Not scientific, very casual Quattro vs. Merrill test

Started Jul 11, 2014 | Discussions
Hng Contributing Member • Posts: 645
Re: Not scientific, very casual Quattro vs. Merrill test
3

Roland Karlsson wrote:

Hng wrote:

Easy to see, clouds with M is 3D object, with Q no more.

You are making life a little too easy.

Our brains are marvellous image processors. So - what we see is not what is there. When taking an image you get a simplification of reality. It is just a flat object on a screen or in a print.

In order to get the feeling of reality back, artists make lots of manipulation of the image.

So, the 3D feeling you get when looking at Merrill images might be a deliberate manipulation - an enhancement not existing in real life.

And Merril can do such trick to our brain, Q not

Roland Karlsson Forum Pro • Posts: 29,102
Re: Not scientific, very casual Quattro vs. Merrill test

Hng wrote:

And Merril can do such trick to our brain, Q not

What the Q (or SPP for Q) does is to enhance local contrast and to add noise.

 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
Hng Contributing Member • Posts: 645
Re: Not scientific, very casual Quattro vs. Merrill test
1

Roland Karlsson wrote:

Hng wrote:

And Merril can do such trick to our brain, Q not

What the Q (or SPP for Q) does is to enhance local contrast and to add noise.

From M to Q seems easy with simple painting action but Q back to M irreversibly I tried; you can go and show the way.

mypic Regular Member • Posts: 151
Re: Not scientific, very casual Quattro vs. Merrill test

Can you explained more about this special effect from the Merrills? Is this the false details some people talked about in the past. I'm just curious about it that's all.

 mypic's gear list:mypic's gear list
Sigma DP2 Merrill Sony a7R II Sony FE 35mm F2.8
Rudi Senior Member • Posts: 1,727
Re: Not scientific, very casual Quattro vs. Merrill test

Let´s wait and see what the final version of SPP6.x gets out of the RAWs. We´re still in some kind of beta release with SPP 6.03 software.

Some things are like good wine - becoming better by the time.

Roland Karlsson Forum Pro • Posts: 29,102
Re: Not scientific, very casual Quattro vs. Merrill test

Hng wrote:

From M to Q seems easy with simple painting action but Q back to M irreversibly I tried; you can go and show the way.

Painting? Painting is not photography.

 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: 29,102
Re: Not scientific, very casual Quattro vs. Merrill test

mypic wrote:

Can you explained more about this special effect from the Merrills? Is this the false details some people talked about in the past. I'm just curious about it that's all.

False or not false? It depends.

We have now seen lots of M vs Q comparisons. In all of them Q have been sharper, resolving more detail. But, also in all of them, there are clear details in the M that are more or less invisible in Q. It is details in low contrast areas.

Now, M is more noisy. So ... some of the detail might be noise? But ... there is undoubtedly also real detail that is more visible.

One way of explaining this is that Merrill do some local contrast enhancement.

Another way of explaining it is that Q simply smooths out detail.

To me it looks like the Q detail level looks more real. Might be wrong.

 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
JackM2 New Member • Posts: 13
Imaging-Resource seems to disagree?

Something's off here. Imaging-Resource's comparison tool:

http://www.imaging-resource.com/IMCOMP/COMPS01.HTM

shows the Quattro beating the pants off the SD1 and DP1M.  And a Canon 5D3 for that matter. They don't have a DP2M though. Hmmmm...

DMillier Forum Pro • Posts: 21,655
Re: Not scientific, very casual Quattro vs. Merrill test
1

Looking at my pictures, the M just seems to have incredibly high contrast that picks out the edges of things very clearly. But it seems unforced and easy:  you can't get the same effect by doing normal processing like adding grain, boosting global contrast or clarity or applying sharpening.

Subjectively, Merrill files respond to increasing contrast or extra sharpening really dramatically. For example, I export tiffs from SPP with sharpening set to -0.8.  In Lightroom if I add say 15 units of sharpening, the detail sharpens up like a Bayer file would if you applied a 100 units of sharpening but it does so without the halos of false look of excessive sharpening. It's remarkable, actually, especially if you want to emphasise the texture of surfaces.  Something like rough stone or plaster leaps off the print at you in a way I have not seen with any other camera. It's a bit like the easy, unforced sharpness and detail of large format film. It doesn't work so well with skin tones though, adding 50 years to your age.

I reckon it is a special effect and an always on one.

i can send you some files to play with if you want to explore this yourself.

-- hide signature --

"...while I am tempted to bludgeon you, I would rather have you come away with an improved understanding of how these sensors work" ---- Eric Fossum
Galleries and website: http://www.whisperingcat.co.uk/
Flickr: http://www.flickr.com/photos/davidmillier/

Filmophile Regular Member • Posts: 306
Re: Not scientific, very casual Quattro vs. Merrill test

Actually it occurs to me there's no need to do this at all - Raw Therape or Raw Developer wouldn't have implemented such a feature so a side by side comparison with a finely textured image in SPP should reveal any processing trick.   Raw Therapee seems to have been broken since I upgraded to OSX 10.9.

Roland Karlsson Forum Pro • Posts: 29,102
Re: Not scientific, very casual Quattro vs. Merrill test

DMillier wrote:

Looking at my pictures, the M just seems to have incredibly high contrast that picks out the edges of things very clearly. But it seems unforced and easy: you can't get the same effect by doing normal processing like adding grain, boosting global contrast or clarity or applying sharpening.

Subjectively, Merrill files respond to increasing contrast or extra sharpening really dramatically. For example, I export tiffs from SPP with sharpening set to -0.8. In Lightroom if I add say 15 units of sharpening, the detail sharpens up like a Bayer file would if you applied a 100 units of sharpening but it does so without the halos of false look of excessive sharpening. It's remarkable, actually, especially if you want to emphasise the texture of surfaces. Something like rough stone or plaster leaps off the print at you in a way I have not seen with any other camera. It's a bit like the easy, unforced sharpness and detail of large format film. It doesn't work so well with skin tones though, adding 50 years to your age.

I reckon it is a special effect and an always on one.

Strange

i can send you some files to play with if you want to explore this yourself.

Hmmmmm ... lets have that as an option. For some reason, my wife have planned both this and that and I seem to have just time to write here and no more. And I am trying to find time to upgrade the X3F Tools for Quattro.

 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
Lord metroid
Lord metroid Regular Member • Posts: 476
Re: Not scientific, very casual Quattro vs. Merrill test

Roland Karlsson wrote:

Hng wrote:

Easy to see, clouds with M is 3D object, with Q no more.

You are making life a little too easy.

Our brains are marvellous image processors. So - what we see is not what is there. When taking an image you get a simplification of reality. It is just a flat object on a screen or in a print.

In order to get the feeling of reality back, artists make lots of manipulation of the image.

So, the 3D feeling you get when looking at Merrill images might be a deliberate manipulation - an enhancement not existing in real life.

I always found unexposed images on my SIGMA DP2s to have a little bit of 3D magic. Perhaps it is the same for the SIGMA DP2 Quattro.

Could someone try to underexpose an image by -1?

 Lord metroid's gear list:Lord metroid's gear list
Sigma DP2s Sigma dp2 Quattro
MarkWW
MarkWW Contributing Member • Posts: 881
Re: Raw files for download .x3fs for download
2

Roland Karlsson wrote:

MarkWW wrote:

Like I said, any sharpening done after SPP can be done to any camera so there's no point to it for the purposes of a comparison.

Actually there is.

The claim is that Merrill has something other cameras lack. If this "something" can be added in post processing, then it is not as bad as if it cant.

By looking at the photos, I have no sense that it can be added back in post. It's as plain as day - one image has bricks, the other does not. No amount of sharpening in post will add in something that the sensor simply did not capture.

And yes, I'm saying that the Quattro sensor absolutely did not capture as much detail as the Merrill sensor. How a Bayer sensor would compare - I don't know, I'd need a direct comparison image.

Now, I'm not saying that this is either good or bad, just that it is. If you want the Merrill look, get the Merrill. If you want the Quattro look, get the Quattro. If you want the Polaroid look, get a Polaroid.

 MarkWW's gear list:MarkWW's gear list
Ricoh GR Digital III Panasonic Lumix DMC-LX5 Sigma DP2 Merrill Sigma DP3 Merrill Ricoh GR +8 more
MarkWW
MarkWW Contributing Member • Posts: 881
Re: Not scientific, very casual Quattro vs. Merrill test
2

Hng wrote:

Roland Karlsson wrote:

Hng wrote:

Easy to see, clouds with M is 3D object, with Q no more.

You are making life a little too easy.

Our brains are marvellous image processors. So - what we see is not what is there. When taking an image you get a simplification of reality. It is just a flat object on a screen or in a print.

In order to get the feeling of reality back, artists make lots of manipulation of the image.

So, the 3D feeling you get when looking at Merrill images might be a deliberate manipulation - an enhancement not existing in real life.

And Merril can do such trick to our brain, Q not

I agree with Hng. The Merrills have a "3d" pop - that I've repeatedly described as "each pixel seeming to have a height to it."

Quattros do not have that. Quattros have the opposite - a very smooth look. Everything is flattened and evened out. This isn't good or bad, it just is. Some will prefer one look, others will prefer the other.

I quite enjoy the Merrill look because it is SO unnatural - so much more than what the eye sees and what we're used to getting out of our cameras. Then it can be toned down to look more natural. It's excessively sharp without the ugly "unsharp mask" look, and with just a touch of unsharp mask, it's even sharper - without looking like it was all faked in photoshop, but looking - to my eyes - more like "reality enhanced". The Quattro could never get the Merrill look because it simply doesn't have all the data. The pixels don't have "height" anymore and there is no adding it back in.

 MarkWW's gear list:MarkWW's gear list
Ricoh GR Digital III Panasonic Lumix DMC-LX5 Sigma DP2 Merrill Sigma DP3 Merrill Ricoh GR +8 more
MarkWW
MarkWW Contributing Member • Posts: 881
Re: Not scientific, very casual Quattro vs. Merrill test

mypic wrote:

Can you explained more about this special effect from the Merrills? Is this the false details some people talked about in the past. I'm just curious about it that's all.

Scroll down about halfway and look at the photo of the model, the one that's closeup of the eyes.

http://www.markwiggin.com/b/Sigma_DP3_Merrill_Review

Look at the texture of the skin, look at the hairs on the nose. It's not "realistic" it's something more than realistic. The later Phase One image looks more realistic. The Phase One image is more like what skin looks like to me. Now it's easy enough to dial this back in post, but you can't fake that look if the detail isn't there in the first place.

 MarkWW's gear list:MarkWW's gear list
Ricoh GR Digital III Panasonic Lumix DMC-LX5 Sigma DP2 Merrill Sigma DP3 Merrill Ricoh GR +8 more
Hng Contributing Member • Posts: 645
Re: Not scientific, very casual Quattro vs. Merrill test

MarkWW wrote:

mypic wrote:

Can you explained more about this special effect from the Merrills? Is this the false details some people talked about in the past. I'm just curious about it that's all.

Scroll down about halfway and look at the photo of the model, the one that's closeup of the eyes.

http://www.markwiggin.com/b/Sigma_DP3_Merrill_Review

Look at the texture of the skin, look at the hairs on the nose. It's not "realistic" it's something more than realistic. The later Phase One image looks more realistic. The Phase One image is more like what skin looks like to me. Now it's easy enough to dial this back in post, but you can't fake that look if the detail isn't there in the first place.

With a bit mastering of ACR, push max Blacks then compensate Shadows, combined with Whites and Exposure, you can get more natural skin texture. Its beauty of ACR that SPP has no such extra controls besides Clarity for micro-contrast adjustment.

Hng Contributing Member • Posts: 645
Re: Not scientific, very casual Quattro vs. Merrill test

MarkWW wrote:

Hng wrote:

Roland Karlsson wrote:

Hng wrote:

Easy to see, clouds with M is 3D object, with Q no more.

You are making life a little too easy.

Our brains are marvellous image processors. So - what we see is not what is there. When taking an image you get a simplification of reality. It is just a flat object on a screen or in a print.

In order to get the feeling of reality back, artists make lots of manipulation of the image.

So, the 3D feeling you get when looking at Merrill images might be a deliberate manipulation - an enhancement not existing in real life.

And Merril can do such trick to our brain, Q not

I agree with Hng. The Merrills have a "3d" pop - that I've repeatedly described as "each pixel seeming to have a height to it."

Quattros do not have that. Quattros have the opposite - a very smooth look. Everything is flattened and evened out. This isn't good or bad, it just is. Some will prefer one look, others will prefer the other.

I quite enjoy the Merrill look because it is SO unnatural - so much more than what the eye sees and what we're used to getting out of our cameras. Then it can be toned down to look more natural. It's excessively sharp without the ugly "unsharp mask" look, and with just a touch of unsharp mask, it's even sharper - without looking like it was all faked in photoshop, but looking - to my eyes - more like "reality enhanced". The Quattro could never get the Merrill look because it simply doesn't have all the data. The pixels don't have "height" anymore and there is no adding it back in.

Feel like mix of multivariable pixel and noise "grains" that you call "height" then in ACR and later noise control, you can get deep sense and elegant images from M and sadly Q no more.

Tiger1 Contributing Member • Posts: 554
Re: Imaging-Resource seems to disagree?
1

The test is unfair. The SD1 had a 17-50mm lens on it which is nowhere near as good as the 30mm F2.8 lens. Even the DP1M lens doesn't compete.

On the other hand if you process the raw files you'll find the results a lot closer.  Even the Quattro's file is much better than the OOC JPEG.

 Tiger1's gear list:Tiger1's gear list
Sony RX1 Canon EOS 5D Mark II Fujifilm FinePix S5 Pro Sigma SD14 Sigma SD1 Merrill +14 more
MarkWW
MarkWW Contributing Member • Posts: 881
Re: Not scientific, very casual Quattro vs. Merrill test

Hng wrote:

MarkWW wrote:

mypic wrote:

Can you explained more about this special effect from the Merrills? Is this the false details some people talked about in the past. I'm just curious about it that's all.

Scroll down about halfway and look at the photo of the model, the one that's closeup of the eyes.

http://www.markwiggin.com/b/Sigma_DP3_Merrill_Review

Look at the texture of the skin, look at the hairs on the nose. It's not "realistic" it's something more than realistic. The later Phase One image looks more realistic. The Phase One image is more like what skin looks like to me. Now it's easy enough to dial this back in post, but you can't fake that look if the detail isn't there in the first place.

With a bit mastering of ACR, push max Blacks then compensate Shadows, combined with Whites and Exposure, you can get more natural skin texture. Its beauty of ACR that SPP has no such extra controls besides Clarity for micro-contrast adjustment.

I just import the TIFF into Photoshop and then use skin smoothing plugins etc. to pull back on the "too much detail."

I plan on (in the next few weeks) doing a shoot designed around the Merrill love of detail that I'm really looking forward to.

 MarkWW's gear list:MarkWW's gear list
Ricoh GR Digital III Panasonic Lumix DMC-LX5 Sigma DP2 Merrill Sigma DP3 Merrill Ricoh GR +8 more
Gesture Veteran Member • Posts: 7,383
Re: Not scientific, very casual Quattro vs. Merrill test

Good thoughts.  I expect Merrills to hold value and even become extra sought after.

Keyboard shortcuts:
FForum MMy threads