Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: Sigma DP2 Merrill colour  (Read 4329 times)
xtopher
Newbie
*
Offline Offline

Posts: 4


« on: September 20, 2013, 12:26:54 PM »
ReplyReply

Trying out a Sigma DP2 Merrill and comparing it to Panasonic GX1 and Nikon D800.

I took some shots of my bookcase, and realised small details in red on blue background on book spines were being reproduced as blue only. I’ve done a few tests since that show the same thing — weak or non-existent reproduction of any small areas of red.

Very surprised, as if anything I would have expected the sensor would have been better at colour detail than a Bayer sensor.

Anybody else noticed this?

Chris
Logged
xtopher
Newbie
*
Offline Offline

Posts: 4


« Reply #1 on: September 20, 2013, 02:52:45 PM »
ReplyReply

Here are a couple of jpegs showing what I am talking about. The blue lines on the red tint and the red on blue in the book by ALAN WOODS are visible in the GX1 image but not in the DP2M one.
« Last Edit: September 20, 2013, 02:55:05 PM by xtopher » Logged
BarbaraArmstrong
Sr. Member
****
Offline Offline

Posts: 288


« Reply #2 on: September 20, 2013, 04:25:22 PM »
ReplyReply

If those are red lines, I couldn't tell it from the enlarged version of the GX1 shot.  What I did particularly notice is that shot is heavily sharpened -- too much so.  Look at the large light halos around the letters in "Alan" on the spine.  At least we don't see that in the overall sharper image from the Sigma.  (I will grant that there is more color differentiation in the GX1 shot, but wouldn't say it was a lot.)  If you're looking for a reason why the Sigma may not pick up red as well, I understand that it has to do with the different depths to which the various light wavelengths have to penetrate the photosites on the Sigma chip.  That is not true of the Bayer array, where different photosites measure different colors.  Nevertheless, the Sigma innards still have to interpolate between/among the photosites to yield the multitude of colors we perceive (just as is also needed with the Bayer array) , so the particulars of the way that interpolation is done may be involved as well.  (I'm sure if I got any of this technology wrong, someone will correct me Smiley.)  --Barbara
Logged
BarbaraArmstrong
Sr. Member
****
Offline Offline

Posts: 288


« Reply #3 on: September 20, 2013, 04:28:39 PM »
ReplyReply

One more comment.  This is either a low-light shot, or somewhat under-exposed, not a strong point of the small Merrills.  I thoroughly enjoy my DP2M and DP3M, but recognize their limitations, and use them in more optimal situations.  --Barbara
Logged
degrub
Sr. Member
****
Offline Offline

Posts: 276


« Reply #4 on: September 20, 2013, 05:47:43 PM »
ReplyReply

What was the lighting?

Have tried the same experiment  with flash ?

Frank
Logged
xtopher
Newbie
*
Offline Offline

Posts: 4


« Reply #5 on: September 21, 2013, 04:39:11 AM »
ReplyReply

Interesting point — it was tungsten lighting —  a relatively long exposure — 1.3 seconds. It may be that the sensor is noisier with long exposures and that the colour is being removed by the chroma noise reduction in SPP.

I'll try again in what passes for daylight.

Chris
Logged
xtopher
Newbie
*
Offline Offline

Posts: 4


« Reply #6 on: September 23, 2013, 05:46:50 AM »
ReplyReply

OK, have done more tests, with improved results. The factors in the loss of colour detail seem to have been:
light temperature — re-photographing in daylight gives better red detail
exposure level — boosting exposure by 1 stop produces stronger red detail — 2 even more
chroma noise reduction — turning it off produces better detail in small areas of colour

It's quite striking how much saturation and detail is lost with underexposure.

On the other hand this is probably a price worth paying for the level of detail — not quite up to the level of the d800 on full frame, but close.

Chris
Logged
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad