G9 studio scene comparison

[TomFid]

...

The conversation thread I was responding to was specifically about the Lenstip results.

Actually the conversation thread has been about the DPR studio scene test shots for the G9, but the lenstip charts were referenced by samtheman2014 as evidence of little difference between the two lenses at f/5.6 (the aperture setting used in DPR shots). The ambiguity was introduced by TomFid in his reference to "they refuse". He can clarify if he wants as to whether he was referring to DPR or Lenstip (see excerpt from your original post below). I interpreted his response as being in reference to DPR refusing to investigate the impact of shutter shock in its studio scene results since the DPR ISO 200 shots have long been known to be in the shutter shock danger zone.

It's no big deal, but I just wanted it to be clear that here on DPR and with reference to the G9 and other comparison studio scene test images being referred to in this thread, shutter shock (and the mitigation of EFCS) are real variables to consider with respect to at least the ISO 200 shots.

I was actually referring to the LensTip chart originally posted by Samtheman. If DPR also isn't using e-shutter or EFCS, they might be implicated as well, but at least they're using modern bodies (obviously, since the test is for the body not the lens).

 

[TomFid]

Lenstip, post: 60470602, member: 1127013"]
By way of digression: after publishing our Panaleica 100-400 mm test we found out our results at the longest focal lengths have been questioned on many foreign websites. It was claimed that they were heavily understated because of vibrations of the mechanical shutter and it was suggested that a Panasonic body with an electronic shutter would make them higher. It wasn’t exactly true. Having very strong lamps in our editorial office we were able to reduce the exposure times even to 1/1000-1/2000 sec. and additionally we used an “anti-shock” option with a value as high as 15 seconds.

[/QUOTE]
This is exactly what I meant by "steadfast refusal" to address the issue. Their solution doesn't deal with the whole problem, specifically 2nd curtain vibration. They could easily have retested on a body with e-shutter, and if they got the same answer I'd believe them. But this is just pure arrogance.

 

[whumber]

By way of digression: after publishing our Panaleica 100-400 mm test we found out our results at the longest focal lengths have been questioned on many foreign websites. It was claimed that they were heavily understated because of vibrations of the mechanical shutter and it was suggested that a Panasonic body with an electronic shutter would make them higher. It wasn’t exactly true. Having very strong lamps in our editorial office we were able to reduce the exposure times even to 1/1000-1/2000 sec. and additionally we used an “anti-shock” option with a value as high as 15 seconds.

This is exactly what I meant by "steadfast refusal" to address the issue. Their solution doesn't deal with the whole problem, specifically 2nd curtain vibration. They could easily have retested on a body with e-shutter, and if they got the same answer I'd believe them. But this is just pure arrogance.

Part of the issue in that case is that testing with another body is just going to muddy the waters as the results will no longer be comparable to their historical measurements. In addition, just because another body shows a large difference between mechanical and electronic shutter doesn't mean that also applies to their reference configuration. The better solution is exactly what they did, test at a high enough shutter speed that shutter shock is no longer a factor as well as comparing against another lens with a similar focal length to see if the same behavior is seen.

 

[samtheman2014]

Yes but a high iso on the mk2 has more dynamic and tonality range than the others most of the noise can cleared at those iso,s which is same for all of them

I don't think it does despite the "different" sensor designations due to the specific customisations required by the brand, be it the CDAF layer on the E-M1 II or the video centric features of the GH5. Looking at the near identical RAW results across the board with the various 20mp m43 model , it is reasonable to suppose that they all { not surprisingly } share the same 20mp Sony base sensor. Hence why other than processing tweaks { NR levels etc } and AA filter strength/weakness the output is so close,

not much progress then in m43 land but then it is same with all sensors and not forgetting the extra pixels will help a little also over the 16 mp ,but your right the only way to get clean images is to use hi Rez but of course this has limitations ,

I am in total agreement Paul , if you look at say the Nikon D3s vs the A7s II at high ISO not much of a difference same story with say the D810 vs D850 little steps forward not huge leaps. My point was aimed at those claiming the E-M1 II is a full stop or more better than its m43 competition which is demonstrably untrue .

I have a g9 on order I am sure I will be pleased with the results jpegs look really pleasing to the eye ,just need to wait for the dynamic range at low iso,s etc but I do so hate test charts they are only ever any good for noise extraction ,or some resolution examples .

DR at low ISO with regards to the significant noise penalty when pushing shadows is for me the single biggest weakness of m43. And thanks mainly to our default and too high 200 base ISO sensors we are disadvantaged more than we need to be in this area

As most have pointed out in print well the size I print at all this pixel peeping and measure bating goes flat on its face you have to elevate your enevelope at best most of people’s images are viewed on 27 -32 inch monitor or a smartphone .

Not an unreasonable claim Paul , I certainly do print very large on occasion . I also use a 40" 4K monitor . If you only look at what is needed for small prints , for typical web sized images or an A4 output m43 is already overkill

Even a printed publication will only be A4 .

 

[JaKing]

Yes but a high iso on the mk2 has more dynamic and tonality range than the others most of the noise can cleared at those iso,s which is same for all of them

I don't think it does despite the "different" sensor designations due to the specific customisations required by the brand, be it the CDAF layer on the E-M1 II or the video centric features of the GH5.

Reference for your assertion about it being a layer? It is my understanding that the focusing pixels are part of the imaging surface of the sensor.

Looking at the near identical RAW results across the board with the various 20mp m43 model , it is reasonable to suppose that they all { not surprisingly } share the same 20mp Sony base sensor. Hence why other than processing tweaks { NR levels etc } and AA filter strength/weakness the output is so close,

Bill Claff's measurements show otherwise, very distinctly.

not much progress then in m43 land but then it is same with all sensors and not forgetting the extra pixels will help a little also over the 16 mp ,but your right the only way to get clean images is to use hi Rez but of course this has limitations ,

I am in total agreement Paul , if you look at say the Nikon D3s vs the A7s II at high ISO not much of a difference same story with say the D810 vs D850 little steps forward not huge leaps. My point was aimed at those claiming the E-M1 II is a full stop or more better than its m43 competition which is demonstrably untrue .

I have a g9 on order I am sure I will be pleased with the results jpegs look really pleasing to the eye ,just need to wait for the dynamic range at low iso,s etc but I do so hate test charts they are only ever any good for noise extraction ,or some resolution examples .

DR at low ISO with regards to the significant noise penalty when pushing shadows is for me the single biggest weakness of m43. And thanks mainly to our default and too high 200 base ISO sensors we are disadvantaged more than we need to be in this area

As most have pointed out in print well the size I print at all this pixel peeping and measure bating goes flat on its face you have to elevate your enevelope at best most of people’s images are viewed on 27 -32 inch monitor or a smartphone .

Not an unreasonable claim Paul , I certainly do print very large on occasion . I also use a 40" 4K monitor . If you only look at what is needed for small prints , for typical web sized images or an A4 output m43 is already overkill

Even a printed publication will only be A4 .

 

[TomFid]

By way of digression: after publishing our Panaleica 100-400 mm test we found out our results at the longest focal lengths have been questioned on many foreign websites. It was claimed that they were heavily understated because of vibrations of the mechanical shutter and it was suggested that a Panasonic body with an electronic shutter would make them higher. It wasn’t exactly true. Having very strong lamps in our editorial office we were able to reduce the exposure times even to 1/1000-1/2000 sec. and additionally we used an “anti-shock” option with a value as high as 15 seconds.

This is exactly what I meant by "steadfast refusal" to address the issue. Their solution doesn't deal with the whole problem, specifically 2nd curtain vibration. They could easily have retested on a body with e-shutter, and if they got the same answer I'd believe them. But this is just pure arrogance.

Part of the issue in that case is that testing with another body is just going to muddy the waters as the results will no longer be comparable to their historical measurements. In addition, just because another body shows a large difference between mechanical and electronic shutter doesn't mean that also applies to their reference configuration. The better solution is exactly what they did, test at a high enough shutter speed that shutter shock is no longer a factor as well as comparing against another lens with a similar focal length to see if the same behavior is seen.

I disagree. Lenses have different mass and stiffness, which may interact with the body to vary the effective shock. And with long lenses, there may be no shutter speed that is high enough. I tried a lot of things to get sharp images out of a GH2 + Meade ETX, but it just wasn't possible with the mechanical shutter.

When there's an easy test that would confirm or refute your results, you do it. You don't just wave your hands and invoke nefarious foreigners.

 

[whumber]

Looking at the near identical RAW results across the board with the various 20mp m43 model , it is reasonable to suppose that they all { not surprisingly } share the same 20mp Sony base sensor. Hence why other than processing tweaks { NR levels etc } and AA filter strength/weakness the output is so close,

Bill Claff's measurements show otherwise, very distinctly.

Not really, Bill plots PDR vs specified ISO setting rather than some sort of measured ISO value. Olympus decided to pull a Fuji with the ISO definitions on the E-M1ii so that plot looks much better than the camera performs in reality.

 

[whumber]

I disagree. Lenses have different mass and stiffness, which may interact with the body to vary the effective shock. And with long lenses, there may be no shutter speed that is high enough. I tried a lot of things to get sharp images out of a GH2 + Meade ETX, but it just wasn't possible with the mechanical shutter.

That may be but I really doubt even 800mm EFL is going to run into that limit. The E-M1 has pretty pronounced shutter shock with the mechanical shutter and I can't see any hint of it at 1/1000s with an EFL of 840mm. Is it enough to be measured? Probably. Is it enough to cause the degree of softness shown in LensTips testing? Definitely not. If any, sample variation is much more likely of a candidate if there's anything off about the test.

When there's an easy test that would confirm or refute your results, you do it. You don't just wave your hands and invoke nefarious foreigners.

But the whole point is that there's not an easy test like you're suggesting, they can't just add an electronic shutter to the E-PL1. If they switch over to, say, an E-M1 now we can't compare the results to any of the historical data and because we're not using the same shutter mechanism we can't even say if the relative performance between the mechanical and electronic shutter would be the same. To do things in a meaningful manner you either need to switch your entire system over to electronic shutter, find a way of minimizing potential shutter shock, or find some other test to say if shutter shock is even an issue. They did both #2 and #3 and drew the same conclusion from both, that shutter shock was not the issue.

 

[JaKing]

Not in my experience.

You also have both the E-M1 MkI and MkII. Haven't you noticed the difference in DR? I particularly notice the difference when the cats are lying half in and half out of the sun. There is little detail in the shaded parts with the MkI, and significantly more with the MkII. Of course, this difference is always there, just harder to see in many images.

 

[samtheman2014]

Looking at the near identical RAW results across the board with the various 20mp m43 model , it is reasonable to suppose that they all { not surprisingly } share the same 20mp Sony base sensor. Hence why other than processing tweaks { NR levels etc } and AA filter strength/weakness the output is so close,

Bill Claff's measurements show otherwise, very distinctly.

Not really, Bill plots PDR vs specified ISO setting rather than some sort of measured ISO value. Olympus decided to pull a Fuji with the ISO definitions on the E-M1ii so that plot looks much better than the camera performs in reality.

That's why I prefer the crazy old fashioned method of comparing RAW images Which unless you have a pair of Olympus brand rose tinted specs clearly show no advantage beyond any minor differences in processing . Between respective 20mp models .

I have read it explained t that Bill's DR testing method reduces the impact of read noise which in essence "cancels the advantage of the large sensor cameras of similar pixel count"

 

[ATP62]

Check out these two links :

Link 1

Link 2

 

[whumber]

Not in my experience.

You also have both the E-M1 MkI and MkII. Haven't you noticed the difference in DR? I particularly notice the difference when the cats are lying half in and half out of the sun. There is little detail in the shaded parts with the MkI, and significantly more with the MkII. Of course, this difference is always there, just harder to see in many images.

--
br, john, from you know where
My gear list and sordid past are here: https://www.dpreview.com/members/1558378718/overview
Gallery: https://www.canopuscomputing.com.au/zen2/page/gallery/

Not really, but let's do an experiment. I hope you can excuse the boring test subject but it's just something dark I happened to have in front of me. I underexposed both shots by 4 stops and then lifted them in LR and then exported both at 20MP. Both shot with the 45 Pro at f/4 | ISO 200 | 1/25s mounted on a tripod. I used the camera neutral profile for both. Can you see a significant difference between the two? Can you identify which is the E-M1 and which is the E-M1ii?




Image 1




Image 2

 

[JaKing]

Thanks for doing what you have done, but I really don't see your point.

I would call this a relatively low DR scene, and I see no point whatsoever in underexposing by 4 stops then lifting the resulting image by 4 stops. I would never do such a thing in taking and processing an image, and fail to see what relationship this has to the DR of any sensor.

You have also resized the images. While I understand why, this also destroys any point that you may have been trying to make.

My example with the cats being partially in full sun and partially in the deep shade does show a distinct difference between the MkI and MkII, and even more so when comparing with similar photos taken with my E-30 and E-510.

However, I do not currently have any such images uploaded to my web site, and do not have the time ATM to do so, in order to show that we seem to be talking about two completely different things.

An example of what I was talking about:

--
br, john, from you know where
My gear list and sordid past are here: https://www.dpreview.com/members/1558378718/overview
Gallery: https://www.canopuscomputing.com.au/zen2/page/gallery/

 

[whumber]

Thanks for doing what you have done, but I really don't see your point.

I would call this a relatively low DR scene, and I see no point whatsoever in underexposing by 4 stops then lifting the resulting image by 4 stops. I would never do such a thing in taking and processing an image, and fail to see what relationship this has to the DR of any sensor.

Umm, because that's the whole point of the dynamic range conversation. How much information can be recover in the shadows before electronic read noise overcomes the details in the scene. The ideal scenario is that the electronic read noise is essentially zero and all you're left with is shot noise. You don't actually need to high contrast scene to get a measurement of sensor dynamic range.

You have also resized the images. While I understand why, this also destroys any point that you may have been trying to make.

Uh, no it doesn't destroy the point at all. If you make the comparison at different magnifications then you're comparison is significantly less meaningful.

My example with the cats being partially in full sun and partially in the deep shade does show a distinct difference between the MkI and MkII, and even more so when comparing with similar photos taken with my E-30 and E-510.

However, I do not currently have any such images uploaded to my web site, and do not have the time ATM to do so, in order to show that we seem to be talking about two completely different things.

An example of what I was talking about:

Ok, so you seem to be confusing dynamic range with default tone curves.

 

[JaKing]

Thanks for doing what you have done, but I really don't see your point.

I would call this a relatively low DR scene, and I see no point whatsoever in underexposing by 4 stops then lifting the resulting image by 4 stops. I would never do such a thing in taking and processing an image, and fail to see what relationship this has to the DR of any sensor.

Umm, because that's the whole point of the dynamic range conversation. How much information can be recover in the shadows before electronic read noise overcomes the details in the scene. The ideal scenario is that the electronic read noise is essentially zero and all you're left with is shot noise. You don't actually need to high contrast scene to get a measurement of sensor dynamic range.

?? would you care to explain how that's the case?

You have also resized the images. While I understand why, this also destroys any point that you may have been trying to make.

Uh, no it doesn't destroy the point at all. If you make the comparison at different magnifications then you're comparison is significantly less meaningful.

My example with the cats being partially in full sun and partially in the deep shade does show a distinct difference between the MkI and MkII, and even more so when comparing with similar photos taken with my E-30 and E-510.

However, I do not currently have any such images uploaded to my web site, and do not have the time ATM to do so, in order to show that we seem to be talking about two completely different things.

An example of what I was talking about:

Ok, so you seem to be confusing dynamic range with default tone curves.

Not how I understand the term "dynamic range".

--
br, john, from you know where
My gear list and sordid past are here: https://www.dpreview.com/members/1558378718/overview
Gallery: https://www.canopuscomputing.com.au/zen2/page/gallery/

 

[whumber]

Umm, because that's the whole point of the dynamic range conversation. How much information can be recover in the shadows before electronic read noise overcomes the details in the scene. The ideal scenario is that the electronic read noise is essentially zero and all you're left with is shot noise. You don't actually need to high contrast scene to get a measurement of sensor dynamic range.

?? would you care to explain how that's the case?

That the ideal scenario is that the shadow noise is shot noise limited or that you don't need a high contrast scene to measure dynamic range?

My example with the cats being partially in full sun and partially in the deep shade does show a distinct difference between the MkI and MkII, and even more so when comparing with similar photos taken with my E-30 and E-510.

However, I do not currently have any such images uploaded to my web site, and do not have the time ATM to do so, in order to show that we seem to be talking about two completely different things.

An example of what I was talking about:

Ok, so you seem to be confusing dynamic range with default tone curves.

Not how I understand the term "dynamic range".

So would you say that in the images below, that the image on the right has better dynamic range than the image on the left?

 

[TomFid]

I disagree. Lenses have different mass and stiffness, which may interact with the body to vary the effective shock. And with long lenses, there may be no shutter speed that is high enough. I tried a lot of things to get sharp images out of a GH2 + Meade ETX, but it just wasn't possible with the mechanical shutter.

That may be but I really doubt even 800mm EFL is going to run into that limit. The E-M1 has pretty pronounced shutter shock with the mechanical shutter and I can't see any hint of it at 1/1000s with an EFL of 840mm. Is it enough to be measured? Probably. Is it enough to cause the degree of softness shown in LensTips testing? Definitely not. If any, sample variation is much more likely of a candidate if there's anything off about the test.

I agree that sample variation is a strong candidate, but neither of us, nor anyone at LensTip, actually knows without running the test!

When there's an easy test that would confirm or refute your results, you do it. You don't just wave your hands and invoke nefarious foreigners.

But the whole point is that there's not an easy test like you're suggesting, they can't just add an electronic shutter to the E-PL1. If they switch over to, say, an E-M1 now we can't compare the results to any of the historical data and because we're not using the same shutter mechanism we can't even say if the relative performance between the mechanical and electronic shutter would be the same. To do things in a meaningful manner you either need to switch your entire system over to electronic shutter, find a way of minimizing potential shutter shock, or find some other test to say if shutter shock is even an issue. They did both #2 and #3 and drew the same conclusion from both, that shutter shock was not the issue.

I'm not proposing that they shift their whole system to another body, just that they repeat the test on another body for verification. If the resolution peak is in the same place, great - they're right. And, it IS easy. I can shoot a series of shots of a resolution target and run them through MTFmapper in the time it would take you to have a leisurely cup of coffee.

Realistically, with more 20mp bodies and HiRes around, LensTip does need to modernize if they want to stay relevant.

 

[JaKing]

Try here:

Loading…

theonlinephotographer.typepad.com

Then follow on to Part 2 ...

Or here:

http://www.cambridgeincolour.com/tutorials/dynamic-range.htm

 

[whumber]

Try here:

http://theonlinephotographer.typepa...namic-range-is-not-exposure-range-part-i.html

Then follow on to Part 2 ...

Or here:

http://www.cambridgeincolour.com/tutorials/dynamic-range.htm

 

[whumber]

I'm not proposing that they shift their whole system to another body, just that they repeat the test on another body for verification. If the resolution peak is in the same place, great - they're right. And, it IS easy. I can shoot a series of shots of a resolution target and run them through MTFmapper in the time it would take you to have a leisurely cup of coffee.

It's extremely unlikely that changing from a mechanical to electronic shutter would have any significant effect on the shape of the MTF curve in this case, even if the shutter shock were significant. Most likely what you would see is that the curve would see some combination of a shift up and stretching in the vertical direction but the aperture at which the peak occurs won't change. You would then need to be able to differentiate the contribution of changing the sensor (resolution, AA filter, etc.) vs the contribution of the shutter shock which is non-trivial without testing a bunch of other lenses to look for relative changes.

Realistically, with more 20mp bodies and HiRes around, LensTip does need to modernize if they want to stay relevant.

Sure, I agree with this.