Just for fun - can you pick the APS-C digital camera and lens?

[Scrambler]

There has been some discussion lately around “how many megapixels is enough” and the like. It got me to thinking…
We recently upgraded our Pentax K100d SLR – mostly because I thought it was dying but in fact it just needed a firmware upgrade to cope with larger capacity cards – Doh!

My wife got a Canon EOS700D and 18-135 super zoom. I bought a Sony a3000 mirrorless for adapting lenses.

I got wondering – how much difference is there really between 6mp, 18mp and 20mp?

All photos were manually focussed (so as not to be able to blame the autofocus spot) or focus selected on the Canon live-view.

All cameras were shot in raw and converted in Lightroom using auto-tone settings and then resized to 16MP and 6MP as these were “sweet spots” in recent discussions.

My question for the forum: who can match the photo with the camera/lens combination? All zooms were shot between 50 and 58mm, the primes are in the same range. All shot at f5.6.

I will start with 6MP and then post the 16MP (16MP IS easier to pick differences, though not necessarily the ones I expected).

The options:
Canon EOS700D with EF-S 18-135
EOS700D with EF 35-80
Pentax K100D with Sigma 18-50
K100D with Super-Takumar (M42) 55/2
Sony a3000 with E-mount 18-55
a3000 with Leica Summitar 50/2


camtest 6m-6 by Scrambler@4350, on Flickr

camtest 6m-5 by Scrambler@4350, on Flickr

camtest 6m-4 by Scrambler@4350, on Flickr

camtest 6m-3 by Scrambler@4350, on Flickr

camtest 6m-2 by Scrambler@4350, on Flickr

camtest 6m-1 by Scrambler@4350, on Flickr

 

[Sejanus.Aelianus]

Great way to show off all your toys.

Enjoy yourself.

 

[Scrambler]

Crops from the 16MP files (not in the same order):


camtest crop 6 by Scrambler@4350, on Flickr


camtest crop 5 by Scrambler@4350, on Flickr


camtest crop 4 by Scrambler@4350, on Flickr


camtest crop 3 by Scrambler@4350, on Flickr


camtest crop 2 by Scrambler@4350, on Flickr


camtest crop 1 by Scrambler@4350, on Flickr

This might not be of much interest, but to me it revealed some interesting results. And before there are any more cracks about my toys, I have no doubt that the results could be applied to much more interesting cameras and lenses - and subjects. The point here is that even at the consumer/ price conscious end of the spectrum there are choices. And that if all you do is post to the web or print 7x5 then you may be better to get an old (or just better) lens rather than a new camera.

One result was that at 6MP there isn't that much in it. And that means at postcard print size 5 of the 6 would be pretty much equal in my eyes.

At 16MP ... well, you be the judge.

 

[jmilkins]

Hmmmn - I think you make a very good point on at the 6mp and postcard to 5x7 print size. But don't tell the public or they'll stick with their current digicam and not upgrade every year ;-) (i'm kicking myself for missing a GRdigital version 1 for for $85 recently - beautiful 8MP files).

At 16MP yep some clear sharpness vs mush,and chromatic aberations that seem to creep in. Eg sharpness of crop 5 vs 4 . Not sure I could pick which is which, though I'm thinking the K100 with sigma zoom is going to be amongst the mushiest when upsized. Or maybe the 35-80 with the newer Canon might not perform too well.
The lower contrast is what 'd generally expect from the uncoated lenses.
Loon forward to the answers. Hoping the summitar gives you a good result on the nex. I keep seeing its signature, esp in mono and really liking it.

 

[Addy101]

Looking at the crops, there is only one I like... Could that be the A3000 with the Leica lens?

 

[Scrambler]

Looking at the crops, there is only one I like... Could that be the A3000 with the Leica lens?

Hard to say if you don't tell me which one ...

 

[Spanik]

Difficult. Certainly in the first batch of photo's. All are resampled to the same small size and I fear that any difference is lost after that. Also I think there is a difference using a 6Mp camera at 6Mp and a 16Mp resampled at 6Mp. But at the same time nothing to worry about when you put them like this on the net.

The second batch is impossible to say anything about. It looks as if there are very different focus points selected. The difference is too large to be attribuated solely to different camera.

 

[Scrambler]

Difficult. Certainly in the first batch of photo's. All are resampled to the same small size and I fear that any difference is lost after that. Also I think there is a difference using a 6Mp camera at 6Mp and a 16Mp resampled at 6Mp. But at the same time nothing to worry about when you put them like this on the net.

The second batch is impossible to say anything about. It looks as if there are very different focus points selected. The difference is too large to be attribuated solely to different camera.

Selected focus point was the left (as viewed) eye in each case. This was harder to hit for some lenses. I agree that the lenses made a bigger difference than the cameras, even though cameras varied from 6 to 20 MP. But camera did make a difference...

 

[Scrambler]

Results follow, with extreme crops.
16MP answers:

camtest -1 Summitar by Scrambler@4350, on Flickr


camtest -3 Canon 18-135 by Scrambler@4350, on Flickr


camtest -4 Canon 35-80 by Scrambler@4350, on Flickr


camtest -2 Sony 18-55 by Scrambler@4350, on Flickr


camtest -5 Pentax 55 by Scrambler@4350, on Flickr


camtest -6 Sigma 18-50 by Scrambler@4350, on Flickr


To my eyes, the a3000/ Summitar looks slightly clearer at 16MP than the Canon 18-135 though lower contrast. It’s pretty close-run (as far as resolution goes). I think you pretty much get your 16MP with these lens/camera combinations.

My subjective sense is that the next three combinations are close – very close. But not too close to call. I think the Canon 35-80 just outshines the Sony 18-55 which also pips the Pentax/ 55mm. The big surprise here is how the quality prime pulls up the low MP camera. It’s hard to argue there are 16MP of real imaging with the Canon and Sony kit lenses when a 6MP camera comes this close.

The Sigma earned the wooden spoon. It was also nearly impossible to manually focus due to a very short throw. I had multiple attempts with bracketing minor changes to get even this mushy example. And no part of the frame is sharper than this crop. The Sigma even looks bad in the 6MP photos.


Lessons? Perhaps if I’d known all this I would have been more inclined to put up with the Pentax’s limitations and maybe get it a decent zoom lens. And Lola is a surprisingly good resolution test.

And I take the criticism re focus variation, though I don’t think it alters the results. It’s not real easy to get the SLR’s spot-on.

 

[michaelwj]

(i'm kicking myself for missing a GRdigital version 1 for for $85 recently - beautiful 8MP files).

For the best John, aren't you minimising your collection anyway (or at least keeping it static)!! 😀

Cheers,
Michael

 

[Scrambler]

Hoping the summitar gives you a good result on the nex. I keep seeing its signature, esp in mono and really liking it.

The Summitar does shine and the full 20MP file is even better. I went for 16MP though to give a "level playing field" at the point most people seemed to think was "enough" resolution for large prints.