DX Auto-Crop

J-see

Senior Member
The crop factor doesn't need to be factored in since I check only the optical sharpness. 10 or 16 would count for that particular lens/sensor. But that's an overall number which I only use as a guideline.

Up till now I always assumed the DX was better for shots that required crop. I always shot macro with the DX. I haven't use the FX much for it since the season ended and I'll have to wait until next spring. For birding I used the D750 simply because I can gain more light by using higher ISO. The D3300 isn't that great compared.

It's that I wondered about it and did some test shots, I noticed that the "DX has better detail" isn't necessarily true. You don't really notice much when doing normal shots but when you crop the same part out of the FX shot and enlarge it to 150% to get the same size as the DX, it is obvious more megapixels doesn't necessarily translate into more detail. In theory yes but it would require both to be identical loss-wise.

It's possible my DX being a slightly worse unit of all sold but anyone having both types wouldn't do bad by doing the same test. If my cam can be bad, so could anyone's. Focusing won't be the issue since I have to manually focus the 200mm which is why I took plenty a shot and selected the best to rule out bad focus. If my DX would deliver better results, I'd shoot it without thinking twice. But as it shows, me just relying on what is written online isn't necessarily beneficial.

I didn't upgrade to a D810 based upon the idea for macro there would be no gain. I think I was wrong.

I checked the shots of the D7100 vs D800 but it's hard to say since they're not the same size. When comparing numbers it's possible it's different for you. It's 19/36 vs 11/24. Mine was 10/24 vs 19/24 using the D610 sensor. That's assuming you would have shot my lens.
 
Last edited:

J-see

Senior Member
I've been thinking why in my case the FX does better and I think it is very simple.

If for a moment we consider the numbers used as correct, we have 10/24 DX vs 19/24 FX.

That implies the lens sharpness only is used at 42% capacity on the DX vs 79% on the FX. Using that for the 24Mp shot vs 12Mpix crop, that's 24*0.42 vs 12*0.79 = 10.08 vs 9.48.
The quality probably isn't equally distributed and is highest at the center which comes as an advantage to the FX here.

This could explain why the shots show the opposite of what I previously assumed.


To add: I'm using a very simple formula now to check when it pays. If my FF/lens (P-Mpix/3)*2 equals or is greater than the DX/lens P-Mpix it might do better on the FF and is worth testing. If it is lower, DX is probably better for crop-shots. But only testing it might show how true that is.

Tamron at 600mm:

002-Edit.jpg

003-2-Edit.jpg

Strange; manual mode same shutter speed but both different data in the EXIF.
 
Last edited:

Geoffc

Senior Member
[MENTION=31330]J-see[/MENTION] are you taking these as jpgs? If yes, what are you picture controls set at as this will affect what the images look like. Yours do not seem like anything I've ever seen before.Whenever I do this kind of thing I tend to use raw and expose them identically so that any sharpening can be applied to both in the same way. If you could post the raw files somewhere I could download them that would be really interesting.

If I get time I might have a look for something meaningful (I often use furry toys due to the texture being a good reference) to shoot, because as you zoom in more and more the lower pixel image will pixelate before the other.
 

J-see

Senior Member
No I always shoot RAW and didn't use any post besides the standard lens adjustment and minor sharpening LR does.

The shots are a crop in LR smaller than 1k*1k so I can show the 100% here. Higher and it gets scaled again. To equal the FX to the DX, I go to PS, copy the DX shot into the FX layers, duplicate the FX, scale it to 150% and with a lower opacity match it with the DX.

I'm using the D750 of which not much data is out there. I haven't got any lens/cam info so I assume it to perform as well as the D610 sensor. Maybe this sensor is better, I wouldn't know. It's not that the D3300 is a bad cam. I have always taken a good shot with it but when purely comparing detail, it shows the DX sensor doesn't make the lens perform as well as the current FX. Or the sensor can't grab as much detail. The money difference has a reason.

Technology changes all the time so what might have been true yesterday not necessarily is today.

All in all it's a luxury problem. You need to have both formats before you can worry which is best. It's also only important for my macro and bird lens. But when looking at the numbers, anyone having a D600-610, D750, D800E or D810 and has a DX they use for macro or wildlife best put their lenses to the test. They might be surprised.
 

J-see

Senior Member
Last edited:

Eyelight

Senior Member
Interesting stuff [MENTION=31330]J-see[/MENTION]

Couple suggestions that may be useful.

The exposure does affect the sharpness of an image, especially when making these type comparisons and when we are pushing the envelope of sharpness. Our eyes are still in this process, and colors and highlights and shadows affect what we can and cannot see. So, I would aim for near identical exposures.

Instead of the PS matching/cropping method, crop the DX and then crop the FX to an equivalence size. Does not need to be identical cropped image as long as we see the same detail in the middle. Example: Crop the DX to 600 pixels wide (6000 * 0.1) and crop the FX to 394 pixels wide (6016 * 0.1 * 0.655). I'd use LRs cropping tool and change the crop until the size is where you want it. You can check the selected crop width by hovering over the thumbnail in the bin before clicking Done in the cropping tool.
 

J-see

Senior Member
I agree the exposure can have a minor effect but the previous shots had the same exposure and the quality difference showed too. It's about 1/3th of a stop which is minimal.

I can scale them in LR. I just used PS because I wanted to be as close to the same size as possible. It's annoying I can't define a crop size in LR. That should have been a standard option.
 

J-see

Senior Member
I cropped it in LR to about the same size but then had to save that again in a bigger format to match the DX shot. I never get that as good as simply matching in PS. In some way it shouldn't matter. If we'd blow up something, we'd do it as good as possible and would use PS anyways.

Anyways, here it is but it isn't as correctly sized as the other.

002.jpg

003-2.jpg

I'll shoot the other lengths of the Tam tonight to see if it is the same for all. It's some work since I have to use the self-timer function on the D3300 to avoid trigger-shake.
 
Last edited:

Horoscope Fish

Senior Member
@J-see: It seems I am, once again, late to the party...

What you're doing here may be of particular interest to me but I'm not sure.

Can you sum up your findings for me please?

....
 

J-see

Senior Member
@J-see: It seems I am, once again, late to the party...

What you're doing here may be of particular interest to me but I'm not sure.

Can you sum up your findings for me please?

....

It was about the idea that DX is better to shoot macro or wildlife because of the crop advantage and thus would show more detail than an FX which inevitably loses plenty a pixel when we use the same area of the shot.

I tested it with my lenses; D3300 vs D750 and apparently it is not true here. When shooting DX and FX and then cropping/scaling the FX area to the same size as the DX, the FX shot is of a better quality even when having used less Mpix for that area.

What makes the difference is not the megapixels but the efficiency of the sensor.
 

Horoscope Fish

Senior Member
It was about the idea that DX is better to shoot macro or wildlife because of the crop advantage and thus would show more detail than an FX which inevitably loses plenty a pixel when we use the same area of the shot.

I tested it with my lenses; D3300 vs D750 and apparently it is not true here. When shooting DX and FX and then cropping/scaling the FX area to the same size as the DX, the FX shot is of a better quality even when having used less Mpix for that area.

What makes the difference is not the megapixels but the efficiency of the sensor.
Thank you... That is, pretty much, what I was getting out of this thread.

Wouldn't this disparity in image quality be more related to the fact the pixels on an FX sensor are larger than those of a DX sensor (assuming, of course, the same total number of MP for each)?

....
 
Last edited:

J-see

Senior Member
Thank you... That is, pretty much, what I was getting out of this thread.

Wouldn't this disparity in image quality be more related to the fact that the pixels on an FX sensor are larger than those of a DX sensor, assuming the same total number of MP for each sensor, than sensor efficiency, though?

....

I wouldn't know what exactly causes that difference.

I was curious because it is like a standard rule "DX is better for crop shots". From what I read online, that seems to be the general consensus but when checking the optical sharpness of my lenses, the differences were huge which made me wonder. When checking the values at DxOmarks it seems to be true for a range of FF but not all. The D800 seems not as efficient as the D800E with my lens so it probably isn't just pixels or size.

In theory my D3300 should win since the shot is 6k*4k and the D750 has to compete with 3.9k*2.6k pixels. Logic tells us therefor the D3300 should contain more detail. When scaling the smaller format to 6*4 it doesn't improve quality either. Yet the D750 still does a better job with some lenses.

That's the reason I didn't opt for the D810 for macro since the D3300 had more pixels and thus would grab more detail. Yet even with the D750 I already grab more detail so when using the D810 which has a larger pixel format even in DX, the quality should even be better.

I haven't got any idea what the reason here is but it goes against photo folklore. ;)
 

gqtuazon

Gear Head
Interesting read. Another late person in the topic. Good reason for me to break-out my camera.

Are the images shown above done using "Mup"? If not, then you should have especially with a 1 second shutter speed.

The other factor when comparing DX vs FX when pixel peeping, you need to use a lens that is mounted on the tripod so that nothing is changed except the camera to compare the image quality. Just my two cents.
 

J-see

Senior Member
Yeah, the D750 is all shot with Mup and the last of the D3300 is shot with the self-timer. I doesn't have Mup and I have no remote. It took me a while to realize the self-timer works too.

Lens was mounted on the tripod all the time, I just replaced the cams. Except the 35mm which cant be attached.
 

Eyelight

Senior Member
I have been lusting for a DX body for ages especially for macro , but now this is making me think again.

Could the issue be around the diffraction limit.
Diffraction Limited Photography: Pixel Size, Aperture and Airy Disks

Diffraction should not be an issue at the apertures in use, and would also only affect the image quality of an otherwise perfect image, if that makes sense.


Edited to add: It just seems to me that we are not seeing what the D3300 should do and it is puzzling. I think there is a point where the FX would surpass a DX in this type of comparison, but it seems to me it's passing the DX ahead of where it should.
 
Last edited:

Eyelight

Senior Member
Yeah, the D750 is all shot with Mup and the last of the D3300 is shot with the self-timer. I doesn't have Mup and I have no remote. It took me a while to realize the self-timer works too.

Lens was mounted on the tripod all the time, I just replaced the cams. Except the 35mm which cant be attached.

Thought about this earlier, but slipped from the mind. Since the D3300 has no MUP, would be a more fair battle at a faster speed. Yes??
 

J-see

Senior Member
Thought about this earlier, but slipped from the mind. Since the D3300 has no MUP, would be a more fair battle at a faster speed. Yes??

I used the self-timer since the mirror of the D3300 isn't really causing quakes. It's not like the D750 that comes with 2 little guys inside that slam it up and down. It's hard to shoot high shutter inside since I lack the light and when upping the ISO, the D3300 would suffer more noise than the D750. Outside it is not the good time a year and even the wind might mess up.

I'll see what I can do later on. Any vibrations picked up would count for both. I took multiple shots with the D3300 to get the best. It's not as if I can accidentally shoot sharper.
 

J-see

Senior Member
If anyone else has a DX and FX and a lens that makes the difference, I'd love to see if you can repeat the results.

Until then I could be blessed with a too good D750 or a too bad D3300. ;)
 
Top