DP1 - Production Test Samples and Charts

Even dcraw will not open them. It also extracts the wrong image as
the embedded thumbnail (getting the tiny 221x147 one) as does
IrfanView (which uses dcraw code.) ThumbsPlus will load the
embedded JPEG correctly although it will also barf on decoding the
full file.

So did Sigma/Foveon do this on purpose to inhibit use of unoptimized
converters?
I don't know everything they changed but I do know they changed around what sections they actually store - gone is the RGB thumbnail, replaced as you saw by the really small JPG. DCRaw just pulls out the first JPG it sees and misses the larger version...

As for the RAW data itself I'm not sure what was done but I'd bet it's more to accommodate the camera than to foil RAW converters. Since they never documented the structure of that section exactly I can see where they would feel free to mess with it.

--
---> Kendall
http://InsideAperture.com
http://www.pbase.com/kgelner
http://www.pbase.com/sigmadslr/user_home
 
Not even just sharpness, but actual detail - you can just see more
detail is present in the RAW converted image. Remember these are
from an ISO 800 image, too. An ISO 100 JPG sample would probably be
much closer, it's just that the in-camera noise reduction for JPG has
a hard choice to make between scrubbing out as much noise as possible
vs. retaining detail, and the computer can do a better job with both
tasks.
Oh - I understand much better now. Thanks for the explanation! :)
 
That Imaging Resource decides to use RAW when determining what the DP1 is capable of. Yes, they should also comment on the challenges with the in-camera JPGs but RAW should be the focus when determining maximum resolving power of the camera.
--Britton
My photos page:
http://www.flickr.com/photos/brittonx/
 
What are the purple marks all over her face and neck at 200asa and
upwards?
Look at her face from the same shot using other cameras - those are
natural marks on her skin. The DP-1 in higher ISO cases tended to
render them a little more purple, which could be a white balance
thing or possibly a JPG conversion.

They are less noticeable in files converted from the RAW images.
Also in conversion from the RAW it seemed like the WB was slightly
better if taken from her shirt.
Yes, I don't see any purple marks when converting from the RAW files they have provided.

--
http://aminfoto.com
http://aminsabet.com
 
I'm always on the market for a great compact - haven't found it yet! ;-)

In the meantime I've been through (or still use) a couple IXUSes, and a couple Panny FZ-5 then FZ-18.

The main weakness of these small sensor beasts, is of course, noise at higher ISOs - and I rarely use my FZ-18 at 200 (and almost never above 200).

Now I've just taken a look at the samples on imaging-resource, and guess what? They're full of horrible noise from ISO 400, I mean my FZ-18 does as well and that's from a tiny chip and Panny not having great reputation in the noise area! And of course if you take a 40D or a D300 w comparable chip size, you get very clean shots up to 800, and shots w less noise (and no colour blobs!) than the DP1 all the way up to 3'200 (available only on the D300, I know).

So, to ignorant me at least, it seems that once more Sigma is doing the nice theoretical idea (sensor captures all 3 colours at once, like colour film / sensor is DSLR-like size) but fails in the execution because the world doesn't stand still in the meantime.

In the DSLR world it seems clear (based on reviews, or sales) that the SD14 never came even close to delivering image quality comparable to that of the contemporaneous Canon / Nikon models - possibly because while the sensor was great, Sigma just couldn't deliver an image processing chain to match. Not to mention that it couldn't deliver a body that could match others in AF performance and you name it.

Now in the compact world, at least based on these samples it seems that there is a big noise problem - and maybe it can be overcome by shooting raw but remains to be seen. And what if we see Canon / Nikon launch comparable models w larger sensors, is it likely they will also be better cameras?

Seems like Sigma have a lot of good / great ideas, but not the capacity to execute on them, and to do so quickly enough so that by the time the product is out, it's not outdated by others.

But all this being said, I'd love for Sigma to prove me wrong, and for the DP1 to deliver nearly noise-free images up to, say, ISO 800. Keep fingers crossed.
 
Now I've just taken a look at the samples on imaging-resource, and
guess what? They're full of horrible noise from ISO 400
Yes, because all the samples you can view are from in-camera JPG. Take a look at this post just above that shows the difference between the JPG's you have been looking at vs. what you get from a RAW conversion without further adjustment:

http://forums.dpreview.com/forums/read.asp?forum=1027&message=27051437

--
---> Kendall
http://InsideAperture.com
http://www.pbase.com/kgelner
http://www.pbase.com/sigmadslr/user_home
 
If this is the real stuff then yes it'll look doable

I personally am not a fan of post-processing - am not a pro, life is short, spend enough time in front of a computer as is, would rather do more fun things - but if conversion using all defaults does work then that's fine by me, it can be done in batch.

It'll be nice to have a review from experts "proving" all this - just to be safe. And to buy from a store w a good returns policy so that, should it not work, one doesn't remain stuck with a nearly $1'000 camera that under-performs! ;-)

Let's wait and see
 
If this is the real stuff then yes it'll look doable
It'll be nice to have a review from experts "proving" all this - just
to be safe.
Well, I can't prove to you that I'm not faking all this, but here is a quick screen grab, still in the PNG format that Macs use for screenshots (by default). On the top is Sigma Photo Pro 3.1 for Mac. As you can see, all settings are at their default, and the ISO information (ISO 800) is specified. On the bottom is my open Firefox browser window showing the corresponding in-camera JPEG file in the Imaging Resource Comparometer.



--
http://aminfoto.com
http://aminsabet.com
 
I recall they called this a 'production version' but I wondered if
it's really the same firmware as the retail DP1s now on sale in Japan.
Sandy, I was wondering the same. Hopefully the production version will do a better job with high ISO in-camera JPEGs. The ones in the Comparometer just don't do justice to this camera.

--
http://aminfoto.com
http://aminsabet.com
 
Anyone else see purple blotches in the fabric, on the left side of Amin's comparison here? They're not in the raw conversion on the right, only in the Imaging Resource crop on the left. They're in the dark blue floral fabric, about 1/4 the size of the flower.

Interestingly enough, I can see it on my work PC, but not on my home laptop. Both are LCDs, and neither one is calibrated.
  • David
DP1 ISO 800 JPEG Crop from the Imaging Resource Comparometer on the
left. Same image processed from RAW using default values (All
sliders set to '0') in Sigma Photo Pro on the right.



That's with zero postprocessing after the RAW conversion. With a
little bit of care (Noise Ninja, etc), it can be made to look even
better.

--
http://aminfoto.com
http://aminsabet.com
 
Anyone else see purple blotches in the fabric, on the left side of
Amin's comparison here? They're not in the raw conversion on the
right, only in the Imaging Resource crop on the left. They're in the
dark blue floral fabric, about 1/4 the size of the flower.
Hi David, for sure I see the blotches you're talking about. There is color blotching in my RAW conversion as well, most noticeably in the black and blue fabrics, but it's not nearly as bad as in the Imaging Resource in-camera JPEG. Check out the difference in color blotchiness in this comparison as well -> http://forums.dpreview.com/forums/read.asp?forum=1027&message=27062288

Regards,
Amin

--
http://aminfoto.com
http://aminsabet.com
 

Keyboard shortcuts

Back
Top