Sigma DP2 Shootout Pt. 3 - Dynamic Range

In Part 3 of the shootout, we'll take a look at how the Sigma DP2 dynamic range compares to that of the Panasonic GH1 and M. Zuiko 17mm f/2.8.

Readers of this site may recall that the DP1 showed superior dynamic range in an earlier shootout, but the GH1 arguably features the latest and greatest Four Thirds sensor technology to date. How will they compare?

I chose a nearly impossible scene for the sake of comparison. At noon, without a cloud in the sky, there is probably no digital camera that can fully handle this dynamic range in a single exposure:



For comparison, we'll be looking specifically at the two regions highlighted by the yellow boxes above.

Test conditions:

  • The GH1 was used in 3:2 aspect ratio.
  • Subject distance was adjusted to match framing with the two systems.
  • GH1 files were processed in Lightroom (latest version) and subsequently downsized using Photoshop Bicubic Sharper to match the native dimensions of the DP2 files.
  • DP2 files were processed in Sigma Photo Pro (latest version).
  • All settings were left on defaults.  White balance was not adjusted to match colors, though I experimented with this and found that doing so did not substantially affect the results.

To begin with, a single RAW file was taken with each camera at ISO 100, f/5.6, and 1/50s.

Here are the crops without any exposure adjustments during RAW processing:



Clearly the DP2 is holding highlights better here. The black points are different, but you can actually see the shadow detail better in the DP2 file. This is evident in the stones under the deck (bottom panel).

Now let's see how much highlight detail can be recovered with a -2EV adjustment during RAW processing:



One possible explanation for the much greater highlight detail in the DP2 file is that what Sigma is calling ISO 100 corresponds to a lower actual ISO than what Panasonic is calling ISO 100. This would result in the DP2 file being underexposed relative to the GH1 file.

If that were the reason for the disparity, we would expect greater usable shadow detail in the GH1 file. Let's see what happens when the same files are pushed to +2EV during RAW conversion:



The greater shadow detail is now even more apparent in the DP2 file.

Conclusion: Under identical circumstances at base ISO, a DP1 RAW file simultaneously delivers greater highlight and greater shadow detail.

Next lets see how the two systems handle the same scene at ISO 400, f/11, 1/50s.

Here are the crops without any exposure adjustments during RAW processing:



The same files after a -2EV adjustment during RAW conversion:



Interesting that the ISO 400 DP2 file with a -2EV adjustment retains more highlight information than the ISO 100 DP2 file with a -2EV adjustment. We saw this in the DP1 shootout as well. It suggests that the high ISO DP files are being (partially or fully) pushed in SPP.

Now the same files after a +2EV adjustment during conversion:



Both files are struggling to produce any usable shadow information here. In the top panel, the DP2 still seems to do better.

The bottom line for ISO 400 is unchanged from the bottom line for ISO 100: better dynamic range from the Sigma.

The RAW files used in this comparison can be downloaded below:




Posted by Amin

Comments (35)

Loading... Logging you in...
  • Logged in as
Commenting Disabled
Further commenting on this page has been disabled by the blog admin.
GH1 is more or less multimedia camera. More is interesting compare with Panasonic G1 (or GF1 or Olympus EP-1).
Reply
1 reply · active 811 weeks ago
Thank you for doing this shootout.

I wasn't interested in the DP-2 until recently. Now, I'm thinking that the slow performance is a low price to pay for a camera that gives better results for the kind of pictures I actually take- very little action. Quality is more important to me than speed.
Reply
1 reply · active less than 1 minute ago
Well, it looks like from the point of view of IQ there is no contest, which is incredible given the already high level of performance of the GH1. I was torn between these two cameras (and Ricoh GRIII and Lumix GF1) but now I have no doubts -- I'll get a DP2. Thanks for the shootout
Reply
improbable's avatar

improbable · 811 weeks ago

Impressive performance by the DP2 -- I remember previous tests showing that it beat the compact cameras in dynamic range, but here it clearly beats a 4/3 sensor too. Nice test this.

Do you recall how it did against bigger SLRs?
Reply
1 reply · active 811 weeks ago
Thanks Amin for the tests, nice job as usual. I used my dp2 at my vacation under harsh lights in similar to your test and I was amazed how much I can get from the images with adjusting dynamic range without too much noise. I had a similar experience with dp1 and Canon 40D comparison shots as DP1 was holding or even better in some conditions due to retention of its nice colors. In comparison Oly EP-1 lags behind in dynamic range with its smaller sensor similar to your test and pushes the noise more when it is adjusted for darker areas. I don't know how EP1 vs GH1 compare but I don't expect too much difference as DXO tests show the similar values for Oly and Pana DSLR/m43 cameras esp if you are shooting raw. I hope these will answer some of the questions above.
Reply
1 reply · active 811 weeks ago
I shoot with Pentax - would not change to either camera or sensor technology, probably not much bias.

So, you shoot with different real exposure, process with different software and withour controllet light and we'll get these results. Not too interesting.

And to make issues worse, you totally ignore the color cast. Due to Foveon sensor construction there will easily be color problems outside of the "normal" range - the dynamic range will be much more limited on some color channels than other. Seems like green channel is well represented, but what about blue and red?

Ironically, foveon sensor is better suited for B&W than color photos - considering DP2 suits street shooting, it's not a big dea, especially, since B&W DR is likely to be quite nice. What is a big deal is low resolution and aliasing it causes, especially with the very sharp lens of DP2.
Reply
1 reply · active 811 weeks ago
Oops delete the previous, please :)

A clarification on the color cast: when you fix it, you'll raise red and blue noise and that reduces DR
Reply
1 reply · active 811 weeks ago
I would not be too concerned about colors: if you look carefully at Jim Radcliffe's photos with the DP2 (and DP1) it's hard not to get impressed. Based on the few serious reviews I see (like this one by Amin) I believe this little camera has among the highest IQ including serious DSLR. The problem is that Sigma is a small company, it doesn't have the reputation and customer recognition of, say, a Canon or a Nikon or an Olympus, and it doesn't get the same press coverage. Stick on the top of the DP2 "Canon" or "Olympus", and the DP2 would be, today, a cult camera, probably the world's bestseller given how wide the quality difference is between it and similarly priced compacts.
Reply
2 replies · active 811 weeks ago
Long Time Reader's avatar

Long Time Reader · 811 weeks ago

For your test conditions, instead of listing:
* GH1 files were processed in Lightroom (latest version)
* DP2 files were processed in Sigma Photo Pro (latest version).
could you instead post the actual versions used?

The way it is written may make sense for "right now", but it becomes much less useful to readers finding this article several months down the road when one or both converters have had new versions released. As was evidenced by the SPP 2.5 to 3.5 update, aspects such as highlight recovery can change dramatically between software versions.

Keep up the great work!
Reply
1 reply · active 811 weeks ago
Very nice article, I think I need to get a DP2. And about the colors well the Fuji Velvia or the Kodak Portra weren't very accurate in color reproduction but they shell a lot.
Reply
I have a DP2 in the mail - now available 349 EUR new in Germany.

I really look very much forward to this cam. I am used to taking fotos with rangefinders, so I guess I won't be troubled by the slowness (hopefully). I just cannot understand the daming review of the DP2 over at DPreview (which appeared online today).
Reply
Well, you open the website and the first thing you notice is a Lumix button that looks very much like an advertisement. I have no idea about who finances DPreview and this is totally speculative, but if they survive thanks to advertising, and the big camera manufacturers advertise heavily, then it's hard to swallow that the little Sigma performs better than the new big thing in photography (micro four thirds). So, that is an option. The other possibility is that the Sigma is an inferior product...
Reply
I'm actually just writing in to say that 'Serious Compacts" is such a good name and theme for a blog! If I'd chosen it -- or something similar -- the content of my own blog might be significantly different.

Keep up the good work.

Gordon
lightdescription.blogspot.com
Reply
1 reply · active 810 weeks ago
DP2 Fan:

Where can you get the DP2 for that price in Germany - I'd be very keen at that level!
Reply
Good article. My observation however would be that "ISO 100, f/5.6, and 1/50s" mean very different things for different cameras and obviously different lenses, so this may not be a totally fair comparison. The GH1 seems hopelessly overexposed from the get-to. I have not handled a DP2, but my experience using the GF1 under similar conditions is very different from your results here.
Reply
3 replies · active 810 weeks ago
Paul H:

Foto Brenner (http://www.fotobrenner.de/) sells DP2 with 599 EUR minus 250 EUR cashback = 349 EUR. Shipping is free. To get the cash back, you need to send them an old camera with your written order. Any crap film camera will do, I sent them a 1 EUR cam from a junk store (so the DP2 has cost me actually 350 EUR).

For that money the DP2 is absolutely competitvely priced, with that lens and the sensor.

BTW, the whole discussion reminds me a bit on the Fuji F Series cameras (F10,11,30,31). Good sensor in a mediocre camera body. The F30 was sold out very cheaply, but only after production stopped people realized how good the F30 actually was. And the prices for F30/31 on ebay today are still more than what I paid when I got mine four years ago...
Reply
1 reply · active less than 1 minute ago
i'm a bit concerned with the different subject distance to ensure similar framing.

for example if camera A is closer than camera B to the subject, but they both have the same exposure settings, camera A would be more likely to overexpose than camera B.

just a thought. :)
Reply
1 reply · active 810 weeks ago
Obviously the GH1 is far more sensitive than the DP2. Match exposure with the histogram- look at the white levels! You can't rate the two cameras at ISO 100 and set the same exposure and get an accurate result for dynamic range. Start by consider the GH1 an ISO 320 when set at ISO 100. You'll get results that seem much more comparable. Really both cameras should be rated before conducting a dynamic range test. You don't do you blog any favors by mishandling a simple test like this, this test is a fail, and nothing meaningful can be gleaned from either camera.
Reply
1 reply · active 810 weeks ago

Comments by

 
Copyright 2007 | Andreas08v2 by GeckoandFly and TemplatesForYou | Design by Andreas Viklund
TFY Burajiru