Sony a7R II vs III

[Lou Jost]

Yes, it seems like a good value, at least as far as I can tell from reading about it. There is a lot of confusion in the camera community about the difference between EFCS and fully electronic shutter. I am not 100% sure, but I am sure that during pixel shift capture, there is no sound, so the camera does have a real fully electronic shutter built in. I understood that the firmware update just extended that feature to non-pixel-shifting photos. That would be an easy thing to do in software.

But it is worrisome that even some well-respected websites like DigiLloyd seem to be confused about this.

I will have mine in hand in ten days, so I can tell you then for sure.

[Beatsy]

I use Sony RMT-DSLR2. Full controll over anything in your camera. Even power-zoom if you have it.

Thanks. Turns out I already have one of those. Money saved

[mjkzz]

There is a lot of confusion in the camera community about the difference between EFCS and fully electronic shutter. I am not 100% sure, but I am sure that during pixel shift capture, there is no sound, so the camera does have a real fully electronic shutter built in.

At least with my A7III, using electronic shutter (Silent mode), there is NO sound at all, no mechanical movements. During stacking, I only see flashes going off and sometimes the flash tube making some sound.

Sometimes I wish there is some indication of shutter, a fake one from speaker, as I am not sure if a picture is taken after pressing the shutter button.

[Lou Jost]

There's no confusion about the Sony or Olympus silent shutter modes, but the Pentax K1 Mark I original documentation doesn't mention this mode; it was added on later in a firmware upgrade, with ambiguous documentation. Pentax calls it "electronic shutter", not "silent shutter" and not "electronic first-curtain shutter".

I am pretty sure the people who have called it "EFCS" are just confused. We know that this camera is completely silent when shooting in pixel-shifting mode.

[Macrero]

Yes, it seems like a good value, at least as far as I can tell from reading about it. There is a lot of confusion in the camera community about the difference between EFCS and fully electronic shutter. I am not 100% sure, but I am sure that during pixel shift capture, there is no sound, so the camera does have a real fully electronic shutter built in. I understood that the firmware update just extended that feature to non-pixel-shifting photos. That would be an easy thing to do in software.

But it is worrisome that even some well-respected websites like DigiLloyd seem to be confused about this.

I will have mine in hand in ten days, so I can tell you then for sure.

Nice! I'll look forward to seeing your impressions. The camera looks certainly promising.

[ray_parkhurst]

So for me the choices for a FF camera are A7Rii and Pentax K1 Mark 1. Both cost the same and both have fully electronic shutters (the Pentax has this only after a firmware update). The Sony has better adaptability to non-native lenses but the Pentax has pixel-shifting.

Lou,

I was not considering a DSLR, but since you mentioned the K1 I looked into it and it seems a very good option as well. And, hmm... its pixel-shifting seems to work, at least in the DPR samples.

Nikon D810 Crop (same sensor resolution):

snip

Pentax K-1 Mk I - Pixel-shift:

snip

There is clearly more/better detail in the K-1 crop.

Are you sure the firmware added e-shutter is fully electronic? I read in a few sites that it is an EFCS.

Best,

- Macrero

Certainly appears more detailed, and the moire is reduced, but is the added detail real? I found that in 3 cameras (including the Pentax) that I tested, the added "detail" was actually just an extreme form of fine sharpening, and in some cases there was less actual detail. I did not publish my study unfortunately, but my method was to shoot with and without the the pixel shift mode, and then compare the "details" that appeared versus shots of the same area of the subject using a microscope objective. I did not test the Olympus, so can't speak to its more extensive sub-pixel sampling effectiveness, but the Pentax and Sony (which both attempt to fill-in RGB info missing from the Bayer array) were not convincing.

It's also pretty easy to use a 2x TC, with 2x downsizing, to simulate what these images "should" look like.

[Lou Jost]

I did see your Sony and Pentax tests, Ray, and they have me worried. But the results on other tests seemed to show real detail (as in the results given here, where the false color moire of the unshifted image was correctly eliminated in the shifted image). The simple one-pixel shift really ought to just reduce color noise introduced by Bayer interpolation. It could be that the algorithms to process the four images are adding excess sharpening, but maybe this can be controlled. Pentax gives us all four images and there are many choices of processing programs, so there is room to fix any over-sharpening.

There's no doubt about the Olympus algorithm revealing real additional detail (eg ridges on butterfly scales) in my shots. That is a far trickier algorithm than the simple one used by Pentax and Sony, so it would surprise me if their simpler versions can't produce useful results, though it doesn't surprise me at all that Sony would ruin it with undocumented pre-RAW sharpening, as they had done in the past with their "star-eater" pre-RAW noise-reduction.

Edit: The software used to process the pixel-shifted images does add optional sharpening, and the alternative programs do produce different results, so some care will be needed:
https://www.dpreview.com/forums/thread/ ... t-61141727

[mjkzz]

The sensor is probably the best one out there.

And it's not from Sony, rumor is it's from the Tower/Jazz fab designed by Nikon.

Best,

Another rumor, reversing TowerJazz rumor that D850 sensor is indeed made by Sony.

https://nikonrumors.com/2018/06/15/new- ... sony.aspx/

Some guy on Youtube is speculating that Sony will release a medium format mirrorless camera ... oh boy.

[mawyatt]

The sensor is probably the best one out there.

And it's not from Sony, rumor is it's from the Tower/Jazz fab designed by Nikon.

Best,

Another rumor, reversing TowerJazz rumor that D850 sensor is indeed made by Sony.

https://nikonrumors.com/2018/06/15/new- ... sony.aspx/

Some guy on Youtube is speculating that Sony will release a medium format mirrorless camera ... oh boy.

Yes just saw this new rumor. So we have a rumor on another rumor. I tried to get some folks at the conference I'm attending to confirm the Tower/Jazz rumor, but not so

Sony seems for logical than Jazz, but who knows

Best,

[Macrero]

Certainly appears more detailed, and the moire is reduced, but is the added detail real? I found that in 3 cameras (including the Pentax) that I tested, the added "detail" was actually just an extreme form of fine sharpening, and in some cases there was less actual detail. I did not publish my study unfortunately, but my method was to shoot with and without the the pixel shift mode, and then compare the "details" that appeared versus shots of the same area of the subject using a microscope objective. I did not test the Olympus, so can't speak to its more extensive sub-pixel sampling effectiveness, but the Pentax and Sony (which both attempt to fill-in RGB info missing from the Bayer array) were not convincing.

It's also pretty easy to use a 2x TC, with 2x downsizing, to simulate what these images "should" look like.

At least in the sample RAWs I played with it most of the detail seems real. Also less moire and color shift per se should mean more/better detail.

*Pics removed due DPR copyright policy. You can download the RAW files from DPR if you want to see how they compares.

From what I've seen so far I like more the 4x1 shifting approach which albeit simpler than the Oly's 8x0.5, delivers better looking/cleaner result. I don't need 80MB pics either.

- Macrero

[ChrisR]

What am I missing . What do we see on a 1012 pixel FF image?

[Macrero]

What am I missing . What do we see on a 1012 pixel FF image?

Those are (obviously) crops of a FF images. The pixel-shifted one shows clearly less moire and more fine detail resolved. Maybe that's what you were missing

[Lou Jost]

Macrero, the Oly 80Mp files, reduced to 50Mb, are quite good. I agree that the Pentax or Sony FF results will be cleaner, since their sensors are twice as big. But if compactness is important for field work, then a small sensor has advantages.

I will probably continue to use my Oly camera for field work, and probably use the Pentax rather than the Oly camera for studio work.

Edit: Macrero, have you found out whether the K1 has fully silent electronic shutter?

[ray_parkhurst]

Just looks like they didn't critically-focus properly on the Sony picture.

[Macrero]

Macrero, the Oly 80Mp files, reduced to 50Mb, are quite good. I agree that the Pentax or Sony FF results will be cleaner, since their sensors are twice as big. But if compactness is important for field work, then a small sensor has advantages.

I will probably continue to use my Oly camera for field work, and probably use the Pentax rather than the Oly camera for studio work.

Sure, compactness/low-weight are important for field work. I'm gonna use the camera in studio only, so that's not a drawback for me though.

It's also true that it is quite unfair to compare m4/3 to FF sensor. Oly is actually doing a great job with their PS tech. I just like better how the FF sensor 4x1 shifted image looks.

Edit: Macrero, have you found out whether the K1 has fully silent electronic shutter?

No, I'm not sure yet Pentax (Ricoh) refer to it as "Electronic shutter" and that's how most sites call it: "Pentax K1 firmware update adds electronic shutter in live view mode", "Firmware Brings Electronic Shutter to Pentax K-1", etc... though in some places they talk about EFCS, e.g.:

https://diglloyd.com/blog/2016/20160920 ... r-fix.html