FX (Crop) VS DX

[Braineack]

I dunno? I'm assuming it's indentical. it's a crop...

 

[Ysarex]

I dunno? I'm assuming it's indentical. it's a crop...

You mean like what I originally said: "The fundamental reason an FX sensor provides better low-light performance is it's bigger. If you crop it so that it's no longer bigger then you're giving up the reason for the advantage."

Joe

 

[JBPhotog]

In the 40+ years I have been shooting, much of it professionally, I have never come across a light meter that has a setting for different sized film or sensors.

Modern sensors are made up of clusters of photosites, photons do not spill over, they are either captured or not.

Think of it like pizza, two pies with the same ingredients, one 8” the other 12”, does the bigger one taste better!

 

[Derrel]

Looking over the results at the site that Joe linked us to, I see that virtually every camera does worse in DX mode than it does when it is used in FX mode. Not by a large degree but by a measurable degree. In the past we had people who compared the Nikon D800 to the Nikon D7200 and in every case the D800 yielded slightly better results when its full frame images were cropped down to the DX equivalent image field of the Nikon D7200

Those comparisons were quite frequent about 6 years ago. The D800 was introduced in 2012, if I remember correctly. It is a particularly good sensor. Last summer I owned both the D800 and the Nikon D610 and in my non-scientific testing both cameras offered roughly equivalent dynamic range, but the D800 with its 36 million pixels offered incredible crop-in capability.

Instead of theory, I would urge you to actually rent a camera and conduct your own tests. I was at first not sold on the 36 million pixels of the D800. 24 million pixels is quite good but 36 million pixels offers incredible crop- in ability, as I said earlier. The difference is hard to put into words. In the field of portrait photography I found that it is possible at ISO 100 to crop from a half-body seated pose and into a Head and Shoulder portrait with no significant loss final image quality, even when using a relatively inexpensive lens. The idea that the D800 requires the highest level of glass is in my opinion, over-stated.

 

[Ysarex]

In the 40+ years I have been shooting, much of it professionally, I have never come across a light meter that has a setting for different sized film or sensors.

What has that to do with it? Did anyone say anything about different exposures? Larger sensors as a rule perform better in low light that smaller sensors. By that we mean less noise and more DR when used at high ISO values (low light). That's what I said. I'm correct about that.

What's that noise? Part one: Shedding some light on the sources of noise

Joe

 

[Ysarex]

In the 40+ years I have been shooting, much of it professionally, I have never come across a light meter that has a setting for different sized film or sensors.

Modern sensors are made up of clusters of photosites, photons do not spill over, they are either captured or not.

Think of it like pizza, two pies with the same ingredients, one 8” the other 12”, does the bigger one taste better!

Let's have a look then.

I happen to have a Nikon camera. So I took two photos with the camera in a low light condition.

FX at ISO 6400, 1/13th sec, f/8.

Same lens, I backed up the tripod so the scene was framed the same side to side.

DX at ISO 6400, 1/13th sec, f/8.

Both raw files default open in C1 with Luminance noise filtering off. Here's the central section of the two side by side at 100% for the DX camera.

Does the bigger one taste better? It sure does. Why? Larger sensor = less noise or if you prefer, a crop from a larger sensor = more noise.

Joe

 

[Derrel]

I did not know you have a Z7@@@!

 

[Ysarex]

I did not know you have a Z7@@@!

Long story, but yes.

Joe

 

[Designer]

I am just wondering in general if it would be better to get 2 cameras (a DX and FX around 20-24mp) or just 1 high mp fx camera.

You are getting caught up in the "megapixel" argument, which is useless as it goes.

There are many side issues involved which all contribute to the final image quality. I think you will not see two different size sensors produce the exact same image, as you seem to assume.

Even if, by some strange coincidence, both sensors had the same pixel pitch, there would still be differences. If the FX was manufactured by a different company using a different design, and at a different level of technology than the DX sensor, then simply using that FX sensor in "cropped mode" would probably not produce the same image as the DX sensor.

As to the "high mp fx", keep in mind that the file size is huge! You will need a fast computer with lots of RAM to do your editing.

Make your decision based on your intended use, and get the system you need, and ignore the "mega-pixels".

 

[PJM]

In the 40+ years I have been shooting, much of it professionally, I have never come across a light meter that has a setting for different sized film or sensors.

Modern sensors are made up of clusters of photosites, photons do not spill over, they are either captured or not.

Think of it like pizza, two pies with the same ingredients, one 8” the other 12”, does the bigger one taste better!

Let's have a look then.

I happen to have a Nikon camera. So I took two photos with the camera in a low light condition.

FX at ISO 6400, 1/13th sec, f/8.

Same lens, I backed up the tripod so the scene was framed the same side to side.

DX at ISO 6400, 1/13th sec, f/8.

Both raw files default open in C1 with Luminance noise filtering off. Here's the central section of the two side by side at 100% for the DX camera.

Does the bigger one taste better? It sure does. Why? Larger sensor = less noise or if you prefer, a crop from a larger sensor = more noise.

Joe

View attachment 190085

I would do your experiment differently. Starting with the image framed the way you want it in DX mode, take both photos, FX and DX modes, without moving the camera. I would then crop the FX down to DX size in post and then compare the photos. When you move the camera you are changing how the light enters and falls on the sensor so I don't think you can make an apples to apples comparison.

 

[Ysarex]

In the 40+ years I have been shooting, much of it professionally, I have never come across a light meter that has a setting for different sized film or sensors.

Modern sensors are made up of clusters of photosites, photons do not spill over, they are either captured or not.

Think of it like pizza, two pies with the same ingredients, one 8” the other 12”, does the bigger one taste better!

Let's have a look then.

I happen to have a Nikon camera. So I took two photos with the camera in a low light condition.

FX at ISO 6400, 1/13th sec, f/8.

Same lens, I backed up the tripod so the scene was framed the same side to side.

DX at ISO 6400, 1/13th sec, f/8.

Both raw files default open in C1 with Luminance noise filtering off. Here's the central section of the two side by side at 100% for the DX camera.

Does the bigger one taste better? It sure does. Why? Larger sensor = less noise or if you prefer, a crop from a larger sensor = more noise.

Joe

View attachment 190085

I would do your experiment differently. Starting with the image framed the way you want it in DX mode, take both photos, FX and DX modes, without moving the camera. I would then crop the FX down to DX size in post and then compare the photos. When you move the camera you are changing how the light enters and falls on the sensor so I don't think you can make an apples to apples comparison.

I made the only comparison that makes sense. What you're suggesting is to compare X with X -- the crop with the crop. I'll bet X will be the same as X.

What I did was use the two different formats to take the same photo. That's the logical comparison. We want to see if there's a performance difference in real usage. If you wanted to avoid moving the camera then it would be necessary to change lenses between the two photos, but the comparison should be: is there a difference when the two different formats are used to take the same photo.

Joe

 

[PJM]

I also expect the two to be the same. So, maybe I didn't understand the question. I thought the question was would the image from an FX format cropped down to DX size be better than one from a DX format. In order to do that I believe you have to take both images from the same spot. My understanding is that assuming you used the same lens the target image would be the same size on each physical sensor. The FX format would include more area around the target. When you crop the FX down to DX they both show the same image area. Then you can compare performance. I think you can only do this with two different cameras, an FX and a DX. Using the the FX and the same FX in DX mode produce the equivalent result.

 

[Derrel]

My gut feeling is that the FX image when cropped down to DX will be slightly better than when the FX camera is used in DX mode....but there is only one way to find out...an actual test. LOOKING at the Photostophotons test result listing, it shows that image quality from a dynamic range standpoint declines when an FX Camera is shot at a smaller capture size, the example listed being DX (or aps-c).

 

[Ysarex]

I also expect the two to be the same. So, maybe I didn't understand the question. I thought the question was would the image from an FX format cropped down to DX size be better than one from a DX format. In order to do that I believe you have to take both images from the same spot. My understanding is that assuming you used the same lens the target image would be the same size on each physical sensor. The FX format would include more area around the target. When you crop the FX down to DX they both show the same image area. Then you can compare performance. I think you can only do this with two different cameras, an FX and a DX. Using the the FX and the same FX in DX mode produce the equivalent result.

Got it. You're thinking of the OP's original question. I answered that and no, cropping the FX sensor basically removes the low light advantage. In my post with the photos I was responding to JBPhotog who disagreed with my statement: "The fundamental reason an FX sensor provides better low-light performance is it's bigger. If you crop it so that it's no longer bigger then you're giving up the reason for the advantage."

Joe

 

[Derrel]

I think the newest, best-performing full frame cameras perform so well that they would be more than adequate with the images cropped down to DX, either in-camera, or in post.

The Nikon D850 is amazing; the Nikon Z7 and Z6 are amazing, the Sony A7 RIII is amazing. The Sony A9 is amazing.