had different materials, and those materials contributed to his art that make up a huge part of the landscape tradition that we are building on.   We have new tools, and new things are possible, including small prints.  Awhile back I saw a exhibit of original Edward Weston prints.  Photos I have seen in books, posters, calendars.... well, the "originals" were 4x5 contact prints. Quite beautiful IMHO, they blew me away.

But back to the video reviews, I found them refreshing.  After reading so much speculation and "math," charts, graphs, etc, finally people are actually taking pictures with these cameras.  Also, consider the source.  They are a camera store, after all.  They have to sell to both camps, keep their respective camera reps happy.  I also found it refreshing that someone will admit to shooting jpgs.  I hope there is not lynch party out for that guy!

I do sympathize with someone trying to sell a 60" print in a gallery. That has to be a tough customer, a tough way to make a living.

It is spring time, where I live, go takes some photos! Make some art, or whatever!

John

Kirk Gittings

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1561
    • http://www.KirkGittings.com
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #23 on: April 20, 2012, 10:46:27 pm »

Indeed! I have been in NY for 20 years, and seen Ansel Adams prints a million times. Not one of them was 60" wide :P

What the hell does AA know about landscape compared to the pixel peepers on this site? :)
Logged
Thanks,
Kirk Gittings

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #24 on: April 20, 2012, 11:11:02 pm »

What the hell does AA know about landscape compared to the pixel peepers on this site? :)

probably because the technical capability didn't exist. New technology opens up new possibilities. If Shakespeare lived today, he'd probably make movies. If Mozart were allive, he'd probably have written music for the electric guitar.
Logged

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #25 on: April 21, 2012, 02:59:11 am »

probably because the technical capability didn't exist. New technology opens up new possibilities. If Shakespeare lived today, he'd probably make movies. If Mozart were allive, he'd probably have written music for the electric guitar.

And AA would undoubtly be a stitcher! :-)

Cheers,
Bernard

John Camp

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2171
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #26 on: April 21, 2012, 02:19:25 pm »

And AA would undoubtly be a stitcher! :-)
Cheers,
Bernard

I know you said that tongue-in-cheek...but it's an interesting comment, and I sort of doubt it -- I think a lot of his most famous pictures are of one precise moment, or, at least, give you that sense. 
Logged

Slobodan Blagojevic

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 18087
  • When everyone thinks the same, nobody thinks
    • My website
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #27 on: April 21, 2012, 02:35:07 pm »

... I do sympathize with someone trying to sell a 60" print in a gallery. That has to be a tough customer, a tough way to make a living...

But that's how Peter Lik makes his millions, not by selling 4x5s, but by bathtub-size ones. Size matters and size sells.

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
On Ansel Adams, stitching and very large prints
« Reply #28 on: April 21, 2012, 03:12:13 pm »

I know you said that tongue-in-cheek...but it's an interesting comment, and I sort of doubt it -- I think a lot of his most famous pictures are of one precise moment, or, at least, give you that sense.  
On the other hand, some of his great 8x10” format pictures involve exposure times of about a minute (f/64, very slow film, shadowed scenes) -- not so "precise" a moment. I guess that AA today would experiment with stitching for some shots. After all, he was a gear-head and lab rat as well as an artist.

By the way, he did produce at least a few very large prints, perhaps mostly for special events like a World's Fair exhibit. But print formats like 20"x16" and less seem more usual for his gallery prints. And with those jumbo prints seemingly intended for viewing from further away in situations with large crowds viewing, they quite likely did not need more resolution that a carefully examined 20x16.

Another note: at f/64 in 10"x8" format, the diffraction limit on resolution is about as for f/8 in 35mm format ---how many pixels worth is that?!
« Last Edit: April 21, 2012, 03:15:16 pm by BJL »
Logged

Guillermo Luijk

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2005
    • http://www.guillermoluijk.com
DxOMark DR: at low ISO D800 wins, at high ISO both are equal
« Reply #29 on: April 21, 2012, 08:44:03 pm »

I took DxOMark's Dynamic Range plot, IMO the most useful since it shows noise performance in the deep shadows (where noise is really an issue), for every ISO value. Canon 5D3, 5D2 and Nikon D800 are compared:




Interpretation to whom may be interested:

At base ISO (typical tripod or well lit applications):
- Both Canons perform the same, so no improvement in the new 5D3's sensor.
- Nikon D800 outperforms the Canons by more than 2 stops, this means deep shadows noise in the Canons is more than 4 times Nikon's level of noise (noise is doubled for every extra stop in the deep shadows). Translating this into a real case, for example in arquitecture or interiors, this may mean one shot is enough in the D800 and at least one extra shot is needed in the Canons in a reasonably high DR scene.

At high ISO (typical handheld low light applications):
- Sony's sensor in the Nikon behaves as the ISOless sensors found in the Pentax K5, Nikon D7000,... This means nearly 1 stop of DR is lost for every 1 stop ISO is pushed. It's almost not worth pushing ISO in order to get proper RAW exposure because noise improvement is minimal.
- Canons on the other side both behave as typical Canons, nearly maintaining their DR capability when pushing ISO, specially in the first steps 100->200, 200->400,... so it's critical in them pushing ISO as soon as RAW underexposure can take place at base ISO.
- The result is that from ISO1600 there is no winner between the D800 and the 5D3, which means a slight improvement over the 5D2.


IMO Canon must improve its DR at base ISO, or they will suffer in the mid/long term against the competition. More than 2 stops is a huge advantage for tripod photographers!.

Regards

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #30 on: April 21, 2012, 09:08:50 pm »

As far as I can tell, the improvement at high Exposure Index but not at low indicates that compared to the 5D2, the 5D3

- Has significantly less read noise from the photosites, this being the noise floor seen at high EI, low light situations.

- Has about the same noise level downstream, entering the signal after variable gain for EI adjustment has been applied, with this downstream noise dominating at low EI, meaning with low gain levels.
This is noise either in the output stages of the variable gain amplifier, or in transportation from there to the ADC, or in the ADC. My guess is the first, with this variable gain applied early, in the signal transfer from photosite to sense capacitors at the end of each column of photosites. That early amplification strategy (if I am right about it) seems to be beneficial for achieving low noise in low light (high EI) situations, slightly better than the D800 achieved, bit at the cost of performance in good light thst is good, but no longer "state of the art".

Maybe Canon judges this to be the best balance that it can currently achieve for expected use patterns. And with its excellent AF and Canon's excellent longer focal length offerings, this might be good for users whose most challenging situations are subjects like wildlife and sports and other action.
Logged

mr purdy

  • Guest
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #31 on: April 21, 2012, 09:11:55 pm »

While looking at the charts and graphs, I am reminded that the 5D2 (and original 5D for that matter) served me well in terms of IQ, and NOT ONCE did I miss my exposure so badly that I had to bring up shadows so far as to reveal the pattern noise (shooting fashion and food.) Not once.

As to DXO: I have seem a number of side-by-side tests with 5D2 vs. 5D3, and the newer camera has a lot less noise and such in the shadows, so I am curious as how it ends up with the same DR measurement according to DXO.

e.g.:




And here is the final image:
http://galleries.stevemelvin.com/670C0973-3.jpg

These are not my images - they were posted by www.stevemelvin.com to show a test of the 5D3's D.R. capabilities.

As you can see in the samples above, even with a HUGE push in the shadows, the new 5D3 seems to have much less of a pattern noise issue when jacking up the shadows ;-) That last image is a 100% crop.

So for my work, I am sure that the new 5D3 will be fine as well. Aside from some"landscape" guys, I find it hard to believe that many people shoot in such a way that they need to rely on bringing up shadows by 3, 4, or 5 stops to get the image they need. What happened to proper exposure? :P

I know these debates will go on until the end of time, but rest assured: MANY HUGE ad campaigns will budgets in the millions (Guess, Disney, Lancome, Avon, Ralph Lauren, Louis Vuitton...) are shot on a daily basis with these cameras and no one seems to be complaining too badly.
« Last Edit: April 21, 2012, 09:18:55 pm by mr purdy »
Logged

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #32 on: April 21, 2012, 09:23:00 pm »

It has less *pattern* noise in the shadows, but DR measurements don't include pattern noise.

It has less shadow noise at very high ISOs, but that isn't where DR is measured - DR is significantly lower at high ISO anyway.

Shadow noise, minus pattern noise, is about the same at base ISO. Which is why DR is similar.
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #33 on: April 21, 2012, 09:24:53 pm »

While looking at the charts and graphs, I am reminded that the 5D2 (and original 5D for that matter) served me well in terms of IQ, and NOT ONCE did I miss my exposure so badly that I had to bring up shadows so far as to reveal the pattern noise (shooting fashion and food.) Not once.
I think the main case for bringing shadows up substantially is not fixing exposure errors, or studio work wth controlled lighting, but some naturally lit scenes with very high subject brightness range, like a mixture of bright sky above and deep shadows below, or an interior with view through a window to a sunlit outside. Or like your example! Situations that caused us old-timers to resort to printing at very low contrast and/or with lots of dodging and burning to fit into the six stop or so range of a print.

P. S. And I do not dispute that the difference might only be relevant to a small proportion of cases, with most extreme SBR challenges. That is the hazard of looking at numbers without context. For example, if the DR scores were 20 stops vs 30, would that ever matter in practice?


By the way, what ISO speed is your example at? [EDIT] sorry, I mean where can I get details of Stephen Melvin's example: I misread this as being your photo and you being him. [END EDIT]
« Last Edit: April 22, 2012, 11:15:31 am by BJL »
Logged

mr purdy

  • Guest
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #34 on: April 21, 2012, 09:27:37 pm »

It has less *pattern* noise in the shadows, but DR measurements don't include pattern noise.

It has less shadow noise at very high ISOs, but that isn't where DR is measured - DR is significantly lower at high ISO anyway.

Shadow noise, minus pattern noise, is about the same at base ISO. Which is why DR is similar.

Whatever you want to call it - the above example looks to me to be a far cry from what boosting shadows on the 5D2 would get you.

AND, importantly, while one file is prettier and more useful than the other, they receive the same "score." Raising the question of how useful such tests are. 

BJL: I know there are times when extreme DR is needed. That is why I mentioned the kinds of work I do. But even when I shoot travel landscapes it is rarely an issue in any practical sense. (again IMO)

Just my opinions :P
« Last Edit: April 21, 2012, 09:44:28 pm by mr purdy »
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #35 on: April 21, 2012, 09:34:46 pm »

Stephen Ron, [sorry also for the name confusion]

Sorry, I was editing my reply while you responded. I think we pretty much agree about putting the extent of the issue in practical context.
« Last Edit: April 22, 2012, 11:16:28 am by BJL »
Logged

shadowblade

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2839
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #36 on: April 21, 2012, 09:44:23 pm »

I'd say the differences are very relevant in my day-to-day photography.

With my 5D2, I need to use multiple blended exposures for almost every shot where the horizon won't allow a GND. Many of my shots have strong backlighting, or otherwise high dynamic range. When using my friend's D3x, I don't need to - I can get great results just by pushing the shadows a bit.

Multiple exposures aren't always practical - often, there will be moving elements in a scene (leaves rustling in the wind, grass moving a bit) which make it impossible to stack them, and many public places (including many national monuments in some countries) won't allow tripods. Moreover, many competitions explicitly forbid photos created from multiple exposures.

And the improved dynamic range make lenses with bulbous front elements - the Nikon 14-24, for example, before the customised Lee filter holder came out - much more useable.
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
5D2 and 5D3 DR: how often is it not enough for you, and in what situations?
« Reply #37 on: April 21, 2012, 10:10:57 pm »

I'd say the differences are very relevant in my day-to-day photography.

With my 5D2, I need to use multiple blended exposures ... When using my friend's D3x, I don't need to - I can get great results just by pushing the shadows a bit.
It might be nice to survey 5D2/5D3 users in this forum, on how often and in what circumstances a few more stops of DR (at base ISO speed) would help, by avoiding ND filters, HDR blending or such. I have no idea of how many are in shadowblades's situation vs how many are in Stephen Melvin's (mr purdy or Ron Purdy's) vs other scenarios.
« Last Edit: April 22, 2012, 11:17:32 am by BJL »
Logged

mr purdy

  • Guest

It might be nice to survey 5D2/5D3 users in this forum, on how often and in what circumstances a few more stops of DR (at base ISO speed) would help,

Don't get me wrong, it would be great to have 20 stops of DR at all times!
I am just making the point that I have had not one shoot which failed due to lack of DR.
(but again, I am no Ansel Adams ;-)



Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: 5D3 vs. D800 - great real world head-to-head tests
« Reply #39 on: April 22, 2012, 12:30:32 am »

As to DXO: I have seem a number of side-by-side tests with 5D2 vs. 5D3, and the newer camera has a lot less noise and such in the shadows, so I am curious as how it ends up with the same DR measurement according to DXO.

There does indeed seem to be a lot of conflicting reports about the relative performance of the 5D2 and 5D3. Unless one is able to examine the RAW files, one is not able to determine or verify the accuracy of the conclusions and the soundness of the methodology. Even when the RAW images are made available, there may be no guarantee that the lighting conditions were the same for both shots, especially when comparing sun-lit, outdoor scenes with the sun blazing through shifting clouds, as in the shot you present from Stephen Melvin.

Nevertheless, as an owner of the original 5D, I'm well aware of the banding problem. I would have thought that Canon would have fixed it in the 5D2. If they haven't, perhaps just improved the situation slightly, they should certainly have fixed the problem in the 5D3.

However, I don't think I've ever seen any comparisons of 5D2 and 5D3 deep shadows showing banding in the 5D2 shot and a lack of it in the 5D3 shot. However, I have seen comparisons on Dpreview demonstrating approximately equal noise in the shadows of shots from both cameras taken inside a church.

I suppose a relevant question might be, do the deep levels of shadows that contain more visible banding in the 5D2 shot, have acceptable noise in other respects that would make such shadows presentable in a print, or more likely, if the shadows are so deep that banding is visible in the 5D2 shot, would it be the case that both the 5D2 and 5D3 shots would still have unacceptable levels of noise.

I never found even with the original 5D that shadow noise in general was acceptable whenever banding reared its ugly head, which is why I never showed much interest in techniques to remove the banding.
Logged
Pages: 1 [2] 3 4 5   Go Up