The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

New Sigma DP Merrill's?

The Ute

Well-known member
Huyless,

From what I can tell it's actually a VERY clever approach to dealing with several issues that have confounded this technology. As I discovered fiddling with SPP monochrome files by using SPP's 'color wheel' to select only top layer information, the files were surprisingly clean and sharp at ISO's as high as ISO 4000. Sigma apparently concluded that it was more important to retain luminance cleanliness and compromise color (R & G) just a tad (and that may turn out to be not a tad at all). This way the Foveon sensor with 19.6 million photo sites on the top layer provide nearly 80% of the image information (and blue channel) and relegating the lower G and R channels for providing ONLY color information and no luminance.

By letting the top layer essentially 'draw' the image while providing the traditional 'blue' channel information and letting the lower 'vertically aligned' layers stacked directly below it to supply the ONLY the corresponding Red and Green information I can see how this might result in some impressive files never thought achievable before with a Foveon sensor at higher ISOs. Black and white, like with a Leica Monochrom, should remain impressive if not even moreso with the increase in surface resolution/pixel count.

In other areas, while many may be aghast with the, shall we say, 'unconventional' product design, I can see the ergonomic benefits (think Dvorak keyboards for computers). Again, only hands on experience will tell but I suspect Sigma's penchant for minimalist engineering will bear them out. Plus, in their own unique way (as always), it appears Sigma may have also addressed the underpowered battery issue by allowing for (1) a larger unit and (2) with the new sensor and processing approach that needs far less image information to process files that are smaller than previously, less power is needed per shot--another battery capacity extending improvement.

Now if only the AF would be a tad swifter in lower light. Progress...progress...

P
Peter-

Kudos to you for having discovered that on your own.

Any concern on the effect the new sensor will have on color rendition ?
 

Quentin_Bargate

Well-known member
I'm unbelievably excited by these new cameras. The original DP2M changed my perception of compact cameras in general; the new ones, if they improve on the old, will simply take me further down a very happy and productive path.

Well done Sigma.
 

The Ute

Well-known member
Don't worry, Sigma's recent track record is pretty good. There is no point in releasing a new product inferior to the old. I hope ... :thumbup:
Your work is wonderful.

We all hope for better.

One nitpick would have been the option for an EVF.

I use a Hoodman. It's adequate but a bit clunky.
 

peterb

Member
Peter-

Kudos to you for having discovered that on your own.

Any concern on the effect the new sensor will have on color rendition ?
Thanks. It was one of those "I wonder what would happen if.." type experiments that surprised me when I noticed the color wheel when the version with SPP with monochrome was issued .

But you raise a good question regarding color rendition. My guess is on the one hand the new sensor will at least be as accurate as pre-DP Merrill cameras since the pixel dimensions for the red and green channels and the topmost blue layer will be sampling equivalently larger spaces (the quad part) for that information as the previous cameras.

But...on the other hand I also wonder whether there may be any inaccuracies from color values determined over a somewhat larger, less defined area from each of the 4.9 pixels per layer (and its equivalent on the top most layer) that will be superimposed or interpolated with the sharper, higher acutance image created by 19.6 million pixels deriving luminance information. Again, for the most part it may not be any worse than what a Bayer approach already does (regarding color) but in the some critical situations there may be some displeasure with the results.

Again only time and testing will tell.
 

The Ute

Well-known member
Thanks. It was one of those "I wonder what would happen if.." type experiments that surprised me when I noticed the color wheel when the version with SPP with monochrome was issued .

But you raise a good question regarding color rendition. My guess is on the one hand the new sensor will at least be as accurate as pre-DP Merrill cameras since the pixel dimensions for the red and green channels and the topmost blue layer will be sampling equivalently larger spaces (the quad part) for that information as the previous cameras.

But...on the other hand I also wonder whether there may be any inaccuracies from color values determined over a somewhat larger, less defined area from each of the 4.9 pixels per layer (and its equivalent on the top most layer) that will be superimposed or interpolated with the sharper, higher acutance image created by 19.6 million pixels deriving luminance information. Again, for the most part it may not be any worse than what a Bayer approach already does (regarding color) but in the some critical situations there may be some displeasure with the results.

Again only time and testing will tell.

I think you posed an interest way of analyzing things.

By comparing the color portion to the original DP's.

We shall see.

Since they are attempting to fix many of the common complaints: Noise at higher ISO's, processing speed, Battery life, and handling I wish they would have dealt with the lack of a viewfinder too.

Guess you can't have everything. At least not yet.

;)
 

Charles2

Active member
The sensor chip does not appear to embody any improved electronics nor semiconductor fabrication. It trades away color resolution for less noise, a design choice that you may or may not like when you see results. Foveon technology has stalled or reached its end.
 

peterb

Member
I dunno, Charles. I'd say the technology may be more evolving or being tweaked to achieve a better result within a particular approach to sensor design no different than say Fuji achieved better results over existing Bayer configured sensors by reshuffling the array of RGB photo sites from a regular 4x4 pattern with 1R, 1B and 2G photo sites per quad to a larger area with more random photo sites. That simple tweak allowed them to have a Bayer-like sensor that no longer needed an AA filter as a fudge factor in interpolating the results. And they even improved high ISO results as well.
 

The Ute

Well-known member
The sensor chip does not appear to embody any improved electronics nor semiconductor fabrication. It trades away color resolution for less noise, a design choice that you may or may not like when you see results. Foveon technology has stalled or reached its end.
I think that's premature but you are entitled to your opinion.

;)
 

The Ute

Well-known member
Another thing that has been taken care of is that the new Quattro comes w a USB cable release.

;)
 

raist3d

Well-known member
The thing is that the GREEN and RED layers have much bigger pixels, 4 pixels binned in one. That is done for the CAPTION part.

The signal after processing is pretty simple, without any interpolations as some said: This is division of the GREEN and RED layers in a certain %age, after the caption.
You actually have to interpolate clor now if you want to keep 19.6 megapixel spatial color resolution. So its not a Bayer interpolation but a different kind of de mosaic interpolation. It should still be easier / better than Bayer but may introduce new artifacts to deal with unique to this sensor.

In that sense its no longer a "true" Fiveon but more like somewhere in the middle of Foveon and Bayer.

The good news is that since the Merrill sensor was so noisy it could be argued that beore you never quite got full resolution in the "red" layer, at least once you hit iso 200 and up.

You do have to interpolate color here to get full spatial color resolution. Note how the full amount of photo detectors is less than the previous foveon (but most likely still an overall better result)

Other benefits: less circuitry in the green and red layer means less interference, less data - faster processing, less noise.

Compared to an ideal noiseless x3 Merrill foveon this is a step back, but its the reality of implementing a current Merrill under current technology what will make this approach a step forward.
 

Tim

Active member
I'm unbelievably excited by these new cameras. The original DP2M changed my perception of compact cameras in general; the new ones, if they improve on the old, will simply take me further down a very happy and productive path.

Well done Sigma.
I agree, I started with the DP2 and what that yielded in its package size was very surprising. I think these Quattro's will be quite small. Its almost as though Sigma have hacked away everything that is not part of the camera. I can't believe they would release a shape design like this without extensive testing, so I am sure it will be good.

The DP's have been incredible value for money, bringing HBlad like IQ to the hands of users with limited budget. The bang for buck is not approached by anything else. Even the Sony RX1 was three times the price - hey! there are three Quattro's ... hmmmm.

The DP is a sensor married to a great optic, I don't think they could perform like they do married to a m43 or other mount.

I continue to remain excited. :watch:
 

neilvan

Well-known member
The new form factor looks interesting to say the least.

Personally, I would be happy with the DP2M as-is if it just was supported in ACR!
 

The Ute

Well-known member
The new form factor looks interesting to say the least.

Personally, I would be happy with the DP2M as-is if it just was supported in ACR!
I am not going to be in a hurry unless their is a huge improvement all the way
round.

As it stands now the Merrill's are a screaming good value.

There prices have come down drastically w the announcement.

:clap:
 

larryk

New member
I'm excited. I own the DP3 Merrill and was about to buy the DP2 when and if the prices dropped a bit. Well, now I can wait for the Quatro. Why? We know the lens is great . . . it's the same lens. The lens should be designed for the Foveon sensor. Here's what looks new to me. If the image quality is 30% better (forget using MP's), then it will have higher resolution than the current Nikon D800. Why? It already close to the resolution of the D800. Second, this may be the second best b&w camera after the Leica MM. Why? The improvement is in the blue channel for luminance. Converting to b&w in the the new SP raw converter will optimize the quality of b&w. This is already true for the DP2/3 Merrills which produce great b&w images. Third, processing should be much faster now. I noticed on my dp3 merrill that the writing speed to the new, ultra fast cards only takes a second or so, but the preliminary image processing is what is slowing down the transfer. So, it should be much faster now. Fourth, the battery should be larger and last longer. Fifth, well high ISO levels remains an unknown. Sixth, economics. Look closely at the photo. Pure design "outside the box." Literally and so quite innovative. Look again and take away the grip and battery on the right hand side. Looks the same as the original dp merrill but thinner. And then someone said, let's just add a battery and grip on the right hands side! Look at the specs: it's lighter than the dp3 merrill. You can wear it around your neck with another camera. Finally, with the 1 and 2 lenses, it looks like it really could fit into a jacket pocket. Post finally, the tripod sits squarely under the middle of the lens. Adds up well. But it will attract attention for a while, won't it?
 

Tim

Active member
I was considering / about to add another Merrill body - the DP3M. Now I think I will wait to see samples, and the price of both Quattro and Merrill (discounts) before making a decision.

I may even now consider the DP1Q, but a lot depends on specs of lens.

To save my fingers could we refer to them as DP2, DP2M, DP2Q etc ?
 
Top