10MP shootout - APS vs Digicam

[*isteve]

Its been argued over before, but the main advantage of larger sensors is noise and DR. But is resolution also affected?

In this test, my Pentax Optio A30 (10MP 1/1.8" sensor) takes on my K10D (10MP and APS sensor with 9X the area). To match more or less the A30 I used a mid range 24-90 F3.5 - F4.5 zoom on the K10D.

Both cameras were shot more or less wide open (F3.1/F3/5) at an equivalent FL of around 42mm at ISO 100. Both shots were taken within 30 seconds of each other.

No resizing was done, but in the RAW conversion from the K10D (using PPL) the sharpening and colour balance was set to match as near as possible the shot from the A30.

Now I am NOT proposing that the A30 takes better pictures, but I was surprised that they were so close in resolution terms, at least in the centre frame (the quality of the A30 drops off a lot more at the edges though due to the lens).

So, see if you can tell the difference....

Here is the full picture (converted to BW so you cant use it to judge!!) and following that are 8 crops. See if you can tell which come from which camera. A big thanks to Falconeye for hosting this comparison...

[falconeye]

Hi Steve,

the differences are indeed surprisingly hard to tell.

Let me try, assuming "better" is K10D ...

K10D:
B D E H

A30:
A C F G

I'm particularly unsure about the "Ryder" sample. Seems like one is seeing differences from the lens and missing RAW more than anything else.

One thing is for sure:
By having the crops from one camera alone, I wouldn't be able to tell if from a P&S or DSLR!


Great job, Steve!

--
P.S.
Other P&S may give worse results as the A30 seems to be a "resolution monster", as has already been discussed here:
Pentax outclasses competition

Hi Steve,

Thank you for the comparation but I guess a third row is needed with K10D and a better lens (maybe a FA43 Ltd, DA 16-45 or a DA* 16-50 to stay in zoom teritory). Then we can draw some conclusion about the quality differences between p&s and dslr.

Best regards,
Radu

[distudio]

Originally Posted by *isteve

So, see if you can tell the difference....

Here is the full picture (converted to BW so you cant use it to judge!!) and following that are 8 crops. See if you can tell which come from which camera. A big thanks to Falconeye for hosting this comparison...

A, D, E & H P&S
B, C, F & G K10

Maybe;

Both show very good and similar resolution but it's difficult to identify each as contrast is being managed differently even though the exposures are very similar. The B... set shows more CA + colour aliasing effects the A... set shows a reduction of shadow detail, hmmm... good test.

[falconeye]

Originally Posted by RaduA

a better lens

Disagree.
First, some of the crops don't look soft at all.
Second, one can always keep asking for better glass. A zoom comparison is ok because the A30 is zoom as well.

And what I've seen from the DA* 16-50 didn't blow me away
look here http://www.pentaxforums.com/forums/p...ntax-last.html
to see how much digital processing is required to make the DA* 16-50 perform decently. I'll say that it won't come any better when using a zoom lens.

Steve, what lens make was it on the K10D?

[konraDarnok]

B, E, and D are the sharpest.

Excellent test BTW!

I'd like to see more of this actually. As the BS can pile pretty high. . .it doesn't matter what people post, there are always several posts about how 'super zooms' suck, or the only way to go is primes, or larger sensors are the ultimate panacea.

Honestly, my old Canon S1 took wonderfully clear and extremely sharp photos -- just as good as my k10d.

But it couldn't do some things, like take good pictures in low light, or have a bulb setting, or have that snappy instantaneous reaction of the shutter, or show any depth of field.

I am much more happy with my Pentax, but IQ isn't why -- control and variety are reasons I bought a DSLR.

This test of course would be much different if taken in low light conditions -- or at least that is my guess.

[konraDarnok]

Actually, I take back my sharpest guesses. Each crop seems to have areas better than others in terms of sharpness and contrast.

Originally Posted by falconeye

Disagree.
First, some of the crops don't look soft at all.
Second, one can always keep asking for better glass. A zoom comparison is ok because the A30 is zoom as well.

And what I've seen from the DA* 16-50 didn't blow me away
look here http://www.pentaxforums.com/forums/p...ntax-last.html
to see how much digital processing is required to make the DA* 16-50 perform decently. I'll say that it won't come any better when using a zoom lens.

Steve, what lens make was it on the K10D?

Well, we'll have to agree to disagree I guess

The lens used is mentioned in the original post by Steve ("I used a mid range 24-90 F3.5 - F4.5 zoom") and I think that even DA 16-45 could be significantly sharper even wide open at f4. As for DA* 16-50 I think is more of a hit or miss especially in conjunction with K10D (for various reasons). But most reports even on K10D that at the tele end and stoped down a bit (f3,5) DA* is very very sharp. More so I put my money where my mouth is and ordered one for my K20D. Not trying to abuse Steve's time but I think a third row of pics with a "better" lens could solve this disagreement faster that an endless debate.

Radu

[drabina]

The pictures are really close. Hard to tell which one is the DSLR. But it is no surprise that P&S can make a good picture if the lens is decent. With DSLR you get all the manual settings and ability to override any camera setting, that you usually do not have with P&S (or have in very limited quantities).

[ricardobeat]

A30: A, D, F, H

[Geekybiker]

Well
B, is much cleaner than A noise and artifact wise. So it much be the k10d. The fine wires along the sign also have less breaks in B

C has less noise than D, also the subtle brick pattern in the wall came through better. C is the k10d

F has smoother rendered tones. As for resolution, E appears sharper in most of it but rendered the black sloped roof as one big smear while F retains detail in that section. F is probably the k10d

G again with the dark roof section. H has the singles rendered as a big smear. G would be my guess for k10d.

B C F G k10d
A D E H a30

How'd I do?

As for absolute resolution, I think we're just getting to the point where the two will start to diverge. It'd be a more interesting test to see something like the k20d vs a 14 mp PS (do they make any yet?)

[FotoPete]

BDFG = K10D

ACEH = Optio A30

Sharpness is not always the deciding factor I think.

1) The top crop G exhibit alot more shadow detail than the second row H which could be the extra sensitivity of the K10D sensor over the much smaller A30 one. This is most apparent in the Fuller's Building roof with more subtle examples on the wall of the Brewery Tap and on the pavement by the red car.

2) Watch the noise pattern in the area behind the Fuller's Building. The noise on the top photo (G) looks worse but is more grain like. The noise pattern on the bottom (H) is less chroma but more mottled and 'smudged'

[Geekybiker]

Well if there is one thing this test is proving is that most people arent immediately picking out the differences. Answers are all over the place. I wonder if that says more about the equipment or our eye for detail as a group?

[drabina]

Originally Posted by Geekybiker

Answers are all over the place. I wonder if that says more about the equipment or our eye for detail as a group?

One thing I can say is that the photos will look 10 times better when I check them at home than they look here at work on my crappy monitor.

[nostatic]

B, C, F, G all from the same camera.

Can't say which is which...once you start doing any PP all bets are off on how it deals with noise, etc. My guess would be those are the K10d with a not-great lens.