Cooked RAW: DP2M vs dp2Q

Started Jul 21, 2014 | Discussions
PrebenR Veteran Member • Posts: 4,164
Cooked RAW: DP2M vs dp2Q
5

See here as I could not upload the dp2Q file: https://www.flickr.com/photos/prebenr/sets/72157645396604688/

And even if you probably do not believe me I tried to make both images to look as best as I could and as close to the real scene. It was easiest with the quattro image though.

There still might be possible to do some colour tweeking, but I have spent 1.5h trying to upload the dp2q image to flickr before I found out that SPP 6.0.4 generates a corrupt jpg and I had to open it in gimp and save it again from there. So as it is 02:29 AM I do not have more time to do any changes now. Only 4 hours sleep before work

And sorry that the photo is not more interesting, but I only had time to take one from my balcony today. It should serve to show the potential of the cameras though.

-- hide signature --

Out lightwriting with Sigma
«Kaos»

Sigma DP2 Merrill Sigma dp2 Quattro
If you believe there are incorrect tags, please send us this post using our feedback form.
OP PrebenR Veteran Member • Posts: 4,164
Re: Cooked RAW: DP2M vs dp2Q

Oops. I just noticed that the dp2q images was set at private. I have changed it to public now. Sorry...

Ah, and just one more thing. The DP2M is still on greek time so it shows one hour more than it should. Forgot to change back. The photos where taken right after each other and the light did not change.

-- hide signature --

Out lightwriting with Sigma
«Kaos»

Truman Prevatt
Truman Prevatt Forum Pro • Posts: 11,329
Re: Cooked RAW: DP2M vs dp2Q

PrebenR wrote:

See here as I could not upload the dp2Q file: https://www.flickr.com/photos/prebenr/sets/72157645396604688/

And even if you probably do not believe me I tried to make both images to look as best as I could and as close to the real scene. It was easiest with the quattro image though.

There still might be possible to do some colour tweeking, but I have spent 1.5h trying to upload the dp2q image to flickr before I found out that SPP 6.0.4 generates a corrupt jpg and I had to open it in gimp and save it again from there. So as it is 02:29 AM I do not have more time to do any changes now. Only 4 hours sleep before work

And sorry that the photo is not more interesting, but I only had time to take one from my balcony today. It should serve to show the potential of the cameras though.

Observations. The DPQ is under exposed compared to the DPM - see house. The house in the DPM looks to be Zone 6 while in the DPM Zone 5.  Also the DPM has more contrast than the DPQ.  For some reason the DPQ has almost lost the cloud in the upper right hand corner while the DPM has it nicely.

Yes the cameras are different.  Which is better.  I prefer the DPQ image looks very flat - just look at the white house that is suppose to be sunlit. The shadows of the roof indicate that but the house is sure flat.

-- hide signature --

Truman
www.pbase.com/tprevatt

 Truman Prevatt's gear list:Truman Prevatt's gear list
Fujifilm X-H1 Fujifilm X-Pro3 Fujifilm XF 18mm F2 R Fujifilm XF 35mm F1.4 R Fujifilm XF 55-200mm F3.5-4.8 R LM OIS +10 more
OP PrebenR Veteran Member • Posts: 4,164
Re: Cooked RAW: DP2M vs dp2Q

Truman Prevatt wrote:

PrebenR wrote:

See here as I could not upload the dp2Q file: https://www.flickr.com/photos/prebenr/sets/72157645396604688/

And even if you probably do not believe me I tried to make both images to look as best as I could and as close to the real scene. It was easiest with the quattro image though.

There still might be possible to do some colour tweeking, but I have spent 1.5h trying to upload the dp2q image to flickr before I found out that SPP 6.0.4 generates a corrupt jpg and I had to open it in gimp and save it again from there. So as it is 02:29 AM I do not have more time to do any changes now. Only 4 hours sleep before work

And sorry that the photo is not more interesting, but I only had time to take one from my balcony today. It should serve to show the potential of the cameras though.

Observations. The DPQ is under exposed compared to the DPM - see house. The house in the DPM looks to be Zone 6 while in the DPM Zone 5. Also the DPM has more contrast than the DPQ. For some reason the DPQ has almost lost the cloud in the upper right hand corner while the DPM has it nicely.

Yes the cameras are different. Which is better. I prefer the DPQ image looks very flat - just look at the white house that is suppose to be sunlit. The shadows of the roof indicate that but the house is sure flat.

Both images were 1EV overexposed. The Quattro image has the sky as it was. The Merrill doesn't.

Do you mean the house in the far back? That is probably due to that highlight recovery gives artifacts in SPP6 so I had to leave it in Neutralise.

Have a look at the closer house and you'll see it pops more than the Merrill. Look at the trees at 100%.

I don't understand where one can see that the quattro image is more flat.

-- hide signature --

Truman
www.pbase.com/tprevatt

-- hide signature --

Out lightwriting with Sigma
«Kaos»

richard stone Veteran Member • Posts: 3,445
Re: Cooked RAW: DP2M vs dp2Q
3

I think the q is better. And different.

But does it matter that much? Even now it's angels on pinheads, and within the range of adjustments in exposure and PP.

What I have observed is that the q seems to have a little better "DR" leading to what looks like "reduced' contrast, and conversely, to a sort of harshness in the M. But everyone seems to like the harshness, when it fits with the image. Similarly, SPP is not yet optimized for the q.

Green sharpening and other SPP and PP techniques should get the q well past the M, in the end. And with a much more versatile camera (people pictures?) in the end.

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
OP PrebenR Veteran Member • Posts: 4,164
Re: Cooked RAW: DP2M vs dp2Q
1

richard stone wrote:

I think the q is better. And different.

But does it matter that much? Even now it's angels on pinheads, and within the range of adjustments in exposure and PP.

What I have observed is that the q seems to have a little better "DR" leading to what looks like "reduced' contrast, and conversely, to a sort of harshness in the M. But everyone seems to like the harshness, when it fits with the image. Similarly, SPP is not yet optimized for the q.

Green sharpening and other SPP and PP techniques should get the q well past the M, in the end. And with a much more versatile camera (people pictures?) in the end.

Yes, but in my eyes the q already shows it can have better resolution in greens etc. At least I find the trees are clearer than in the M. No need to do sharpening in other software.

-- hide signature --

Out lightwriting with Sigma
«Kaos»

maple Veteran Member • Posts: 3,349
Re: Cooked RAW: DP2M vs dp2Q
2

Thank you very much for these two images, Preben. They are excellent images and good examples of what each camera is capable of. They reconfirm: 1) Q is just as capable of sharp and fine details as M; 2) They are different, Q’s more subtle and subjectively more truthful, while M has character that some insist to be the “standard” of Foveon. 3) The difference in fine details is insignificant.

At first I actually found M to be better between these two, then realized I had confused Q for M. Q looks more natural, and the 30% more pixels just manages to makes a visible difference.

-- hide signature --

Maple

jennyrae Senior Member • Posts: 2,690
Re: Cooked RAW: DP2M vs dp2Q

thank you. I like the DP2Q saturated neutral look here. contrast difference seem negligible and possibly more headroom with the Quattro. it's possible future iteration of SPP would make Quattro files much better result and handling.  do not kow yet how it would handle some of those green blotch in future but so far, result here is very good and seem to manage with exposure set.

dr.noise
dr.noise Veteran Member • Posts: 3,749
Re: Cooked RAW: DP2M vs dp2Q

Awesome work. Thank you very much!
At last I see the pleasing Q pic, without obvious grain and rope-thick sharpening halos, no strange uneven pixel artefacts, at last all details are looking natural and precise - Foveon way!

 dr.noise's gear list:dr.noise's gear list
Sigma DP2s Sony DSC-RX0 Sigma SD10 Sigma fp
MOD rick decker Forum Pro • Posts: 17,119
Re: Cooked RAW: DP2M vs dp2Q
1

Very easy to pick out the Q - cooler greens, bluer skies and less-saturated red on the one house. The M tends to go too warm in overcast and you can see it in this image. Blues got pushed out of the way by the warmer tones.  Yes clouds are better with the M - the trees are there but the forest is worse.  In addition since  there are warmer tones with the M, I am guessing that the color is off in the house on the right.  Can you confirm or deny that?

OP PrebenR Veteran Member • Posts: 4,164
Re: Cooked RAW: DP2M vs dp2Q

rick decker wrote:

Very easy to pick out the Q - cooler greens, bluer skies and less-saturated red on the one house. The M tends to go too warm in overcast and you can see it in this image. Blues got pushed out of the way by the warmer tones. Yes clouds are better with the M - the trees are there but the forest is worse. In addition since there are warmer tones with the M, I am guessing that the color is off in the house on the right. Can you confirm or deny that?

As I mentioned the colours could probably have been tweeked further for both images. The quattro a bit less on the blue side and the merrill less warm. I had some strange thing with SPP6 in that I thought I had gotten the colour correct and then when I opened the jpeg it wasn't the same. Not sure if it is a problem on my system. I may revisit the colours later today, but as it was in the middle of the night and I struggled with uploads I gave up on adjusting them.

What I did in the end was to pick the same spot to adjust the colour balance in both images.

As in the landscape images on the blog linked in a couple of other threads, one can see the same seperation. Merrill goes warm and Quattro is more controlled.

My main focus was not on the colours, but to process the images to see how both cameras handles resolution and foilage. For both images I adjusted that towards the scene.

As to the house colour the quattro is very close to the correct colour.

-- hide signature --

Out lightwriting with Sigma
«Kaos»

(unknown member) Senior Member • Posts: 1,857
Re: Cooked RAW: DP2M vs dp2Q

Thanks for the comparison.
The foilage on the quattro looks surprisingly good. But these are also perfect, clear conditions with strong sunlight from the right that gives lots of contrast.
I prefer the clouds from the Merrill. The quattro has blown out parts and look flatter.
Also the yellow house on the far right is blown out with the quattro.
But all in all, both images look good with slighty more resolution for the quattro.

PrebenR wrote:

That is probably due to that highlight recovery gives artifacts in SPP6 so I had to leave it in Neutralise.

Actually "Restore" is the normal position and this is what the sensor captured, and "Neutralize" is trying to make a softer gradation to harsh clipping. So "Neutralize" is a post processing effect.

Overexposure-Correction is also a post effect, trying to fill blown out parts with color.

Johan Borg Veteran Member • Posts: 3,005
Re: Cooked RAW: DP2M vs dp2Q

maceoQ wrote:

Thanks for the comparison.
The foilage on the quattro looks surprisingly good. But these are also perfect, clear conditions with strong sunlight from the right that gives lots of contrast.
I prefer the clouds from the Merrill. The quattro has blown out parts and look flatter.
Also the yellow house on the far right is blown out with the quattro.
But all in all, both images look good with slighty more resolution for the quattro.

PrebenR wrote:

That is probably due to that highlight recovery gives artifacts in SPP6 so I had to leave it in Neutralise.

Actually "Restore" is the normal position and this is what the sensor captured, and "Neutralize" is trying to make a softer gradation to harsh clipping. So "Neutralize" is a post processing effect.

I think both are a post processing to some extent, that the Restore function is like Overexposure-Correction in that it tries to guess the right color even if a channel or two are blown.

Overexposure-Correction is also a post effect, trying to fill blown out parts with color.

Except that it may be less of a post effect than Restore since we have reason to believe that Overexposure Correction is based on actual color readings from darker pixels in the RAW file (darker by themselves or by different processing in an analog front end).

OP PrebenR Veteran Member • Posts: 4,164
Re: Cooked RAW: DP2M vs dp2Q

maceoQ wrote:

Thanks for the comparison.
The foilage on the quattro looks surprisingly good. But these are also perfect, clear conditions with strong sunlight from the right that gives lots of contrast.
I prefer the clouds from the Merrill. The quattro has blown out parts and look flatter.
Also the yellow house on the far right is blown out with the quattro.
But all in all, both images look good with slighty more resolution for the quattro.

OK I'll try some non perfect conditions. Quattro meters differently than Merrill. I forced them to be the same because otherwise people would complain about that.

PrebenR wrote:

That is probably due to that highlight recovery gives artifacts in SPP6 so I had to leave it in Neutralise.

Actually "Restore" is the normal position and this is what the sensor captured, and "Neutralize" is trying to make a softer gradation to harsh clipping. So "Neutralize" is a post processing effect.

No in SPP 6.0.4 Neutralise is the default.

Overexposure-Correction is also a post effect, trying to fill blown out parts with color.

-- hide signature --

Out lightwriting with Sigma
«Kaos»

(unknown member) Senior Member • Posts: 1,857
Re: Cooked RAW: DP2M vs dp2Q

PrebenR wrote:

No in SPP 6.0.4 Neutralise is the default.

Interesting, this is definitely not the case on my computer (X3F position in SPP 6.0.4 and 5.5.3)

Truman Prevatt
Truman Prevatt Forum Pro • Posts: 11,329
Re: Cooked RAW: DP2M vs dp2Q

PrebenR wrote:

Truman Prevatt wrote:

PrebenR wrote:

See here as I could not upload the dp2Q file: https://www.flickr.com/photos/prebenr/sets/72157645396604688/

And even if you probably do not believe me I tried to make both images to look as best as I could and as close to the real scene. It was easiest with the quattro image though.

There still might be possible to do some colour tweeking, but I have spent 1.5h trying to upload the dp2q image to flickr before I found out that SPP 6.0.4 generates a corrupt jpg and I had to open it in gimp and save it again from there. So as it is 02:29 AM I do not have more time to do any changes now. Only 4 hours sleep before work

And sorry that the photo is not more interesting, but I only had time to take one from my balcony today. It should serve to show the potential of the cameras though.

Observations. The DPQ is under exposed compared to the DPM - see house. The house in the DPM looks to be Zone 6 while in the DPM Zone 5. Also the DPM has more contrast than the DPQ. For some reason the DPQ has almost lost the cloud in the upper right hand corner while the DPM has it nicely.

Yes the cameras are different. Which is better. I prefer the DPQ image looks very flat - just look at the white house that is suppose to be sunlit. The shadows of the roof indicate that but the house is sure flat.

Both images were 1EV overexposed. The Quattro image has the sky as it was. The Merrill doesn't.

Do you mean the house in the far back? That is probably due to that highlight recovery gives artifacts in SPP6 so I had to leave it in Neutralise.

The light colored house in the right foreground of the scene.  It looks light tan in the Q image and almost a light rose in the M image.  It is clear that the sun was coming from the right side slightly behind you by the roof shadows.  The M image has a lot more pop on that house. The Q images of the house is somewhat flat with little local contrast.

Have a look at the closer house and you'll see it pops more than the Merrill. Look at the trees at 100%.

I don't understand where one can see that the quattro image is more flat.

-- hide signature --

Truman
www.pbase.com/tprevatt

-- hide signature --

Out lightwriting with Sigma
«Kaos»

-- hide signature --

Truman
www.pbase.com/tprevatt

 Truman Prevatt's gear list:Truman Prevatt's gear list
Fujifilm X-H1 Fujifilm X-Pro3 Fujifilm XF 18mm F2 R Fujifilm XF 35mm F1.4 R Fujifilm XF 55-200mm F3.5-4.8 R LM OIS +10 more
Truman Prevatt
Truman Prevatt Forum Pro • Posts: 11,329
Re: Cooked RAW: DP2M vs dp2Q

PrebenR wrote:

Truman Prevatt wrote:

PrebenR wrote:

See here as I could not upload the dp2Q file: https://www.flickr.com/photos/prebenr/sets/72157645396604688/

And even if you probably do not believe me I tried to make both images to look as best as I could and as close to the real scene. It was easiest with the quattro image though.

There still might be possible to do some colour tweeking, but I have spent 1.5h trying to upload the dp2q image to flickr before I found out that SPP 6.0.4 generates a corrupt jpg and I had to open it in gimp and save it again from there. So as it is 02:29 AM I do not have more time to do any changes now. Only 4 hours sleep before work

And sorry that the photo is not more interesting, but I only had time to take one from my balcony today. It should serve to show the potential of the cameras though.

Observations. The DPQ is under exposed compared to the DPM - see house. The house in the DPM looks to be Zone 6 while in the DPM Zone 5. Also the DPM has more contrast than the DPQ. For some reason the DPQ has almost lost the cloud in the upper right hand corner while the DPM has it nicely.

Yes the cameras are different. Which is better. I prefer the DPQ image looks very flat - just look at the white house that is suppose to be sunlit. The shadows of the roof indicate that but the house is sure flat.

Both images were 1EV overexposed. The Quattro image has the sky as it was. The Merrill doesn't.

Skies are always a challenge. It is very difficult to get the color of the sky to exactly how one perceives it for several reasons.  One is water vapor or other components of haze in the atmosphere that scatter light and lighten the sky.  Another is sensitivity to UV.  Our eyes tend to process that out since we know the color we want - the sky is blue.  The camera doesn't know the sky is blue.  In B&W film filters were used to deal with the issue - from light yellow through red.  In color film, most people spent a lot of time with color balancing filters in the dark room to "darken the sky."  Skies are tough.

Digital is no different.  The same phenomena exist.  However, the tools today are a little easier in many cases to work with.  It is not just the sky that is darker on the Q image.  The light tan house in the right center is darker on the Q image as is the red house. The red is not as pronounce as the tan house. Although you took the pictures at the same EV setting - that doesn't mean the base ISO is the same of the two sensors.  ISO standard 12232:2006 gives camera makers five different methods to the determine the base ISO.  The ISO is also measured in the sRGB color space by definition of the standard. Here is a decent explanation of how the ISO standard is applied to digital cameras. In my teaching days the first thing every student did was to measure their camera/film and if B&W development and processing true ISO.  It often differed from that of the film manufacturer and it differed from film to film. However, to get consistent results you need to know that and compensate for it.

A camera true ISO test is an easy test using a white sheet on an outside wall.  It might even be the case that the base ISO for one or the other or both is slightly different for different colors.  That is quite common for film and for digital.

In the foliage I would expect that there is sufficient aliasing to make any comparison moot.

Do you mean the house in the far back? That is probably due to that highlight recovery gives artifacts in SPP6 so I had to leave it in Neutralise.

Have a look at the closer house and you'll see it pops more than the Merrill. Look at the trees at 100%.

I don't understand where one can see that the quattro image is more flat.

-- hide signature --

Truman
www.pbase.com/tprevatt

-- hide signature --

Out lightwriting with Sigma
«Kaos»

-- hide signature --

Truman
www.pbase.com/tprevatt

 Truman Prevatt's gear list:Truman Prevatt's gear list
Fujifilm X-H1 Fujifilm X-Pro3 Fujifilm XF 18mm F2 R Fujifilm XF 35mm F1.4 R Fujifilm XF 55-200mm F3.5-4.8 R LM OIS +10 more
OP PrebenR Veteran Member • Posts: 4,164
Re: Cooked RAW: DP2M vs dp2Q

maceoQ wrote:

PrebenR wrote:

No in SPP 6.0.4 Neutralise is the default.

Interesting, this is definitely not the case on my computer (X3F position in SPP 6.0.4 and 5.5.3)

In 5.5.3 Restore is default, but Neutralise was default in 6.0.4 on my computer. Strange

-- hide signature --

Out lightwriting with Sigma
«Kaos»

OP PrebenR Veteran Member • Posts: 4,164
Re: Cooked RAW: DP2M vs dp2Q

Truman Prevatt wrote:

The light colored house in the right foreground of the scene. It looks light tan in the Q image and almost a light rose in the M image. It is clear that the sun was coming from the right side slightly behind you by the roof shadows. The M image has a lot more pop on that house. The Q images of the house is somewhat flat with little local contrast.

I don't agree at all. There is less gritt in Q yes, but the colour and look is more natural in the Q image.

-- hide signature --

Out lightwriting with Sigma
«Kaos»

OP PrebenR Veteran Member • Posts: 4,164
Two more samples in flat light

Added two with flat light, no overexposure. I took the Merrill shot fast after the Quattro and I had problems viewing the Merrill screen so it didn't become straight (I really like the Q screen and the level).

https://www.flickr.com/photos/prebenr/sets/72157645396604688/

-- hide signature --

Out lightwriting with Sigma
«Kaos»

Keyboard shortcuts:
FForum MMy threads