fredmiranda.com Login
This topic is in the 2015 archive. You can read it and like its posts here. It is read only.

Archive 2015Canon 5DS dynamic range: zero improvement (RAWs analyzed)

  
 
gdanmitchell
Offline
Details
p.14 #1Canon 5DS dynamic range: zero improvement (RAWs analyzed)
ben egbert wrote:


Guilty as charged... If I had to do an ROI for photography it would be ended.

You are in good company. For example, Vivian Maier's ROI was awful. ;-)

Dan
Mar 27, 2015 at 10:02 AM
retrofocus
Offline
Details
p.14 #2Canon 5DS dynamic range: zero improvement (RAWs analyzed)
ben egbert wrote:
And when shooting into the sun, 14 stops is not enough.


The photo below taken against the sun was a single photo (not bracketed) taken with the A7R. You can imagine that normally with 11 stops DR you would have a hard time getting either a good structure of the ground (too dark) or the sky would be overexposed. Bracketing could give you a similar result here but would require more post processing work. At least I prefer to have it simple and in one shot .

http://martinbluhm.zenfolio.com/img/s7/v168/p900498622-4.jpg
Mar 27, 2015 at 10:20 AM
ben egbert
Offline
Details
p.14 #3Canon 5DS dynamic range: zero improvement (RAWs analyzed)
retrofocus wrote:
ben egbert wroteAnd when shooting into the sun, 14 stops is not enough.
The photo below taken against the sun was a single photo (not bracketed) taken with the A7R. You can imagine that normally with 11 stops DR you would have a hard time getting either a good structure of the ground (too dark) or the sky would be overexposed. Bracketing could give you a similar result here but would require more post processing work. At least I prefer to have it simple and in one shot .

http://martinbluhm.zenfolio.com/img/s7/v168/p900498622-4.jpg


Yes, this is fine, but I am speaking of a sun still high enough to be white, usually behind clouds. This is when you get the ring around the clouds effect. I have a single shot of Mesa arch complete with a sunburst that I was able to do in one image not unlike this one. But after I learned some HDR I prefer the effect it has on tone mapping the mid range.
Mar 27, 2015 at 10:25 AM
retrofocus
Offline
Details
p.14 #4Canon 5DS dynamic range: zero improvement (RAWs analyzed)
ben egbert wrote:
Yes, this is fine, but I am speaking of a sun still high enough to be white, usually behind clouds. This is when you get the ring around the clouds effect. I have a single shot of Mesa arch complete with a sunburst that I was able to do in one image not unlike this one. But after I learned some HDR I prefer the effect it has on tone mapping the mid range.


I agree - there are still situations as I also mentioned earlier where the contrast is too big to avoid bracketing or HDR. Just wanted to demonstrate above that more DR makes it simply easier to get the shot right out of the box (with little processing). More DR is something which doesn't negatively affect you if you don't need it, but can make a big difference in some situations as above when you make use of it.
Mar 27, 2015 at 10:30 AM
arbitrage
Offline
Details
p.14 #5Canon 5DS dynamic range: zero improvement (RAWs analyzed)
I have an honest question about how the extra 2-3 stops of dynamic range shows up in the image. Say I have a scene and I set an exposure on the 5D3 that allows the brightest whites to be just below clipping. If I had the same scene with a 14 stop DR camera, would I be able to add 3 stops of light and still not clip the highlights (thus bringing the darker areas 3 stops brighter) or is it more that the exposure would stay the same but the shadows would be able to be pushed easier?
Mar 27, 2015 at 10:35 AM
retrofocus
Offline
Details
p.14 #6Canon 5DS dynamic range: zero improvement (RAWs analyzed)
arbitrage wrote:
I have an honest question about how the extra 2-3 stops of dynamic range shows up in the image. Say I have a scene and I set an exposure on the 5D3 that allows the brightest whites to be just below clipping. If I had the same scene with a 14 stop DR camera, would I be able to add 3 stops of light and still not clip the highlights (thus bringing the darker areas 3 stops brighter) or is it more that the exposure would stay the same but the shadows would be able to be pushed easier?


I didn't try this here for this image, but from experience in general with this sensor I would say the latter is the case - the shadows can be pushed more easily without losing anything. In the composition above I didn't need to push the shadows severely though - just to an extent that I saw the grass and the pumpkins better in the front.

If I had taken the photo just exposing the foreground correctly, even with the 14 stops of DR I would have clipped the highlights likely.

The difference to 11 stops is that I wouldn't have been able without bracketing to get the exposure as above in a single shot and keeping the details in the darker and also in the brighter areas.
Mar 27, 2015 at 10:42 AM
NCAndy
Offline
Details
p.14 #7Canon 5DS dynamic range: zero improvement (RAWs analyzed)
StillFingerz wrote:
GEORGE LEPP TESTS THE NEW CANON 5DS R IN THE FIELD
http://www.georgeleppimages.com/new-page-2/?cm_mmc=TW-_-Explorer+of+Light-_-2015-03-27_1123AM_TWPRO_Here%27s-a-look-at-_-129336%3A2238809


Nice images. I would never have the patience to shoot and stack so many frames but the result is amazing.
Mar 27, 2015 at 11:12 AM
jimmy462
Offline
Details
p.14 #8Canon 5DS dynamic range: zero improvement (RAWs analyzed)
arbitrage wrote:
I have an honest question about how the extra 2-3 stops of dynamic range shows up in the image. Say I have a scene and I set an exposure on the 5D3 that allows the brightest whites to be just below clipping. If I had the same scene with a 14 stop DR camera, would I be able to add 3 stops of light and still not clip the highlights (thus bringing the darker areas 3 stops brighter) or is it more that the exposure would stay the same but the shadows would be able to be pushed easier?


Canon 5DS dynamic range: zero improvement (RAWs analyzed) - FM Forums: Page 11
https://www.fredmiranda.com/forum/topic/1355280/10#12932685

Perhaps the link I provide in that post helps answer your question?!
Mar 27, 2015 at 11:28 AM
Scott Stoness
Offline
Details
p.14 #9Canon 5DS dynamic range: zero improvement (RAWs analyzed)
retrofocus wrote:
ben egbert wroteBut I doubt one image will meet my needs even with the better DR, so I would probably continue the habit formed back in my film days of using a 3 shot bracket. And a 3 image blend takes about 10 minutes for the blend part. Being retired, I actually need more stuff to do.
I can ensure you that you won't if you have the option - this was a big eye-opener for me when I got my A7R that I simply didn't need to bracket landscape shots anymore. In fact the result from a single photo and some simple post processing is better than the bracketed final composition. I am only using bracketing with this sensor for indoor architectural shots with very bright and dark areas in it.
Avoiding bracketing for landscape shots is a big plus - no issues with moving branches or grass in the wind between bracketed shots. The DR
...Show more →

Retrofocus and I have different conclusions from the same data.

I bought the a7r when it first came out and continue to use it pending the 5dsr becuase of the 36mpx. I still prefer AEB (better colors in dark areas, 14 stops does not mean good quality in the last 2 stops, and its only 14 stops) and blending or hdr/blending with the a7r. The dynamic range is still not enough to achieve pictures that look like what you see, and require AEB to fix. The big advantage of a7r is mpx, not dynamic range.

Clearly when it comes to images that only work with 1 shot (waves, moving objects) the A7r has an advantage in dynamic range however, the ergonomics of the camera are not as nice, it does not function well in the cold, the evf is too dim in bright light, the ergonomics of the menu suck, the need to use a adapter sucks, the battery life is awful, the shutter chugs along like a freight train, and the dust control is not as good as the canon. In short its a better sensor but everything else is weaker.

I can't wait to ditch the a7r to avoid the multiple systems.


Mar 27, 2015 at 11:45 AM
Arun Gupta
Offline
Details
p.14 #10Canon 5DS dynamic range: zero improvement (RAWs analyzed)
AJSJones wrote:
Here is an illuminating post by David Hull from the discussion on "Canon rumors" board on "Canon and will they develop a first class FF sensor(!)" ( http://www.canonrumors.com/forum/index.php?topic=20935.40 )
......


In that same discussion, someone wrote this: "Companies makes a bigger profits, if the leader changes with time, this makes users go to one side or another.". I tend to agree.
Mar 27, 2015 at 11:50 AM
AJSJones
Offline
Details
p.14 #11Canon 5DS dynamic range: zero improvement (RAWs analyzed)
arbitrage wrote:
I have an honest question about how the extra 2-3 stops of dynamic range shows up in the image. Say I have a scene and I set an exposure on the 5D3 that allows the brightest whites to be just below clipping. If I had the same scene with a 14 stop DR camera, would I be able to add 3 stops of light and still not clip the highlights (thus bringing the darker areas 3 stops brighter) or is it more that the exposure would stay the same but the shadows would be able to be pushed easier?


The exposure would be pretty much the same, but you could boost the shadow levels by 3 additional steps before seeing the noise level comparable to the other camera. Net result, shadows have 3 stops less noise.
Mar 27, 2015 at 12:03 PM
jimmy462
Offline
Details
p.14 #12Canon 5DS dynamic range: zero improvement (RAWs analyzed)
ben egbert wrote:
ggreene wrote
Beni wroteTo be frank 50 megapixels as a necessity is rather rarefied territory. I wonder how many people urgently need this camera.
Canon only cares about how many people "want" it. Still in the top 20 selling DSLR's on Amazon so somebody is buying.
Guilty as charged, I don't even need a camera. All my photography is a hobby, and hobbies are optional. But I do enjoy it and I have no clue what other hobby I could pursue. If I had to do an ROI for photography it would be ended.

I could make a similar analysis for DR. How much time are you willing to spend to increase DR? If camera A can do it with one 36 mp image versus (3) 50 MP images, obviously more time is required per image and maybe more storage. But I doubt one image
...Show more →

+1 @ being retired and needing more things to do! (Ben, you got me chuckling on that one! )

Being a member of that same, er, "free time club" as you, my choices as to what those "things" are are different. I'd prefer spending more of my time in the field gathering images and less time at the screen stacking or NRing or otherwise having to massage my images into shape. FWIW, my choices for replacing my 5D2 and 7D are going to be about sensor IQ (particularly high ISO performance, DR and noise characteristics) first and foremost. Given a choice between 10 noisy frames per second at ISO 3200 and 5 clean frames per second at ISO 6400, I'll take the latter. Given a choice between a camera that can capture all the lighting detail that I would like in a single sunset exposure and a camera that can't, well, you see where I'm going with this...
Mar 27, 2015 at 12:13 PM
Eric Gottesman
Offline
Details
p.14 #13Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Approximately how many stops do the sliders in Lightroom recover/push?

AJSJones wrote:
arbitrage wroteI have an honest question about how the extra 2-3 stops of dynamic range shows up in the image. Say I have a scene and I set an exposure on the 5D3 that allows the brightest whites to be just below clipping. If I had the same scene with a 14 stop DR camera, would I be able to add 3 stops of light and still not clip the highlights (thus bringing the darker areas 3 stops brighter) or is it more that the exposure would stay the same but the shadows would be able to be pushed easier?
The exposure would be pretty much the same, but you could boost the shadow levels by 3 additional steps before seeing the noise level comparable to the other camera. Net result, shadows have 3 stops less noise.

Mar 27, 2015 at 12:17 PM
gdanmitchell
Offline
Details
p.14 #14Canon 5DS dynamic range: zero improvement (RAWs analyzed)
arbitrage wrote:
I have an honest question about how the extra 2-3 stops of dynamic range shows up in the image. Say I have a scene and I set an exposure on the 5D3 that allows the brightest whites to be just below clipping. If I had the same scene with a 14 stop DR camera, would I be able to add 3 stops of light and still not clip the highlights (thus bringing the darker areas 3 stops brighter) or is it more that the exposure would stay the same but the shadows would be able to be pushed easier?


It is the latter — typical current reproduction methods (screens and print) cannot display display a 14 stop dynamic range, and final images must have their DR compressed to fit the range of the reproduction media.

If you expose for the highlights (often but not quite always a good starting point) you will use post-processing techniques to lighten up darker areas in your large DR image. This is already try with virtually all current DSLRs and similar cameras and has been so for some time.

Dan
Mar 27, 2015 at 12:22 PM
RustyBug
Offline
Details
p.14 #15Canon 5DS dynamic range: zero improvement (RAWs analyzed)
AJSJones wrote:

birding is often at high ISO where Canon's DR is just fine. For landscape/tripod shots, ETTR to the max and a couple of backup bracket shots should cover the DR thing for a while longer



Pretty much how I see it.

Pulling up the shadows with less noise is fine for what it can do as more convenient, but I'm inclined to think that when you DO bracket for the shadows, your greater exposure gets you more information to work with (akin to the reason for ETTR) and affords better tonality latitude.

Convenience will always have a place ... action shooters / event shooters likely will really like the extra DR for pulling up the shadows (assuming their camera & glass are up to the task as well) ... compared to those shooting more static subjects. Personally, unless I place my camera in single shot mode, I frequently rip off two frames faster than I get my finger off the shutter release ... it's that quick, so having a three shot bracket set to 0, +2, +4 or -1, +1.5, +4 etc. or conversely @ 0, -2, -4, etc. is a piece of cake.

Some folks use three or more, but I typically only use two of the frames, bracketing when I want 4 stops or more (adding my personal pref @ 2.5 EV push / pull), I get 6.5+ with the bracketing. The extra 1.5 or 2 stops that the SoNikon affords over the Canon ... while more is better, still doesn't cover > 6 EV range ... thus doesn't tempt me away from my glass and AF performance.

More is better, little argument there, but the amount of more @ 1.5 - 2 stops didn't have people jumping over to Fuji S5 from SoNiCanon ... but, suddenly now that same extra 1.5 - 2 EV is a game changer.

The real irony there is that the Fuji S3 is a 10 year old camera that is an F-Mount and accepts Nikon glass ... yet, only since SoNikon outpaced Canon has it become such an en vogue "game changer". If DR was the holy grail ... Fuji should have put SoNiCanon out of business 10 years ago.

Personally, I thought the Super CCD with large & small pixels was a superb concept and would have liked for Fuji to take that into the CMOS / FF realm, but they went X-Trans instead. One can only imagine what they might have achieved with 10 years advancement upon the Super CCD, given the vastly superior DR it had 10 years ago. Heck it still outpaces plenty of cameras in the DR department ... just not very many MP's by today's standards.

BTW, note that the Fuji ISO's are mostly spot on, while the others are overly exaggerated. Also, note that the X-over @ Fuji vs. Nikon is at EXACTLY 200 ISO (hmmm @ base) and beyond that the 10 year old Fuji is better @ 400 & 800. Just sayin' ... for a 10 year old camera, it is pretty exciting for the SoNikon innovation to "almost" reach its DR @ 400 & 800 ISO (or be a little better with more light @ ISO 73).








Edited on Mar 27, 2015 at 01:54 PM · View previous versions
Mar 27, 2015 at 12:33 PM
arbitrage
Offline
Details
p.14 #16Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Thanks for the replies about the DR and how it shows up. I don't do a lot of landscape stuff and usually I'm up in the higher ISOs anyways because I'm doing birds and wildlife and need the shutter speed to be fairly fast. When I expose my bald eagle shots in contrasty light I try to just get tiny blinkies in the white head and then in LR I push the shadows up a bit to get more details in the dark feathers. This method is usually enough for the eagles and the brown feathers don't show too much noise after the shadow lifting. I guess that means they actually don't need 14 stops of DR because I'm not having to push the shadows too far to look natural.

Shooting in a canyon with sky up above would be a different scenario and I would gather those are the types of shots where the extra stops help.
Mar 27, 2015 at 12:42 PM
retrofocus
Offline
Details
p.14 #17Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Scott Stoness wrote:
retrofocus wrote
ben egbert wroteBut I doubt one image will meet my needs even with the better DR, so I would probably continue the habit formed back in my film days of using a 3 shot bracket. And a 3 image blend takes about 10 minutes for the blend part. Being retired, I actually need more stuff to do.
I can ensure you that you won't if you have the option - this was a big eye-opener for me when I got my A7R that I simply didn't need to bracket landscape shots anymore. In fact the result from a single photo and some simple post processing is better than the bracketed final composition. I am only using bracketing with this sensor for indoor architectural shots with very bright and dark areas in it.
Avoiding bracketing for landscape shots is a big plus - no issues with moving branches or grass in the wind between bracketed shots. The DR of the A7R sensor simply allows to cope with a single shot. It is impossible to do the same with my 5D MkII (and I tried!).
Retrofocus and I have different conclusions from the same data.

I bought the a7r when it first came out and continue to use it pending the 5dsr becuase of the 36mpx. I still prefer AEB (better colors in dark areas, 14 stops does not mean good quality in the last 2 stops, and its only 14 stops) and blending or hdr/blending with the a7r. The dynamic range is still not enough to achieve pictures that look like what you see, and require AEB to fix. The big advantage of a7r is mpx, not dynamic range.

Clearly when it comes to
...Show more →

We discussed this earlier, so I am not going again in discussing with you your complaints about the performance of this camera (and this is not what this discuiussion is about). It is a difference what you personally prefer and what the camera potentially can do - the 14 stops make a big difference as many Sony and Nikon shooters can confirm.
Mar 27, 2015 at 01:10 PM
RustyBug
Offline
Details
p.14 #18Canon 5DS dynamic range: zero improvement (RAWs analyzed)
StillFingerz wrote:
GEORGE LEPP TESTS THE NEW CANON 5DS R IN THE FIELD
http://www.georgeleppimages.com/new-page-2/?cm_mmc=TW-_-Explorer+of+Light-_-2015-03-27_1123AM_TWPRO_Here%27s-a-look-at-_-129336%3A2238809


Interesting to note his three image focus stacking of the red rocks & moon to ensure optimal sharpness of the varying components of the image to compensate for the inability of the lens to capture the entire scene optimally. Kinda the same thing as bracketing for high DR scenes to compensate for the inability of the camera to capture the entire scene optimally (key concept @ optimally). Really, the same concept if you think about.

Sure, he could have gone with the hyperfocal (read convenient compromise), single image technique ... but he chose to shoot for optimal sharpness of his entire scene (via multi-shot), instead of the sacrificially convenient method of hyperfocus. Sure, we can pull up 5 EV, but just because he could use hyperfocus, doesn't mean he saw that as being the method for producing his best results. If he had been rushed for time or only had one frame of film left in his camera for a single last shot ... hyperfocal single shot could have "saved the day", but that wasn't the case.

It just seems silly to me to put forth the time, money, effort and energy to produce a 50MP (or 36MP) file with $$$ glass and camera ... but to use a technique of pulling up 5EV in post, instead of bracketing ... all in the name of convenience. Kinda like painting over rust in some regard. If you're gonna spend all that $$$ for quality ... but you can't spend a little time to make it your best ... IDK, just seems a bit contradictory in philosophical terms.

Edited on Mar 27, 2015 at 01:49 PM · View previous versions
Mar 27, 2015 at 01:39 PM
ggreene
Offline
Details
p.14 #19Canon 5DS dynamic range: zero improvement (RAWs analyzed)
arbitrage wrote:
Shooting in a canyon with sky up above would be a different scenario and I would gather those are the types of shots where the extra stops help.


The one area where it would be helpful to me is shooting bright noon time games with deep shadows (very small % of my work though). I just have my doubts about how it will handle skins tones/colors as that is the one area critical to me. Media relations wants faces.
Mar 27, 2015 at 01:40 PM
StillFingerz
Offline
Details
p.14 #20Canon 5DS dynamic range: zero improvement (RAWs analyzed)
RustyBug, bracketing, blending, stitching works for static subjects just fine, but if you introduce movement into your subject, wind, water, whatever, then a single shot is your best if not only option and that's when extra DR; exposure latitude, is a huge benefit.

Bright shy, dark shadows, which do you choose, not having to would be nice, freeing in some respects, like when light is fleeting; at sunrise/sunset
Mar 27, 2015 at 02:36 PM
FM ForumsCanon ForumJoin Upload & Sell
  
 

You are not logged inLog inRegister