[Klaus]

It seems to be available now.

The APS-C model goes for 800$

 

I am mildly tempted.

[JJ_SO]

I was decided to go for the sd H quattro, get a 18-35 and 50-100 plus the grip. I like the simplicity, the design and the results of the dp 0 quattro are to my likening, so the results of the new quattros should be not far away from the D810.

 

Now it all depends what Sigma is replying to my complaint about their new Photo Pro 6.4.0 version. With 6.3.3, I could open a file and wait to be ready to zoom in to 100% in 30 seconds, That already is close to my limit of impatience. The new version now needs 90 sec on the iMac (2.8 GHz / 12 GB) and 70 sec on the MacBook (2.5 GHz / 8 GB). This is not acceptable for me, so I went back to the old version.

 

Which cannot handle the sd quattro files. The only other RAW-converter on the Mac is Iridient and it's developer told me he will not support the quattro sensor, not enough users are asking for and it has become more complicated to handle. Btw., Iridient opens a Merrill RAW in something like 7 secs and toggles to 100% view in 3 sec - those times Photo Pro users can only dream of.

 

But it will need some time in Switzerland to become available, maybe Sigma manages to bring this bloody piece of stone-age 32-bit (really....) software to the old standard in terms of processing time. Otherwise there's no point for me to invest in a camera with a super-exotic lens-mount.

[Brightcolours]

What I think:

 

• Very refreshing styling, reminiscent of english or danish cottage industry audio
  
• Moronic tha they went the Pentax K01 way. Mirrorless without the mirrorless advantage (built in mirrorbox without a mirror). They could have easily gone for a short flange distance and used an adapter to mount the SLR lenses, and still give the option to allow mounting of shorter flange distance (and for instance make it possible to use the Sigma 30mm DN). So now you have a mirrorless camera,, and the smallest normal you can use is the big and heavy 35mm f.4 Art (for the H version, at least the other version has that one small prime (30mm f1..4 DC).
  
• No AA-filter. Not a fan of the false detail/fake sharpness look at all.
  
• I do not get the Quattro idea totally? They still work in layers, so you still have that massive light loss resulting in truly low DR and very noisy results ate moderate ISO settings, still the same colour issues, yet without capturing all colours at all pixels. 
  
• At least they used the blue channel for the actual detail, meaning that they found an effective way of minimising diffraction softening effects with smaller apertures (kinda.. Only in parts where there is a blue component in the light).
  
• Same thing as with other Sigmas: limited RAW converter choice.
  

So that sums it up really. Nice and quirky for good light low ISO shooting, disliking the AA-lessness of it and not getting the silly choice of DSLR lenses only.

[Brightcolours]

I was decided to go for the sd H quattro, get a 18-35 and 50-100 plus the grip. I like the simplicity, the design and the results of the dp 0 quattro are to my likening, so the results of the new quattros should be not far away from the D810.

 

Is that the way to go, APS-C zooms with the bigger APS-H sensor? Won't you run into some limitations of the image circle?

[JJ_SO]

After downloading the Windows version and installing it on my office PC, I'm looking forward to open the same file here. It appears, the Windows version is already running on 64 bit while I checked the state of the process monitor on the Mac: Of 255 processes, the Sigma software was the only one (!) running on 32 bit.  :angry:

 

No wonder, it sucks  <_<

[Brightcolours]

It does not really matter much if it runs as 32 bit or 64bit application, speed wise that will not make much of a difference at all.

[Klaus]

I was surprised that images are becoming fairly noisy from ISO 800 already.

Thus yes - this seems to be a low ISO camera.


The choice of mount is good and bad I reckon. It is a low confidence vote though - they don't believe that they can be more than a niche vendor so they can't afford a dedicated lineup. A better conclusion would have been MFT then though.

[JJ_SO]

With the new software, they have something like "binning function":

 

 

 

It incorporates a “Binning function” that can improve the color rendering of RAW data with ISO 800 or higher.

 

But as they put it nicely: "can improve" ... and it's about color rendering, not noise.

 

I will not give up on my Fujis, they can be the "fast", "high ISO" "better (?) featured" ones, but quality-wise for landscapes and portraits, I don't see them beating the Sigma. Anyway, soon the X-T2 becomes available and suits very well to my lenses.

[Brightcolours]

I was surprised that images are becoming fairly noisy from ISO 800 already.

Thus yes - this seems to be a low ISO camera.


The choice of mount is good and bad I reckon. It is a low confidence vote though - they don't believe that they can be more than a niche vendor so they can't afford a dedicated lineup. A better conclusion would have been MFT then though.

The thing is that of they made a "native" mount with short enough flange distance, they would already have 4 DN primes for the non-H version of the camera. They introduced a Sony FE mount adapter for DSLR lenses at the same time as this camera, which made it all the more obvious? But yes, the H version makes that more complicated for buyers, and a real native lens line up would probably be not doable.

 

With the light loss through the 3 layers (the blue layer gets contaminated by red and green, which need to be calculated out, the green layer gets contaminated by red and misses an amount of green that got lost in the top layer, the "red" layer actually captures  yellow, which then has to be calculated to red, combined with the loss in the top 2 layers) I would have been very surprised if they had better high ISO performance..

[Klaus]

Albeit the size of the photodiodes should be bigger than on conventional sensors.


However, the APS-C one is effectively a 19mp sensor. Thus with the bayer losses in mind this would be equivalent to a 27-30mp-ish conventional sensor.