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)

  
 
Monito
Offline
Details
p.16 #1Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Also merge pano AND do HDR.

I made a photo in Yosemite at dawn with 2 rows of 12 panels and 3 shots HDR per panel. With mirror lockup on a 5D classic, which of course requires two clicks per shot. That was a total of 144 clicks in the snow, before the sun rose where I stood, taking care not to lose track or miss an overlap. It took four hours of CPU time to merge the HDRs and then blend the panels.

It's what we do as photographers when necessary.
Mar 27, 2015 at 08:15 PM
retrofocus
Offline
Details
p.16 #2Canon 5DS dynamic range: zero improvement (RAWs analyzed)
RustyBug wrote:
The notion that your "fleeting light" is very responsible for your low level portions of the scene is likely not as prominent as is often thought. I'm not saying that there is no instance in which a single exposure doesn't have its place ... just that I think folks aren't developing well thought, strategic more DR)bracketing approaches ... and thus are very relieved to have the added DR.

As noted many times, more DR is welcome, but I still find that 5+ EV pull is less than ideal ... even if possible.


I read your posts, and I think it is quite funny to see how defensive you are to find excuses why 11 stops of DR are sufficient. This said, I am fully convinced that you would say 5 stops of DR are not enough if Sony's current sensor had 5 stops more DR. On the other hand, if Canons sensor had the 14 stops of DR, everything would be great and you wouldn't talk how great and advantageous bracketing is. Yes, bracketing can lead to the same and in some situations even to a better result, but if it can be avoided it is simply easier with more DR. I highly recommend that you simply rent an A7R or D810 to check it out yourself before making all those claims how less important DR is if you never tried the competitive sensor (and many arguing this way simply stick to Canon sensors and talk here without experiencing the other product's capacity).

Same story with the high resolution - in 2012 until now most here found all kind of excuses why 36 MP are not needed.....now after Canon announced its release, many of the same people who were finding all those excuses are now very interested in this camera and pre-order. Same story now with DR...discussion just to protect the favorite brand for whatever reason.
Mar 27, 2015 at 08:27 PM
Monito
Offline
Details
p.16 #3Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Give it a rest, retrofocus.

You aren't adding anything new to this thread, but RustyBug did.
Mar 27, 2015 at 08:56 PM
RustyBug
Offline
Details
p.16 #4Canon 5DS dynamic range: zero improvement (RAWs analyzed)
I've said a multitude of times that more DR is better and that there is a disappointment that Canon's DR in the lowest ISO range isn't further improved. OTOH, the DR differential @ ISO 400 or greater, really isn't that much (appx 1 EV), if at all at higher ISO's.

What I'm trying to get across is that a correct (or closer to correct) exposure is going to produce a better result than a 5 stop underexposure pulled up. If that warrants additional effort to do so, then so be it when you are aspiring for your best.

All I'm saying is that bracketing ... exposure stacking if you will ... can produce a different result from a 5 stop underexposure. If that is being defensive, then yes I'm defending a different process. This has nothing to do with brand loyalty.

I'm glad to see that you do acknowledge it can produce a better result ... that's all I've been saying is that the extra DR isn't the holy grail to producing your best results possible ... and in context to the OP, the disappointment is noted, but not sufficient to warrant a platform change (maybe that's the sufficiency you were suggesting I made) for a variety of reasons.

I never said that 11 stops was sufficient (noting that the 6D has appx. 12 stops) ... that's kind of self-evident by the advocacy for bracketing. My point is merely that while 14 stops is better than 12 stops (to which I've never disagreed with), bracketing techniques can still produce better results than pull up 5 EV (even with 14 stops available).

But, as "funny" things go ... where in the world was everyone clamoring about DR when Fuji came out with the SuperCCD with DR that smoked everything else 10 years ago and still can hold its own against everything in existence in DSLR (to certain ISO limits).

Best I recall, folks in all camps made "excuses" for not using the Fuji, despite its superior DR ... or were those just personal preferences, that outweighed the benefits of the additional DR back then? How would that be any different from today with someone preferring Canon's glass offerings or AF performance, etc. over the DR different @ ISO 100 - ISO 200 for folks who routinely shoot @ ISO 400 or greater?

I'm not bashing SoNikon ... but DR is just one piece of the puzzle.





Edited on Mar 27, 2015 at 09:24 PM · View previous versions
Mar 27, 2015 at 08:56 PM
mttran
Offline
Details
p.16 #5Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Sony sensor is 3 times (stops) more than canon sensor in the bracketing process. Obviously, sony sensor shall collect wider range data than canon sensor in any cases so what is the points of all these discussions
Mar 27, 2015 at 09:20 PM
Monito
Offline
Details
p.16 #6Canon 5DS dynamic range: zero improvement (RAWs analyzed)
mttran wrote:
what is the points for all these discussions


The point that RustyBug made, and it is a good one, is that consideration of the lighting of the scene is more important than whatever equipment might be at hand or on the dealer's shelf. Lighting is important in photography, mttran.
Mar 27, 2015 at 09:22 PM
gdanmitchell
Offline
Details
p.16 #7Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Monito wrote:
Also merge pano AND do HDR.


How about stitched pano, focus stacking, and exposure blending in one shot? Let's add moving water, trees blowing in strong winds, moving clouds, and, oh, a herd of deer!

I think we may have just identified pure hell. ;-)

Dan

Edited on Mar 27, 2015 at 09:29 PM · View previous versions
Mar 27, 2015 at 09:27 PM
RustyBug
Offline
Details
p.16 #8Canon 5DS dynamic range: zero improvement (RAWs analyzed)
mttran wrote:
Obviously, sony sensor shall collect wider range data than canon sensor


Honest

How does a sony sensor collect a wider range of data?

I understand that it introduces less noise in the ISO 100-400 range, but how does it collect a wider range of data?

Are we not still on a 0-255 (really 0-1) scale.


BTW, last I checked (A7R) 14.07 - (6D) 12.11 = 1.96 ... not 3
(DXO results. If others have other data, presentation is welcome.)

Edited on Mar 27, 2015 at 09:38 PM · View previous versions
Mar 27, 2015 at 09:27 PM
Monito
Offline
Details
p.16 #9Canon 5DS dynamic range: zero improvement (RAWs analyzed)
gdanmitchell wrote:
How about stitched pano, focus stacking, and exposure blending in one shot/
I think we may have just identified pure hell. ;-)


Not exactly one shot, is it? But taking your meaning as "one photograph" ... you have just explained how to replicate Edward Weston's famous pears, ... replicate without film, without an 8x10 camera, without the tent, and without f/64, in less than the 36 hours it took Weston to make the one shot.

Note to sharpness freaks: during the 36 hour exposure thermal expansion and contraction blurred the image slightly. That did not damage the beauty of the image.

I think I'd try the 8x10 first

Mar 27, 2015 at 09:32 PM
chez
Offline
Details
p.16 #10Canon 5DS dynamic range: zero improvement (RAWs analyzed)
EB-1 wrote:
chez wroteI under exposed the shot by three stops allowing me to shoot at ISO 1600
RustyBug wroteAnd how much more DR does the A7R have @ ISO 1600 than the Canon 6D?

Ummm, if DXO measurements are any indication ... that would be .09 EV difference @ 11.21 vs. 11.12

It's a good thing you weren't stuck with Canon's DR ... you'd have never gotten the shot.
(Sorry, couldn't resist.)
What Is that for sure? I thought the A7r was supposed to have the best IQ. If not, then the other crippling factors in using the camera are very difficult to justify.

EBH


Best is very subjective...but it beats the IQ I've gotten from my 5d2...that's a fact.

Mar 27, 2015 at 10:13 PM
chez
Offline
Details
p.16 #11Canon 5DS dynamic range: zero improvement (RAWs analyzed)
StillFingerz wrote:
Chez, do checkout the 6D, it is most difficult to find 'any' banding and as you move above ISO 100 it stays pretty darn clean. Canon has found something, I'm not sure if it is sensor based or in firmware but it's there. I'm foregoing the 5D3 and all it's goodies because of the 5D3's minimal banding when pushed hard.

I do like the idea of 50MPs as I often crop, my wheels make my up-close-and-personal options sometimes very limited. But there'll be no 5Ds or sr in my kit, a 6D is nearing and later perhaps a 5D4 if one
...Show more →

Jerry, it's a very rare occasion that I'll shoot anything above base ISO. My work is basically off a tripod 90% of the time. I understand the 6d is good at high ISO which really does not matter to me. I'm extremely pleased with Sony's results to date.
Mar 27, 2015 at 10:16 PM
RustyBug
Offline
Details
p.16 #12Canon 5DS dynamic range: zero improvement (RAWs analyzed)
chez wrote:
I understand the 6d is good at high ISO which really does not matter to me. I'm extremely pleased with Sony's results to date.


You mentioned shooting @ base ISO by choice ... you'd love shooting an SLR/C ... base ISO is pretty much a one-trick pony.

That's interesting ... I'm thinking the 6D is good @ low ISO too ... likely better than the 5D II at least. That's not to take away from any SoNikon (or other) user's satisfaction. Although it does me how the OP in @ Canon's results in a Canon Forum turns into a brand comparison. We might as well tack on platform changes to m4/3, MF, LF and video as well to the discussion.

Ford uses larger bed bolts in their trucks than Chevy or Dodge (and make it a point of contention to ALWAYS ensure they do so). The strength of that bed bolts might be greater in the Ford ... but that doesn't mean folks should all sell their Dodges and Chevy's to go buy Fords. Some folks just don't like the seats in a Ford and the "disadvantage" of smaller bed bolts in the Chevy or Dodge are not a poison to them. For others the seating in the Ford isn't a problem.


Somehow, I tend to think that Adams or HCB would do just fine with a 6D or an A7R ... no matter which one you put in their hands. Loading a mule with plates to go into back country, etc. ... I'm pretty sure, they wouldn't complain about some extra processing time. That would be horrible compared to remaking a print in the darkroom 100's of times over a span of years to get the final image how you want it.

While harder doesn't necessarily make for better, the tenacity and willingness to think, plan and execute has been a hallmark of photographic excellence throughout its history. Likewise, technological advances have served to reduce the need for the same. The convenience of auto-exposure is nice, as is auto-focus ... yet we still recognize the shortcomings that each of those incur and still rely on manual exposure and manual focus to achieve optimal results at times, i.e. we didn't abandon them just because technology advanced. Technology still can't out think people in all instances.

More DR is nice, but that doesn't mean we should abandon thinking about our light, planning and executing accordingly. At the end of the day, it looks to me like the latest gen of SoNikon has a 2 stop DR over the latest Canon @ ISO 100, 1 stop DR @ ISO 200, 1/2 stop DR @ ISO 400 and it's a dead heat from thereafter.

I enjoy seeing the work of those who use the SoNikon sensor. I've demo'd Pentax to Sony to Nikon ... often times at the insistence of fellow FM'ers of how great they find the non-Canon systems to be. At the end of the day, it comes down to me asking myself if those gains are worth me changing platforms from the reasons why I (formerly a Nikon shooter) invested in the Canon platform.

That's a personal choice that each one of us has to make for ourselves ... and it includes a multitude of factors, DR being but one of them. Knowing your style & flow is key to how much weight those factors carry. I applaud those using their system of choice and producing good works. Currently my good works come from either my SLR/C, my FZ1000 or a Rebel T5i (inherited). If your good works come from a SoNikon while pulling up 5EV, that's your good works, and you should be proud of your good works ... particularly when you know it is your very best effort.

That said ... horses for courses. Sometimes I shoot for convenience, sometimes for excellence. YMMV



Edited on Mar 28, 2015 at 06:13 AM · View previous versions
Mar 28, 2015 at 06:07 AM
jctriguy
Offline
Details
p.16 #13Canon 5DS dynamic range: zero improvement (RAWs analyzed)
mttran wrote:
Sony sensor is 3 times (stops) more than canon sensor in the bracketing process. Obviously, sony sensor shall collect wider range data than canon sensor in any cases so what is the points of all these discussions


Please write in a way that makes your point clear. What you wrote makes absolutely no sense...
Mar 28, 2015 at 06:09 AM
jctriguy
Offline
Details
p.16 #14Canon 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


Would be curious to see the unprocessed photo here. How much did you pull the shadows or drop the highlights?
Mar 28, 2015 at 06:12 AM
shirozina
Offline
Details
p.16 #15Canon 5DS dynamic range: zero improvement (RAWs analyzed)
11 stops of DR is what we are getting so why argue about it. Buy Sony or Nikon if you need more ( I did) from a single exposure or use bracketing workarounds ( as I had to) if you are going to stick with the brand. Even with 14 stops I still need to bracket occasionally so if the next Sony sensor doesn't have 16 stops of DR I'm going to go 😝
One reason many still bracket is because the histogram and clipping are not representative of the RAW data but just the JPEG preview so you often need to bracket anyway to nail the perfect ETTR exposure i'd like to see RAW histograms and 3 channel clipping in future cameras of any brand.
Mar 28, 2015 at 06:32 AM
RustyBug
Offline
Details
p.16 #16Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Charlie N wrote:
that said, I do take a ton of bracket shots with the sony just in case.


"just in case" ...

That sounds a bit like "guessing".

Light values are pretty much a known entity and the Sunny 16 rules of exposure or incident metering or other tools, such as an ExpoDisk can really take the guesswork out of things. I'm a strong advocate of bracketing ... but I tend to only bracket when I know my light values and then choose a bracket value that supports what I want to do with those portions of the scene.

http://en.wikipedia.org/wiki/Exposure_value

Even committing to memory just a few of the EV values (say 3 EV increments) can be helpful to convert "just in case" toward a targeted strategy. That, combined with the inherently available DR, can afford a very efficient bracketing strategy. Even with the ability to do a 5 EV pull ... knowing your 3 EV and 6 EV points can be useful in assessing whether you want to bracket or are confident in the DR's ability to deliver your desired results.

Take particular note of the relationships of 15 EV, 12 EV & 9 EV. 15 EV includes direct sunlight, 12 EV (open shade) does not and you source illumination is the open sky.

Similarly, 12 EV @ sunset vs. 9 EV after sunset is once again the difference between direct sun inclusion and exclusion. Essentially (other factors equal), removing the sun from the equation of illumination on your subject is a 3 EV reduction. Take an incident meter reading of an object facing the sun during the day, then meter the opposite side of the subject where its shadow is produced by the subject blocking the sun, and hmmm @ 3 EV variance.

THIS is the reason that for me that even with the greater DR of the SoNikon that allows for 5 EV pull, that I would still bracket. When I have 3 EV or 6 EV variance in lighting (which certainly exists when including a light source, such as the sun), I'm bracketing to better capture the differing light ... regardless of the camera's ability in DR. Of course, that in part is determined by my goals for the image for each area of differing light. Sometimes, I just want my shadows to be shadows.

A little homework might help change that "ton of bracketing", into a well-thought strategic bracketing approach ... i.e. fewer shots, less culling, etc.
Mar 28, 2015 at 06:37 AM
gdanmitchell
Offline
Details
p.16 #17Canon 5DS dynamic range: zero improvement (RAWs analyzed)
There is a myth that says really good photographers do not guess and that they get it exactly right the first time.

It ain't true.

I do not disparage in the least the importance of being able to make smart and accurate exposure choices, even when working under some degree of pressure. But the truth is that even the greatest photographers often have and often do operate on hunches (experience helps make those hunches better), and they regularly make extra exposures "just in case."

Educated "guessing" is part of the game. It is nothing to sneer at and, in fact, being able to make good guesses quickly and sometimes intuitively is often the mark of an experienced photographer... and it frequently leads to better photographs.

Take care,

Dan

RustyBug wrote:
Charlie N wrotethat said, I do take a ton of bracket shots with the sony just in case.
"just in case" ...

That sounds a bit like "guessing".

Light values are pretty much a known entity and the Sunny 16 rules of exposure or incident metering or other tools, such as an ExpoDisk can really take the guesswork out of things. I'm a strong advocate of bracketing ... but I tend to only bracket when I know my light values and then choose a bracket value that supports what I want to do with those portions of the scene.

http://en.wikipedia.org/wiki/Exposure_value

Even committing to memory just a few of the EV values (say 3 EV increments) can be helpful to convert
...Show more →
Mar 28, 2015 at 07:10 AM
Paul Mo
Offline
Details
p.16 #18Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Okay, okay, but if I shoot a scene at 100iso, 500th, f4 with a 5D3 and a D800 I will still have more detail and malleability in post in preparing for screen or print from the D800.

The SoNikon file is simply nicer to work with.

And that is what Canon users are missing out on and are getting bored/annoyed with Canon because of it.

Edited on Mar 28, 2015 at 07:23 AM · View previous versions
Mar 28, 2015 at 07:21 AM
RustyBug
Offline
Details
p.16 #19Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Dan,

I "guess" frequently too ... but it is rooted in the knowledge and experience (as you mention) and heavily on the Sunny 16 rule or the EV table. I hope that I didn't infer that I was "sneering" at anything (that would be rude, bad form for FM'ers). I was just offering something that I thought might help lighten the load for a fellow FM'er from the burden of a "ton of bracket shots" ... adding, instead to the joy of our beloved craft.

No more than Adams took the time, nor need, to meter for "Moonrise Over Hernandez" do we always have to meter. My point is simply that coupling our knowledge of lighting conditions, combined with our technology can turn a "ton of bracket shots" into just a few strategic ones, with only a little more forethought.

+1 @ experienced shooters go by "hunch" or "gut" ... that's really just subliminal or informal knowledge (same as formal knowledge, but without the explanation). How a person comes to "know" anything can occur from a variety of ways ... all of which are valuable. The EV table is merely another tool available to us. I tend to think that Adams (et al) was quite aware of it ... which (imo) is why his "hunch" @ Moonrise was as effective as it turned out to be.

Whether we build our own table, based on our observations or testing ... or whether we write it down of just store it in our "hunches" ... or we memorize it from another source ... imo, understanding varying lighting conditions is a key component to understanding extreme DR situations, and how that might play out with whatever camera we are using, yesterday, today or tomorrow.

Edited on Mar 28, 2015 at 07:50 AM · View previous versions
Mar 28, 2015 at 07:23 AM
RustyBug
Offline
Details
p.16 #20Canon 5DS dynamic range: zero improvement (RAWs analyzed)
Paul Mo wrote:
Okay, okay, but if I shoot a scene at 100iso, 500th, f4 with a 5D3 and a D800 I will still have more detail and malleability in post in preparing for screen or print from the D800.


More detail as a byproduct of the MP, the 5Ds / 5DsR probably now addresses that.

More malleability ... I can't speak to that since the D700 was the last Nikon I used. I bought it as a byproduct of claims for its greater DR, expecting it to be a miracle cure for crappy lighting. Well, crappy lighting is still crappy lighting, so I stayed the course with Canon.

It is possible for the Sony files to have more malleability ... when I switched from Nikon to Canon years back, I found the Canon files to be more malleable, so I can appreciate that perspective. I just don't know the degree it exists in current models, so I can neither refute, nor support the magnitude of the assertion, but I will trust that good FM'ers telling me they are more malleable is likely rooted in truth to at least a degree and for their workflow.

Mar 28, 2015 at 07:35 AM
FM ForumsCanon ForumJoin Upload & Sell
  
 

You are not logged inLog inRegister