DP3M photos

Started Jan 29, 2013 | Discussions
Roland Karlsson Forum Pro • Posts: 28,156
Re: Please

Please, be civil.
No need for name calling.

RonJG's comment was quite correct, and without any hatred in it at all.

The coat of the lady is full of blotches and banding.
It was good information, and an information several here also have given.

You only have to try to decide what this means for you.
Given that almost all other are free from it, the image is probably treated differently.

-- hide signature --
 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
Roland Karlsson Forum Pro • Posts: 28,156
Re: On Banding and Noise
1

maceoQ wrote:

Nevertheless...i also see faint banding in the original picture of the two ladies, and my monitor is calibrated. (Eizo)
The picture itself is anyway terrible, wrong white balance, ugly skin tones, photographed trough a window..

The picture Laurence was talking about was the image with a lady and a green bag. And I agree ther, very hard to see any problems in the original.

Of course, the problems are easily visible in the image with the two ladies, both banding and splotches. Nemas problemas.

But ... we dont know what the photographer has made with the image. Maybe it was slightly under exposed and then the darker parts compensated for? Maybe even fill light. We dont know that.

-- hide signature --
 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
Roland Karlsson Forum Pro • Posts: 28,156
Re: This discussion has become evidence for Roger's Law of New Product Intros

BarrytheB wrote:

Best post in this thread:)

Yeah - funny. But --- as most funny stuff ... not true at all.

  • Real fans think its the best thing since sliced bread. They are up at the top constantly.
  • Conservative fans think SD15, or SD10 or even SD9 are superior.
  • Some people cant decide, like me .. they swing up and down.
  • Some people sees problems, but still use the cameras. They probably own a DP2 or something, and say its fantastic all the time when its not a disaster.
  • Some people think everything Sigma/Foveon sux.

I cant really match this scenario to the graph

-- hide signature --
 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
(unknown member) Senior Member • Posts: 1,857
Re: On Banding and Noise

The picture Laurence was talking about was the image with a lady and a green bag. And I agree ther, very hard to see any problems in the original.

yes, two ladies, one with a green bag.

Of course, the problems are easily visible in the image with the two ladies, both banding and splotches. Nemas problemas.

But ... we dont know what the photographer has made with the image. Maybe it was slightly under exposed and then the darker parts compensated for? Maybe even fill light. We dont know that.

we know it, it looks like this:

White Balance: Overcast

Exposure Compensation : +0.2
Contrast : +0.0
Shadow : +0.0
Highlight : +0.0
Saturation : +0.0
Sharpness : +0.0
X3 Fill Light : +0.0
Chrominance Noise Reduction : +0.250
Luminance Noise Reduction : +0.750

Maceo

Roland Karlsson Forum Pro • Posts: 28,156
Re: Banding, blotches and skin tones?
2

So -- many here say the images shows banding, blotches and yellow skin tones.

Yes, there are two images with banding. The two ladies is evident. The dog in the window needs some amplification to get bad.

Yes, there is one image with bad blotches. The two ladies.

Yes, the two ladies does not have attractive skin tones. But (unfortunately guys) there is not much skin to look upon in those images.

So - I would say that the critique is slightly exaggerated. What do you think?

Maybe I have missed the problems in other images. Anyone?

-- hide signature --
 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
Raist3d Forum Pro • Posts: 38,652
Re: Banding, blotches and skin tones?

http://forums.dpreview.com/forums/post/50769986

ill e checking some of the other shots later but that was a start last night.

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - George Orwell

Raist3d Forum Pro • Posts: 38,652
Best reply I to the...

so called best post in this thread

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - George Orwell

Roland Karlsson Forum Pro • Posts: 28,156
Re: On Banding and Noise

maceoQ wrote:

yes, two ladies, one with a green bag.

Oops --- you are right. I mixed things up.

I had a downloaded copy of the ladies that easily showed the banding. Hmmmm ... must have edited that one ...

Because ... now when I look at the ladies .... then the banding is not so obvious. Or have Sigma replaced the image?

Never mind - the image with the green bag and the two ladies was the same as Laurence showed and was talking about. And he showed the original and the manipulated. The manipulated with very evident banding, and the original without.

The original I can see now on the page has banding that can be seen on my calibrated monitor, but its quite faint. I would not have thought about it if no one have told me to look.

The color splotches are more evident. They are easy to see above the ladies arm.

...........

Regarding the EXIF meta data you showed with lots of zeroes.

That does not say so much. We dont know what editing really have been done. There might have been post processing in some other editor, and the new camera/SPP might not dump correct values yet.

-- hide signature --
 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
mike earussi Veteran Member • Posts: 8,125
Re: Banding, blotches and skin tones?

Roland Karlsson wrote:

So -- many here say the images shows banding, blotches and yellow skin tones.

Yes, there are two images with banding. The two ladies is evident. The dog in the window needs some amplification to get bad.

Yes, there is one image with bad blotches. The two ladies.

Yes, the two ladies does not have attractive skin tones. But (unfortunately guys) there is not much skin to look upon in those images.

So - I would say that the critique is slightly exaggerated. What do you think?

Maybe I have missed the problems in other images. Anyone?

For those who may have not noticed, the shot of the two ladies with yellow skin tones are shot a with a WB setting of "overcast" which tends to artifically add yellow to the scene to compensate for the blue of overcast light. This doesn't tell you that the DP3 can't do good skin tones, only that the WB settings on the camera need to be improved. I suggest reserving judgment until some portraits taken under regular 5500K lighting are shown.

(unknown member) Senior Member • Posts: 1,857
Re: Banding, blotches and skin tones?

This doesn't tell you that the DP3 can't do good skin tones, only that the WB settings on the camera need to be improved.

You're right:

Roland Karlsson Forum Pro • Posts: 28,156
Re: Banding, blotches and skin tones?

Nice - what did you do?

-- hide signature --
 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
(unknown member) Senior Member • Posts: 1,857
Re: On Banding and Noise
1

Roland Karlsson wrote:

Regarding the EXIF meta data you showed with lots of zeroes.

That does not say so much. We dont know what editing really have been done. There might have been post processing in some other editor, and the new camera/SPP might not dump correct values yet.

Yes, you're right. We don't know for sure. But actually, i think the EXIFs are correct.
Usually if another editor was used after SPP, "Exiftool" loses the information like "fill-light" etc.

Like it happened in the first picture:
Software                        : Adobe Photoshop CS3 Macintosh

...and no more specific SPP informations.
When i was opening the picture of the two ladies, i already thought that there must be too much luminace noise reduction applied. 
And Exiftool shows it:

Chrominance Noise Reduction : +0.250 (this is one step below "middle" position)
Luminance Noise Reduction : +0.750 (this is one step above the "middle" position and 3 steps above "zero")
This destroys details and increases "posterization".

(sorry, i guess the topic is already too technical for my english..)

Maceo

(unknown member) Senior Member • Posts: 1,857
Re: Banding, blotches and skin tones?

Roland Karlsson wrote:

Nice - what did you do

Well...i changed the white balance to my taste, and then made a small correction in the reds and yellows in photoshop. 
User "cinefeel" (where is he?) was showing it once:

http://forums.dpreview.com/forums/thread/3176779

Looks complicated first, but its done in less than a minute..

Maceo

Raist3d Forum Pro • Posts: 38,652
Another one I like

http://www.sigma-dp.com/DP3Merrill/samplephoto/SDIM1205.jpg

-- hide signature --

Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - George Orwell

richard stone Veteran Member • Posts: 3,068
The usual suspects
3

As usual, the same people show up to claim that the "color blotches" are a big problem. Between those and the banding people and the responses about two thirds of the thread is taken up with the photo equivalent of a food fight. And is about as significant.

Personally I thought the images posted were dreary, although with very even lighting, probably due to overcast skys, which in turn leads to underexposure and, with underexposure and the lighting, to odd and not so pleasing colors. They were amazingly sharp and crisp images. I was looking for a portrait, too, because 50mm would be (in theory) a more useful lens focal length in that regard than a wide or normal lens length. Was the city depopulated that day? Mysterious. The lens looks superb. As we would expect. The sensor is (in theory?) the same as on the other cameras in the series.

However, given that it was an overcast day most of the colors look excellent considering that the images "reflect" the colors from the sky as the light source. Images of people on an overcast day would likely have the same color issues.

But all of this is more to question why Sigma chose those images to illustrate the use of the camera than anything else. The images show a very sharp 50mm lens attached to a sensor that is about the equivalent of  30MP Bayer sensor. So at this point we have a wide array of digital camera reviewers and users who consider the DPM cameras to be superb, and yet, on this site, in this thread, we have our own usual small but determined group of doubters posting that the images and cameras are no good.

Hmmm. What to think? Who to believe?

Second, banding. This issue was raised by the same people in the DP2M initial images, and yet.... Banding is not a DP2M issue now. The DP3M will undoubtedly use the same processing as the DP2M, which now has minimal (unobservable) banding.

So the posts about banding are what? Another chance for some posters to prove how observant they are? That they would buy the DP3M if only it were perfect? Another chance to show that M. Reichman et al are wrong about who a DPM camera is for?

There is no doubt that the Bayer sensor cameras are probably easier to work with and more developed in terms of the issues presented in this thread. It may be easier to get the colors right (in the sense that the colors appeal to what people think they saw and want to see) and "banding" may indicate that the manufacturing and processing of the sensor/image is more difficult for a small firm like Sigma/Foveon, rather than indicating that the final results will not be superb.

As for skin tones, I take H. Bowman seriously, but I do not see the issue quite the way he does. Skin tones are difficult, but when the DPM cameras produce superb colors from everything else, what are we to think? Since the beginning the Sigma skin tones have been an issue, not because they are so inaccurate, but because they are very sensitive to how the scene is illuminated. And very sharp. Just like the rest of the image.

Richard

-- hide signature --
 richard stone's gear list:richard stone's gear list
Sigma SD10 Sigma sd Quattro Sigma 17-50mm F2.8 EX DC OS HSM Sigma 30mm F1.4 DC HSM Art
Roland Karlsson Forum Pro • Posts: 28,156
Re: The usual suspects
1

Personally I think those images looks rather good, or rather very good.

Yeah ... there are some problems if you look closely. But ... you have to really search for it.

And except for the image with the two ladies and the green bag, I cant find any major problems.

Some here is clearly exaggerating IMHO.

Usual suspects? I dont really know. Dont think so though.

-- hide signature --
 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
Veijo Vilva
Veijo Vilva Regular Member • Posts: 393
Re: Banding, blotches and skin tones?
1

Roland Karlsson wrote:

So -- many here say the images shows banding, blotches and yellow skin tones.

Yes, there are two images with banding. The two ladies is evident. The dog in the window needs some amplification to get bad.

Yes, there is one image with bad blotches. The two ladies.

Yes, the two ladies does not have attractive skin tones. But (unfortunately guys) there is not much skin to look upon in those images.

So - I would say that the critique is slightly exaggerated. What do you think?

Maybe I have missed the problems in other images. Anyone?

How about the bottom of the second photo in the series, i.e., the wall and the floor? The banding is more evident here than on the photo with the ladies. Whether it would be visible on a decent size print is another matter.

-- hide signature --

Veijo

 Veijo Vilva's gear list:Veijo Vilva's gear list
Sigma DP2s Sigma DP2 Merrill Sony RX100 Sigma DP3 Merrill Nikon 1 V1 +7 more
Roland Karlsson Forum Pro • Posts: 28,156
Re: Banding, blotches and skin tones?

Veijo Vilva wrote:

How about the bottom of the second photo in the series, i.e., the wall and the floor? The banding is more evident here than on the photo with the ladies. Whether it would be visible on a decent size print is another matter.

You are absolutely right. That is the most evident example so far. Missed it - thought the image was too boring

Hmmmmm ... I would really, really, really want to have the RAW to that one.

If you look at 200%, then you see that the stripes are broken, like if there was some noise reduction applied. It would be very nice to know if the RAW stripes are just one pixel stripes. Because, if they are one pixel stripes, then it would be very easy to write a program that can remove them before doing RAW->RGB. But of they really are broken like that ... thats another story.

-- hide signature --
 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
3dreal Senior Member • Posts: 2,271
Skin-Calibration

Fotowand has a skin-calibration-card. for portraits, nude, fashion. Have used other cards of him. Its a small one man-company.

-- hide signature --

www.stereopan.org

JL Salvignol
JL Salvignol Senior Member • Posts: 1,663
Voltaire in Candide - After Gottfried Wilhelm von Leibniz...

'All's for the best, in the best of all possible worlds'

-- hide signature --

JLS

 JL Salvignol's gear list:JL Salvignol's gear list
Sigma DP2 Merrill Nikon Coolpix A Nikon 1 V1 Nikon 1 J1 Pentax K-01 +16 more
Keyboard shortcuts:
FForum MMy threads