Why Canon made 1Dx best specification ... II

Started May 24, 2013 | Discussions
jjnik
Senior MemberPosts: 1,286Gear list
Like?
Re: Nice colors
In reply to qianp2k, May 25, 2013

qianp2k wrote:

Mako2011 wrote:

qianp2k wrote:

  I reduced uncropped original photo to 2000-pixel wide. 

5D3+100-400L full scene

- Heavy Cropped Image (not reduced in size - so you can view at 100% if you want to):

Now here is 100% cropped on the pelican ONLY.

above 5D3+100-400L 100% cropped

there is no difference by showing a small portion of photo and entire photo at original size that may not fit into your monitor.

I think your wording might be adding to your confusion

Not that is your part.

Obviously less resolution in the 5D pic

less size. But it contains more textual details.  Check both birds' bodies, it's pretty clear my one contains more feather details.  Please be fair not because you're a Nikon shooter and many times I don't share with your opinions.

but I'm a little surprised at the noise in the blue channel at ISO 240.

ISO 250, blue channel is from reflection of blue ocean.  Think about, his photo taken at 1/4000 while my one taken at 1/1000.  Sure it's one stop difference of aperture f2.8 vs f5.6.  But still his photo received more light and obviously the bird got sunlight lit.  My pelican unfortunately was in shadow that together adding a bit of noise, not much and I can NR more that I can do later.

Not as much light as you think - I shot shutter priority at 1/4000 as these birds were lightning fast and it was underexposed wide open.  I added +0.67 stop in post and added 75 (out of 100) in shadow recovery - both during RAW conversion in Capture NX2.

To compare, here's the processed image I posted:

And here's the much darker original capture:

 jjnik's gear list:jjnik's gear list
Nikon D4s Nikon D700 Nikon D800E Nikon 1 AW1 Nikon AF-S Nikkor 24-70mm f/2.8G ED +11 more
Reply   Reply with quote   Complain
Mako2011
Mako2011 MOD
Forum ProPosts: 15,594
Like?
What do you see
In reply to qianp2k, May 25, 2013

qianp2k wrote:

I don't expect you will be fair as you seldom are.  So whatever you said..

Check two birds. My one obviously has more feather details on both bird's body and wings.  Sure my one is a little bit noisy (higher ISO and under shadow) that I can add a bit more smart NR, and that little noise will not show up on prints that effectively NR.

What do you see? Here is the comparison at 100% I'm having an easier time distinguishing individual feather detail in the top pic...though the lower pic does show more feathers

- From the 22 MP 5D3:

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

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

Reply   Reply with quote   Complain
Mako2011
Mako2011 MOD
Forum ProPosts: 15,594
Like?
Bigger than I thought
In reply to jjnik, May 25, 2013

jjnik wrote:

qianp2k wrote:

Mako2011 wrote:

qianp2k wrote:

  I reduced uncropped original photo to 2000-pixel wide. 

5D3+100-400L full scene

- Heavy Cropped Image (not reduced in size - so you can view at 100% if you want to):

Now here is 100% cropped on the pelican ONLY.

above 5D3+100-400L 100% cropped

there is no difference by showing a small portion of photo and entire photo at original size that may not fit into your monitor.

I think your wording might be adding to your confusion

Not that is your part.

Obviously less resolution in the 5D pic

less size. But it contains more textual details.  Check both birds' bodies, it's pretty clear my one contains more feather details.  Please be fair not because you're a Nikon shooter and many times I don't share with your opinions.

but I'm a little surprised at the noise in the blue channel at ISO 240.

ISO 250, blue channel is from reflection of blue ocean.  Think about, his photo taken at 1/4000 while my one taken at 1/1000.  Sure it's one stop difference of aperture f2.8 vs f5.6.  But still his photo received more light and obviously the bird got sunlight lit.  My pelican unfortunately was in shadow that together adding a bit of noise, not much and I can NR more that I can do later.

Not as much light as you think - I shot shutter priority at 1/4000 as these birds were lightning fast and it was underexposed wide open.  I added +0.67 stop in post and added 75 (out of 100) in shadow recovery - both during RAW conversion in Capture NX2.

To compare, here's the processed image I posted:

And here's the much darker original capture:

Much bigger dif than I though from just looking at the EXIF. Thanks for the comparison.

-- hide signature --

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

Reply   Reply with quote   Complain
jjnik
Senior MemberPosts: 1,286Gear list
Like?
Re: Bigger than I thought
In reply to Mako2011, May 25, 2013

Mako2011 wrote:

Much bigger dif than I though from just looking at the EXIF. Thanks for the comparison.

No problem - I forgot about the EXIF not reflecting the exposure and shadow pulling during RAW conversion and didn't mention that when I originally posted.

 jjnik's gear list:jjnik's gear list
Nikon D4s Nikon D700 Nikon D800E Nikon 1 AW1 Nikon AF-S Nikkor 24-70mm f/2.8G ED +11 more
Reply   Reply with quote   Complain
qianp2k
Forum ProPosts: 10,000Gear list
Like?
Re: What do you see
In reply to Mako2011, May 25, 2013

Mako2011 wrote:

qianp2k wrote:

I don't expect you will be fair as you seldom are.  So whatever you said..

Check two birds. My one obviously has more feather details on both bird's body and wings.  Sure my one is a little bit noisy (higher ISO and under shadow) that I can add a bit more smart NR, and that little noise will not show up on prints that effectively NR.

What do you see? Here is the comparison at 100% I'm having an easier time distinguishing individual feather detail in the top pic...though the lower pic does show more feathers

What kind of details on feathers, and of course that bird has less feathers

Let's check entire body details that would be much clearer

- From the 22 MP 5D3:

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

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

-- hide signature --
 qianp2k's gear list:qianp2k's gear list
Sony RX100 Canon EOS-1D Mark III Canon EOS 5D Mark II Canon EOS 5D Mark III Sony Alpha 7R +20 more
Reply   Reply with quote   Complain
Great Bustard
Forum ProPosts: 23,567
Like?
Re: *Sigh*
In reply to qianp2k, May 25, 2013

qianp2k wrote:

Great Bustard wrote:

qianp2k wrote:

Great Bustard wrote:

qianp2k wrote:

There are too many sentences and I don't want to reply one on one basis as most of them are the same regarding below the key point.

Finally something we can both agree on! 

Great Bustard wrote:

...or print to their supposed sizes such as 20x30" vs 40x60".

When printed at the same size (and why would we be comparing systems in any other matter), more pixels will yield more detail every time (all else equal).

Yeah a big ONLY if all else equal, that's the point.

Well, I mean, duh.  The point is, everything can be made equal except the lenses (although the lenses are often, if not usually, close enough).

But how to make everything else equals?

You can't.  You compare the systems on the basis of the photos you intend to take with the systems and the lenses you expect to use.

That's the point that you cannot.  I'd agree with you if truly everything equals.  But in reality they don't.  Canon 24-70L II is better than Nikon counterpart...

According to DxOMark, not according to PZ.  The problem with these tests, among other things, is that they test only a single copy of the lens:

http://www.slrgear.com/articles/variation_canon50f14/canon50f14.htm

...Canon 70-200L II is better than Nikon counterpart, Canon 24mm TS-E II is better, Canon 17mm TS-E is unique and better than 14-24 at 17mm in edges/corners clearly.

Talk about cherry-picking examples.  The 17 TSE vs the 14-24 / 2.8?  Really?

Do you have to use a better technique such as tripod and the best lens on D800?

No.  Do you think I use a tripod with my 20 MP 6D when I came from a 13 MP 5D?

So you can see difference at 2000-pixel wide or at 13x19" print? I cannot between my 5D3 and 5D1 except former is slightly better in DR.  I don't think you can tell difference between 5D3 and D800 at 3000-4000 pixel wide or at 18x26" or even 20x30" prints.  I read several due system owners said so.  In order to print really well at 60" wide, you do need to put D800 on tripod while you have a good chance to print hand-held photos from 5D3 at 30" wide.  To see clear difference at 40" wide, I do believe you'd need put D800 on tripod or still very small if both shoot hand-held.  5D3 files can upsampling thru special software and be pretty comparable to D800 to print at 40" wide.

What makes you think the 5D3 files will look nicer by upsampling but the D800 files will not?

Otherwise the difference is much smaller due to imperfect techniques and lens.

You know, it really isn't.  Indeed, you are quite the hypocrite, as I made this exact point:

Please don't get excited and take personal.

http://www.dpreview.com/forums/post/51518848

and in your response to that point, you added more support to it.

That is your post.  I am not sure what you talking about.  But 5D2 that OP talking about is not a high density camera compared to D800.  it affected much less by shooting in hand-held.

I actually read you're hypocrite a few weeks' ago in Open Forum that you boated your 6D shallower DOF and you can shoot wide open in low light not to worry about DOF much, that contradicts to your posts in mFT forum that you keep emphasizing to shoot on eq. DOF, that I disagreed.  So it seems you talked differently in different forums.

What you read is entirely different from what you wrote.  You can find the link and quote, but if I said anything about it, it would have been "Since I love shallow DOF photography, the added bonus of getting less noise at the wider apertures on FF works well for me".

To achieve all else equal such as similar per-pixel level IQ...

And again, Peter, per-pixel IQ is meaningless.  It's the photo we're talking about here.  I mean, if we're not talking photos, then there's no point to this "conversation".


But that directly affects cropping capability and print quality. Don't they?

What I'm saying, Peter, is that it's not the per-pixel IQ that matters, but the IQ of the resulting photo.  A 100% crop from the 5D3 will not necessarily look as good as a 100% crop from a 5D, but that doesn't mean that the 5D3 doesn't resolve significantly more.

It's matter absolutely.  Actually 5D2 and 5D3 per-pixel quality is pretty close to 5D1 while D800 is obvious less.  Better per-pixel quality, better cropping capability and better print quality.

Peter, if System A has better per-pixel quality than System B, that information, by itself, tells us nothing about the IQ of the photo.

...you' need better technique on D800 such as the best lens (such as Sigma 35/1.4) on tripod, otherwise the advantage is disappearing quickly.

Yes -- the *advantage* will disappear quickly with poor technique (or narrow apertures where diffraction is dominant), but the *advantage* will always remain.

Becomes much smaller or even noticeable.But I am glad you finally agreed up the value of a good technique

Peter, I acknowledged the value of good technique with my very first sentence of my very first post in this thread.  Please, stop mistrepresenting, whether it be of deliberate intent or ignorance of what I've said.

So you agreed your own acknowledgement.  What's the problem?  My point is that higher pixel amounts, better technique you needed not to export to the same size of lower MP camera, but to its own designed potential.  D800 is not designed to print to 20x30" or 4000-pixel on screen, but 40x60" print and 7000-pixel on screen.  Otherwise you wasting 36mp and better to buy a 24mp D600.

Peter, many people are better off with a good compact than a DSLR.  What's your point?

See, DXOMark shows with a better lens such as 24-70L II 5D3 actually has better perceptual sharpness that roughly based on MTF 30 resolution in beating D800 with 24-70G, whoops, no mention if D800 not take on a sturdy tripod.

24-70L II on 5D3 vs 24-70G on D800

So DxOMark shows, with the particular lenses they tested, that the Canon lens is significantly better than the Nikon lens.

No, DXOMark tested entire system not lens.  Sensor or lens doesn't take photos alone but must work together.

Yes.  But since the D800 sensor resolves signficantly more than the 5D3 sensor (36 MP vs 22 MP, and the AA filters are not that different), then for the 24-70 / 2.8L II to resolve more on the 5D3 than the 24-70 / 2.8G on the D800, the 24-70 / 2.8L II must be significantly sharper.

Exactly.  D800 sensor advantage quickly reduced by imperfect lens, so-called diminishing return.  Even Canon releases a 46mp camera, it will NOT have double resolution with the same 24-70L II.  DXOMark perceptual sharpness tests that roughly based on MTF-30 resolution explain these very well.  It's only matter form sensor-lens resolution/sharpness not sensor's paper resolution.

DxOMark's PMP measure is no more meaningful than DPR's overall rating for a camera.

Did you read Hans' post?

http://www.dpreview.com/forums/post/51219749

I did.  Like I said, the D800 will resolve a maximum of 28% more linear detail than the 5D3, all else equal.  If it has a better lens, then more.  If it has a lesser lens, then it can even resolve less.

28% only on paper but in reality as Hans and others experienced, the difference is very small.

Again, a maximum of 28% -- let's say 14% on average.  I wonder how many could see a difference of 14% more linear detail.  More to the point, I wonder how many more would even care if they could notice.

system resolution.  Lens is the bottleneck.  You can squeeze 100mp into, but the resolution that human eyes can see are greatly reduce by imperfect lenses.

And you think Canon lenses are oh so much better than Nikon lenses, right?

However, this is neither here nor there.  If the 5D3 had the D800 sensor (and, by the way, the 5D3 with the D800 sensor in the 6D body is nearly my ideal camera) then the 5D3 would have produced better photos still.

In theory yes. In reality if you only print to 20x30" or view to 4000-pixel wide, then the difference is hardly noticeable. Personally I don't believe cropping but native length lens.

Again, I refer you to the FF photographers going mFT for the exact same reasons (with the added bonus of a significantly smaller kit).

That is another topic.  Crop format has much bigger impact on IQ than amount of pixels difference.  The difference between 5D2/5D3 and mFT is much bigger than between 5d3/5D2 and D800.

Except:

http://www.dpreview.com/forums/post/50428206

http://www.dpreview.com/forums/post/50665397

Loved your BS posts in that second thread.

Whether or not the photos would have been "enough better" to be worth the operational downsides is another matter entirely.

Agreed. So as DPR review said, you do need to use the best technique such as the best lenses and on tripod to fully leverage D800 resolution in order to print in excellent quality to billboard size 40x60", LOL.

But you can shoot handheld and still enjoy a substantial advantage.  I refer you again to my post that you not only agreed with, but added to:

http://www.dpreview.com/forums/post/51518848

P.S.:  I thought you were done with line-by-line replies.

Yes but just on different degree.  As I said, with hand-held you can get pretty sharp full size 5D3 files and excellent print quality to 20x30".  With D800, you don't get the same level of full size D800 files and similar print quality at 40x60".

So what it boils down to is that you're saying the 36 MP of the D800 is a waste since photos at 40x60 inches don't have the same IQ as photos from the 5D3 at 20x30 inches?

My, my.  It's not education that you need, but...

Reply   Reply with quote   Complain
Mako2011
Mako2011 MOD
Forum ProPosts: 15,594
Like?
no need
In reply to qianp2k, May 25, 2013

qianp2k wrote:


Let's check entire body details that would be much clearer

No need when the 100% crop tells the real truth

- From the 22 MP 5D3:

Reply   Reply with quote   Complain
qianp2k
Forum ProPosts: 10,000Gear list
Like?
Re: no need
In reply to Mako2011, May 25, 2013

Mako2011 wrote:

qianp2k wrote:


Let's check entire body details that would be much clearer

No need when the 100% crop tells the real truth

You can pickup whatever you like but real truth is pretty clear here.  Which one has more feather details, which one is sharper?  Despite my photo received less light, and despite I used much inferior lens, a prosumer zoom vs a pro prime.  Also if you check closely, what noises inside the dark shadow?  How much details smeared? 

- From the 22 MP 5D3:

-- hide signature --
 qianp2k's gear list:qianp2k's gear list
Sony RX100 Canon EOS-1D Mark III Canon EOS 5D Mark II Canon EOS 5D Mark III Sony Alpha 7R +20 more
Reply   Reply with quote   Complain
jjnik
Senior MemberPosts: 1,286Gear list
Like?
Re: no need
In reply to qianp2k, May 25, 2013

qianp2k wrote:

Mako2011 wrote:

qianp2k wrote:


Let's check entire body details that would be much clearer

No need when the 100% crop tells the real truth

You can pickup whatever you like but real truth is pretty clear here.  Which one has more feather details, which one is sharper?  Despite my photo received less light, and despite I used much inferior lens, a prosumer zoom vs a pro prime.  Also if you check closely, what noises inside the dark shadow?  How much details smeared? 

- From the 22 MP 5D3:

-- hide signature --

All I can say is.... wow - I thought alternate realities were just science fiction...

 jjnik's gear list:jjnik's gear list
Nikon D4s Nikon D700 Nikon D800E Nikon 1 AW1 Nikon AF-S Nikkor 24-70mm f/2.8G ED +11 more
Reply   Reply with quote   Complain
jjnik
Senior MemberPosts: 1,286Gear list
Like?
Re: no need
In reply to jjnik, May 25, 2013

The point you won't seem to admit is that a high resolution camera (in this case a 36 MP Nikon, in the future likely an even higher MP Canon) can deliver very sharp images at the pixel level even handheld.  Here's another example for you to not understand (and this one features a Canon 7D and 70-200/2.8 IS  ).  Handheld at 1/400th:

- Full frame image (showing focus point):

- Cropped full size image (to get under the 20 MP gallery upload limit):

100% crop (800 pixels x 500 pixels - note this is slightly behind the actual focus point):

 jjnik's gear list:jjnik's gear list
Nikon D4s Nikon D700 Nikon D800E Nikon 1 AW1 Nikon AF-S Nikkor 24-70mm f/2.8G ED +11 more
Reply   Reply with quote   Complain
qianp2k
Forum ProPosts: 10,000Gear list
Like?
Re: no need
In reply to jjnik, May 25, 2013

jjnik wrote:

qianp2k wrote:

Mako2011 wrote:

qianp2k wrote:


Let's check entire body details that would be much clearer

No need when the 100% crop tells the real truth

You can pickup whatever you like but real truth is pretty clear here.  Which one has more feather details, which one is sharper?  Despite my photo received less light, and despite I used much inferior lens, a prosumer zoom vs a pro prime.  Also if you check closely, what noises inside the dark shadow?  How much details smeared? 

- From the 22 MP 5D3:

-- hide signature --

All I can say is.... wow - I thought alternate realities were just science fiction...

Do you actually realize that ?

1) You used the best lit upper wing from your bird to compare to the wing in the shadow from my bird.  How about you show us the lower wing that is in shadow from your bird in 100%? I already see noises and red-color (banding) in your 50% cropped

2) You used 1/4000 shutter in your photo vs 1/1000 shutter in my photo.  Aka, isn't that a better technique as from a tripod at 1/4000?

3) Check birds' bodies that have similar lit, then the difference is pretty clear that my one has more feather details and obvious sharper.

So theory is one thing, and reality is another thing.

-- hide signature --
 qianp2k's gear list:qianp2k's gear list
Sony RX100 Canon EOS-1D Mark III Canon EOS 5D Mark II Canon EOS 5D Mark III Sony Alpha 7R +20 more
Reply   Reply with quote   Complain
qianp2k
Forum ProPosts: 10,000Gear list
Like?
Re: no need
In reply to jjnik, May 25, 2013

jjnik wrote:

The point you won't seem to admit is that a high resolution camera (in this case a 36 MP Nikon, in the future likely an even higher MP Canon) can deliver very sharp images at the pixel level even handheld.  Here's another example for you to not understand (and this one features a Canon 7D and 70-200/2.8 IS  ).  Handheld at 1/400th:

- Full frame image (showing focus point):

- Cropped full size image (to get under the 20 MP gallery upload limit):

100% crop (800 pixels x 500 pixels - note this is slightly behind the actual focus point):

Sharp and clean photo.  You miss my point.  You should not read I mean you cannot get sharp D800 photos at 100% cropped level thru hand-held (on what shutter speed?).  I just said (read carefully), you'd need better technique such as using higher shutter than lower pixel amount 24mp D600 and 22mp 5DIII at respective per-pixel level.  I'd argue that 5DIII with 300L/2.8 II can take the same photo at 1/200 or even 1/150 for the same per-pixel level sharpness.  After of all, you used 1/400 from 300G/2.8 VR lens.

More pixels, higher crop factor needs higher shutter to get the same per-pixel level sharpness.

-- hide signature --
 qianp2k's gear list:qianp2k's gear list
Sony RX100 Canon EOS-1D Mark III Canon EOS 5D Mark II Canon EOS 5D Mark III Sony Alpha 7R +20 more
Reply   Reply with quote   Complain
qianp2k
Forum ProPosts: 10,000Gear list
Like?
Re: *Sigh*
In reply to Great Bustard, May 25, 2013

Great Bustard wrote:

qianp2k wrote:

Great Bustard wrote:

qianp2k wrote:

Great Bustard wrote:

qianp2k wrote:

There are too many sentences and I don't want to reply one on one basis as most of them are the same regarding below the key point.

Finally something we can both agree on! 

Great Bustard wrote:

...or print to their supposed sizes such as 20x30" vs 40x60".

When printed at the same size (and why would we be comparing systems in any other matter), more pixels will yield more detail every time (all else equal).

Yeah a big ONLY if all else equal, that's the point.

Well, I mean, duh.  The point is, everything can be made equal except the lenses (although the lenses are often, if not usually, close enough).

But how to make everything else equals?

You can't.  You compare the systems on the basis of the photos you intend to take with the systems and the lenses you expect to use.

That's the point that you cannot.  I'd agree with you if truly everything equals.  But in reality they don't.  Canon 24-70L II is better than Nikon counterpart...

According to DxOMark, not according to PZ.  The problem with these tests, among other things, is that they test only a single copy of the lens:

Only if all those review sites said Canon 24-70L II and 70-200L II are better than Nikon counterparts.  So how coincidence?  Not sure if DXOMark uses only one copy.  How about owners' opinions?  I read several due system owners said the same that Canon copies are better.

http://www.slrgear.com/articles/variation_canon50f14/canon50f14.htm

...Canon 70-200L II is better than Nikon counterpart, Canon 24mm TS-E II is better, Canon 17mm TS-E is unique and better than 14-24 at 17mm in edges/corners clearly.

Talk about cherry-picking examples.  The 17 TSE vs the 14-24 / 2.8?  Really?

Yes.  17mm TS-E is my UWA lens while 14-24 is widely choice for Nikon UWA purpose.  17mm has much less distortion, clear better in edges/corners.

Do you have to use a better technique such as tripod and the best lens on D800?

No.  Do you think I use a tripod with my 20 MP 6D when I came from a 13 MP 5D?

So you can see difference at 2000-pixel wide or at 13x19" print? I cannot between my 5D3 and 5D1 except former is slightly better in DR.  I don't think you can tell difference between 5D3 and D800 at 3000-4000 pixel wide or at 18x26" or even 20x30" prints.  I read several due system owners said so.  In order to print really well at 60" wide, you do need to put D800 on tripod while you have a good chance to print hand-held photos from 5D3 at 30" wide.  To see clear difference at 40" wide, I do believe you'd need put D800 on tripod or still very small if both shoot hand-held.  5D3 files can upsampling thru special software and be pretty comparable to D800 to print at 40" wide.

What makes you think the 5D3 files will look nicer by upsampling but the D800 files will not?

Please don't twist what I meant.  I mean D800 needs to use better technique such as on tripod or faster shutter in order to show its best potential otherwise the difference is smaller.

Otherwise the difference is much smaller due to imperfect techniques and lens.

You know, it really isn't.  Indeed, you are quite the hypocrite, as I made this exact point:

Please don't get excited and take personal.

http://www.dpreview.com/forums/post/51518848

and in your response to that point, you added more support to it.

That is your post.  I am not sure what you talking about.  But 5D2 that OP talking about is not a high density camera compared to D800.  it affected much less by shooting in hand-held.

I actually read you're hypocrite a few weeks' ago in Open Forum that you boated your 6D shallower DOF and you can shoot wide open in low light not to worry about DOF much, that contradicts to your posts in mFT forum that you keep emphasizing to shoot on eq. DOF, that I disagreed.  So it seems you talked differently in different forums.

What you read is entirely different from what you wrote.  You can find the link and quote, but if I said anything about it, it would have been "Since I love shallow DOF photography, the added bonus of getting less noise at the wider apertures on FF works well for me".

Same to me, so why you keep pushing BS eq DOF to others while yourself don't follow your own bible?

To achieve all else equal such as similar per-pixel level IQ...

And again, Peter, per-pixel IQ is meaningless.  It's the photo we're talking about here.  I mean, if we're not talking photos, then there's no point to this "conversation".


But that directly affects cropping capability and print quality. Don't they?

What I'm saying, Peter, is that it's not the per-pixel IQ that matters, but the IQ of the resulting photo.  A 100% crop from the 5D3 will not necessarily look as good as a 100% crop from a 5D, but that doesn't mean that the 5D3 doesn't resolve significantly more.

It's matter absolutely.  Actually 5D2 and 5D3 per-pixel quality is pretty close to 5D1 while D800 is obvious less.  Better per-pixel quality, better cropping capability and better print quality.

Peter, if System A has better per-pixel quality than System B, that information, by itself, tells us nothing about the IQ of the photo.

Well if case a)  A = B in amount of pixel on the same format, then A's IQ clear better than B's.  However if case b) A < B in amount of pixels, then the difference is smaller than case a).  Is that clear now?

...you' need better technique on D800 such as the best lens (such as Sigma 35/1.4) on tripod, otherwise the advantage is disappearing quickly.

Yes -- the *advantage* will disappear quickly with poor technique (or narrow apertures where diffraction is dominant), but the *advantage* will always remain.

Becomes much smaller or even noticeable.But I am glad you finally agreed up the value of a good technique

Peter, I acknowledged the value of good technique with my very first sentence of my very first post in this thread.  Please, stop mistrepresenting, whether it be of deliberate intent or ignorance of what I've said.

So you agreed your own acknowledgement.  What's the problem?  My point is that higher pixel amounts, better technique you needed not to export to the same size of lower MP camera, but to its own designed potential.  D800 is not designed to print to 20x30" or 4000-pixel on screen, but 40x60" print and 7000-pixel on screen.  Otherwise you wasting 36mp and better to buy a 24mp D600.

Peter, many people are better off with a good compact than a DSLR.  What's your point?

My point is that if you use D800 mostly hand-held with so-so lenses for example, then why not to get D600? What's point if per-pixel D800 photos look soft and mushy and print not that good at 40x60" with inferior technique, so you print 20x30" and only show at 4000-pixel wide instead?

See, DXOMark shows with a better lens such as 24-70L II 5D3 actually has better perceptual sharpness that roughly based on MTF 30 resolution in beating D800 with 24-70G, whoops, no mention if D800 not take on a sturdy tripod.

24-70L II on 5D3 vs 24-70G on D800

So DxOMark shows, with the particular lenses they tested, that the Canon lens is significantly better than the Nikon lens.

No, DXOMark tested entire system not lens.  Sensor or lens doesn't take photos alone but must work together.

Yes.  But since the D800 sensor resolves signficantly more than the 5D3 sensor (36 MP vs 22 MP, and the AA filters are not that different), then for the 24-70 / 2.8L II to resolve more on the 5D3 than the 24-70 / 2.8G on the D800, the 24-70 / 2.8L II must be significantly sharper.

Exactly.  D800 sensor advantage quickly reduced by imperfect lens, so-called diminishing return.  Even Canon releases a 46mp camera, it will NOT have double resolution with the same 24-70L II.  DXOMark perceptual sharpness tests that roughly based on MTF-30 resolution explain these very well.  It's only matter form sensor-lens resolution/sharpness not sensor's paper resolution.

DxOMark's PMP measure is no more meaningful than DPR's overall rating for a camera.

Why?  DXO PMP is basically based on MTF 30 resolution.  DPR overall rating is subjective.  BTW, 5D3 and D800 have the same DPR score while D800E is only hairy higher.

Did you read Hans' post?

http://www.dpreview.com/forums/post/51219749

I did.  Like I said, the D800 will resolve a maximum of 28% more linear detail than the 5D3, all else equal.  If it has a better lens, then more.  If it has a lesser lens, then it can even resolve less.

28% only on paper but in reality as Hans and others experienced, the difference is very small.

Again, a maximum of 28% -- let's say 14% on average.  I wonder how many could see a difference of 14% more linear detail.  More to the point, I wonder how many more would even care if they could notice.

system resolution.  Lens is the bottleneck.  You can squeeze 100mp into, but the resolution that human eyes can see are greatly reduce by imperfect lenses.

And you think Canon lenses are oh so much better than Nikon lenses, right?

In those lenses I mentioned as I own and use, yes Canon copies are better.

However, this is neither here nor there.  If the 5D3 had the D800 sensor (and, by the way, the 5D3 with the D800 sensor in the 6D body is nearly my ideal camera) then the 5D3 would have produced better photos still.

In theory yes. In reality if you only print to 20x30" or view to 4000-pixel wide, then the difference is hardly noticeable. Personally I don't believe cropping but native length lens.

Again, I refer you to the FF photographers going mFT for the exact same reasons (with the added bonus of a significantly smaller kit).

That is another topic.  Crop format has much bigger impact on IQ than amount of pixels difference.  The difference between 5D2/5D3 and mFT is much bigger than between 5d3/5D2 and D800.

Except:

http://www.dpreview.com/forums/post/50428206

No idea, no opinion how OP did the test.  But if DXOMark ever tested E-M5 sensor+lens resolution, 5D2's setup will be obvious better.  I read a different story from a true owners of both 5D2 and EM-5.

http://www.dpreview.com/forums/thread/3491683?page=7

http://www.dpreview.com/forums/post/50665397

Loved your BS posts in that second thread.

LOL, the OP used such high shutters in both 5D2 and EM-5.  What I said there are pretty meaningful. That's what I'd do, and that's what you actually did despite you said differently in that post.

Whether or not the photos would have been "enough better" to be worth the operational downsides is another matter entirely.

Agreed. So as DPR review said, you do need to use the best technique such as the best lenses and on tripod to fully leverage D800 resolution in order to print in excellent quality to billboard size 40x60", LOL.

But you can shoot handheld and still enjoy a substantial advantage.  I refer you again to my post that you not only agreed with, but added to:

http://www.dpreview.com/forums/post/51518848

P.S.:  I thought you were done with line-by-line replies.

Yes but just on different degree.  As I said, with hand-held you can get pretty sharp full size 5D3 files and excellent print quality to 20x30".  With D800, you don't get the same level of full size D800 files and similar print quality at 40x60".

So what it boils down to is that you're saying the 36 MP of the D800 is a waste since photos at 40x60 inches don't have the same IQ as photos from the 5D3 at 20x30 inches?

Sigh.  You love to cut my words out of meaning and twist what I actually meant.  I never said and meant by exact words you described.  I simply mean D800 does need better techniques such as on tripod or faster shutter, or best lenses in order to print big such as 40x60".  5D3 will be more forgivable to have similar print quality at 20x30".  40x60" not necessarily better than 20x30" prints, just bigger size and size alone is not the only factor.  Does 60" TV must be better than a 40" TV?

My, my.  It's not education that you need, but...

yeah, please read what I meant as a whole.  I don't deny D800 resolution advantage but only suggested as DPR review also said, that it demands higher technique to fully leverage its 36mp.  Sure 5D3 also needs better technique but demands less to fully leverage 22mp.  By using the same lower level technique the difference is less than by using the same higher technique.

-- hide signature --
 qianp2k's gear list:qianp2k's gear list
Sony RX100 Canon EOS-1D Mark III Canon EOS 5D Mark II Canon EOS 5D Mark III Sony Alpha 7R +20 more
Reply   Reply with quote   Complain
Mako2011
Mako2011 MOD
Forum ProPosts: 15,594
Like?
Understatement
In reply to qianp2k, May 25, 2013

qianp2k wrote:

jjnik wrote:

The point you won't seem to admit is that a high resolution camera (in this case a 36 MP Nikon, in the future likely an even higher MP Canon) can deliver very sharp images at the pixel level even handheld.  Here's another example for you to not understand (and this one features a Canon 7D and 70-200/2.8 IS  ).  Handheld at 1/400th:

- Full frame image (showing focus point):

- Cropped full size image (to get under the 20 MP gallery upload limit):

100% crop (800 pixels x 500 pixels - note this is slightly behind the actual focus point):

You miss my point.

"miss" might not be the right word.

-- hide signature --

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

Reply   Reply with quote   Complain
Great Bustard
Forum ProPosts: 23,567
Like?
Um, Peter?
In reply to qianp2k, May 25, 2013

qianp2k wrote:

Great Bustard wrote:

qianp2k wrote:

That's the point that you cannot.  I'd agree with you if truly everything equals.  But in reality they don't.  Canon 24-70L II is better than Nikon counterpart...

According to DxOMark, not according to PZ.  The problem with these tests, among other things, is that they test only a single copy of the lens:  http://www.slrgear.com/articles/variation_canon50f14/canon50f14.htm

Only if all those review sites said Canon 24-70L II and 70-200L II are better than Nikon counterparts.  So how coincidence?  Not sure if DXOMark uses only one copy.  How about owners' opinions?  I read several due system owners said the same that Canon copies are better.

Um, Peter?  Take a read, please:

http://www.lensrentals.com/blog/2013/01/a-24-70mm-system-comparison

...Canon 70-200L II is better than Nikon counterpart, Canon 24mm TS-E II is better, Canon 17mm TS-E is unique and better than 14-24 at 17mm in edges/corners clearly.

Talk about cherry-picking examples.  The 17 TSE vs the 14-24 / 2.8?  Really?

Yes.  17mm TS-E is my UWA lens while 14-24 is widely choice for Nikon UWA purpose.  17mm has much less distortion, clear better in edges/corners.

Um, duh.  How ridiculously embarrassing would it be for Canon if their latest and greatest UWA TSE lens couldn't outperform Nikon's UWA zoom?

Do you have to use a better technique such as tripod and the best lens on D800?

No.  Do you think I use a tripod with my 20 MP 6D when I came from a 13 MP 5D?

So you can see difference at 2000-pixel wide or at 13x19" print? I cannot between my 5D3 and 5D1 except former is slightly better in DR.  I don't think you can tell difference between 5D3 and D800 at 3000-4000 pixel wide or at 18x26" or even 20x30" prints.  I read several due system owners said so.  In order to print really well at 60" wide, you do need to put D800 on tripod while you have a good chance to print hand-held photos from 5D3 at 30" wide.  To see clear difference at 40" wide, I do believe you'd need put D800 on tripod or still very small if both shoot hand-held.  5D3 files can upsampling thru special software and be pretty comparable to D800 to print at 40" wide.

What makes you think the 5D3 files will look nicer by upsampling but the D800 files will not?

Please don't twist what I meant.  I mean D800 needs to use better technique such as on tripod or faster shutter in order to show its best potential otherwise the difference is smaller.

How many times have I told you that the first sentence of my first post in this thread was to say that this is true for all systems, and not unique to the D800.  Seriously -- go back and count'em.

Peter, if System A has better per-pixel quality than System B, that information, by itself, tells us nothing about the IQ of the photo.

Well if case a)  A = B in amount of pixel on the same format, then A's IQ clear better than B's.

Are we comparing systems with equal pixel counts?  No?  I didn't think so.

However if case b) A < B in amount of pixels, then the difference is smaller than case a).  Is that clear now?

No, Peter, it's not clear.  What's clear is that per-pixel measurements are irrelevant if we are interested in the IQ of the final photo.  What we compare, if we are interested in the IQ of the photo, is equal portions of the scene.

So you agreed your own acknowledgement.  What's the problem?  My point is that higher pixel amounts, better technique you needed not to export to the same size of lower MP camera, but to its own designed potential.  D800 is not designed to print to 20x30" or 4000-pixel on screen, but 40x60" print and 7000-pixel on screen.  Otherwise you wasting 36mp and better to buy a 24mp D600.

Peter, many people are better off with a good compact than a DSLR.  What's your point?

My point is that if you use D800 mostly hand-held with so-so lenses for example, then why not to get D600?

Because if I'm shooting landscapes at 24mm f/5.6, for example, then the shutter speed will usually be so high that no special care will need to be taken with the D800.  I worked the math for you before, but, alas, it appears that I have to do it again.  If you need 1/100 to get a pixel-level sharp pic with 22 MP, then you need 1/125 to do the same with 36 MP.  How often is that a problem, Peter?

What's point if per-pixel D800 photos look soft and mushy and print not that good at 40x60" with inferior technique, so you print 20x30" and only show at 4000-pixel wide instead?

That's a strawman argument, which only accomplishes yet another display of your ignorance.  As I've also told you more times than I can count, if 40x60 inches looks bad with 36 MP, it will look worse still with 22 MP.  If 20x30 inches looks good with 22 MP, it will look better with 36 MP.

So what's with your bogus strawman arguments rather than simply accepting such an intuitively simple fact?

Again, I refer you to the FF photographers going mFT for the exact same reasons (with the added bonus of a significantly smaller kit).

That is another topic.  Crop format has much bigger impact on IQ than amount of pixels difference.  The difference between 5D2/5D3 and mFT is much bigger than between 5d3/5D2 and D800.

Except:

http://www.dpreview.com/forums/post/50428206

No idea, no opinion how OP did the test.

Not surprising, since it was only spelled out in the link I gave.

Yes but just on different degree.  As I said, with hand-held you can get pretty sharp full size 5D3 files and excellent print quality to 20x30".  With D800, you don't get the same level of full size D800 files and similar print quality at 40x60".

So what it boils down to is that you're saying the 36 MP of the D800 is a waste since photos at 40x60 inches don't have the same IQ as photos from the 5D3 at 20x30 inches?

Sigh.  You love to cut my words out of meaning and twist what I actually meant.  I never said and meant by exact words you described.  I simply mean D800 does need better techniques such as on tripod or faster shutter, or best lenses in order to print big such as 40x60".

And what you ignore is that many landscape photographers use a tripod regardless and that there is usually sufficient light to get more than a fast enough shutter speed regardless.

In fact, you *actively* ingore those simple facts.  Whether it's due to ignorance or dishonesty, I cannot comment.

5D3 will be more forgivable to have similar print quality at 20x30".

No, it will not.  All else equal, 36 MP will *always* record at least as much detail as 22 MP.

My, my.  It's not education that you need, but...

yeah, please read what I meant as a whole.

I have, and it just gets more and more disappointing each time to see you either not getting the simplest of concepts or...

I don't deny D800 resolution advantage but only suggested as DPR review also said, that it demands higher technique to fully leverage its 36mp.

As I've said multiple times (please go back and count), there are many non-trivial situations where no special care is needed to take advantage of the full potential of 36 MP over 24 MP.  For example, a landscape taken with Sigma 35 / 1.4 at f/5.6 1/200.

More to the point, for someone like me, who shoots ultra shallow DOFs all the time even in good light, the higher pixel count will render more detail on the portions of the scene within the DOF, with shutter speed to spare in good light, such as for pics like these:

http://www.dpreview.com/forums/post/51523737

Now, before you ask again why I didn't get the D800 instead of the 6D, I implore you to go find where I already answered that for you, as you are in the habit of not remembering what I have said.

P.S:  How embarrassing is this post of yours?

http://www.dpreview.com/forums/post/51524821

Look a bit further up in that thread to see why.  That said, the "reveal" also presents a conundrum for me as well.

Reply   Reply with quote   Complain
jjnik
Senior MemberPosts: 1,286Gear list
Like?
Re: no need
In reply to qianp2k, May 25, 2013

qianp2k wrote:

Do you actually realize that ?

1) You used the best lit upper wing from your bird to compare to the wing in the shadow from my bird.  How about you show us the lower wing that is in shadow from your bird in 100%? I already see noises and red-color (banding) in your 50% cropped

Look at my original unprocessed  image - the wing was not well lit - it was underexposed and was bumped 0.67EV and 75/100 pushed on shadow recovery inthe brighter image you're comparing to!  The far wing was essentially pitchblack.  You subject actually appears to be much more evenly lit than mine, and if you had pushed exposure and shadow recovery in post with a 5D3, it would look far worse than what you've presented - so I'm not seeing your point here at all?

BTW, as I posted a crop of the full size image, you could have checked the 100% 1:1 pixel view already - it seems you still don't get the 100% thing even though you're fixated on it being a major IQ driver?

2) You used 1/4000 shutter in your photo vs 1/1000 shutter in my photo.  Aka, isn't that a better technique as from a tripod at 1/4000?

I"ll guarantee that the bird in my picture was travelling MUCH faster thatn the bird in yours, so I needed that shutter speed to freeze it's motion - I needed the same with my D4 - so I didn't pick it because of the D800's 36 MP.  It's not better technique - it's just proper technique for the subject!

3) Check birds' bodies that have similar lit, then the difference is pretty clear that my one has more feather details and obvious sharper.

The bird's bodies are absolutely not similary lit as I indicated above and the bird's feathers are too different to make any definitive judgment on sharpness - but you seem to gloss over the fact that you are looking at a significantly more magnified view of a much smaller object in the frame in my shot.  If I had shot your pelican side by side with you - the D800 image would be much sharper for the same FOV/image size you captured with your 5D3 guaranteed.

So theory is one thing, and reality is another thing.

....depends on your version of reality

 jjnik's gear list:jjnik's gear list
Nikon D4s Nikon D700 Nikon D800E Nikon 1 AW1 Nikon AF-S Nikkor 24-70mm f/2.8G ED +11 more
Reply   Reply with quote   Complain
jjnik
Senior MemberPosts: 1,286Gear list
Like?
Re: no need
In reply to qianp2k, May 25, 2013

qianp2k wrote:

jjnik wrote:

The point you won't seem to admit is that a high resolution camera (in this case a 36 MP Nikon, in the future likely an even higher MP Canon) can deliver very sharp images at the pixel level even handheld.  Here's another example for you to not understand (and this one features a Canon 7D and 70-200/2.8 IS  ).  Handheld at 1/400th:

- Full frame image (showing focus point):

- Cropped full size image (to get under the 20 MP gallery upload limit):

100% crop (800 pixels x 500 pixels - note this is slightly behind the actual focus point):

Sharp and clean photo.  You miss my point.  You should not read I mean you cannot get sharp D800 photos at 100% cropped level thru hand-held (on what shutter speed?).  I just said (read carefully), you'd need better technique such as using higher shutter than lower pixel amount 24mp D600 and 22mp 5DIII at respective per-pixel level.  I'd argue that 5DIII with 300L/2.8 II can take the same photo at 1/200 or even 1/150 for the same per-pixel level sharpness.  After of all, you used 1/400 from 300G/2.8 VR lens.

More pixels, higher crop factor needs higher shutter to get the same per-pixel level sharpness.

-- hide signature --

I didn't choose 1/400th - the camera did as I was shooting aperture priority - Any camera would do the same to get a proper exposure in that scene with the chosen aperture.

But since you seem fixated on your quest to make everyone align with your belief that high resolution DSLR's need high shutter speeds and the best glass, let's try this:

1/10th with the 70-200 VR2 (which as you say is worse than the Canon counterpart) and with the 1.4x Teleconverter mounted to further dragrade the pixel level IQ:

- Full frame to show focus point:

-  Cropped full size image (to get under the 20 MP gallery upload limit):

Gee - I guess you're right:  a high MP camera needs the best glass and high shutter speeds or tripod use, otherwise the the pixel level (100% crop) is soft and mushy.....

 jjnik's gear list:jjnik's gear list
Nikon D4s Nikon D700 Nikon D800E Nikon 1 AW1 Nikon AF-S Nikkor 24-70mm f/2.8G ED +11 more
Reply   Reply with quote   Complain
Great Bustard
Forum ProPosts: 23,567
Like?
People don't need a TV at all.
In reply to qianp2k, May 25, 2013

qianp2k wrote:

Similarly not everyone needs a 120" TV, or even a 60" TV. To many a 40" or even a 32" TV is big enough.

What will look better, a 1080p HDTV or a 720p HDTV?

Oh yeah deepens on your viewing distance. A 40" TV could have higher resolution than a sloppy 50" TV. Sure I know you'd ask question why not down to 1mp? That's similar as why you cannot have a 1" TV? The issue is that can you view 1mp photo or watch on 1" TV? So everything is relative.

What are you talking about?  A 1200x800 photo on a computer monitor is 1 MP.  1920 x 1080 is 2 MP.  And?

For me, 20x30" prints or max 30" monitor is only I needed on my needs at least in foreseen future.

No one is trying to tell you want your needs are, or trying to tell you what your standards should be.

Therefore I'd only need a 46mp camera in future not for 40x60" print or on 60" monitor but view on future 30" 4K/8K ultra HD monitor or print to 20x30" ultra HD prints when they become affordable.

Well, 4K video is 8 MP, and 8K video is 33 MP.  Either way, the more pixels that the photo is captured with, the more resolution that is going to remain when downsampled to the final display size.

For example, 36 MP downsampled to 8 MP will result in a photo with higher IQ than if it were captured at 8 MP.  How much sharper?  Don't know.  Does it matter?  Would depend on a number of factors, I would imagine.

I predict I will have such ultra HD monitor first than ultra HD printer. Monitor or digital display technology will move way ahead of old print technology years down to the road.

The bottom line is that more pixels result in more captured detail, all else equal.  However, every quadrupling of pixels does not result in a doubling of linear detail.  That is, if 12 MP resolved 2000 lw/ph, 48 MP will not resolve 4000 lw/ph, all else equal.

Let's say that 12 MP to 48 MP takes the resolution from 2000 lw/ph for a particular scene and setting to 3000 lw/ph -- a 50% increase.  If we now quadruple the pixel count again to 192 MP, we won't see another 50% increase in linear resolution.  Maybe 25%.  If we quadruple again...

The point is, more pixels give more resolution, all else equal.  How much resolution does a person "need" and whether or not it is worth the operational downsides is another matter all together.

In my opinion, for the vast majority, we passed the point of "good enough" long ago, and many of those giving up FF for smaller format systems, like mFT, seem to agree.

Reply   Reply with quote   Complain
Apewithacamera
Senior MemberPosts: 2,161Gear list
Like?
Place a 1/64" ruler near one of your large prints if you have any LOL
In reply to Dave Luttmann, May 25, 2013

From 5' away let us know if you can resolve that scale with your eyes. I know I can't but if you can, you are indeed

Superman

End

 Apewithacamera's gear list:Apewithacamera's gear list
Canon EOS 7D Mark II Tamron SP 150-600mm F/5-6.3 Di VC USD Tamron 16-300mm F/3.5-6.3 Di II VC PZD Macro Tamron 28-300mm F/3.5-6.3 Di VC PZD Sigma 18-300 F3.5-6.3 DC Macro OS C +3 more
Reply   Reply with quote   Complain
Mako2011
Mako2011 MOD
Forum ProPosts: 15,594
Like?
Can you
In reply to Apewithacamera, May 25, 2013

Apewithacamera wrote:

From 5' away let us know if you can resolve that scale with your eyes. I know I can't but if you can, you are indeed

Superman

End

Can you  go here http://www.dpreview.com/members/settings/profile
And under the "allow download of originals"  check  "Allow visitors to download full size original images from your gallery" That way we can see the image better. Looks like a 34x24 image from here.

-- hide signature --

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads