New Sigma Quattro DP2?

[GaryLH]

Post number 37 Frank.

Jon posted the official announcement from Sigma global. Forbes, cnet and dpreview also are carrying the info as official announcement.

The Sigma site has a lot of more detail then what is in the announcements if u go thru all the tabs.

Gary

 

[GaryLH]

Hopefully they are going to have working demonstration models at cp+.

Rff members that are going to cp+, can u go by the Sigma both. Give us any additional info.

Gary

 

[shadowfox]

The design is a bit too weird for me.
I'll just watch the local ads for those who are selling their DP3 Merrill to get this camera.

😀

 

[FA Limited]

If this new approach really works well, think of what they could do with full frame..

If the dp2q has
- better af
- good color at iso 3200, monochrome up to at least 6400 w/o playing the blue channel game
- no real issues w/ overall iq in real life shooting situations
This could become my default vacation camera as well.

Gary

i think that's a bit too optimistic there gary 😀
maybe after a few more iterations

 

[GaryLH]

I can hope, can't i?
Gary

 

[GaryLH]

The design is a bit too weird for me.
I'll just watch the local ads for those who are selling their DP3 Merrill to get this camera.

😀

Europe already has dp Merrill's at 40% off. Haven't seen any price drops from any of the US websites yet. Used prices currently around 600-700 from places like b&h or KEH.

Gary

 

[Michael Markey]

Yes Gary ... all models $572.

 

[Eric T]

I am intrigued. All I need are faster processing times and better low light performance.
Better battery life is good but I can replace a battery pretty quickly.
I also don't care about full frame. The current 1.5x sensors are fine.
On the Sigma dSLR, I would like to have live view but my sense is that would really kill battery life - faster than i can replace them.
I suppose and SD1q or SD2 is around the corner.

 

[tsiklonaut]

All's nice and I really dig the body design - reminds me a camera with a proper grip which is much better and firm to hold than a conventional camera body, but WTF is going on with the Foveon sensor in "Quattro" version??? 😕

If it's less pixels on other channels it means straight-out interpolating and other various forms of mathematical cheatings that comes along with it to bring all channels equal on pixel count and pixel location. It's basically now a Foveon-Bayer hybrid.

"Sigma claims this reduces the data processing overhead without compromising image fidelity." Well of course the marketing department kicks in...

Always liked the proper complete Foveons because there's arguably a lot less cheating mathematics involved than any bayer- or x-trans sensor having all three channels represented on each pixel location (no de-bayer, no interpolation, no-anti-alias, no line jiggle smoothing, no microcontrast manipulations etc etc cheatings elliminated by the 3X layer design), mostly just tonal, color mixture and noise shaping mathemathics with the TRUEII engine.

With the Quattros it looks to me Sigma has decided to go the same cheap tricks manipulative mathematical route too like all the non-high-end camera manufacturers today: to produce cheaper sensors and make it up by running a more complex (TRUEIII) mathematical cheating engine inside the camera.

I really hope to be proven wrong on those concerns though...

 

[ssmc]

Agreed. If they were dead-set on doing this, I cannot fathom why they didn't make the green layer have 4x the pixels, since most of the information in an image (as perceived by humans) is in the luminance channel (if you look at it in the sense of "compatible color tv"). I understand that the stacking order is based on spectral sensitivity and other properties of the sensor material, but having twice the resolution in the blue channel seems completely counterintuitive.

As for the ergonomics, I'm having a hard time visualizing how thie Quattro is in any way better than the existing "box and a lens" design. They could have added a blocky grip section to the Merrills and ended up with something like a fixed-lens NEX-6, giving plenty of room for a bigger battery (even better if it included an EVF like the Sony). Finally, OS would have gone some way to helping with low-light shooting.

Despite their idiosyncrasies I suspect the outgoing models will become "classics" as the interpolation-free images provided by the existing sensor design are soon to be history. I've been a fence-sitter as far as the DP2/DP2M goes for a long time but if the prices drop I might just grab one.

In other words... they "fixed" the one part (other than the lens) of these cameras that didn't need it!

 

[dfatty]

from what i can tell from the patent, it sounds like merrill tried the green layer with 4x pixels, but simulations indicated that the results were "indistinguishable" from the blue 4x model. but the blue 4x model requires a lot fewer contacts in the semiconductor design, so they went with that. can't wait to see how it performs, the proof will be in the pudding as they say.

 

[ssmc]

Ah, I was wondering if fabrication issues had something to do with it.

As you say, we will have to wait and see how this turns out in actual use.

 

[nlubis]

Just found out about this.
I might be a willing prey for Sigma's next trap.

 

[noimmunity]

but WTF is going on with the Foveon sensor in "Quattro" version??? 😕
...

"Sigma claims this reduces the data processing overhead without compromising image fidelity." Well of course the marketing department kicks in...

Exactly my concerns voiced in posts #77, #87 and #107 in this thread.

The new sensor's write times will be faster, files will be smaller and easier to handle. The body brings significant improvements to ergonomics, operation, and battery life. But we can expect there will be a difference in look compared to the old Merrill sensor (just as there was between the earlier DP models and the Merrill ones).

Sounds roughly analogous to the transition that Leica made going from CCD to CMOS in a redesigned body with the Typ 240.

Each generation of the Sigma Foveon cameras seems to have a different look, kind of like different film. The original DP1 has a look that the DP Merrill doesn't quite capture, either.

For those who don't need the operational improvements and who really like the output from the Merrills, a move to the Quattros might not make any sense.

And just like Leica M Typ 240 vs the M9, I don't expect to see a definitive conclusion to this debate. It really comes down to subjective factors and priorities, no?

For myself, even though I love the design, I'm just bummed by the huge increase in the thickness of the bodies, and the continuing lack of a super wide.

 

[Archiver]

The modern European-style design looks very intriguing, and it might just prove to be an ergonomic wonder.

The new sensor design, well, I'd prefer to wait until samples are in the hands of the shooting community. Bigger battery, faster write times, and possibly improved focus and operation times: it all looks pretty good from there.

I'm still rocking the original DP1 and DP2, as they have a unique look to their images that isn't easy to reproduce in other cameras. Prices on the Merrills have bottomed out in Australia at $780, so I might pick one or two up in the future.

 

[Keith]

While this Quattro may be a stunning design with some image improvements along with write speed etc I'm very locked into the look of the current sensor so would be a very unlikely candidate for a switch. Not to mention I've just dropped a bundle on an SD1M and some Sigma primes and I'm not really interested in colour.

Via the SD1M I now have a lot more control over the foveon sensor as it stands and the DSLR basically solves all the issues the DPMs (although I love them) have given me. I have lens speed to spare (f1.4 with the 50mm) a real viewfinder and a shutter that will go to 1/8000 sec. The DSLR also focuses effectively in low light which can be a PITA with the DPMs.

IMO many of the issues that make the current DPM series pretty clunky shooters will still be there with the Quattro.

 

[GaryLH]

All's nice and I really dig the body design - reminds me a camera with a proper grip which is much better and firm to hold than a conventional camera body, but WTF is going on with the Foveon sensor in "Quattro" version??? 😕

If it's less pixels on other channels it means straight-out interpolating and other various forms of mathematical cheatings that comes along with it to bring all channels equal on pixel count and pixel location. It's basically now a Foveon-Bayer hybrid.

"Sigma claims this reduces the data processing overhead without compromising image fidelity." Well of course the marketing department kicks in...

Always liked the proper complete Foveons because there's arguably a lot less cheating mathematics involved than any bayer- or x-trans sensor having all three channels represented on each pixel location (no de-bayer, no interpolation, no-anti-alias, no line jiggle smoothing, no microcontrast manipulations etc etc cheatings elliminated by the 3X layer design), mostly just tonal, color mixture and noise shaping mathemathics with the TRUEII engine.

With the Quattros it looks to me Sigma has decided to go the same cheap tricks manipulative mathematical route too like all the non-high-end camera manufacturers today: to produce cheaper sensors and make it up by running a more complex (TRUEIII) mathematical cheating engine inside the camera.

I really hope to be proven wrong on those concerns though...

Only time will tell. I feel the interpretation/manipulations will happen as well, but my guess is that it will be when two or more blue channels directly above the bigger green and red channels has absolutely zero info to contribute. But i suspect in most real world pictures, this will not happen a lot... Whether it is good enough as compared to their prior cameras, we shall all c.

In the Quattro, I think the new low size raw (4.9 mp) setting essentially maps the 4 blue to green and red cells together to emulate a 1-1-1 relationship by averaging the value of the 4 blues.

I definitely agree the new Quattro colors are going to have a different look.. The more important question, will it look pleasing to your eye. Was Kodachrome or Etachrome or Fujichrome or Velvia accurate or was really the fact they pleased your eyes?

Gary

 

[GaryLH]

While this Quattro may be a stunning design with some image improvements along with write speed etc I'm very locked into the look of the current sensor so would be a very unlikely candidate for a switch. Not to mention I've just dropped a bundle on an SD1M and some Sigma primes and I'm not really interested in colour.

Via the SD1M I now have a lot more control over the foveon sensor as it stands and the DSLR basically solves all the issues the DPMs (although I love them) have given me. I have lens speed to spare (f1.4 with the 50mm) a real viewfinder and a shutter that will go to 1/8000 sec. The DSLR also focuses effectively in low light which can be a PITA with the DPMs.

IMO many of the issues that make the current DPM series pretty clunky shooters will still be there with the Quattro.

That is exactly the reasons I have a sd1m. But if I could have a dp w/ faster af and better high iso for times I want to go light, especially I family vacations. The dp Merrill's are close, but right now I need to take another camera for times the Merrill can't handle the situation.

I think the Quattro has potential to removal most of my issues. Anyway, I am willing to give it a try, but I not going to get rid of my Merrill's.

Gary

 

[GaryLH]

From getdpi - raist3d - pretty well matches what I am currently thinking

From getdpi - raist3d - pretty well matches what I am currently thinking

The following is a cut and paste from getdpi member raist3d.. It pretty well comes close to my current thinking. In this section he is answering another member and added his comments about the low size raw mode. This is where I am also pretty excited as well.

----------------- cut and paste ---------------



Re: New Sigma DP Merrill's?

Originally Posted by Quentin_Bargate
So to review where we are (rounded numbers):

Old Foveon: 15 + 15 + 15 mp = circa 45mp

New sensor: 20 + 5 + 5mp = circa 30mp

What I am not sure about is how much influence the blue (20mp) layer will have with predominantly red or green images. One of the problems with mosaic (Bayer) sensors is the relative lack of red pixels. With the revised "Foveon" design, are we not looking at the exact same problem?
Correct but I would call it a "similar problem." The sigma Quattro still has the advantage of not having a grid with holes in color like CfA sensors, but yes- you are correct there will be cases the full resolution for color won't be quite realized.
And some interpolation is now required.

This is why I say this design is no longer really x3 foveon. That said there is one nice surprise hidden in the specs- the small raw format.

I missed it and a kind soul pointed it out to me elsewhere- the dp2 Quattro can capture a 4.9 megapixel raw shot. In that mode its a true x3 foveon color capture and this time with a 4x sampled "blue layer" which may improve iq even a bit further.

I am excited about this mode because it means you could use it for high iso and keep better color/slightly better DR, etc along with less ram storage in the memory card and possibly speeding up saving images to the sd card.

Once one moves away from the pure foveon RGB design, I can see issues with interpolation and possibly with aliasing etc.

In short, I don't see how this can be described as higher resolution sensor when the total number of pixels across the three layers is lower.

I agree on paper. In reality because the Merrill sensor has so much noise and the issue that diffraction presents by the time you hit the red layer, this sensor may very well realize in real world implementation a higher resolution or about same with potential iso gains. But its true it is a set of pros and cons like all engineering designs in the world and will have its own quirks.

This is why I said it will require interpolation at 19.6 mp to get color and yes, it's not a true foveon x3 design that way.

 

[GaryLH]

So does he have inside scoop or is he guessing?

So does he have inside scoop or is he guessing?

Another cut and paste from getdpi. This time from Bowman again.

---------------- cut and paste ------------------

Re: New Sigma DP Merrill's?

Originally Posted by biglouis

I only have two questions:

1. How much?
2. When?


LouisB

around 999$ launch date start or middle summer 2014.
Share
__________________
Best regards - Hulyss - www.hulyssbowman.com