Question about in camera image size reduction

[crimbfighter]

Does anyone know exactly how DSLRs capture the image data when you choose a reduced image size? If, for example, I select the smaller image size (ie selecting the 24mp option) in my D800, does the camera simply ignore or turn off some of the in between pixels, or does it capture the full 36mp image, then eliminate pixels using the camera processor?

I'm curious because when I am using my telescope for astrophotography, there is a phenomenon known as over or under sampling, meaning the resolution of the camera is either too much, or too little to properly resolve the image. Over sampling causes the image to he soft, and under sampling causes the image to be pixelated in the fine detail. The pixel pitch of my D800 causes over sampling, and I want to try using the a smaller image size (more like the 24mp option instead of the 36mp) to simulate a camera with the proper pixel pitch. I think if it effectively turns off pixels to do this, it would work, but if it does it by eliminating pixels after the fact, it may not.

 

[Ysarex]

Does anyone know exactly how DSLRs capture the image data when you choose a reduced image size? If, for example, I select the smaller image size (ie selecting the 24mp option) in my D800, does the camera simply ignore or turn off some of the in between pixels, or does it capture the full 36mp image, then eliminate pixels using the camera processor?

I'm curious because when I am using my telescope for astrophotography, there is a phenomenon known as over or under sampling, meaning the resolution of the camera is either too much, or too little to properly resolve the image. Over sampling causes the image to he soft, and under sampling causes the image to be pixelated in the fine detail. The pixel pitch of my D800 causes over sampling, and I want to try using the a smaller image size (more like the 24mp option instead of the 36mp) to simulate a camera with the proper pixel pitch. I think if it effectively turns off pixels to do this, it would work, but if it does it by eliminating pixels after the fact, it may not.

When you switch the D800 between FX and DX the pixel pitch stays the same -- you just use a smaller area of the sensor.

 

[crimbfighter]

Does anyone know exactly how DSLRs capture the image data when you choose a reduced image size? If, for example, I select the smaller image size (ie selecting the 24mp option) in my D800, does the camera simply ignore or turn off some of the in between pixels, or does it capture the full 36mp image, then eliminate pixels using the camera processor?

I'm curious because when I am using my telescope for astrophotography, there is a phenomenon known as over or under sampling, meaning the resolution of the camera is either too much, or too little to properly resolve the image. Over sampling causes the image to he soft, and under sampling causes the image to be pixelated in the fine detail. The pixel pitch of my D800 causes over sampling, and I want to try using the a smaller image size (more like the 24mp option instead of the 36mp) to simulate a camera with the proper pixel pitch. I think if it effectively turns off pixels to do this, it would work, but if it does it by eliminating pixels after the fact, it may not.

When you switch the D800 between FX and DX the pixel pitch stays the same -- you just use a smaller area of the sensor.

I'm not referring to FX vs DX. There is an option to change the image size, in pixels, while maintaining the FX format. Image area (FX vs DX) and image size are different. For example, mlthe D800, under FX image are, can record large, medium, or small images, being 7,360 x 4,912 pixels, 5,520 x 3,680 pixels, and 3,680 x 2,456 pixels respectively, all while still being FX format.

 

[crimbfighter]

These are all the options to choose from in the various area modes. I'm basically wondering how it goes about recording the image if I select, say the medium image size as opposed to the large. Does it turn off or ignore the unneeded pixels?

 

[Ysarex]

Does anyone know exactly how DSLRs capture the image data when you choose a reduced image size? If, for example, I select the smaller image size (ie selecting the 24mp option) in my D800, does the camera simply ignore or turn off some of the in between pixels, or does it capture the full 36mp image, then eliminate pixels using the camera processor?

I'm curious because when I am using my telescope for astrophotography, there is a phenomenon known as over or under sampling, meaning the resolution of the camera is either too much, or too little to properly resolve the image. Over sampling causes the image to he soft, and under sampling causes the image to be pixelated in the fine detail. The pixel pitch of my D800 causes over sampling, and I want to try using the a smaller image size (more like the 24mp option instead of the 36mp) to simulate a camera with the proper pixel pitch. I think if it effectively turns off pixels to do this, it would work, but if it does it by eliminating pixels after the fact, it may not.

When you switch the D800 between FX and DX the pixel pitch stays the same -- you just use a smaller area of the sensor.

I'm not referring to FX vs DX. There is an option to change the image size, in pixels, while maintaining the FX format. Image area (FX vs DX) and image size are different. For example, mlthe D800, under FX image are, can record large, medium, or small images, being 7,360 x 4,912 pixels, 5,520 x 3,680 pixels, and 3,680 x 2,456 pixels respectively, all while still being FX format.

In that case the camera image processor simply re-samples the size while creating the output JPEG. Raw files and so pixel pitch would remain unchanged from the standard FX size.

 

[crimbfighter]

Does anyone know exactly how DSLRs capture the image data when you choose a reduced image size? If, for example, I select the smaller image size (ie selecting the 24mp option) in my D800, does the camera simply ignore or turn off some of the in between pixels, or does it capture the full 36mp image, then eliminate pixels using the camera processor?

I'm curious because when I am using my telescope for astrophotography, there is a phenomenon known as over or under sampling, meaning the resolution of the camera is either too much, or too little to properly resolve the image. Over sampling causes the image to he soft, and under sampling causes the image to be pixelated in the fine detail. The pixel pitch of my D800 causes over sampling, and I want to try using the a smaller image size (more like the 24mp option instead of the 36mp) to simulate a camera with the proper pixel pitch. I think if it effectively turns off pixels to do this, it would work, but if it does it by eliminating pixels after the fact, it may not.

When you switch the D800 between FX and DX the pixel pitch stays the same -- you just use a smaller area of the sensor.

I'm not referring to FX vs DX. There is an option to change the image size, in pixels, while maintaining the FX format. Image area (FX vs DX) and image size are different. For example, mlthe D800, under FX image are, can record large, medium, or small images, being 7,360 x 4,912 pixels, 5,520 x 3,680 pixels, and 3,680 x 2,456 pixels respectively, all while still being FX format.

In that case the camera image processor simply re-samples the size while creating the output JPEG. Raw files and so pixel pitch would remain unchanged from the standard FX size.

The image size options are available when shooting in RAW only. I could be wrong, and I will have to test this to be sure, but I believe it records the RAW file in the specified image size without creating a jpeg.

 

[Ysarex]

Does anyone know exactly how DSLRs capture the image data when you choose a reduced image size? If, for example, I select the smaller image size (ie selecting the 24mp option) in my D800, does the camera simply ignore or turn off some of the in between pixels, or does it capture the full 36mp image, then eliminate pixels using the camera processor?

I'm curious because when I am using my telescope for astrophotography, there is a phenomenon known as over or under sampling, meaning the resolution of the camera is either too much, or too little to properly resolve the image. Over sampling causes the image to he soft, and under sampling causes the image to be pixelated in the fine detail. The pixel pitch of my D800 causes over sampling, and I want to try using the a smaller image size (more like the 24mp option instead of the 36mp) to simulate a camera with the proper pixel pitch. I think if it effectively turns off pixels to do this, it would work, but if it does it by eliminating pixels after the fact, it may not.

When you switch the D800 between FX and DX the pixel pitch stays the same -- you just use a smaller area of the sensor.

I'm not referring to FX vs DX. There is an option to change the image size, in pixels, while maintaining the FX format. Image area (FX vs DX) and image size are different. For example, mlthe D800, under FX image are, can record large, medium, or small images, being 7,360 x 4,912 pixels, 5,520 x 3,680 pixels, and 3,680 x 2,456 pixels respectively, all while still being FX format.

In that case the camera image processor simply re-samples the size while creating the output JPEG. Raw files and so pixel pitch would remain unchanged from the standard FX size.

The image size options are available when shooting in RAW only. I could be wrong, and I will have to test this to be sure, but I believe it records the RAW file in the specified image size without creating a jpeg.

Oops! I forgot about that. You're talking about sRaw or Nikons Raw S format. I think there are some differences in how various camera makers do that. This is probably helpful about what Nikon does: Digging into Nikon RAW Size S NEFs (Hope I got the right one this time!)

 

[greybeard]

It depends on the camera brand and model. On a D850, according to Tom Hogan, when you select (medium) or (small) the raw data is resampled twice. Medium looks much like a 24mp with anti-aliasing. (D750) and Small is like a 12mp image (D700). The down side is that the files are not all that much smaller. The medium setting is still 33mp. I'm not sure how your D800 does this as Tom wrote that this double resampling was something new.

 

[JBPhotog]

The image size options are available when shooting in RAW only. I could be wrong, and I will have to test this to be sure, but I believe it records the RAW file in the specified image size without creating a jpeg.

You might want to check your manual again for the page you included in your OP, that page refers to Large, Medium and Small only for JPEG and TIFF files. These adjustments are made via the QUAL button and sub and command dials.

I don’t recall the the actual options for the D800 but I seem to remember you had two sizes for Raw images, Raw Large and Raw Small. The Raw Small is by pixel dimensions, smaller and thus the file size is small even though the frame capture size is FX rather than cropping, unlike what can be done by choosing the image area and the view finder will show the crop setting. The cameras processor is going to reduce the pixel dimensions of the FX image in Raw Small and record it in 12 bit no matter what you have set your NEF recording options to be. Additionally it does not record in RGB and the colour information is shared with a couple of adjacent pixels. If you are seeking the largest Dynamic range and the most accurate colour, shoot in Raw Large. I never use Raw Small.

FWIW, the camera creates a JPEG for every file even if you haven’t set it to create one since the LCD only displays JPEG’s, it just doesn’t send that file to the memory card if you only want Raw NEF’s.