Why does cropping reduce dynamic range? (Page 4): Good.

Ysarex Loc: St. Louis

bleirer wrote:

Thanks for all the replies. I'm satisfied with the answers given.

Good.

selmslie Loc: Fernandina Beach, FL, USA

Ysarex wrote:

I did not make an original error. This statement is correct and as I said it's common knowledge: "Larger sensors are less noisy than smaller sensors because larger sensors record more total light than smaller sensors. If you crop and only use the center portion of a sensor you're using less total light and so increasing noise and so decreasing DR." ....

The OP's question was not about two different sized sensors. It was about the same sensor with two different crops.

You didn't understand the question. You are doubling down because you went off on a tangent. You don't want to admit you were wrong.

Ysarex Loc: St. Louis

selmslie wrote:

The OP's question was not about two different sized sensors. It was about the same sensor with two different crops.

Same principle applies whether it's two different sensors or two sections of one sensor and it is even more clear that what I said is correct because the pixels are of course the same and so there's no complicating other issue. Read the article I linked that substantiates what I said: https://www.dpreview.com/articles/5365920428/the-effect-of-pixel-and-sensor-sizes-on-noise/2

Again this is common knowledge. I've provided the reference that proves I'm right.

selmslie Loc: Fernandina Beach, FL, USA

Ysarex wrote:

Same principle applies whether it's two different sensors or two sections of one sensor and it is even more clear that what I said is correct because the pixels are of course the same and so there's no complicating other issue. Read the article I linked that substantiates what I said: https://www.dpreview.com/articles/5365920428/the-effect-of-pixel-and-sensor-sizes-on-noise/2

Again this is common knowledge. I've provided the reference that proves I'm right.

No, it's not the same principle.

You either didn't read the original post, didn't understand it or simply ignored it so you could talk about something else that you wanted to talk about.

bleirer wrote:

When I go to the dynamic range section of the photonstophotos website and plot any full frame camera in full frame mode next to the same camera in crop mode, the dynamic range of the crop mode is always less by the same amount at every point. ...

So why does cropping alone reduce dynamic range?

Bill Claff gave the ultimate answer https://www.uglyhedgehog.com/t-687091-3.html#12031170 which you either missed or ignored. It's his web site. He should know.

Several of us gave pretty much the same answer.

You are still trying to defend your statements which were off topic. Man up and admit it.

Ysarex Loc: St. Louis

selmslie wrote:

No, it's not the same principle.

Yes it is the same principle. Prove how it isn't. I've provided appropriate references that substantiate what I said.

selmslie Loc: Fernandina Beach, FL, USA

Ysarex wrote:

Yes it is the same principle. Prove how it isn't. I've provided appropriate references that substantiate what I said.

But not relevant to this simple topic.

Ysarex Loc: St. Louis

selmslie wrote:

But not relevant to this simple topic.

OK, so you were wrong about that -- obviously. And you're wrong about this. It most certainly is relevant. It answers the OP's question. Switching an FX sensor to DX results in a DR drop. That's because DR is noise limited and the smaller SX sensor is noisier. It's noisier because it collects less total light. That's what I said. It is common knowledge and I provided references.

bclaff Loc: Sherborn, MA (18mi SW of Boston)

bleirer wrote:

That makes sense, so as Ysarex said, if one just took a crop out of a full frame sensor image without resizing there would be no impact?

Of course there would, whether you crop in camera or later the effect is the same, you still have to enlarge more to make the same image size.

bleirer

So thinking of some real world cropping scenarios, let's say I have a 4000x6000 sensor:

1. If I only need a final image of 1000x1500 I could fill the frame and downsample or not fill the frame and use a cropped portion of the image. More dynamic range with fill the frame?

2. If I need 4000x6000 I could fill the frame and use all pixels or not fill the frame and resize the crop up to 4000x6000. More dynamic range with fill the frame?

Ysarex Loc: St. Louis

bleirer wrote:

So thinking of some real world cropping scenarios, let's say I have a 4000x6000 sensor:

1. If I only need a final image of 1000x1500 I could fill the frame and downsample or not fill the frame and use a cropped portion of the image. More dynamic range with fill the frame?

2. If I need 4000x6000 I could fill the frame and use all pixels or not fill the frame and resize the crop up to 4000x6000. More dynamic range with fill the frame?

More DR with fill the frame. Or better, say less noise with fill the frame and so more DR since noise limits DR.

All of these comparisons rest on a foundation of "take the same photo," and view the same photo the same way at the same size. That's necessary to establish an even playing field.

bleirer

bclaff wrote:

Of course there would, whether you crop in camera or later the effect is the same, you still have to enlarge more to make the same image size.

I mean if I only needed a final image of say 1000x1500 out of a 4000x6000 sensor. Just taking that slice out and using it as is wouldn't reduce dynamic range would it?

selmslie Loc: Fernandina Beach, FL, USA

bleirer wrote:

I mean if I only needed a final image of say 1000x1500 out of a 4000x6000 sensor. Just taking that slice out and using it as is wouldn't reduce dynamic range would it?

Yes, it would. That's because the smaller crop would need 4x as much enlargement to make the same result. That would make the noise easier to see.

selmslie Loc: Fernandina Beach, FL, USA

Ysarex wrote:

OK, so you were wrong about that -- obviously. And you're wrong about this. It most certainly is relevant. It answers the OP's question. Switching an FX sensor to DX results in a DR drop. That's because DR is noise limited and the smaller SX sensor is noisier. It's noisier because it collects less total light. That's what I said. It is common knowledge and I provided references.

It was a simple question. It called for a simple answer.

If you understood the question you would have been able to supply that simple answer as Biull Claff and I have.

Since you didn't understand the issue you chose to hijack the thread to talk about other issues. That's where you went wrong.

Ysarex Loc: St. Louis

selmslie wrote:

It was a simple question. It called for a simple answer.

And I provided a simple and correct answer. DR is noise limited. Larger sensors are less noisy than smaller sensors primarily because of the difference in total light. That is simple and that is correct. In what specific way is that not correct and please provide references.

larryepage Loc: North Texas area

bclaff wrote:

Photographic Dynamic Range (PDR) is normalized to a standard image size and viewing conditions.
The DX (APS-C) capture must be enlarge more than the FX (FF) one.
For some details see DX Crop Mode Photographic Dynamic Range

First off...thank you for all the work you have done and continue to do on the Photons to Photos site. And thanks for clarifying here, even if I might handle the dynamic range part of the equation differently. But...I'm still a little confused that the D500 shows up better than the D850 in crop format. Even though my experience has shown that images from the two cameras are pretty much indistinguishable in real life, I still would expect the D850 to come out at least a little bit ahead in a test environment given equal size sensors...