D800E vs. 5D3: Diminishing Returns or Reversal of Returns?

Started Jul 19, 2013 | Discussions
Rick Knepper
OP Rick Knepper Forum Pro • Posts: 15,619
First off, thank you for actually looking at the RAWs...

LMCasey wrote:

Rick,

After playing with your raws, I really can't see any significant difference at all. I think it's all in how you got from your raws to your 1800x1200 jpegs.

...before forming an opinion. If you are suggesting I did it wrong, I would be very interested in your technique. I would then re-conduct the procedure. I am not above learning.

And yes, I've got to believe you are way into the "diminishing returns" at this amount of downsizing.

Thank you again. I do not see what is so hard for some to come out and agree with the outcome. I believe there are more than a few D800 owners who were oblivious to this. In fact, I think there are some of those folks posting here today that are posing as "previously-informed" but in reality just got a big jolt.

If you end up with "less" when starting out with "more" then it's on you.

Is this a reference to my characterization "reversal of return"? If so, I will repeat that I would be very interested in a workflow that avoids this when working with files of this size. If you are inclined to answer but would like to avoid the mindless criticisms others might dump on you, please shoot me a line privately.

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +8 more
Rick Knepper
OP Rick Knepper Forum Pro • Posts: 15,619
Re: Why want to reduce to 1800x1200 as default
1

qianp2k wrote:

rondhamalam wrote:

My preference is to have full scale as DEFAULT.

From there you have the freedom to crop or reduce or even enlarge.

For prints I agreed. But most don't view photos on computer monitors on full-size unless when 4K/8K monitors getting cheaper and popular so full resolution can fit into monitor frame (not directly related to monitor's size but resolution) mostly or even entirely without having to move horizontal bars.

Here are my processed results in CS6 - everything default/zero setting but just upsampling 5DIII file to match to D800E full size.

https://docs.google.com/file/d/0B4tNPhTNWSf3cEZVTVBXUXpTeWc/edit?usp=sharing

https://docs.google.com/file/d/0B4tNPhTNWSf3N3laV3lwNkV5a0k/edit?usp=sharing

Some side-by-side comparisons after upsampling 5D3 file to match to D800E full size.

Sure you can tell difference here or there by pixel-peeping at such D800 full size level. But the difference is pretty small, certainly not significant. 5D3 file suffers from pixelation a bit after upsampling. I don't believe you will notice much difference at 30x20" print or even 40x30" print. Personally I don't print more than 30" wide (very occasionally I print) while 99% of photos are viewed on my computer monitor. I usually downsize to 2000-pixel wide to view on my 24" monitor. Resolution difference certainly is not a big concern to me until I buy a good quality 4K/8K monitor when becomes affordable.

Amidst the rush to discredit me and this comparison, many of the folks in this thread failed to read or conveniently omitted my future plans for a larger editing monitor and a 4k display dedicated to presenting my images. The 1800x1200 file size will be a thing of the past for me but not for many who will continue to view their images on even smaller displays than I have who will have to decide on such matters as a D600 vs a D800 or whatever, not necessarily a Canon product.

In comparison DR is a bigger factor but from what I have seen, D800/D600 DR advantage is only obvious in pushing up a severely underexposed photo but not under normally no mention overexposed photo. I don't see much DR difference in Rick's samples. I guess many bought high-end DSLRs is not just for shooting landscape but for multiple purposes. Then 5D3 may have advantages in action photograph and seems has clear advantage in video ( personally have not used much in shooting video) especially with ML firmware RAW video and now dual-ISO RAW video. Many prefer Canon colors and skin tone although they are subjective and could be processed to be close. Canon has some lenses are better than anybody else such as 17mm and 24mm TS-E, 24-70L II, 70-200L II (slightly better), some primes (not really much sharper but unique look) such as 50L and 85L, super-tele lenses (sharper and much lighter but more expensive).

You cannot get wrong with either camera. Both are excellent cameras and have advantages in certain areas.

P.S. Thanks Rick for doing such wonderful well-controlled tests.

You are most welcome.

-- hide signature --
-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +8 more
Rick Knepper
OP Rick Knepper Forum Pro • Posts: 15,619
Thank you for asking.
2

reginalddwight wrote:

Rick,

Unfortunately, I have not had a chance to read every post in these threads.

What was your primary objective of these tests? What have you concluded?

If my first attempt to explain the process left folks confused, I appreciate the opportunity in a non-combative setting to make myself clearer.

First off, I just acquired both cameras recently, the 5D3 very very recently. I was going to do this comparison whether or not I posted the results. I ran a series of comaprisons:

1.) full-size vs full-size both unprocessed

2.) full-size 5D3 vs down-sized D800E to match the file dimensions of the 5D3 both unprocessed

3.) reducing both files to 1800x1200 px (my most common application but not necessarily anyone else's) both unprocessed

4.) same as number three and processing with my normal workflow

5.) making an extreme crop and downsizing to 1800x1200 px unprocessed

6.) making an extreme crop and downsizing to 1800x1200 px and processing with my normal workflow

There were some other noodlings I did but I chose to post images from Nos. 1, 2, 4 & 5. You'll have to look at both threads to find all of the images. Link located in the OP.

Here's what I concluded and decided to share with the forums of both brands

1.) In the past, at the file size 1800x1200, I observed that 21/24 MP files showed visibly more detail than 12 MP files without throwing away too much data. I knew there would be a diminishing return for the D800E but I didn't know how much there would be until I actually had a D800E in hand. I wasn't prepared for the possibility that too much information could be thrown away. I thought this finding might be of some interest to folks who are trying to make a decision between differently spec'ed sensors. I have a D3x and could have used that camera in order to eliminate brand predjudice that's rampant here but I wasn't thinking about that at the time of the comparison which was a personal project to start off with.

(If my methodology sux, I am man enough to admit a mistake though I don't think anything I did or did not do would impact the result dramatically. However, I am open to visual demonstration of a better workflow or a well-documented description of said workflow to replace the one I used.)

2.) I also wanted to show that a diminishing return doesn 't mean there is no return at all and challenged folks, who haven't already, to find that tipping point for their OWN applications, 95% percent of which I believe involve reductions.

3.) Express the idea that down-sizing to a given dimension seems to have a greater impact on the larger file as proximity to this tipping point nears just as the opposite must be true for enlargements.

4.) I was careful to say that the slight degradation in MY D800E sample caused by the severe down-sampling didn't harm the photo from being viewed at normal distances. In fact, I am enjoying a full set of 1800x1200 shots from a recent trip to Zion NP without it eating at me.

I was hoping to get more friendly and helping support from folks in order to arrive at a truthful conclusion for all of us. By the way, there is one lingering doubt that I documented in the original thread and mentioned above briefly but no one has latched on to it yet.

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +8 more
Rick Knepper
OP Rick Knepper Forum Pro • Posts: 15,619
Re: D800E vs. 5D3: Diminishing Returns or Reversal of Returns?

Jon Rty wrote:

Rick Knepper wrote:

And yet, is it your job to advise me as to what my job is? I suggest that you debug your logic before posting.

Do you have these hardware calibrated profiles handy? if so, send them over and I will be glad to re-conduct the comparison.

I used the "former", same settings including the same profile, Adobe Standard, which is available to everyone who uses ACR. However, I would like for TOF guy to confirm that he was refrring to camera profiles when he used the term "Curve".

Suffice it to say that the images have been neutralized to the best of my ability and consistent with other resolution comparisons posted here in the past and where you haven't brought into question the camera profile used.

For the first set of images in this thread, I used profiles developed for each individual model.

I have provided the RAWs. You are welcomed to show me how to neutralize the files for the purpose of comparing resolutions. Hint: a camera profile should not get in the way of comparing resolutions. I think most objective readers understand the difference between resolution and a tone curve.

I'll repeat the following:

"a camera profile should not get in the way of comparing resolutions. I think most objective readers understand the difference between resolution and a tone curve."

Not trying to be combative here so bear with me. I do not want the thread to get locked. Do you know the difference between a tone curve and resolution when you see it? I bet you do. So, I bet most everyone looking in can make the distinction too - eventually.

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

Nothing wrong with my logic. I'd suggest you don't use strawman arguments.

Are you kidding me? Your post is the epitome of Strawman.

http://www.nizkor.org/features/fallacies/

I'm one of the readers who's behavior you presume to control. I'm certainly entitled to inform you that you presume too much.

Presume to control? Look at the link man. Help yourself.

I can't provide a hardware calibrated profile for *your* cameras, as I don't have access to them. You need to have a camera to be able to properly profile it, just like with a display. If you want to do that, invest in a ColorChecker or similar.

"Adobe Standard" is a different profile for each camera model. Hence the delay between release of a new camera and ACR support. You can think of it like Adobe 5DIII Standard and Adobe D800E Standard. They're two separate profiles producing differing results. Adobe doesn't provide profiles that match the output across camera models, you'll have to do that yourself.

Contrast affects resolution, as resolution is measured at a contrast level. Curves affect contrast. As a profile directly affects the tone curve and thus contrast, saying that a profile "shouldnt?" affect resolution just doesn't make any sense. That's basically saying that a profile shouldn't affect the tone-curve or contrast, which is impossible. All conversions need to apply a tone curve.

Here are the two RAWs you posted, quickly matched by eye-balling. Far from perfect,

Though I am being held by many in this thrfead to a standard nearing perfection, I'll let this imperfection slide.

but as good as I'll get them with the time I'm willing to invest:

Okay, great. You need this. And, by the way, this is what I hoped folks would do for themselves.

So, what is your conclusion regarding resolution now that you've equalized the files properly.

One

Two

All I did was adjust curves and white balance. And here's one of the above with contrast removed.

Flat

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +8 more
reginalddwight Senior Member • Posts: 2,003
Re: Thank you for asking.
1

Many thanks for the detailed explanation.

After reading the many confrontational posts in this thread, I was expecting at some point the mods to lock this thread.

I appreciate your efforts in the comparisons and think your workflow and processing of the series of files are legit.

Try not to let some of the jokesters on this forum get to you.

Cheers.

Kaj E Veteran Member • Posts: 9,346
Re: Overkill

marcio_napoli wrote:

Comparisons in DP review forums:

Present some test chart comparion: "nahh... you need a real world scenario. This test chart won't do it".

Present some real world comparison: "nahh... too many variables here. You need to equalize colors, brightness, details, enlarge 300% or maybe downsize 400%, then normalize DR, apply lens, moire correction, etc etc etc etc etc.

In other words, please present some test chart, will ya?"

And things get trapped forever in that boring, absolutely pointless loop

-- hide signature --

Marcio Napoli

www.marcionapoli.com

Replying to me?

-- hide signature --

Kind regards
Kaj
http://www.pbase.com/kaj_e
WSSA member #13
It's about time we started to take photography seriously and treat it as a hobby.- Elliott Erwitt

Kaj E Veteran Member • Posts: 9,346
Re: You are wrong and missed the point.

Rick Knepper wrote:

Kaj E wrote:

If you only need 1800x1200 pixel output both cameras are overkill for your needs.

Reread.

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

Why buy a high resolution camera if all you are intersted in is size reduction. You wrote:

>"The criticism came mainly from those whose applications apparently call for enlargement. I mean, >do we even need to see such a comparison? Can’t we just stipulate that more MPs will >always help in Printing Large.

>It is that other 95% of photographers and their applications (myself included) which >involve reductions that need some special illumination. I am sure this percentage could be >hotly debated but for example, 100% of my applications involve reductions.

>My primary application: create 1800x1200 images for display on my NEC 2690 monitor. There are >future applications I am consideing but I will save that discussion for the end of this post. >1800x1200 isn’t a typical web-sized image, it’s nearly as physically large as a 13x19 print when >viewed on my monitor though the dimensions are significantly reduced from its originals. I don’t >know what pixel pitch folks are viewing with or how DPI factors into how others view these files.

>So, we know information is being thrown away as we down-size, but what is the point >at which too much information has been thrown away?"

Downsizing always reduces IQ (resolution), there is no magic point of dimishishing returns. When you print you should never downsize, only upsize if needed to acheive max printer native resolution (720ppi for Epson 600ppi files for Canon printers). This means you don't even have to print large.

http://www.luminous-landscape.com/forum/index.php?topic=48894.0

See Jeff schewe's comment at end of page 1.

If you need to downsize for your output the D800 and 5D3 are overkill, provided your lenses and technique (inlcuding diffraction and lens aberrations) are up to snuff.

-- hide signature --

Kind regards
Kaj
http://www.pbase.com/kaj_e
WSSA member #13
It's about time we started to take photography seriously and treat it as a hobby.- Elliott Erwitt

Bills nature photo Regular Member • Posts: 377
Re: Way to put 5D III curve into D800E or PS for raw conversion?

Rick Knepper wrote:

Bills nature photo wrote:

TOF guy wrote:

Rick Knepper wrote:

TOF guy wrote:

yray wrote:

. The D800 image in this example appears washed out even though I believe that at the pixel level it contains more fine detail.

That's because different curves were used for the images. It has nothing to do with "fine details" and "micro-contrast".

-- hide signature --

Thierry - posted as regular forum member

While that is true for the first set, it is not true of the 2nd set.

It is true for both images.

-- hide signature --

Thierry - posted as regular forum member

I agree that the curves are quite different and I must say that I like the curves applied to the 5D III image much better than the standard Nikon curves. Is there an easy way to upload a Canon like curve into D800 or Photoshop for processing D800E raw files to have the contrast curve of the 5D III?

-- hide signature --

Too much glass...too little time

TOF guy failed to answer my question. Maybe you'll be so kind as to tell me what type of curve are you referring to i.e. a PPing Curve as found in Photoshop?

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

Hi Rick,

Yes I guess.  I am referring to the tone curve which you adjust in Photoshop post processing.  You can change it in part by selecting "camera profile" versus "Adobe profile" under the camera tab of ARC raw converter.  In the old days you could upload tone curves, but I haven't seen anyone offer a set of Tone curves, that will make a Nikon have a similar tone curve to the standard Canon default tone curve.    As many have pointed out,   The Canon default curve appears to have higher contrast and maybe saturation setting than does the Nikon default curve which looks quite bland.  You can fix it in post, but would be nice it started a little snappier without going to the over the top setting like Vivid.

Bill

-- hide signature --

Too much glass...too little time

 Bills nature photo's gear list:Bills nature photo's gear list
Canon EOS-1D X Canon EOS 5D Mark III Canon EF 400mm f/5.6L USM Canon EF 70-300mm f/4-5.6L IS USM Canon EF 500mm f/4.0L IS II USM +3 more
SubPrime Senior Member • Posts: 1,238
A comparison using Capture One - yields different results

Thanks for taking the time to do this comparison.

Based on my own test, it seems the outcome has much to do with the choice of raw converter and the resizing algorithm. Anyone can download a copy of Capture One 7 and repeat this if they want. The D800E images look sharper and more contrasty to me and there is visibly more detail.

5DIII

D800E

Whether you downsize the whole image or a crop, you are still throwing away detail are you not?

The raw files you originally provided obviously show more detail in the D800E file.

anotherMike Veteran Member • Posts: 8,999
Quick thoughts after using another raw converter (to add fuel to fire)

One item I might describe as a "flaw" (excuse my choice of words, it's late) in your test was your choice of ACR for raw conversion. I've always found it to be a bit better for Canon files than for Nikon, and I think that might be impacting your results.  Obviously if you are set on ACR as your converter, then my experiment would be of no use to you.

So I used something I deem a bit nicer than ACR for this comparison - version 1.05 of photo ninja and have provided the 1800x1200 (or so) max quality jpegs from the conversion. Frankly at this much reduction, it's a toss. There is some very slight contrast difference remaining, although I think this is a closer comparison, and some color differences, but I still would tend to say at this point (1800x1200 downsize) neither camera really does anything the other can't do. Downsizing done in CS5 using bicubic sharper.

Anyway, certainly not enough difference for all the arguing and venom in this thread - which I won't participate in other than to provide these photo-ninja converted jpegs.

Nikon at 1800x1201 pixels, converted by photo-ninja at default settings, using high quality demosaic option

Nikon version (conversion: photo ninja. downsize: bicubic sharper in cs5)

Canon version, also converted in photo ninja, default settings, using highest quality demosaic option. Downsized in cs5 the same way.

Canon version (conversion: photo ninja. Downsize: bicubic sharper in cs5)

anotherMike Veteran Member • Posts: 8,999
Re: A comparison using Capture One - yields different results

Ah man, you beat me to it. I had the same thoughts - try a different (and IMO better) raw converter, and I went with photo ninja. Like you, found the nikon raw files to be definitely superior, but at 1800x1200, man, at that point you've tossed a lot away, you know? Not sure it's worth arguing about at that much of a downsize.

-m

Jon Rty Veteran Member • Posts: 3,838
Re: D800E vs. 5D3: Diminishing Returns or Reversal of Returns?
1

Rick Knepper wrote:

Though I am being held by many in this thrfead to a standard nearing perfection, I'll let this imperfection slide.

Okay, great. You need this. And, by the way, this is what I hoped folks would do for themselves.

So, what is your conclusion regarding resolution now that you've equalized the files properly.

No, not near perfection, not just completely different. As others already showed, even standard settings produce shots much closer to each other than what you posted.

My conclusion is the same everyone has been saying the whole thread, if you downsize to 2Mp, there are no differences in resolution to be seen anymore. You could do the same with a 5D and the D800, and reach the same conclusion.

Flashlight Veteran Member • Posts: 7,294
Wow, easy Rick...

You can play the 'fanboy' card any way you like but I'd have more respect for you if you could gracefully handle some criticism.

Below I compared an ISO 100 shot of the 12MP D90 (still available new) to the 36MP D800E both resized to 1800 pixels wide:

The price difference between the shots is €2849-€459=€2390. Waste of money IMO.

Rick Knepper wrote:

You and the rest of the Nikon forum fanbois are making no sense to me. You have the RAWs at your fingertips to do whatever testing you deem necessary in whatever manner you desire and yet, you insist on "proving" I did something wrong - because - this reaction is an exercise in discrediting the source. To what end? People have eyes.

So, you belong to the brigade senselessly attacking the un-processed pair while others take the flank and attack the processed pair. Nice.

I wonder if anyone is wondering by now why any of the detractors haven't pulled the RAWs to show me how it is done.

Flashlight wrote:

In the final crops I notice the following (see my 200% crop):

- In the top left the 5D3 has moire; This is of course not a 'fault' of the camera, the sensor just happened to match the frequency.

- The images have massive CA (see right bottom of the 200% crop); If that doesn't bother you why start these kind of comparisons in the first place? Both CaptureNX2 and Photoshop have a solution for this.

I am going to repeat something for you.

...you belong to the brigade senselessly attacking the un-processed pair while others take the flank and attack the processed pair. Nice.

So if I remove the CA, I am criticized by some, if I leave the CA, I am criticized by some.

- The contrast (and overall processing) is totally different between the two images which makes the impression of 'sharpness/detail' different.

Not if one knows what they are doing and seeing.

In a more common scope, logic and experience dictate that if you start with more information and downsample you end up with more acuity in the image. Trying to prove the opposite with sloppy PP is a waste of time and will convince no one.

This is a ridiculous statement devoid of foundation and particulars. Even if someone wants to be a Nikon fanboi, they'll recognize the vast emptiness of this assertion.

But thank you for pumping a whole lot of unnecessary money into the R&D funds of both Canon and Nikon ( as you could just as well have used an old 10MP camera for these shots and your end use). We respect you for that.

You have got a lot education ahead of you in digital imaging if you think a FF 10 MP camera can produce this kind of detail of either camera.

Rick Knepper wrote:

5D3 cropped & downsized to 1800x1200

View: original size

D800E cropped & downsized to 1800x1200

View: original size

-- hide signature --

Philip

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

-- hide signature --

Philip

Rick Knepper
OP Rick Knepper Forum Pro • Posts: 15,619
D700 vs. D800

Jon Rty wrote:

Rick Knepper wrote:

Though I am being held by many in this thrfead to a standard nearing perfection, I'll let this imperfection slide.

Okay, great. You need this. And, by the way, this is what I hoped folks would do for themselves.

So, what is your conclusion regarding resolution now that you've equalized the files properly.

No, not near perfection, not just completely different. As others already showed, even standard settings produce shots much closer to each other than what you posted.

The point of the first set of images posted in this leg of the comparison which you seem to keep missing isn't to produce identical looking images for a casual viewer. It was intended to show the end results of a similar workflow. It is but one of several ways to compare images I believe an intelligent person would or should conduct if they have the opportunity. There are equalized images elsewhere.

To me, equalizing means to equalize the settings, to the degree that is available to the typical user, in order to see what the native differences are. You seem to be using the term like it is used in audio for non-comparison purposes. Fair enough. All you've done is processed the image differently than I have for a given environment. We all know that one can use Photoshop to give an image any look desired. Thanks for illustrating that. Most people want to see the native differences though.

Without references of my own, I can't, in all good conscience, call into question your statement regarding the Adobe Standard Profile. Suffice it to say that going 6 pages deep into a Google search, I could find no documentation one way or another from Adobe that Adobe Standard is individually calibrated for each camera which is not to say it isn't there somewhere. If you have that reference handy, please post it.

But, let's say you are right, I could find no one else on the Internet besides yourself who believes one can't use the Adobe Standard profile to compare camera/lens combos of different cameras. In fact, it seems it is the standard profile for this type of thing outside of review sites. Aside: if you are right, one wonders how a native difference in sensors can be produced by the typical consumer.

IWO, your processing while valid on one level could be bogus on another depending on what the goals are. IMO of course. Your goal seems to be discrediting my posts. I had to otherwise drag a conclusion out of you.

My conclusion is the same everyone has been saying the whole thread, if you downsize to 2Mp, there are no differences in resolution to be seen anymore.

Let me help you with this last part: "...there are no differences in resolution that I [Jon Rty] can see anymore." I believe your conclusion is a valid one, by the way. It is what you are seeing.

Though I understand the "safety in numbers" fall back (again, check the fallacy website), what "everyone" else thinks is irrelevant (why? because we don't understand the motivations involved - and this "everyone" else entity has grossly miscalculated my motivations).

I would urge all to conduct their own comparisons as you have.

I would disagree wholeheartedly with a blanket statement that every down-sizing to 2 MB would produce "no differences in resolution".

You could do the same with a 5D and the D800, and reach the same conclusion.

This might be true for certain down-sizings but not for the reasons you are trying to imply. We would just have to see the results. You have to prove your assertions here buddy with a process we can all agree on. If you have a D700 and a D800, let's keep this in the family and post up some RAWs of a scene with reasonable detail and let's have a look-see.

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +8 more
Rick Knepper
OP Rick Knepper Forum Pro • Posts: 15,619
Re: Way to put 5D III curve into D800E or PS for raw conversion?

Bills nature photo wrote:

The Canon default curve appears to have higher contrast and maybe saturation setting than does the Nikon default curve which looks quite bland. You can fix it in post, but would be nice it started a little snappier without going to the over the top setting like Vivid.

One wonders then, how to arrive at the base difference between the two sensors for the purposes of comparing and contrasting with a method agreeable to everyone outside of a lab.

Aside: During a Google search for another response in this thread, I ran across mention of a 3rd party software providing camera profiles that can turn your Nikon RAW image into one with a Canon look, Leica, many cameras. There were several links in the search leading to threads on DPR among other sites in addition to the maker's website. I wonder if using one of these profiles for both RAWs could quiet the objections without manually manipulating the files. By the way, I've purchased 3rd party profiles before and while they are handy for a standardized starting point, they provide nothing that can't be done (or undone?) manually further into the process. What I can't do is replicate a Leica camera look (unless it is by accident) which I think might be cool to do. I'll have to dig up that name again.

Vivid is horrid.

Thanks for the courteous and helpful response.

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +8 more
Rick Knepper
OP Rick Knepper Forum Pro • Posts: 15,619
Re: Thank you for asking.

reginalddwight wrote:

Many thanks for the detailed explanation.

After reading the many confrontational posts in this thread, I was expecting at some point the mods to lock this thread.

I appreciate your efforts in the comparisons and think your workflow and processing of the series of files are legit.

Try not to let some of the jokesters on this forum get to you.

Cheers.

Thanks. I am trying to tone down my own reactions to prevent a lock down.

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +8 more
LMCasey Contributing Member • Posts: 851
Re: First off, thank you for actually looking at the RAWs...

Rick,

I opened the raws using the same process and camera profile in ACR 6.7. The parameters used in ACR were the same for both files. No sharpening was applied during conversion. Both files were then resized in PS using bicubic to a long side of 1800. At this point, I looked at sharpening parameters using both smart sharpen and USM. It's at this point that I've got to say that I would probably be lying if I said I saw a difference. If anything, it may be possible that the D800E version has very slightly lower contrast. In the end you have images that are extremely similar.

And yes, I was referring to reversal in my last comment. In an earlier peak at your two raws (from your other thread), when looking at 100% after converting, you can see the D800 file does indeed contain more information (has resolved more detail), and that the Canon image looks a bit muddy in spots (AA?) Therefore, if you see something "less" in the final converted Nikon jpeg than you do in the final Canon jpeg, then it's something you've done along the way.

I don't know if I've been helpful at all, but I've tried. Yes these forums can be challenging, I agree with you on that.

 LMCasey's gear list:LMCasey's gear list
Canon EOS 5D Nikon D800 Canon EF 35mm F1.4L USM Canon EF 135mm F2L USM Canon EF 17-40mm f/4.0L USM +11 more
Rick Knepper
OP Rick Knepper Forum Pro • Posts: 15,619
ACR vs. Capture One

anotherMike wrote:

One item I might describe as a "flaw" (excuse my choice of words, it's late) in your test was your choice of ACR for raw conversion. I've always found it to be a bit better for Canon files than for Nikon, and I think that might be impacting your results. Obviously if you are set on ACR as your converter, then my experiment would be of no use to you.

Yes, I am sort of landlocked with ACR but an excellent point nonetheless. The squabbling elsewhere is coming down to my use of Adobe Standard for both brands and not equalizing further. I used to own capture One back in the day but I don't remember much about it.

So I used something I deem a bit nicer than ACR for this comparison - version 1.05 of photo ninja and have provided the 1800x1200 (or so) max quality jpegs from the conversion. Frankly at this much reduction, it's a toss.

Legitimate observation. Thanks so much. While I tend to agree with this assessment generally, I will disagree with critical viewing in mind. Bear with me just a moment.
I downloaded both images and have examined them at full size on my monitor, an NEC2690, at closer than normal viewing distances. To my eye, your D800 rendering has more resolution than your 5D3 rendering or my D800 rendering. I would specifically direct you to the street at the horizon for comparison purposes. This is a critical distance in my images for my tastes.

You did not mention all of the steps involved in your downsizing. You alluded to downsizing in CS5. Did you convert at full size so there was but one step to 1800x1200? This perhaps could be where our paths diverge.

There is some very slight contrast difference remaining, although I think this is a closer comparison, and some color differences, but I still would tend to say at this point (1800x1200 downsize) neither camera really does anything the other can't do. Downsizing done in CS5 using bicubic sharper.

Anyway, certainly not enough difference for all the arguing and venom in this thread - which I won't participate in other than to provide these photo-ninja converted jpegs.

Nikon at 1800x1201 pixels, converted by photo-ninja at default settings, using high quality demosaic option

Nikon version (conversion: photo ninja. downsize: bicubic sharper in cs5)

Canon version, also converted in photo ninja, default settings, using highest quality demosaic option. Downsized in cs5 the same way.

Canon version (conversion: photo ninja. Downsize: bicubic sharper in cs5)

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +8 more
Rick Knepper
OP Rick Knepper Forum Pro • Posts: 15,619
The crops have been the forgotten comparison.

SubPrime wrote:

Thanks for taking the time to do this comparison.

Based on my own test, it seems the outcome has much to do with the choice of raw converter and the resizing algorithm. Anyone can download a copy of Capture One 7 and repeat this if they want. The D800E images look sharper and more contrasty to me and there is visibly more detail.

5DIII

D800E

Whether you downsize the whole image or a crop, you are still throwing away detail are you not?

The raw files you originally provided obviously show more detail in the D800E file.

This is absolutely true. I made it very clear that for up-sampling purposes, there is no need for a comparison which was my way of saying that the D800E (a 36 MP sensor to keep it generic) captures more information, a point I have hammered home time and time again here and at the Canon forum.
I intended to make the following the subjet of a separate post with a subject line to the effect: cropping a section out of an image and then down-sizng the crop to 1800x1200 is not the same thing as down-sizng the entire image. But you've offered me a seque to point out that not all 1800x1200 images are equal and a fine example of such.

I've downloaded anothermike's down-sized uncropped ff images processed with Capture One and have determined that the D800E image has the most resolution by a hair though it has to be discerned at closer than normal viewing distances but which is still consistent with the main theme: diminishing return.

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +8 more
Rick Knepper
OP Rick Knepper Forum Pro • Posts: 15,619
Sorry...

I am trying to tone it down from yesterday's "passion".

Flashlight wrote:

You can play the 'fanboy' card any way you like but I'd have more respect for you if you could gracefully handle some criticism.

Below I compared an ISO 100 shot of the 12MP D90 (still available new) to the 36MP D800E both resized to 1800 pixels wide:

The price difference between the shots is €2849-€459=€2390. Waste of money IMO.

Rick Knepper wrote:

You and the rest of the Nikon forum fanbois are making no sense to me. You have the RAWs at your fingertips to do whatever testing you deem necessary in whatever manner you desire and yet, you insist on "proving" I did something wrong - because - this reaction is an exercise in discrediting the source. To what end? People have eyes.

So, you belong to the brigade senselessly attacking the un-processed pair while others take the flank and attack the processed pair. Nice.

I wonder if anyone is wondering by now why any of the detractors haven't pulled the RAWs to show me how it is done.

Flashlight wrote:

In the final crops I notice the following (see my 200% crop):

- In the top left the 5D3 has moire; This is of course not a 'fault' of the camera, the sensor just happened to match the frequency.

- The images have massive CA (see right bottom of the 200% crop); If that doesn't bother you why start these kind of comparisons in the first place? Both CaptureNX2 and Photoshop have a solution for this.

I am going to repeat something for you.

...you belong to the brigade senselessly attacking the un-processed pair while others take the flank and attack the processed pair. Nice.

So if I remove the CA, I am criticized by some, if I leave the CA, I am criticized by some.

- The contrast (and overall processing) is totally different between the two images which makes the impression of 'sharpness/detail' different.

Not if one knows what they are doing and seeing.

In a more common scope, logic and experience dictate that if you start with more information and downsample you end up with more acuity in the image. Trying to prove the opposite with sloppy PP is a waste of time and will convince no one.

This is a ridiculous statement devoid of foundation and particulars. Even if someone wants to be a Nikon fanboi, they'll recognize the vast emptiness of this assertion.

But thank you for pumping a whole lot of unnecessary money into the R&D funds of both Canon and Nikon ( as you could just as well have used an old 10MP camera for these shots and your end use). We respect you for that.

You have got a lot education ahead of you in digital imaging if you think a FF 10 MP camera can produce this kind of detail of either camera.

Rick Knepper wrote:

5D3 cropped & downsized to 1800x1200

View: original size

D800E cropped & downsized to 1800x1200

View: original size

-- hide signature --

Philip

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

-- hide signature --

Philip

-- hide signature --

Rick Knepper, photographer, non-professional, shooting for pleasure, check my profile for gear list and philosophy. TJ said, "Every generation needs a new revolution".

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW Pentax DA645 28-45mm F4.5 +8 more
Keyboard shortcuts:
FForum MMy threads