APS-C vs Full Frame RAW files

Real World APS-C vs Full Frame IQ Differences

You can find much written about the image quality superiority of full frame over aps-c cameras.  DxoMark.com does scientific analysis. Click the image below to see their full comparison of the full frame Sony A7riii and the aps-c Sony A6300:

DXOMark Sony A7riii vs A6300

Looking at the score differences, one might expect pretty dramatic differences in the results.  I’ve primarily shot full frame for over 5 years now.  I’ve always “felt” like I was getting better image quality out of full frame, but I never really did a side by side test before.

When you have plenty of light, sensor size actually means little.   With enough light, shooting at ISO 100, you should be able to get equally spectacular image quality out of just about any camera.   In theory, full frame starts to shine in more challenging lighting situations.

Sony Artisan Manny Ortiz put together a great video shooting aps-c side by side with full frame and demonstrated that except for depth of field differences, image quality could be identical.  (His video here).   But he was shooting portraits at low ISO.

When shooting landscapes, especially as the light gets lower, one would expect full frame advantages to really shine.   In theory, the greater dynamic range and noise performance should give visible advantages.  Dynamic range should give more ability to manipulate shadows and highlights in post processing raw files.   Better noise and color performance should produce images with better detail and color rendition at higher ISO.

As I created an experiment, I didn’t know what the results would be.   Have I been wasting my money investing heavily in full frame?  Despite the “score” differences, are there differences in real world shooting?  Or will there be night and day differences, which should convince all serious shooters to switch to full frame?

The Experiment

I went for a late afternoon of landscape shooting.   I equipped the Sony A6300 with the Sony 35mm F/1.8 lens while equipping the Sony A7riii with Sony 55mm F/1.8 lens (my review here), so that both cameras would give nearly equivalent field of view.   For landscapes, I shoot raw and I tend to slightly underexpose images in order to protect highlights.  I set both cameras manually, to always use identical settings.   I shot both the A6300 and A7riii at the same aperture, ISO and shutter speed, so they were getting the same amount of light.  In theory, there were slight differences in depth of field due to the focal length differences but as I shot everything at small aperture (around F8), the depth of field differences simply weren’t significant.

Though you will find a couple of jpegs below, I primarily processed RAW files in lightroom.  (See how I process images here).  For each pair of images, I applied identical post-processing.   When exporting final images, I re-sized them to about 15 megapixels.  You can see larger images by clicking each set.   I re-sized them in order to better evaluate “real world” differences instead of pixel peeping at 42 megapixels.  (Though you will find some cropped and pixel peeped comparisons below).

Set 1 ISO 125 underexposed

 

This set was shot as RAW+JPEG.  The first image, the A6300 JPEG, shows you how it was underexposed to protect the highlights.  This then required significant exposure and shadow lifting in post processing.   A gradient kept the exposure of the sky down but the rest of he image had shadows lifted by about +95 and exposure lifted by about 3 EV.   So how did the aps-c and the full frame raw files handle these massive adjustments?

I’ll admit the results were much closer than I expected but my eye has a clear preference for the full frame shot.   When not pixel peeping, they are close, but the APS-C image has more noise which results in less contrast in the image.  The full frame image is noticeably cleaner with better contrast, even without pixel peeping.   But the difference is not huge.  At a quick glance, you might not notice the differences.

Set 2 – Properly exposed at ISO 3200

I metered a “correct” exposure for this set, and ended up shooting at ISO 3200.  In my opinion, ISO 3200 isn’t a low ISO but not especially high.  In real world shooting, many photographers may often find themselves needed ISO 3200.

These images had some post processing but no extreme adjustments to exposure or shadows.  I confess:   I don’t see a difference.   Bravo to the aps-c A6300.

Set 3:  ISO 2000 with moderate processing

 

These images were just slightly underexposed and shot at ISO 2000.  Shadows were lifted a bit, exposure was lifted a bit, in post processing.  But nothing extreme.

Both camera produced similar quality.  Look carefully at the rocks in the foreground and the trees in the background, the A7riii does produce better and more detailed images.  That’s probably partially the noise performance and partially the benefit of 42 megapixels, even when downsized.

Set 4:  High Contrast Images at ISO 4000

 

First take a look at the unprocessed image from the A6300 shot at ISO 4000.  Despite underexposing, the sky is still pretty blown out, while all the detail of the building is hidden in shadow.   So this image took significant post processing.

With lots of highlight and shadow adjustment, we get usable images from both cameras.   I was surprised at the amount of shadow recovery and highlight recovery I got out of the Sony A6300, I was expecting it to fall further behind the A7riii.  But the difference in the quality of got recovered is pretty evident to my eye.   Noise softens the aps-c image, colors are less true.   Let’s take an opportunity to pixel peep:

Upon close examination, the differences are indeed very significant.   The A6300 is a soft hazy mess in comparison to the A7riii image, there you can now clearly make out the beams under the roof.  (Again, this is partially an ISO performance difference and partially the benefit of more megapixels).

Set 5 — ISO 2500, normal processing

In the first image, you can see how it looked before post-processing.   Maybe slightly underexposed but much.   After processing, the images are very very close.   I could probably print a 8×12 of each and most people wouldn’t be able to tell them apart.

I set the focus point on the rocks in the foreground.   If you look at those rocks, you will see that the color and contrast is better in the full frame.  A critical eye will notice it.

Set 6 — Cropped at ISO 8000

There were some geese in the distance that I decided to photograph.  As many of us have experienced, I didn’t have a telephoto lens handy so I had to resort to cropping.   Cropping effectively is pixel peeping.  At the sun was setting, I used ISO 8000 to get properly exposed images.

First, the out of camera jpegs:

Honestly, neither is great but the A63000 APS-C jpeg is particularly ugly.  It looks like an old flip-phone camera photo, grainy and filled with artifact.  The A7riii image is “ok.”  Looking at the processed raw files:

With more careful processing, the A7riii image can be made to look acceptable, even good, at this high ISO.   The A6300 image is simply unsalvageable.   But remember, we are now shooting at pretty high ISO and cropping on top of it!

Set 7:  ISO 320 protesting the highlights

 

My last set of the evening.  In the first, see the image before processing.   Protecting the blue sky and the clouds, requiring a fair amount of processing.  But at least we were starting near base ISO.   Looking at the results at normal size, it’s pretty comparable.  The full frame retains slightly more contrast overall, as noise has started to affect the A6300 image.  But views at normal size, it isn’t very noticeable.  Let’s pixel peep for the last time:

To me, the difference is obvious in the pixel peeped images but still not an earth shattering difference.  The full frame 42 megapixel A7riii, even downsized, still produces better detail and contrast in the final processed image.   But if we aren’t pixel peeping, then these differences are indeed very minor.

Lessons Learned

Honestly, the differences between full frame and APS-C shooting were not as extreme as I expected, at least when shooting with the Sony A6300 and Sony A7riii.   The aps-c sensor was very close to the full frame sensor in terms of shadow and highlight recovery at low and medium ISOs.   Still, to the critical eye, the full frame sensor consistently produced better images.

If cropping, pixel peeping, or printing large, these differences are even more obvious, making full frame a worthy investment.   Especially with a high megapixel camera like the Sony A7riii or the Nikon D850, you can get a level of crisp detail not possible with aps-c sensors (which currently top out at about 24 megapixels).   With relatively affordable full frame cameras like the Sony A7iii, Nikon D750 and Canon 6Dii, full frame should produce noticeably cleaner images with more flexibility in post processing.

For those with less critical eyes or less critical needs,modern APS-C cameras can produce more-than-adequate image quality.  When printing smaller sizes or viewing at normal sizes, it can be hard to even see the difference compared to full frame, at least at moderate ISOs.  At very high ISOs, differences will be far more obvious.   But not everyone really needs to be able to shoot at ISO 12,800+!

So for myself, I’m glad that I primarily shoot full frame.  The experiment left me with no regrets.   But I may be slower to recommend full frame to others.  If you are considering switching for image quality improvements, get a realistic idea of how much the image quality may improve.

If you are considering any of the gear used in these tests, you can find them on Amazon:

Sony A6300

Sony E-mount 35mm F/1.8

Sony A7riii

Sony 55mm F/1.8 starting

The Best Value Full Frames:

Sony A7iii

Nikon D750

Canon 6Dii