Phone Arena does blind comparison camera test between DSLR, Note 4, and 6 plus...

[magomago]

I'm not sure how good of a comparison it was.

some of those images looked quite soft on the dslr - was there something wrong (way too wide open of an aperture? Even then it was quite fuzzy)

Also, lets do some more real indoor shots...

 

[destrekor]

I was going to complain about the "minimally-processed" RAW photos being used against processed JPEGs, but... I see that a few others have covered that topic with enough detail.

I'm glad others understood that!

 

[magomago]

I was going to complain about the "minimally-processed" RAW photos being used against processed JPEGs, but... I see that a few others have covered that topic with enough detail.

I'm glad others understood that!

Even then RAW shouldn't be that soft out of the camera

 

[rcpratt]

To break it down into something you might understand...its like a Honda Civic pulling up to a McLaren P1 at an intersection. The McLaren doesn't want to race. The light turns green and the Civic pulls away and is ahead of a much faster car. The McLaren isn't racing, the driver is yawning, looking at the sites, the Civic is and gave it the beans. The bystander can say the Civic crossed the line first. But that doesn't mean the Civic is a faster a car.

To break it down into something you might understand - a more expensive car is not necessarily a faster car.

 

[destrekor]

Even then RAW shouldn't be that soft out of the camera

My D600 usually is. There might be settings to apply some sharpening or other trickery to the RAW file, but I purposefully shoot pure RAW and post-process to my liking.

Some time I'll actually mess around with the in-camera JPEG, because there are too many times I'd prefer to just take photos and give them to people without spending any more time on them.

It depends on the sensor itself and also the filters placed in front of the sensor. Some recent cameras, like the D800, had a variant with the anti-aliasing filter removed (D800E), and that produces sharper RAW photos. I think smaller sensors with high MP settings may look sharper out of the gate in RAW compared to larger sensors at the same MP (smaller pixels at full resolution), so some of the higher-tier consumer 24MP+ cameras with APS-sized sensors may produce a slightly sharper RAW than my 24MP Full Frame D600.
I might have that backward, but that sounds right in my head.

 

[magomago]

My D600 usually is. There might be settings to apply some sharpening or other trickery to the RAW file, but I purposefully shoot pure RAW and post-process to my liking.

Some time I'll actually mess around with the in-camera JPEG, because there are too many times I'd prefer to just take photos and give them to people without spending any more time on them.

It depends on the sensor itself and also the filters placed in front of the sensor. Some recent cameras, like the D800, had a variant with the anti-aliasing filter removed (D800E), and that produces sharper RAW photos. I think smaller sensors with high MP settings may look sharper out of the gate in RAW compared to larger sensors at the same MP (smaller pixels at full resolution), so some of the higher-tier consumer 24MP+ cameras with APS-sized sensors may produce a slightly sharper RAW than my 24MP Full Frame D600.
I might have that backward, but that sounds right in my head.

I suppose its entirely possible, but its been a long time since I looked at an unsharpened raw file; lightroom usually applies a +20-25 default sharpening value which is more than enough for me.

Hmm I'm not so sure on the second part, I think there is a balance between MP and sensor area. I remember reading a site that reviewed images from different MP sized cameras (This was a few years back, maybe tech has changed) and showed that with high MP cameras, as your effective surface area is reduced you can get "hazy" photos that aren't nice and sharp. I'm sure there is a some parameter ( MP/effective surface area) that better illustrates where the gains of MP alone are limited.

 

[Midwayman]

I suppose its entirely possible, but its been a long time since I looked at an unsharpened raw file; lightroom usually applies a +20-25 default sharpening value which is more than enough for me.

Hmm I'm not so sure on the second part, I think there is a balance between MP and sensor area. I remember reading a site that reviewed images from different MP sized cameras (This was a few years back, maybe tech has changed) and showed that with high MP cameras, as your effective surface area is reduced you can get "hazy" photos that aren't nice and sharp. I'm sure there is a some parameter ( MP/effective surface area) that better illustrates where the gains of MP alone are limited.

Sounds like you're thinking of the diffraction limit. There is a physical limit on how small a pixel on the sensor can be even with perfect optics. Cell phone cameras are almost certainly exceeding that, and 18mp on a APS-C sensor is probably pushing it.

If you rule out a defect (wrong backspacing on lens, etc) there are a couple other things that could be an issue.

-If they're shooting by hand it creates a non-repeatable test that *will* show up on 1:1 enlargements. You need to shoot on a tripod to effectively compare cameras.

-The DOF is different due to the different focal length at the same FOV. Smaller sensors tend to have greater sharpness because of this, and focus isn't as critical. They seem to be cropping way off center where the focal point may not be anywhere near where they cropped. Or they just got lucky on the other cams.

Since they can't really control what the smartphones are doing very well, they should be cropping center of frame, taking 10 photos on each scene and comparing the best one to give them an equal chance.

 

[Blanky]

Super silly test. Now try all three taking a picture inside of somebody standing in front of a window.