Photons to photos dynamic range ratings

On the site, the A7riv has about +1/2 stop more dynamic range than the 50MP GFX sensor used in 35mm crop mode.

https://www.photonstophotos.net/Charts/PDR.htm#FujiFilm GFX 50S(FF),Sony ILCE-7RM4

Q1: Does half a stop DR improvement accurately reflect the difference between the old and new underlying sensor generations?

On the same site, switching to full 44x33 mode improves the 50MP sensor to + 1/3 stop advantage over the newer 61MP sensor, so the difference from the sensor size increase is about +3/4 of a stop.

https://www.photonstophotos.net/Charts/PDR.htm#FujiFilm GFX 50S,Sony ILCE-7RM4

The figs for the Sony vs 100MP GFX (same sensor tech) in 35mm crop mode are essentially identical:

https://www.photonstophotos.net/Charts/PDR.htm#FujiFilm GFX 100S(FF),Sony ILCE-7RM4

The advantage that comes purely from increasing sensor size to 44x33 is just under +2/3 of a stop.

https://www.photonstophotos.net/Charts/PDR.htm#FujiFilm GFX 100S,Sony ILCE-7RM4

To summarise, roughly speaking there is about 1/2 stop difference in DR between the old and new sensor tech, and about +2/3 to 3/4 stop attributable to the larger sensor in both generations.

Have I got that right?

What is missing from this is the detail improvement from the extra pixels in the 100MP sensor, of course. You can enlarge further without losing quality.

Q2: In terms of FF to MF comparisons, is there a simple way of writing down the improvements in noise, DR and extra enlargeability as a percentage for each?

eg

DR: + x%

Reduced noise: x%

Extra enlargeability: + x%

Basically, some simple numbers we can memorise and quote whenever the question arises to save us having to repeat the research and write another 150 posts. At the moment all I have is a handwavy "MF is a bit better than FF".

It would nice to have a convenient number "MF is 10% better than FF" or whatever the number is



--
Photo of the day: https://whisperingcat.co.uk/wp/photo-of-the-day/
Website: http://www.whisperingcat.co.uk/ (2022 - website rebuilt, updated and back in action)
DPReview gallery: https://www.dpreview.com/galleries/0286305481
Flickr: http://www.flickr.com/photos/davidmillier/ (very old!)

DMillier said:

On the site, the A7riv has about +1/2 stop more dynamic range than the 50MP GFX sensor used in 35mm crop mode.

https://www.photonstophotos.net/Charts/PDR.htm#FujiFilm GFX 50S(FF),Sony ILCE-7RM4

Q1: Does half a stop DR improvement accurately reflect the difference between the old and new underlying sensor generations?

Click to expand...

Note that the graph represents PDR, a variant of DR measurements. A 1/2 stop difference may or may not reflect in improved IQ. The difference may not be representative of the difference in sensor generation. Here, you have four cameras, all with the "same" sensor, all with slightly different PDR:

DavidMillier said:

On the same site, switching to full 44x33 mode improves the 50MP sensor to + 1/3 stop advantage over the newer 61MP sensor, so the difference from the sensor size increase is about +3/4 of a stop.

https://www.photonstophotos.net/Charts/PDR.htm#FujiFilm GFX 50S,Sony ILCE-7RM4

The figs for the Sony vs 100MP GFX (same sensor tech) in 35mm crop mode are essentially identical:

https://www.photonstophotos.net/Charts/PDR.htm#FujiFilm GFX 100S(FF),Sony ILCE-7RM4

The advantage that comes purely from increasing sensor size to 44x33 is just under +2/3 of a stop.

https://www.photonstophotos.net/Charts/PDR.htm#FujiFilm GFX 100S,Sony ILCE-7RM4

To summarise, roughly speaking there is about 1/2 stop difference in DR between the old and new sensor tech, and about +2/3 to 3/4 stop attributable to the larger sensor in both generations.

Have I got that right?

What is missing from this is the detail improvement from the extra pixels in the 100MP sensor, of course. You can enlarge further without losing quality.

Click to expand...

The biggest improvement in the 100MP sensor's DR is the inclusion of dual conversion gain.

DavidMillier said:

Q2: In terms of FF to MF comparisons, is there a simple way of writing down the improvements in noise, DR and extra enlargeability as a percentage for each?

eg

DR: + x%

Reduced noise: x%

Extra enlargeability: + x%

Basically, some simple numbers we can memorise and quote whenever the question arises to save us having to repeat the research and write another 150 posts. At the moment all I have is a handwavy "MF is a bit better than FF".

It would nice to have a convenient number "MF is 10% better than FF" or whatever the number is

Click to expand...

Saying camera A is x% better than camera B has no meaning, except in the forum's flame wars.

DavidMillier said:

Click to expand...

SrMi said:

SrMi said:

Q2: In terms of FF to MF comparisons, is there a simple way of writing down the improvements in noise, DR and extra enlargeability as a percentage for each?

eg

DR: + x%

Reduced noise: x%

Extra enlargeability: + x%

Basically, some simple numbers we can memorise and quote whenever the question arises to save us having to repeat the research and write another 150 posts. At the moment all I have is a handwavy "MF is a bit better than FF".

It would nice to have a convenient number "MF is 10% better than FF" or whatever the number is

Click to expand...

Saying camera A is x% better than camera B has no meaning, except in the forum's flame wars.

SrMi said:

Click to expand...

Click to expand...

No, but it would be useful to be able to say something convenient like:

• GFX50 DR is +3% higher than A7Riv @ ISO100
• GFX50 noise is ?% less than A7Riv @ISO100
• GFX50 can be printed ?% larger than A7Riv @ equal image quality

if this information can be gathered reliably.

Then I could look at my GFX50 (or my A7Riv) and say to myself "see, it's good I have you, you are worth it because you can print 1.5" larger prints than that camera over there!"

There you go again.

Now you are trying to go all mad-scientists and start crunching numbers in order to further convince yourself that FF and MF are about the same (which they are not so it is doomed to be a failed effort).

Now you are focusing on DR as a strict number and want some ridiculous percentage to draw a conclusion for yourself about the exact quantifiable differences between the two systems as if DR was some kind of overall quantifier in your search for sensor size happiness.

You could just go shoot it and look... Or just let me and I'll tell you the result. I just did it again today! I was out testing all afternoon in fact.

Go shoot some GFX images with your GFX 50sII.

Then do post processing and you will immediately feel the difference in latitude. So many have reported that immediate impact - even guys who still prefer FF for various practical reasons.

You will feel the latitude differences between MF and FF. It doesn't matter what DR number or percentage you try to slap on it.

Let me clear it up for you David. The image fidelity of our GFX images is better than any FF system can produce.

If that does something for you like it does for me, then buy it and shoot it.

If it doesn't then don't. Stepping down to FF is a viable alternative, given that 99.9999 Percent of humanity is shooting daily with a sensor the size of half a peanut (the nut, not the shell).

This morning you were telling me that the reason you like your GFX system was for night shots and low light capability. Now on the same day you are trying to come up with some math that says otherwise.

Instead of nit-picking all the reasons you say FF = (or very nearly =) MF, which I think you are way off on as a conclusion, tell us the 5 reasons you think we spend big bucks on GFX and what it is that we see that we like.

That is your homework assignment for this evening.

I'll grade it on the train to Venice tomorrow.

Give it your best effort and don't turn it in late.

David, if you mess up this assignment and don't give it a good effort, there guys here are not going to be happy. If you are really stumped and just can't get started, I can help you by listing the 7 reasons I love shooting GFX and why I think it is superior image fidelity by a good margin, and then I'll tell you why my 60MP Q3 FF is my favorite camera of all time and why I enjoy shooting it more than anything else. This will motivate you.

If you want to know how much better the GFX Image Fidelity is than any FF system, I have already proven that it is 37.675 percent better.

Your search for a percentage that is closer will only lead to further frustration and stimulate ridicule here on the MF Board, even as you are applauded by the boys over on the FF Sony Forum as they search for self-content as they struggle with their inferiority complex.

Remember the formula:

MF > FF.

Or as Fuji likes to say.... MF.... More than FF.

OK now enough talk. Do the Homework assignment.

DMillier said:

SrMi said:

DavidMillier said:

Q2: In terms of FF to MF comparisons, is there a simple way of writing down the improvements in noise, DR and extra enlargeability as a percentage for each?

eg

DR: + x%

Reduced noise: x%

Extra enlargeability: + x%

Basically, some simple numbers we can memorise and quote whenever the question arises to save us having to repeat the research and write another 150 posts. At the moment all I have is a handwavy "MF is a bit better than FF".

It would nice to have a convenient number "MF is 10% better than FF" or whatever the number is

Click to expand...

Saying camera A is x% better than camera B has no meaning, except in the forum's flame wars.

DavidMillier said:

Click to expand...

Click to expand...

No, but it would be useful to be able to say something convenient like:

• GFX50 DR is +3% higher than A7Riv @ ISO100

Click to expand...

Which DR? EDR, PDR, DxO’s DR?

DavidMillier said:

• GFX50 noise is ?% less than A7Riv @ISO100

Click to expand...

Which noise? Mid-tone, deep shadow?

DavidMillier said:

• GFX50 can be printed ?% larger than A7Riv @ equal image quality

Click to expand...

Doesn’t max print size depend on resolution alone?

DavidMillier said:

if this information can be gathered reliably.

Then I could look at my GFX50 (or my A7Riv) and say to myself "see, it's good I have you, you are worth it because you can print 1.5" larger prints than that camera over there!"

Click to expand...

To me, the numbers are the least important elements when deciding which camera to use, except maybe the resolution numbers.

DavidMillier said:

--
Photo of the day: https://whisperingcat.co.uk/wp/photo-of-the-day/
Website: http://www.whisperingcat.co.uk/ (2022 - website rebuilt, updated and back in action)
DPReview gallery: https://www.dpreview.com/galleries/0286305481
Flickr: http://www.flickr.com/photos/davidmillier/ (very old!)

Click to expand...

Greg7579 said:

There you go again.

Now you are trying to go all mad-scientists and start crunching numbers in order to further convince yourself that FF and MF are about the same (which they are not so it is doomed to be a failed effort).

Now you are focusing on DR as a strict number and want some ridiculous percentage to draw a conclusion for yourself about the exact quantifiable differences between the two systems as if DR was some kind of overall quantifier in your search for sensor size happiness.

You could just go shoot it and look... Or just let me and I'll tell you the result. I just did it again today! I was out testing all afternoon in fact.

Go shoot some GFX images with your GFX 50sII.

Then do post processing and you will immediately feel the difference in latitude. So many have reported that immediate impact - even guys who still prefer FF for various practical reasons.

You will feel the latitude differences between MF and FF. It doesn't matter what DR number or percentage you try to slap on it.

Let me clear it up for you David. The image fidelity of our GFX images is better than any FF system can produce.

If that does something for you like it does for me, then buy it and shoot it.

If it doesn't then don't. Stepping down to FF is a viable alternative, given that 99.9999 Percent of humanity is shooting daily with a sensor the size of half a peanut (the nut, not the shell).

This morning you were telling me that the reason you like your GFX system was for night shots and low light capability. Now on the same day you are trying to come up with some math that says otherwise.

Instead of nit-picking all the reasons you say FF = (or very nearly =) MF, which I think you are way off on as a conclusion, tell us the 5 reasons you think we spend big bucks on GFX and what it is that we see that we like.

That is your homework assignment for this evening.

I'll grade it on the train to Venice tomorrow.

Give it your best effort and don't turn it in late.

David, if you mess up this assignment and don't give it a good effort, there guys here are not going to be happy. If you are really stumped and just can't get started, I can help you by listing the 7 reasons I love shooting GFX and why I think it is superior image fidelity by a good margin, and then I'll tell you why my 60MP Q3 FF is my favorite camera of all time and why I enjoy shooting it more than anything else. This will motivate you.

If you want to know how much better the GFX Image Fidelity is than any FF system, I have already proven that it is 37.675 percent better.

Your search for a percentage that is closer will only lead to further frustration and stimulate ridicule here on the MF Board, even as you are applauded by the boys over on the FF Sony Forum as they search for self-content as they struggle with their inferiority complex.

Remember the formula:

MF > FF.

Or as Fuji likes to say.... MF.... More than FF.

Click to expand...

I wonder why they never say: MF … More than APS-C , or APS-C … less than FF?

Greg7579 said:

OK now enough talk. Do the Homework assignment.

--
Greg Johnson, San Antonio, Texas
https://www.flickr.com/photos/139148982@N02/albums

Click to expand...

Greg7579 said:

There you go again.

Now you are trying to go all mad-scientists and start crunching numbers in order to further convince yourself that FF and MF are about the same (which they are not so it is doomed to be a failed effort).

Now you are focusing on DR as a strict number and want some ridiculous percentage to draw a conclusion for yourself about the exact quantifiable differences between the two systems as if DR was some kind of overall quantifier in your search for sensor size happiness.

You could just go shoot it and look... Or just let me and I'll tell you the result. I just did it again today! I was out testing all afternoon in fact.

Go shoot some GFX images with your GFX 50sII.

Then do post processing and you will immediately feel the difference in latitude. So many have reported that immediate impact - even guys who still prefer FF for various practical reasons.

You will feel the latitude differences between MF and FF. It doesn't matter what DR number or percentage you try to slap on it.

Let me clear it up for you David. The image fidelity of our GFX images is better than any FF system can produce.

If that does something for you like it does for me, then buy it and shoot it.

If it doesn't then don't. Stepping down to FF is a viable alternative, given that 99.9999 Percent of humanity is shooting daily with a sensor the size of half a peanut (the nut, not the shell).

This morning you were telling me that the reason you like your GFX system was for night shots and low light capability. Now on the same day you are trying to come up with some math that says otherwise.

Instead of nit-picking all the reasons you say FF = (or very nearly =) MF, which I think you are way off on as a conclusion, tell us the 5 reasons you think we spend big bucks on GFX and what it is that we see that we like.

That is your homework assignment for this evening.

I'll grade it on the train to Venice tomorrow.

Give it your best effort and don't turn it in late.

David, if you mess up this assignment and don't give it a good effort, there guys here are not going to be happy. If you are really stumped and just can't get started, I can help you by listing the 7 reasons I love shooting GFX and why I think it is superior image fidelity by a good margin, and then I'll tell you why my 60MP Q3 FF is my favorite camera of all time and why I enjoy shooting it more than anything else. This will motivate you.

If you want to know how much better the GFX Image Fidelity is than any FF system, I have already proven that it is 37.675 percent better.

Your search for a percentage that is closer will only lead to further frustration and stimulate ridicule here on the MF Board, even as you are applauded by the boys over on the FF Sony Forum as they search for self-content as they struggle with their inferiority complex.

Remember the formula:

MF > FF.

Or as Fuji likes to say.... MF.... More than FF.

OK now enough talk. Do the Homework assignment.

Click to expand...

Why, thank you, Greg.

A small reminder, I already bought my 50s and my A7Riv, so no decision making required.

I haven't yet got around to doing those direct comparisons you urge, but I will at some point. I'll be comparing the 50s to the Sony at 4:3 and 1:1 to make the sensors a bit more directly comparable. Same aspect ratios and very similar pixel counts. It will be a shoot out between old vs new sensor generations and 24x32 vs 33x44 areas. But I know I lack the testing skills to reliably detect small differences, and for the same reasons, field results will likely show no obvious differences. And even if there are differences, they'll be too small to be visible in 12" square prints, so it is all a bit moot for me, TBH. But still good to try.

I found reviews on Imaging Resource that said that both cameras made excellent 40"x30" prints, the max size they test. Which is nice to know, but still not measuring the difference between them, unfortunately.

I'd use my 50s as my main camera if it wasn't so big and heavy. The 35-70 is not too bad, comparable to a FF lens, but the body with the P645 primes and adaptors is too much. I'm such a feeble wimp with a bad back and bad shoulder tendons, unfortunately. So it is my LE tripod mounted option instead.

I'd still like those simple measured numbers, though. I'm sure with all the testing that's been done around here, it's just a case of recasting the numbers into the right format...

--
Photo of the day: https://whisperingcat.co.uk/wp/photo-of-the-day/
Website: http://www.whisperingcat.co.uk/ (2022 - website rebuilt, updated and back in action)
DPReview gallery: https://www.dpreview.com/galleries/0286305481
Flickr: http://www.flickr.com/photos/davidmillier/ (very old!)

Any DR measurement scheme is fine as long as everything is measured using the same scheme. Surely, they're proportionate, even if not calibrated the same?

I've read print comparisons before where pixel count/resolution wasn't the only determinate of print size limits. Miles Hecker did a juried print quality comparison years ago and they concluded that the FF had the edge in large prints even thought the APS-C had more pixels.

Sensor size, even at the same pixel count seems to makes a difference, I can guess reasons why.

You have two great high-res cameras.

The alphasevenarefour is great. The GFX 50 is great.

Now go shoot them. You have the prefect tool for all kinds of shooting with that Sony with that fast focusing and great focus tracking.

GFX? You can kill it on scenic shots, night scenes, landscapes and all kinds of fun slow shooting.

You already have way better cameras than almost anyone and you are a fine photographer.

Now go shoot and quit worrying about the percentage difference of DR between the two.

You have some pretty awesome DR with those two cameras.

DMillier said:

Any DR measurement scheme is fine as long as everything is measured using the same scheme. Surely, they're proportionate, even if not calibrated the same?

Click to expand...

P2P does not have calibrated ISOs, so the graph may need to be moved left or right when comparing cameras. This does not affect the max values.

DavidMillier said:

I've read print comparisons before where pixel count/resolution wasn't the only determinate of print size limits. Miles Hecker did a juried print quality comparison years ago and they concluded that the FF had the edge in large prints even thought the APS-C had more pixels.

Sensor size, even at the same pixel count seems to makes a difference, I can guess reasons why.

Click to expand...

Greg7579 said:

You have two great high-res cameras.

The alphasevenarefour is great. The GFX 50 is great.

Now go shoot them. You have the prefect tool for all kinds of shooting with that Sony with that fast focusing and great focus tracking.

GFX? You can kill it on scenic shots, night scenes, landscapes and all kinds of fun slow shooting.

You already have way better cameras than almost anyone and you are a fine photographer.

Now go shoot and quit worrying about the percentage difference of DR between the two.

You have some pretty awesome DR with those two cameras.

Click to expand...

Not worrying at all, Greg, no need for concern. But I haven't been shooting since Feb because I have been stuck at home being a househusband and nurse since my wife had her left knee replaced. She's up and about now, which is great, but I'm still stuck at home revising for my psychology exam until May 23. My batts are charged and ready and I shall be off the leash in a few weeks but until them I'm left posting on DPR forums as a procrastination technique. The course was just an experiment and I learned never to sign up for an exam ever again. In your 60s, revision doesn't work like it used to (cue innuendos...).

Now, let me tell you all about Sapir-Whorf theory... (at least I would if I could remember any of it)!

--
Photo of the day: https://whisperingcat.co.uk/wp/photo-of-the-day/
Website: http://www.whisperingcat.co.uk/ (2022 - website rebuilt, updated and back in action)
DPReview gallery: https://www.dpreview.com/galleries/0286305481
Flickr: http://www.flickr.com/photos/davidmillier/ (very old!)

If modern day MF (33x44) is about the same as FF (36x24) would that not suggest APS-C is about the same as FF (similar proportional difference in size between the two) and raise the question, is MF any better than APS-C?

Greg7579 said:

There you go again.

Now you are trying to go all mad-scientists and start crunching numbers in order to further convince yourself that FF and MF are about the same (which they are not so it is doomed to be a failed effort).

Click to expand...

DMillier said:

Q2: In terms of FF to MF comparisons, is there a simple way of writing down the improvements in noise, DR and extra enlargeability as a percentage for each?

eg

DR: + x%

Reduced noise: x%

Click to expand...

DR and noise are two sides of the same coin. At same print size with the FF image dropped to 3:4 aspect ratio, the improvement in photon-noise-limited signal to noise ratio at full exposure, we can calculate the improvement as follows:

SNR Ratio = (33/24) - 1 = 0.375.

Want that in stops?

log2(1.375) = 0.46 stops.

DavidMillier said:

Extra enlargeability: + x%

Click to expand...

Assuming sensor sampling limited, that boils down to resolution.

Here's a calculation that finesses the aspect ratio, which makes the math simpler, but is not my favorite was to do this comparison:

sqrt(102/61) - 1 = 29% larger print.

DavidMillier said:

Basically, some simple numbers we can memorise and quote whenever the question arises to save us having to repeat the research and write another 150 posts. At the moment all I have is a handwavy "MF is a bit better than FF".

It would nice to have a convenient number "MF is 10% better than FF" or whatever the number is

Click to expand...

The comparison should be a vector, not a scalar.

DMillier said:

Greg7579 said:

You have two great high-res cameras.

The alphasevenarefour is great. The GFX 50 is great.

Now go shoot them. You have the prefect tool for all kinds of shooting with that Sony with that fast focusing and great focus tracking.

GFX? You can kill it on scenic shots, night scenes, landscapes and all kinds of fun slow shooting.

You already have way better cameras than almost anyone and you are a fine photographer.

Now go shoot and quit worrying about the percentage difference of DR between the two.

You have some pretty awesome DR with those two cameras.

Click to expand...

Not worrying at all, Greg, no need for concern. But I haven't been shooting since Feb because I have been stuck at home being a househusband and nurse since my wife had her left knee replaced. She's up and about now, which is great, but I'm still stuck at home revising for my psychology exam until May 23. My batts are charged and ready and I shall be off the leash in a few weeks but until them I'm left posting on DPR forums as a procrastination technique. The course was just an experiment and I learned never to sign up for an exam ever again. In your 60s, revision doesn't work like it used to (cue innuendos...).

Now, let me tell you all about Sapir-Whorf theory... (at least I would if I could remember any of it)!

Click to expand...

Best wishes to the patient. Tell her you are leaving for 2 hours a day to walk and go shoot something that looks terrible and flat, then change it to B&W to disguise the bad light by punching up the contrast like I did today.



View attachment 38113fb05a4345a5899cca28e27cd364.jpg
Genoa Old Port in the haze and rain over-processed because I felt like it....

--
Greg Johnson, San Antonio, Texas

JimKasson said:

JimKasson said:

Q2: In terms of FF to MF comparisons, is there a simple way of writing down the improvements in noise, DR and extra enlargeability as a percentage for each?

eg

DR: + x%

Reduced noise: x%

Click to expand...

DR and noise are two sides of the same coin. At same print size with the FF image dropped to 3:4 aspect ratio, the improvement in photon-noise-limited signal to noise ratio at full exposure, we can calculate the improvement as follows:

SNR Ratio = (33/24) - 1 = 0.375.

Want that in stops?

log2(1.375) = 0.46 stops.

Click to expand...

About 97% of the time I'd defer to you on this, but I'm coming up with somewhat different numbers--and what I'm coming up with now agrees with what I think I remember being the agreed values previously. By my math, comparing the GFX-size sensor to a theoretical FF sensor*--and assuming that all else is equal, which often it isn't--I get:

● for final output at a 4:3 aspect ratio or more toward square, the GFX sensor has a 0.91-stop advantage;

● for final output at an intermediate / format-neutral aspect ratio similar to the ISO paper sizes like A2, the GFX sensor has a 0.68-stop advantage; and

● for final output at a 3:2 aspect ratio or more toward elongated (e.g. 2:1 panorama), the GFX sensor has a 0.57-stop advantage.

*Theoretically FF is 36x24mm, but the specs on most of them are a hair under that, and they vary by a hair. The GFX sensors all seem to be 43.8x32.9mm.

NAwlins Contrarian said:

NAwlins Contrarian said:

NAwlins Contrarian said:

Q2: In terms of FF to MF comparisons, is there a simple way of writing down the improvements in noise, DR and extra enlargeability as a percentage for each?

eg

DR: + x%

Reduced noise: x%

Click to expand...

DR and noise are two sides of the same coin. At same print size with the FF image dropped to 3:4 aspect ratio, the improvement in photon-noise-limited signal to noise ratio at full exposure, we can calculate the improvement as follows:

SNR Ratio = (33/24) - 1 = 0.375.

Want that in stops?

log2(1.375) = 0.46 stops.

Click to expand...

About 97% of the time I'd defer to you on this, but I'm coming up with somewhat different numbers--and what I'm coming up with now agrees with what I think I remember being the agreed values previously. By my math, comparing the GFX-size sensor to a theoretical FF sensor*--and assuming that all else is equal, which often it isn't--I get:

* for final output at a 4:3 aspect ratio or more toward square, the GFX sensor has a 0.91-stop advantage;

Click to expand...

Photon noise SNR is proportional to the square root of the signal level. Does that help?

NAwlins Contrarian said:

* for final output at an intermediate / format-neutral aspect ratio similar to the ISO paper sizes like A2, the GFX sensor has a 0.68-stop advantage; and

* for final output at a 3:2 aspect ratio or more toward elongated (e.g. 2:1 panorama), the GFX sensor has a 0.57-stop advantage.

*Theoretically FF is 36x24mm, but the specs on most of them are a hair under that, and they vary by a hair. The GFX sensors all seem to be 43.8x32.9mm.

Click to expand...

JimKasson said:

JimKasson said:

JimKasson said:

Want that in stops?

log2(1.375) = 0.46 stops.

Click to expand...

About 97% of the time I'd defer to you on this, but I'm coming up with somewhat different numbers--and what I'm coming up with now agrees with what I think I remember being the agreed values previously. By my math, comparing the GFX-size sensor to a theoretical FF sensor*--and assuming that all else is equal, which often it isn't--I get:

* for final output at a 4:3 aspect ratio or more toward square, the GFX sensor has a 0.91-stop advantage;

Click to expand...

Photon noise SNR is proportional to the square root of the signal level. Does that help?

Click to expand...

Thanks for the quick response. I'm going to admit being out toward the edges of at least my comfort zone and maybe my understanding. I basically get the noise relationship issue and the noise - square root point. I may well have (okay, apparently did) make an unwarranted assumption. Let me put it this way:

Is it correct to say that for final output at a 4:3 aspect ratio or more toward square, compared with FF, the GFX requires stopping down an additional 0.91 stop to get the same depth of field, but the GFX sensor has (only) half that advantage, 0.46 stop, in noise and dynamic range?

NAwlins Contrarian said:

NAwlins Contrarian said:

NAwlins Contrarian said:

NAwlins Contrarian said:

Want that in stops?

log2(1.375) = 0.46 stops.

Click to expand...

About 97% of the time I'd defer to you on this, but I'm coming up with somewhat different numbers--and what I'm coming up with now agrees with what I think I remember being the agreed values previously. By my math, comparing the GFX-size sensor to a theoretical FF sensor*--and assuming that all else is equal, which often it isn't--I get:

* for final output at a 4:3 aspect ratio or more toward square, the GFX sensor has a 0.91-stop advantage;

Click to expand...

Photon noise SNR is proportional to the square root of the signal level. Does that help?

Click to expand...

Thanks for the quick response. I'm going to admit being out toward the edges of at least my comfort zone and maybe my understanding. I basically get the noise relationship issue and the noise - square root point. I may well have (okay, apparently did) make an unwarranted assumption. Let me put it this way:

Is it correct to say that for final output at a 4:3 aspect ratio or more toward square, compared with FF, the GFX requires stopping down an additional 0.91 stop to get the same depth of field,

Click to expand...

Yes.

NAwlins Contrarian said:

but the GFX sensor has (only) half that advantage, 0.46 stop, in noise and dynamic range?

Click to expand...

If you're photon noise limited, yes.

The Sapir-Whorf hypothesis takes me back to the Hardy-Weinberg principle. I was sitting for a 3 question essay test to get my science teaching credential. First question was easy. Last question was doable. Second question? Hardy-Weinberg. Never heard of it. Absolutely no clue in the question about what it might relate to. Went back to #1 and #3 to make them as perfect as possible, as anything less than 60% would be a fail, and the best I could hope for (with no answer to #2) was 66%. Somehow, I passed.

JimKasson said:

JimKasson said:

JimKasson said:

Photon noise SNR is proportional to the square root of the signal level. Does that help?

Click to expand...

Thanks for the quick response. I'm going to admit being out toward the edges of at least my comfort zone and maybe my understanding. I basically get the noise relationship issue and the noise - square root point. I may well have (okay, apparently did) make an unwarranted assumption. Let me put it this way:

Is it correct to say that for final output at a 4:3 aspect ratio or more toward square, compared with FF, the GFX requires stopping down an additional 0.91 stop to get the same depth of field,

Click to expand...

Yes.

JimKasson said:

but the GFX sensor has (only) half that advantage, 0.46 stop, in noise and dynamic range?

Click to expand...

If you're photon noise limited, yes.

Click to expand...

Thanks--and right, and that is an important limitation / qualification: these figures only apply where there's plenty of light, i.e., you can shoot at base or optimal gain (ISO 100 or whatever) and use a shutter speed and aperture such that you get exposure very slightly less than what would cause clipping somewhere other than in specular highlights. That is often true outdoors on a sunny day, in a studio with powerful strobes, and with a totally-static subject and a camera on a sturdy tripod. Except under plenty-of-light circumstances, the effective noise and dynamic range advantage is often less than 0.46-stop, or even nothing--I think!