GFX 100 vs a7RIV landscape IQ Locked

Started 5 months ago | Discussions
This thread is locked.
Macro guy
Macro guy Veteran Member • Posts: 4,566
Re: GFX 100 vs a7RIV landscape IQ

JimKasson wrote:

Jim

Hi Jim,

Very good demo!

How much of that difference is due to the GFX 100 having more pixels?

A lot of it. Probably most of it. Both those lenses are really so good that they could use 800 MP sensors. But at the end of the day, it's easier to make an equivalent lens for a larger sensor.

Jim

Jim,

If the lenses in both formats are capable if resolving up to 800mp, would it be accurate to assume that the sensor size makes no difference from a resolution perspective if both cameras have the same megapixel count? If that's the case, what would be the reason to spend more on a camera with a slightly bigger sensor if a camera with a slightly smaller sensor can deliver identical (or at least very similar results) for less?

 Macro guy's gear list:Macro guy's gear list
Canon EOS 5D Mark II Canon EF 50mm F1.8 II +5 more
Luvwine Regular Member • Posts: 289
Re: Using a more sophisticated upsampling method

JimKasson wrote:

Using a more sophisticated upsampling method yields quite different results:

GFX 100

a7RIV

https://blog.kasson.com/gfx-100/sony-a7riv-vs-fuji-gfx-100-landscape-iq-with-gigapixel-ai/

I admit to being quite surprised at how well GigaPixel AI did with the Sony file.

Jim

I don't understand this result.  If Fuji's image provided more detail prior to uprezzing, then the difference seems to come from the Gigapixel engine doing better with the Sony file than the Fuji file.  Why would this be?  Is the result subject matter dependent?  Would it do as well with a portrait or a rock face as it does in this instance with leaves on a tree?  Not trying to make you do a bunch of testing, but if the previous testing is correct (I have no reason to dispute it) and file from the Fuji GFX shows more detail/resolution than the Sony A7rIV prior to being plugged into Gigapixel, then what is happening?  If Gigapixel is not creating data and Sony has less data, it must be how Gigapixel handles the file.  Perhaps the software is optimized for the Sony file somehow?  More sharpening?

-- hide signature --

Luvwine
Sony A7rIV (x2), Zeiss Loxia 21/2.8, Sigma Art 35/1.2, Voigtlander 40/1.2, Voigtlander 50/1.2, Voigtlander 65/2 apo-lanthar, Zeiss Loxia 85/2.4, Sony GM 85/1.4, Voigtlander 110/2.5 apo lanthar, Leica R 180/2.8 apo, Leica R 280/4 apo, Leica R 1.4x Extender, bags by Thinktank, Ona, and others, Canon IPF 8300

 Luvwine's gear list:Luvwine's gear list
Fujifilm GFX 100 Zeiss Loxia 21mm F2.8 Fujifilm GF 32-64mm F4 Fujifilm GF 110mm F2 Fujifilm GF 23mm F4 +4 more
left eye
left eye Senior Member • Posts: 2,020
Re: Using a more sophisticated upsampling method

Luvwine wrote:

JimKasson wrote:

Using a more sophisticated upsampling method yields quite different results:

GFX 100

a7RIV

https://blog.kasson.com/gfx-100/sony-a7riv-vs-fuji-gfx-100-landscape-iq-with-gigapixel-ai/

I admit to being quite surprised at how well GigaPixel AI did with the Sony file.

Jim

I don't understand this result. If Fuji's image provided more detail prior to uprezzing, then the difference seems to come from the Gigapixel engine doing better with the Sony file than the Fuji file. Why would this be? Is the result subject matter dependent? Would it do as well with a portrait or a rock face as it does in this instance with leaves on a tree? Not trying to make you do a bunch of testing, but if the previous testing is correct (I have no reason to dispute it) and file from the Fuji GFX shows more detail/resolution than the Sony A7rIV prior to being plugged into Gigapixel, then what is happening? If Gigapixel is not creating data and Sony has less data, it must be how Gigapixel handles the file. Perhaps the software is optimized for the Sony file somehow? More sharpening?

if you read Jim's blog article that he linked to, he offers a likely explanation.

 left eye's gear list:left eye's gear list
Olympus E-300 Fujifilm GFX 50S Panasonic GH5 Fujifilm GF 50mm F3.5 +5 more
Erik Kaffehr
Erik Kaffehr Veteran Member • Posts: 3,754
Re: GFX 100 vs a7RIV landscape IQ

Hiphopapotamus wrote:

So you don't agree with the science of MTF charts and you prefer to come to your own conclusions that support your own views. This is human fallibility in action. In technical terms its called...

Hi,

I do agree with MTF charts that are correctly handled. A website called 'Photodo' did tests at the Hasselblad factory, using their MTF testing bed:

Mamiya N 150mm f/4.5 L for Mamiya 7

Hasselblad Sonnar 150/4 CF

The Mamiya is a bit sharper at the center.

I also did a very simple calculation on MTF from Velvia:

But that analysis does not involve the MTF of the enlarger or of the scanner used.

Now, some guys stated that Phase One's IQ180 (an 80 MP 54x41 mm CCD back) performed comparably to 8"x10" film. Tim Parkin, who shoots large format didn't agree, so he investigated. Here are some samples from his testing.

Mamiya T-Max 100 compared, drum scanned to Phase One IQ 180

Mamiya 7, T-Max 4800 PPI Epson V750, not very close.

4"x5" on Provia scanned at 4000 PPI yields more detail than the IQ180.

The IQ180 tested here is an old back. How does it compare to the GFX100?

It seems that GFX 100 (blue curve) is a bit better than the IQ1380MP

Comparing the IQ 180 with the A7rIV the IQ 180 (red curve) is a bit better.

So, I would say that the IQ 180 results are a bit below what can be expected from the GFX 100.

Confirmation bias

Well, I think I looked at the facts.

If you're prepared to actually look at the history books, Mamiya got to where it was by copying and then exacting Schneider and Zeiss optical designs. Their rangefinders were copies of military-grade European rangefinders...

They not only made copies but they improved upon them which is why the Mamiya 7 is still regarded as one of the best rangefinders ever made.

Most Mamiya lenses are actually excellent, rather than good... And I have it on the authority of the types of people I actually know who are trained to rebuild the things and have been working on the tools for more than 50 years now.

Nothing that contradicts that. But lens design has developed a bit since 1990...

Best regards

Erik

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

Rick Knepper
Rick Knepper Forum Pro • Posts: 17,664
Re: Jim’s next article with gigapixel upscale

Joohan wrote:

Seems to zero out the benefits of gfx100. Interesting ....

The GFX 100 & the a7rIV have the same pixel pitch which is one reason I allowed myself to backslide to FF.

-- hide signature --

Once you've done fifty, anything less is iffy.

 Rick Knepper's gear list:Rick Knepper's gear list
Pentax 645Z Canon EOS 5DS R Fujifilm GFX 50S Sony a7R IV Pentax smc D FA 645 55mm F2.8 AL (IF) SDM AW +10 more
Greg7579
Greg7579 Veteran Member • Posts: 8,495
Re: Jim’s next article with gigapixel upscale

Joohan wrote:

Seems to zero out the benefits of gfx100. Interesting ....

Joohan,

You and I know each other well from years on the Fuji Board. I have not been on that Board since January of 2011, much to my displeasure.

But, if you are here on the MF Board, like I was starting about a year ago, trying to decide whether or not to venture out from wonderful Fuji X gear (APSC) and go towards high-res FF Sony and the alphasevenarefour or Fuji GFX, just do what I did.

GFX.

Do not draw a conclusion like the one you said above because it is wrong and is a common fake talking point of the anti MF crowd.  Too easy to spout off, but always wrong.

I see that you are on the Sony Board claiming that there is no reason to upgrade from the alphasevenarethree to the alphasevenarefour. I'm sure some alphasevenarefour owners would disagree with that conclusion (as I do), but I never comment on that Board because I do not own Sony gear. Therefore I spare them my brilliant opinions about their gear.

So if you are here to make decisions and thinking about MF as an option, you are in the right place.

But if you are here to regurgitate anti-MF talking point quips, then I'm not sure what that accomplishes.

Best wishes. PM me and I will give you my phone number. Call me on Whats App from Sweden and I will tell you everything I know about why it is really awesome to shoot MF and what you are going to experience moving from Fuji APSC.

But, if you decide on the alphasevenarefour, that is such a great camera that it is very often the right decision. It is way more versatile and it is outstanding - it is one of the best 3 or 4 cameras in the world.

I may still get one.

But the IQ results of the GFX 100 are on a different planet. Do not believe the false narrative and anti-MF talking points about alphasevenarefour IQ matching GFX 100 IQ (or even GFX 50).

The alphasevenarefour is not MF.

-- hide signature --
 Greg7579's gear list:Greg7579's gear list
Fujifilm XF 16mm F1.4 R WR XF 90mm Fujifilm X100F Leica Q2 Fujifilm X-H1 +16 more
Greg7579
Greg7579 Veteran Member • Posts: 8,495
Backslide..... 😂

Rick Knepper wrote:

…. I allowed myself to backslide to FF.

Rick, you are a teller of truth and are a wise sage of what is true and good.

You almost went GFX 100, but decided on the great alphasevenarefour.  That is understandable, especially in your case where you want some serious reach and action shots.

But it is also a backslide.  😎

-- hide signature --
 Greg7579's gear list:Greg7579's gear list
Fujifilm XF 16mm F1.4 R WR XF 90mm Fujifilm X100F Leica Q2 Fujifilm X-H1 +16 more
Erik Kaffehr
Erik Kaffehr Veteran Member • Posts: 3,754
Re: Jim’s next article with gigapixel upscale

Rick Knepper wrote:

Joohan wrote:

Seems to zero out the benefits of gfx100. Interesting ....

The GFX 100 & the a7rIV have the same pixel pitch which is one reason I allowed myself to backslide to FF.

Same pixel pitch and probably the same pixel design, but the GFX 100 has more of them.

Quantity has a quality of it's own.

To that comes that it is entirely possible that the GF lenses are better.

Best regards

Erik

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

Greg7579
Greg7579 Veteran Member • Posts: 8,495
Re: Jim’s next article with gigapixel upscale

Erik Kaffehr wrote:

Rick Knepper wrote:

Joohan wrote:

Seems to zero out the benefits of gfx100. Interesting ....

The GFX 100 & the a7rIV have the same pixel pitch which is one reason I allowed myself to backslide to FF.

Same pixel pitch and probably the same pixel design, but the GFX 100 has more of them.

Quantity has a quality of it's own.

To that comes that it is entirely possible that the GF lenses are better.

Best regards

Erik

I approve of this message!

-- hide signature --
 Greg7579's gear list:Greg7579's gear list
Fujifilm XF 16mm F1.4 R WR XF 90mm Fujifilm X100F Leica Q2 Fujifilm X-H1 +16 more
JimKasson
OP JimKasson Forum Pro • Posts: 28,219
Re: Jim’s next article with gigapixel upscale

Greg7579 wrote:

Erik Kaffehr wrote:

Rick Knepper wrote:

Joohan wrote:

Seems to zero out the benefits of gfx100. Interesting ....

The GFX 100 & the a7rIV have the same pixel pitch which is one reason I allowed myself to backslide to FF.

Same pixel pitch and probably the same pixel design, but the GFX 100 has more of them.

Quantity has a quality of it's own.

To that comes that it is entirely possible that the GF lenses are better.

Best regards

Erik

I approve of this message!

WRT the test that is the subject of this thread, I don't think the 110/2 is appreciably better than the Otus 85. Or appreciably worse, either.

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Sony a7 III Nikon Z7 Fujifilm GFX 100 Sony a7R IV +2 more
JimKasson
OP JimKasson Forum Pro • Posts: 28,219
Further thoughts on the GigaPixel AI results

I've thought some more about the GigaPixel AI results. The raw images received the same amount of sharpening on a per-pixel level in Lightroom: amount 20, radius 1, detail 0. If we're trying to compensate for the light-sensitive area of the pixels, that's appropriate. The pixel pitch of the two cameras is the same, the pixel design appears to be very similar, and so do the microlenses. But if we're compensating for depth of field and for diffraction, the radius for the larger sensor should be about 1.4 times that of the smaller one. As it is, viewed in relation to the picture height, the Sony image is getting about 1.4 times the sharpening that we're giving the Fuji one.

As I said in the blog post, GigaPixel AI seems to sharpen more the more it upsamples. That also would benefit the Sony image, since the two images were somewhat undersharpened. This is probably worth more experimentation using upsampling algorithms that are less of a black box than the Topaz software. Because the software is essentially inventing information, its efficacy will depend on the content of the image. As an example of that, as we saw with the Siemens Star target tests that I showed in an earlier post, at some point, GigaPixel AI -- quite properly, in my opinion -- gives up and stops trying to make up detail. So before drawing any general conclusions, we should look at disparate example crops.

Finally (at least finally for now), I earlier observed that I thought the biggest advantage of the GFX 100 over the GFX 50x was not increased sharpness, but decreased aliasing. There is no aliasing in the crops above that is immediately obvious to me, but I'm sure that there is aliasing there, and one of the functions of aliasing is to turn high spatial frequency hard-to-see small details into lower frequency easier-to-see -- but wrong -- details. That may be a partial reason for the Sony images surprising sharpness.

Jim

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Sony a7 III Nikon Z7 Fujifilm GFX 100 Sony a7R IV +2 more
JimKasson
OP JimKasson Forum Pro • Posts: 28,219
Re: Jim’s next article with gigapixel upscale

Joohan wrote:

Seems to zero out the benefits of gfx100. Interesting ....

Two things:

-- hide signature --
 JimKasson's gear list:JimKasson's gear list
Nikon D5 Sony a7 III Nikon Z7 Fujifilm GFX 100 Sony a7R IV +2 more
Teila Day
Teila Day Veteran Member • Posts: 4,987
Relevance?

Macro guy wrote:

JimKasson wrote:

Jim

Hi Jim,

Very good demo!

How much of that difference is due to the GFX 100 having more pixels?

A lot of it. Probably most of it. Both those lenses are really so good that they could use 800 MP sensors. But at the end of the day, it's easier to make an equivalent lens for a larger sensor.

Jim

Jim,

If the lenses in both formats are capable if resolving up to 800mp, would it be accurate to assume that the sensor size makes no difference from a resolution perspective if both cameras have the same megapixel count? If that's the case, what would be the reason to spend more on a camera with a slightly bigger sensor if a camera with a slightly smaller sensor can deliver identical (or at least very similar results) for less?

Macro guy

I'm curious what 'how much a lens can resolve' has anything to do with the inherent differences between sensor sizes (even with the same number of pixels).  Off the top of my head (I am not an engineer) I would think that there are enough attribute differences between cameras to make the advantage(s) evident in even a casual test at the counter in a camera store.

Isn't the difference (advantage or disadvantage) between say, a Canon 5Ds (50mp) and a 50mp Hasselblad/Pentax/Fuji rather evident without much fiddling or investigation?

Below is the kind of real-word difference that I'd see between the Canon and the MF offerings of any kind, especially when it came to certain foliage when the sun was out.. leaves were quicker to return to white, less malleable detail/options in post with the Canon... etc..  but I didn't need DPR to tell me that, I could readily see it with my old eyes, especially the difference in processing raws  

Look at the sheen / reflected highlights and you get the idea of what I'm talking about.  Larger sensor = softer, especially the newer MF cameras even if they use the same 'base' 44x33 sensor.

Throw in other differences (higher ISO options, etc..) and whether those differences are worth it to you or not, can really be an easy decision making process.

Best in photography to you

-- hide signature --
Macro guy
Macro guy Veteran Member • Posts: 4,566
Re: Relevance?

Teila Day wrote:

Macro guy wrote:

JimKasson wrote:

Jim

Hi Jim,

Very good demo!

How much of that difference is due to the GFX 100 having more pixels?

A lot of it. Probably most of it. Both those lenses are really so good that they could use 800 MP sensors. But at the end of the day, it's easier to make an equivalent lens for a larger sensor.

Jim

Jim,

If the lenses in both formats are capable if resolving up to 800mp, would it be accurate to assume that the sensor size makes no difference from a resolution perspective if both cameras have the same megapixel count? If that's the case, what would be the reason to spend more on a camera with a slightly bigger sensor if a camera with a slightly smaller sensor can deliver identical (or at least very similar results) for less?

Macro guy

I'm curious what 'how much a lens can resolve' has anything to do with the inherent differences between sensor sizes (even with the same number of pixels). Off the top of my head (I am not an engineer) I would think that there are enough attribute differences between cameras to make the advantage(s) evident in even a casual test at the counter in a camera store.

Actually, no. The only inherent difference is the difference in depth of field at a given aperture.

Isn't the difference (advantage or disadvantage) between say, a Canon 5Ds (50mp) and a 50mp Hasselblad/Pentax/Fuji rather evident without much fiddling or investigation?

Not at all. I am struggling to see any advantages at a given pixel count.

Below is the kind of real-word difference that I'd see between the Canon and the MF offerings of any kind, especially when it came to certain foliage when the sun was out.. leaves were quicker to return to white, less malleable detail/options in post with the Canon... etc.. but I didn't need DPR to tell me that, I could readily see it with my old eyes, especially the difference in processing raws

Look at the sheen / reflected highlights and you get the idea of what I'm talking about. Larger sensor = softer, especially the newer MF cameras even if they use the same 'base' 44x33 sensor.

Throw in other differences (higher ISO options, etc..) and whether those differences are worth it to you or not, can really be an easy decision making process.

Best in photography to you

You are assuming that what you're seeing are differences in sensor size. You could be seeing the differences in lenses or the differences in sensor tech or implementation or a combination of the above. In other words, you really don't know what you're looking at, or rather the causes of whatever differences you're seeing.

 Macro guy's gear list:Macro guy's gear list
Canon EOS 5D Mark II Canon EF 50mm F1.8 II +5 more
Erik Kaffehr
Erik Kaffehr Veteran Member • Posts: 3,754
Re: Relevance?

Macro guy wrote:

Below is the kind of real-word difference that I'd see between the Canon and the MF offerings of any kind, especially when it came to certain foliage when the sun was out.. leaves were quicker to return to white, less malleable detail/options in post with the Canon... etc.. but I didn't need DPR to tell me that, I could readily see it with my old eyes, especially the difference in processing raws

Look at the sheen / reflected highlights and you get the idea of what I'm talking about. Larger sensor = softer, especially the newer MF cameras even if they use the same 'base' 44x33 sensor.

Throw in other differences (higher ISO options, etc..) and whether those differences are worth it to you or not, can really be an easy decision making process.

Best in photography to you

You are assuming that what you're seeing are differences in sensor size. You could be seeing the differences in lenses or the differences in sensor tech or implementation or a combination of the above. In other words, you really don't know what you're looking at, or rather the causes of whatever differences you're seeing.

Hi,

I don't think DPReview's studio set up is really constant. I think that both objects and lights are moved between shots. Also, even with the lights in the same position, reflection will move if the camera position changes and the shots are taken with some variance in focal lengths and thus also camera to subject distance.

Just to say, the DPReview shots are good to have, I have not seen a better alternative. But, they are not the same between shots.

Also, cameras are customer devices, they are not build to reproducible specifications. Some samples are better than others, or just different.
Best regards

Erik

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

Teila Day
Teila Day Veteran Member • Posts: 4,987
Re: Relevance (reiterated.. slowly)

Macro guy wrote:

Teila Day wrote:

Macro guy wrote:

JimKasson wrote:

Jim

Hi Jim,

Very good demo!

How much of that difference is due to the GFX 100 having more pixels?

A lot of it. Probably most of it. Both those lenses are really so good that they could use 800 MP sensors. But at the end of the day, it's easier to make an equivalent lens for a larger sensor.

Jim

Jim,

If the lenses in both formats are capable if resolving up to 800mp, would it be accurate to assume that the sensor size makes no difference from a resolution perspective if both cameras have the same megapixel count? If that's the case, what would be the reason to spend more on a camera with a slightly bigger sensor if a camera with a slightly smaller sensor can deliver identical (or at least very similar results) for less?

Macro guy

I'm curious what 'how much a lens can resolve' has anything to do with the inherent differences between sensor sizes (even with the same number of pixels). Off the top of my head (I am not an engineer) I would think that there are enough attribute differences between cameras to make the advantage(s) evident in even a casual test at the counter in a camera store.

Which I've done between the 645z and the Canon offering; don whatever lens you like and you pretty much get the same k-i-n-d of difference. You'll see more of a contrasty processed look in the Canon's raw in comparison to the 645z. Compare with the older Hassy and you often see the same.

Actually, no. The only inherent difference is the difference in depth of field at a given aperture.

Oh? ISO sensitivity? Do the raws take the same abuse and yield the same results across the ISO range? Practical differences. They didn't to my eyes.

Isn't the difference (advantage or disadvantage) between say, a Canon 5Ds (50mp) and a 50mp Hasselblad/Pentax/Fuji rather evident without much fiddling or investigation?

Not at all. I am struggling to see any advantages at a given pixel count.

I think it would be moot today which is why I wouldn't purchase a 44x33 camera today, but when comparing cameras years ago I noted differences. I was eager to snap up the Canon but it wasn't to my liking.

Below is the (additional input)

** kind of real-word difference that I'd see between the Canon (meaning the 5ds series) and the MF offerings of any kind (with the aging 33x44 CMOS found in the 645z et al, or current MF cameras with an even better sensor) ** (end input)

especially when it came to certain foliage when the sun was out.. leaves were quicker to return to white, less malleable detail/options in post with the Canon... etc.. but I didn't need DPR to tell me that, I could readily see it with my old eyes, especially the difference in processing raws

Look at the sheen / reflected highlights and you get the idea of what I'm talking about. Larger sensor = softer, especially the newer MF cameras even if they use the same 'base' 44x33 sensor.

Throw in other differences (higher ISO options, etc..) and whether those differences are worth it to you or not, can really be an easy decision making process.

Best in photography to you

You are assuming that what you're seeing are differences in sensor size.

I'm not assuming. I'm basing my assertion on what I commonly notice between shooting MF and FF across the brands. Process thousands of files of both and the difference is pretty easily noticeable. I don't think anyone who's processed thousands of files of MF (CMOS) and compared them to processing their FF raw files cannot tell a difference.

You could be seeing the differences in lenses or the differences in sensor tech or implementation or a combination of the above. In other words, you really don't know what you're looking at, or rather the causes of whatever differences you're seeing.

Lenses? I can take a old-as-dirt 400mm or 120mm lens from the film days and get the same general result that I'm talking about- especially when it comes to resolution / pixels which is a primary reason for shooting MF for many photographers manipulating raw files whether the output be intended for screen or print.

Slap whatever lens you have on the Canon 5Ds series - do the same for the Pentax or same vintage Hasselblad and you get the same basic result irrespective of lenses used. What that means to me is that the lenses aren't the cause of the slightly-cooked difference in the Canon's raw files that I was seeing back then. Switch cameras (same models), same thing.

Now since I can't buy a 5dsr, 645z or Hasselblad without whatever electronic seasoning that comes from the factory, in layman's terms, that means that the 5ds/r didn't perform the same as the 33x44 based MF offerings. Better or worse is up to the photographer, but my bet is that it's hard to find an owner of both to attest that they can't tell a difference, ** especially ** after shooting MF for months, then going back to the same era FF Canon/Nikon.

In that vein - 50mp doesn't look like 24mp just like 150mp doesn't look like 50mp, especially when shooting small subjects, cooking the file in your favorite processor, a bit of cropping, then sending to output.. the practical advantages are pretty obvious.

Today the difference between 50mp MF and a 65mp FF camera isn't worth quibbling about in most cases.. and I think it goes without saying that most photographers worried about practical advantages and a bottom line won't shell out tens of thousands in gear unless the difference is a decisive one; not one predicated on emotion.

I sold a BW shot taken at 51k ISO, and several series at 12k ISO... don't think I could've done the former, or that to my liking in the case of the latter, with the 5Ds series.

Let me reiterate... the example that I gave (DPR Graphic above) isn't the point. The point isn't how DPR does their testing. The point is that the graphic represents the k-i-n-d of difference that I've seen between the FF Canon and same-era MF offerings is all. How DPR does their testing isn't relevant here.

Kind regards and best in photography to all of you!

-- hide signature --

Teila K. Day
http://teiladay.com

Erik Kaffehr
Erik Kaffehr Veteran Member • Posts: 3,754
Re: Jim’s next article with gigapixel upscale

JimKasson wrote:

Greg7579 wrote:

Erik Kaffehr wrote:

Rick Knepper wrote:

Joohan wrote:

Seems to zero out the benefits of gfx100. Interesting ....

The GFX 100 & the a7rIV have the same pixel pitch which is one reason I allowed myself to backslide to FF.

Same pixel pitch and probably the same pixel design, but the GFX 100 has more of them.

Quantity has a quality of it's own.

To that comes that it is entirely possible that the GF lenses are better.

Best regards

Erik

I approve of this message!

WRT the test that is the subject of this thread, I don't think the 110/2 is appreciably better than the Otus 85. Or appreciably worse, either.

Hi Jim,

Thanks for the remark. My comment did not regard the Otus, but the GF lens system compared to the Sony lens system.

Best regards

Erik

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

Macro guy
Macro guy Veteran Member • Posts: 4,566
Re: Relevance (reiterated.. slowly)

Teila Day wrote:

You are assuming that what you're seeing are differences in sensor size.

I'm not assuming. I'm basing my assertion on what I commonly notice between shooting MF and FF across the brands. Process thousands of files of both and the difference is pretty easily noticeable. I don't think anyone who's processed thousands of files of MF (CMOS) and compared them to processing their FF raw files cannot tell a difference.

You could be seeing the differences in lenses or the differences in sensor tech or implementation or a combination of the above. In other words, you really don't know what you're looking at, or rather the causes of whatever differences you're seeing.

Lenses? I can take a old-as-dirt 400mm or 120mm lens from the film days and get the same general result that I'm talking about- especially when it comes to resolution / pixels which is a primary reason for shooting MF for many photographers manipulating raw files whether the output be intended for screen or print.

Slap whatever lens you have on the Canon 5Ds series - do the same for the Pentax or same vintage Hasselblad and you get the same basic result irrespective of lenses used. What that means to me is that the lenses aren't the cause of the slightly-cooked difference in the Canon's raw files that I was seeing back then. Switch cameras (same models), same thing.

Now since I can't buy a 5dsr, 645z or Hasselblad without whatever electronic seasoning that comes from the factory, in layman's terms, that means that the 5ds/r didn't perform the same as the 33x44 based MF offerings. Better or worse is up to the photographer, but my bet is that it's hard to find an owner of both to attest that they can't tell a difference, ** especially ** after shooting MF for months, then going back to the same era FF Canon/Nikon.

Ok, so you don't like the Canon sensor.  That's fine, but would that sensor have been any better scaled up to 33x44?  Would the Sony sensor have been any worse scaled down to 24x36? (aside from the resolution differences of course)

In that vein - 50mp doesn't look like 24mp just like 150mp doesn't look like 50mp, especially when shooting small subjects, cooking the file in your favorite processor, a bit of cropping, then sending to output.. the practical advantages are pretty obvious.

Today the difference between 50mp MF and a 65mp FF camera isn't worth quibbling about in most cases.. and I think it goes without saying that most photographers worried about practical advantages and a bottom line won't shell out tens of thousands in gear unless the difference is a decisive one; not one predicated on emotion.

We have a great opportunity to actually compare the same sensor in the Sony A7R4 and in the Fuji GFX100.  As Jim's test have shown, aside from the obvious resolution difference that comes from having a bigger sensor, both cameras are on par with regard to IQ.  And that brings me to the point that I've been trying to make all along and that is that the ONLY advantage that 33x44 has over 24x36 format is greater resolution at a given pixel pitch.

I sold a BW shot taken at 51k ISO, and several series at 12k ISO... don't think I could've done the former, or that to my liking in the case of the latter, with the 5Ds series.

You're absolutely right.  The 5DS doesn't go above ISO 6400

 Macro guy's gear list:Macro guy's gear list
Canon EOS 5D Mark II Canon EF 50mm F1.8 II +5 more
Erik Kaffehr
Erik Kaffehr Veteran Member • Posts: 3,754
Re: Further thoughts on the GigaPixel AI results

JimKasson wrote:

I've thought some more about the GigaPixel AI results. The raw images received the same amount of sharpening on a per-pixel level in Lightroom: amount 20, radius 1, detail 0. If we're trying to compensate for the light-sensitive area of the pixels, that's appropriate. The pixel pitch of the two cameras is the same, the pixel design appears to be very similar, and so do the microlenses. But if we're compensating for depth of field and for diffraction, the radius for the larger sensor should be about 1.4 times that of the smaller one. As it is, viewed in relation to the picture height, the Sony image is getting about 1.4 times the sharpening that we're giving the Fuji one.

I see your point, I think. It may make some sense to look at slanted edges and check the effect of sharpening, radius vs. amount, etc.

As I said in the blog post, GigaPixel AI seems to sharpen more the more it upsamples. That also would benefit the Sony image, since the two images were somewhat undersharpened. This is probably worth more experimentation using upsampling algorithms that are less of a black box than the Topaz software. Because the software is essentially inventing information, its efficacy will depend on the content of the image. As an example of that, as we saw with the Siemens Star target tests that I showed in an earlier post, at some point, GigaPixel AI -- quite properly, in my opinion -- gives up and stops trying to make up detail. So before drawing any general conclusions, we should look at disparate example crops.

Finally (at least finally for now), I earlier observed that I thought the biggest advantage of the GFX 100 over the GFX 50x was not increased sharpness, but decreased aliasing. There is no aliasing in the crops above that is immediately obvious to me, but I'm sure that there is aliasing there, and one of the functions of aliasing is to turn high spatial frequency hard-to-see small details into lower frequency easier-to-see -- but wrong -- details. That may be a partial reason for the Sony images surprising sharpness.

Jim

That aliasing stuff may be interesting. It is entirely possible that GigaPixel AI picks up aliases and turns them into fake detail. I would presume that the A7rIV may have more aliasing as it is used closer to it's optimal aperture.

Comparing images at same aperture, aotbe, my be a good test?

Best regards

Erik

-- hide signature --

Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles

epeorusjim Forum Member • Posts: 69
Re: I Do Not Accept That Talking Point...

I’m going with delusional on this one.

Keyboard shortcuts:
FForum MMy threads