[popo]

Overlooked it earlier, they also announced MC-11 adapter to allow EF or SA lenses onto Sony E mount. I know we've had adapters from various companies small or smaller already, but one coming from Sigma is stepping it up a bit!

[JJ_SO]

popo, an adapter to the Sigma Merills is not necessary, if one decides to go mirrorless and let Sigma do the conversion of their Art lenses. I seriously consider that, although I'm aware of throwing away some FF advantages while keep the disadvantage of heavy, chunky lenses. 39 MP is enough for landscapes and productshots, as well as for studio portraiture. Together with the new 50-100/1.8 Art it will be awesome.

 

Given Sigmas slowliness from product announcement to delivery into real life I don't expect to see those cameras in a shopping window within a year from now. So time enough to get more informations. Slow AF? I think at Sigma still is development, I will wait until I can see for myself - and if thats a vaild point, BC, then start trashing your manual focus legacy lenses. High ISO? For that I have the Fuji and will get another body in time.

 



 

Given what the Fuji X-E2 together with Capture One 9 already is capable of I don't feel in a hurry. To me, the Sigmas are what I hoped for from Sigma. No bloody "new" DSLR, those times are over soon. Enough buttons, clean menu, great design - if they managed to get a state of the art EVF with at least 2MP, better more.

 

The adapter MC-11 is a cool move for all Canonistas trying to wet their toes with mirrorless. Poor Nikon, the trains are moving and they still keep their locomotive in the shed.

[JJ_SO]

The more I read about the new Sigma mirrorless, the more I need to take care not to run in the opposite direction of BC superfast and supernegative predictions. The specs are promising:

 

51MP on a sensor with 1.3× cropfactor, 2.36MP EVF, phase- and contrast AF, USB 3.0 and maybe bigger battery (the one for the quattro is already bigger and this new one hopefully gives enough energy for one day.

 

A lot of Sigma users complain about the Quattro sensor. I've not done a direct comparison, but so far I like the output

[Brightcolours]

My manual focus lenses have their use, as have my auto focus lenses. So no need to trash any for being old.

On the sensor of the H version of the SD Quattro: 6192×4128 makes for a 25.6mp sensor.

 

About the output at low ISO: the high contrast, low DR, punchy tonal curves and saturated colours look nice. I always have had the opinion that high DR with lifting shadows is a strange fad, to me.

Just would be nice if the camera would not blow highlights that fast, so less images would look under exposed, from what I have seen from the quattro sensors. Still not a fan of the AA-less-ness though, I could do without the false detail and fake sharpness. 

 

Not sure still why the quattro sensors have such limits like the foveon ones before them, because they do interpolate and in theory should have less noise issues. Well, I think only just a bit less noise, as the 3 layers still will be the source of a lot of noise.

 

The design of the camera has a very "european cottage industry audio" look to it, which is charming.

 

What is cute is that they call the HDR bracketing feature "Super-Fine Detail exposure mode".

 

No word on if the 50-100mm f1.8 will have a big enough image circle for APS-H, for now.

[JJ_SO]

I didn't refer to them as "old", I just don't think you're as fast and precise in manual focussing as even a slow CDAF is these days? But let's wait and see what testers have to tell about. For sure I don't expect the fastest AF in mirrorless or better than the best DSLR - I'd be happy with fast enough, as long as it's as precise as possible.

 

Sigma makes some artistic maths with the max resolution and counts everything looking like a pixel. Then it's in JPG 3:2 8,768×5,840 = 51.205 MP. Whatever. I got some pictures printed on A2 and I can't tell the difference between the Sigma Merrill and the D810 in terms of resolution, although there is a spec gap. I just don't see it's effect.

 

In aspects of design I totally agree, I also find it charming. It looks also quite well thought. The little detail of diopter setting on top of the body and also the place of the EVF - I know some "left-eye" photographer which might like it.

 

Now I can enjoy the very looooong waiting period. In the spec list there's no weight ("TBD") for the H-version, so I guess, this will come later.

 

50-100/1.8 big enough circle? It's not announced or specified. Yet. But it would be quite odd, if one could not make a use of it? That one and the 18-35/1.8 (which is quite bigger than APS-C) could make a decent pack.

[Brightcolours]

I didn't refer to them as "old", I just don't think you're as fast and precise in manual focussing as even a slow CDAF is these days? But let's wait and see what testers have to tell about. For sure I don't expect the fastest AF in mirrorless or better than the best DSLR - I'd be happy with fast enough, as long as it's as precise as possible.

 

Sigma makes some artistic maths with the max resolution and counts everything looking like a pixel. Then it's in JPG 3:2 8,768×5,840 = 51.205 MP. Whatever. I got some pictures printed on A2 and I can't tell the difference between the Sigma Merrill and the D810 in terms of resolution, although there is a spec gap. I just don't see it's effect.

 

In aspects of design I totally agree, I also find it charming. It looks also quite well thought. The little detail of diopter setting on top of the body and also the place of the EVF - I know some "left-eye" photographer which might like it.

 

Now I can enjoy the very looooong waiting period. In the spec list there's no weight ("TBD") for the H-version, so I guess, this will come later.

 

50-100/1.8 big enough circle? It's not announced or specified. Yet. But it would be quite odd, if one could not make a use of it? That one and the 18-35/1.8 (which is quite bigger than APS-C) could make a decent pack.

Of course MF is slow. That is why I only use them... when it is appropriate. 

I also use fast AF on my 6D, and CD AF on the EOS M.

[popo]

The Quattro sensors confuse me... I can see the technical reasons for doing so, but still it seems like a step back to bayer pattern in terms of worse case resolution. Consider a 2x2 block of final output pixels. Tradition bayer would compose that from 1 red, 2 green, 1 blue samples. Quattro would compose that from 1 red, 1 green, 4 blue samples. So worst case e.g. for red in both cases, would still be 1/4 the area resolution. At least bayer is modelled on human vision with emphasis on green. Quattro seems more like convenient physics, with blue on the top stack, and also the short wavelength is more resistant to diffraction. As such I think a Merrill era sensor would be more fun to play with.

Oh, back to the earlier point about possible adapters for the new body, I always thought the best case scenario would be to have a Foveon sensor with some universal mount (via adapter) so you can play with the sensor with any lens you like. Unfortunately this new body doesn't satisfy me in either respect.

[JJ_SO]

That is true (the basic resolution difference) Just check for yourself, how often you need max resolution in red and green? Most details are in the blue channel. Or maybe not most, but a lot of them which draw the microstructures of a picture.

 

And if the body doesn't satisfy you - well, there are dozens of dozens others. It would be unwise from Sigma to make a body to satisfy all - Foveon users already proved their tolerance against the obvious disadvantages. I'd say, as Medium format also is no option for most of us, the same or even less of the same goes for Foveon niche.

 

I'd also losing some money if I send in my Sigmas to make a mount change and I don't consider each of them would be cool in front of the SD body. If I see the 35/1.4 and compare it to the 20/1.4 or 50/1.4  :unsure:

[Brightcolours]

Most detail is in the green channel, for our eyes. That is why Bayer and whatever pattern Fuji uses put the emphasis on green. 

But, contrary to what some believe, the red and the blue pixels in Bayer and Fujiwhateverpattern also contain detail (luminance) information, which every good RAW converter "extracts".

 

For them, only pure red to black and blue to black subjects only will be half the resolution, every other mix will show higher resolution. Same for Sigma with pure green to black and pure red to black then. 

 

Sigma/Foveon is forced to go for blue, as it samples first blue (tainted with the whole spectrum of light), a kinda green after that (tainted with the whole spectrum minus blue short wave lengths) and finally a yellow (which has lost 2/3rds of its photons along the way already). 

After that they need to do a lot of calculating to end up with cleaned up blue, green and red values.

 

All that loss of light is the reason why the "high" ISO results are so noisy.

[JJ_SO]

It might be both: loss of light = higher amplification and at the same time three layers instead of one = higher temperature after a couple of shots = higher noise.

 

But as the quattro structure of red and green is different in resolution, bigger pixels also mean less noise. At least that's what I see - 800 ISO are less of a noise with the quattro.

 

I just like the results