Return of Interesting article on DxO about 5D III and D800...

Started Apr 10, 2013 | Discussions
schmegg
schmegg MOD
Senior MemberPosts: 5,002
Like?
Re: resolution
In reply to Just another Canon shooter, Apr 14, 2013

Just another Canon shooter wrote:

Mako2011 wrote:

Just another Canon shooter wrote:

The distance depends on the FOV, not on the format. Equivalent FL = same FOV = same distance to the same target.

Note that I am talking about equivalent lenses, not about the same lens.

We were discussing resolution. To compare resolution across different sensor formats accurately...you have to maintain the same distance to target or your resolution results will not be comparable.

When you shoot with equivalent FL's, you do shoot from the same distance.

"Equivalent lenses" is a bit of an oxymoron when you think about it. I can't think of a single pair of 'equivalent' lenses that would be suitable for determining the resolving power of a sensor. Any duo you pick will likely have one a little stronger or weaker than the other.

So, to remove this variance, simply shoot using the exact same lens at the exact same subject distance. That way, exactly the same level of detail is available at the sensor plane for the sensor to record. Then look closely to see which sensor recorded more of it. It's easy to do and the only variable is the sensor - making it a very good test, as it's the sensor that you are testing.

It's easy to see that the DxO tests (and most other MTF tests) are primarily concerned with edge sharpness - not resolution. One look at the test target will tell you that. And, though they are related for sure, they are not the same thing.

BTW, even if you don't, the results will be comparable. Read the Imatest explanations. But this is off topic anyway.

True that.

Reply   Reply with quote   Complain
Just another Canon shooter
Senior MemberPosts: 2,542Gear list
Like?
Re: You bet
In reply to Mako2011, Apr 14, 2013

Mako2011 wrote:

Just another Canon shooter wrote:

Mako2011 wrote:

Just another Canon shooter wrote:

The distance depends on the FOV, not on the format. Equivalent FL = same FOV = same distance to the same target.

Note that I am talking about equivalent lenses, not about the same lens.

We were discussing resolution. To compare resolution across different sensor formats accurately...you have to maintain the same distance to target or your resolution results will not be comparable.

When you shoot with equivalent FL's, you do shoot from the same distance.

Yes...for example... 40mm f/4 1/100 ISO 200 on 4/3 is equivalent to 50mm f/5 1/100 ISO 320 on 1.6x which is equivalent to 80mm f/8 1/100 ISO 800 on FF when the photos of the same scene are taken from the same position and displayed at the same size.

Not sure our friends at DxO are so picky

BTW, even if you don't, the results will be comparable.

Not when comparing resolution across different formats.

Yes, always. The slanted edge test which DXO uses is largely independent of the shooting distance. If they used some periodic pattern, then changing the distance would require changing the spatial frequency.

Again, this is of topic, really. The reason FF resolves more at medium to high contrast (with EF lenses, at least) is the crop factor. There is no conspiracy, no distance cheating. The good lenses resolve more on FF than on crop despite the FL; again, at medium to high contrast. At low contrast, more mp win. I am not sure why we need long threads for this.

 Just another Canon shooter's gear list:Just another Canon shooter's gear list
Samsung Galaxy S III
Reply   Reply with quote   Complain
Just another Canon shooter
Senior MemberPosts: 2,542Gear list
Like?
Re: resolution
In reply to schmegg, Apr 14, 2013

schmegg wrote:

When you shoot with equivalent FL's, you do shoot from the same distance.

"Equivalent lenses" is a bit of an oxymoron when you think about it. I can't think of a single pair of 'equivalent' lenses that would be suitable for determining the resolving power of a sensor.

We are not talking about that, whatever that means. The question was why similar lenses perform better in the DXO tests on FF bodies - which they do. The explanation is - DXO measures basically medium contrast, and then FF wins due to less enlargement. The extinct resolution of the sensor plays very little role for that specific metric.

 Just another Canon shooter's gear list:Just another Canon shooter's gear list
Samsung Galaxy S III
Reply   Reply with quote   Complain
MAC
MAC
Forum ProPosts: 12,268Gear list
Like?
Re: resolution
In reply to Just another Canon shooter, Apr 14, 2013

Just another Canon shooter wrote:

schmegg wrote:

When you shoot with equivalent FL's, you do shoot from the same distance.

"Equivalent lenses" is a bit of an oxymoron when you think about it. I can't think of a single pair of 'equivalent' lenses that would be suitable for determining the resolving power of a sensor.

We are not talking about that, whatever that means. The question was why similar lenses perform better in the DXO tests on FF bodies - which they do. The explanation is - DXO measures basically medium contrast, and then FF wins due to less enlargement. The extinct resolution of the sensor plays very little role for that specific metric.

dxo is measuring sharpness, not resolution.

To do any kind of comparison across formats, one must keep the distance the same.

One needs a better optic, the farther away from the subject.

 MAC's gear list:MAC's gear list
Canon EOS 5D Canon EOS 650D Canon EOS 100D Canon EF 70-200mm f/2.8L USM Canon EF 100mm f/2.8L Macro IS USM +5 more
Reply   Reply with quote   Complain
qianp2k
Senior MemberPosts: 9,794Gear list
Like?
Re: resolution
In reply to Just another Canon shooter, Apr 14, 2013

Just another Canon shooter wrote:

schmegg wrote:

When you shoot with equivalent FL's, you do shoot from the same distance.

"Equivalent lenses" is a bit of an oxymoron when you think about it. I can't think of a single pair of 'equivalent' lenses that would be suitable for determining the resolving power of a sensor.

We are not talking about that, whatever that means. The question was why similar lenses perform better in the DXO tests on FF bodies - which they do. The explanation is - DXO measures basically medium contrast, and then FF wins due to less enlargement. The extinct resolution of the sensor plays very little role for that specific metric.

it's fundamental essential to understand crop format and crop penalties in photography world.  For those denied there is ever crop penalty, I use a simple case that nobody should buy a bulky and super extensive 600mm lens to use on FF, those sport PJs or wildlife photogs should only need much lighter and cheaper 300mm/F4.0 lens on 2.0x crop, or even better a 30mm lens on 20x crop.  I know they are unable to answer this question

Anyway this thread now slipping into old epic 5D vs 7D debate that can continue last another 10 years, LOL.  Nevertheless thank you for your explanations and posts.  You're one of those really understand these tests and terms.  Personally I really don't care much terminology but only care what matter to my eyes in real world photos  To me, I know my 5D takes better photos in landscape, studio and portraiture while I love my 60D in zoo, safari and wildlife.  I own both cameras and have taken 35K and 25K photos from them respectively.  They complement each other very well.  Now 5D is sitting on my shelf in my museum of cameras (along Nikon N70) while 60D is still one of my active cameras

****

To all, let's return to OP - let's discuss what DXOMark Perceptual Sharpness really means to you in real world photography.

-- 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
Senior MemberPosts: 9,794Gear list
Like?
Re: resolution
In reply to MAC, Apr 14, 2013

MAC wrote:

Just another Canon shooter wrote:

schmegg wrote:

When you shoot with equivalent FL's, you do shoot from the same distance.

"Equivalent lenses" is a bit of an oxymoron when you think about it. I can't think of a single pair of 'equivalent' lenses that would be suitable for determining the resolving power of a sensor.

We are not talking about that, whatever that means. The question was why similar lenses perform better in the DXO tests on FF bodies - which they do. The explanation is - DXO measures basically medium contrast, and then FF wins due to less enlargement. The extinct resolution of the sensor plays very little role for that specific metric.

dxo is measuring sharpness, not resolution.

DXOMark sharpness is not regular "sharpness" that you can achieve thru software sharpening.  It's based on MTF resolution but translate with a factor that human eyes can perceive that likely around 20% MTF.  DXOMark uses two words interchangeably sometime.

To do any kind of comparison across formats, one must keep the distance the same.

This is a disputable part.  All those labs test by framing the test subjects (such as resolution chart or test subjects) into the same AOV either by eq FL (such as zoom) or by adjusting distance (such as prime).  It represents in real world how we use the camera-lens system.  Such as we use the same 85mm lens on FF or on crop to shoot half-body portraiture, so 7D/60D shooter need to move 1.6x away.  If shoot from the same distance, you will end with half body on FF vs shoulder above on APS-C crop.  Such as we use 24mm on FF or 15mm on APS-C for the same AOV in landscape photos, otherwise the same 15mm or 24mm on two different crop formats will end with two different AOV scenes.  What you compare, apple to orange?  You could setup a website to promote your test by shooting from the same distance and the same FL in your own lab  But I know which lab tests I will follow, hehe.

One needs a better optic, the farther away from the subject.

It's crop cameras need much better lens and much more pixels to overcome more crop penalties.

-- 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
MAC
MAC
Forum ProPosts: 12,268Gear list
Like?
Re: resolution
In reply to qianp2k, Apr 14, 2013

qianp2k wrote:

MAC wrote:

Just another Canon shooter wrote:

schmegg wrote:

When you shoot with equivalent FL's, you do shoot from the same distance.

"Equivalent lenses" is a bit of an oxymoron when you think about it. I can't think of a single pair of 'equivalent' lenses that would be suitable for determining the resolving power of a sensor.

We are not talking about that, whatever that means. The question was why similar lenses perform better in the DXO tests on FF bodies - which they do. The explanation is - DXO measures basically medium contrast, and then FF wins due to less enlargement. The extinct resolution of the sensor plays very little role for that specific metric.

dxo is measuring sharpness, not resolution.

DXOMark sharpness is not regular "sharpness" that you can achieve thru software sharpening.  It's based on MTF resolution but translate with a factor that human eyes can perceive that likely around 20% MTF.  DXOMark uses two words interchangeably sometime.

To do any kind of comparison across formats, one must keep the distance the same.

This is a disputable part.  All those labs test by framing the test subjects (such as resolution chart or test subjects) into the same AOV either by eq FL (such as zoom) or by adjusting distance (such as prime).  It represents in real world how we use the camera-lens system.  Such as we use the same 85mm lens on FF or on crop to shoot half-body portraiture, so 7D/60D shooter need to move 1.6x away.  If shoot from the same distance, you will end with half body on FF vs shoulder above on APS-C crop.  Such as we use 24mm on FF or 15mm on APS-C for the same AOV in landscape photos, otherwise the same 15mm or 24mm on two different crop formats will end with two different AOV scenes.  What you compare, apple to orange?  You could setup a website to promote your test by shooting from the same distance and the same FL in your own lab  But I know which lab tests I will follow, hehe.

One needs a better optic, the farther away from the subject.

It's crop cameras need much better lens and much more pixels to overcome more crop penalties.

yeah right, the crop blows away the 5d3 "at the same distance with the same lens, thereby measuring resolution

http://www.dpreview.com/forums/post/51145798

-- hide signature --
 MAC's gear list:MAC's gear list
Canon EOS 5D Canon EOS 650D Canon EOS 100D Canon EF 70-200mm f/2.8L USM Canon EF 100mm f/2.8L Macro IS USM +5 more
Reply   Reply with quote   Complain
Just another Canon shooter
Senior MemberPosts: 2,542Gear list
Like?
Re: resolution
In reply to MAC, Apr 14, 2013

MAC wrote:

Just another Canon shooter wrote:

schmegg wrote:

When you shoot with equivalent FL's, you do shoot from the same distance.

"Equivalent lenses" is a bit of an oxymoron when you think about it. I can't think of a single pair of 'equivalent' lenses that would be suitable for determining the resolving power of a sensor.

We are not talking about that, whatever that means. The question was why similar lenses perform better in the DXO tests on FF bodies - which they do. The explanation is - DXO measures basically medium contrast, and then FF wins due to less enlargement. The extinct resolution of the sensor plays very little role for that specific metric.

dxo is measuring sharpness, not resolution.

They measure the MTF curve. Then they convert it in some unknown way to their metric.

To do any kind of comparison across formats, one must keep the distance the same.

This is exactly what they do, say with 50mm on crop and 85mm on FF.

One needs a better optic, the farther away from the subject.

Who said anything about being farther away?

Let me repeat what I said:

BTW, the real test should be at equivalent FL and equivalent f-stops with common lenses.

Do you understand that with equivalent FL, the distance will be the same, or not?

 Just another Canon shooter's gear list:Just another Canon shooter's gear list
Samsung Galaxy S III
Reply   Reply with quote   Complain
Just another Canon shooter
Senior MemberPosts: 2,542Gear list
Like?
Re: resolution
In reply to MAC, Apr 14, 2013

MAC wrote:

It's crop cameras need much better lens and much more pixels to overcome more crop penalties.

yeah right, the crop blows away the 5d3 "at the same distance with the same lens, thereby measuring resolution

Why would you use the same lens? He was clearly talking about equivalent ones.

 Just another Canon shooter's gear list:Just another Canon shooter's gear list
Samsung Galaxy S III
Reply   Reply with quote   Complain
MAC
MAC
Forum ProPosts: 12,268Gear list
Like?
Re: resolution
In reply to Just another Canon shooter, Apr 14, 2013

Just another Canon shooter wrote:

MAC wrote:

Just another Canon shooter wrote:

schmegg wrote:

When you shoot with equivalent FL's, you do shoot from the same distance.

"Equivalent lenses" is a bit of an oxymoron when you think about it. I can't think of a single pair of 'equivalent' lenses that would be suitable for determining the resolving power of a sensor.

We are not talking about that, whatever that means. The question was why similar lenses perform better in the DXO tests on FF bodies - which they do. The explanation is - DXO measures basically medium contrast, and then FF wins due to less enlargement. The extinct resolution of the sensor plays very little role for that specific metric.

dxo is measuring sharpness, not resolution.

They measure the MTF curve. Then they convert it in some unknown way to their metric.

To do any kind of comparison across formats, one must keep the distance the same.

This is exactly what they do, say with 50mm on crop and 85mm on FF.

?? - are you sure?

One needs a better optic, the farther away from the subject.

Who said anything about being farther away?

they move the crop back with the same prime to get the same fov

Let me repeat what I said:

BTW, the real test should be at equivalent FL and equivalent f-stops with common lenses.

Do you understand that with equivalent FL, the distance will be the same, or not?

I understand equivalence between formats - GB's white paper - the distance must be kept the same.

 MAC's gear list:MAC's gear list
Canon EOS 5D Canon EOS 650D Canon EOS 100D Canon EF 70-200mm f/2.8L USM Canon EF 100mm f/2.8L Macro IS USM +5 more
Reply   Reply with quote   Complain
qianp2k
Senior MemberPosts: 9,794Gear list
Like?
Re: resolution
In reply to MAC, Apr 15, 2013

MAC wrote:

qianp2k wrote:

MAC wrote:

Just another Canon shooter wrote:

schmegg wrote:

When you shoot with equivalent FL's, you do shoot from the same distance.

"Equivalent lenses" is a bit of an oxymoron when you think about it. I can't think of a single pair of 'equivalent' lenses that would be suitable for determining the resolving power of a sensor.

We are not talking about that, whatever that means. The question was why similar lenses perform better in the DXO tests on FF bodies - which they do. The explanation is - DXO measures basically medium contrast, and then FF wins due to less enlargement. The extinct resolution of the sensor plays very little role for that specific metric.

dxo is measuring sharpness, not resolution.

DXOMark sharpness is not regular "sharpness" that you can achieve thru software sharpening.  It's based on MTF resolution but translate with a factor that human eyes can perceive that likely around 20% MTF.  DXOMark uses two words interchangeably sometime.

To do any kind of comparison across formats, one must keep the distance the same.

This is a disputable part.  All those labs test by framing the test subjects (such as resolution chart or test subjects) into the same AOV either by eq FL (such as zoom) or by adjusting distance (such as prime).  It represents in real world how we use the camera-lens system.  Such as we use the same 85mm lens on FF or on crop to shoot half-body portraiture, so 7D/60D shooter need to move 1.6x away.  If shoot from the same distance, you will end with half body on FF vs shoulder above on APS-C crop.  Such as we use 24mm on FF or 15mm on APS-C for the same AOV in landscape photos, otherwise the same 15mm or 24mm on two different crop formats will end with two different AOV scenes.  What you compare, apple to orange?  You could setup a website to promote your test by shooting from the same distance and the same FL in your own lab  But I know which lab tests I will follow, hehe.

One needs a better optic, the farther away from the subject.

It's crop cameras need much better lens and much more pixels to overcome more crop penalties when frame subject into the same AOV.

yeah right, the crop blows away the 5d3 "at the same distance with the same lens, thereby measuring resolution

http://www.dpreview.com/forums/post/51145798

LOL. I don't take 5D photos from the same FL lens from the same distance then crop out to match to my 60D's AOV in landscape, studio or portraiture I guess nobody else doing in that way

1) In reality what lens on your T4i or 60D that you can achieve FF eq 24mm AOV to match my 24-70L II on my 5D3 in landscape photos?   From 15-85 at 15mm?

2) In reality what lens on your T4i or 60D that you can achieve FF eq 85mm AOV to match 85L on 5D3 in portraiture? Hehe.

3) In NFL sport and Olympic venues, now professional PJs use 400L/2.8 II, or 500L/4.0 IS II or rarely 600L/4.0 IS II on 1DX/5D3 that absolutely beat 7D with 300L or 400L as with those tele lenses, PJs can frame players into sport fields without cropping. No chances 7D or 2.0x crop E-5 could win.  It's the reality in sport fields these days.  Actually it's also one of reasons why Canon was forced to move to FF sport cameras as its 1.3x crop 1D lines cannot compete to Nikon FF D3/D4 lines in IQ especially in high ISOs.

Sure, you can quote in birding case.  But don't forget 1DX owners can use 1.4x or 2.0x TC on 800L that usually have enough reach most times.  Or 1D4 owners (1.3x crop is much close to FF in IQ than APS-C) have the best balance of IQ and reach in birding.  7D is unable to AF at F8.0 with TC.

That simply reflected in reality of photography world.

-- 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
Keyboard shortcuts:
FForum MMy threads