Pentax 645Z vs A7RII IQ and bang for the buck

[Jon Warwick]

The difference is quite striking, even on an internet post. I'm a bit surprised.

I agree, and the differences look very obvious to me too - in favour of the Pentax. IMHO, details from the Pentax have more texture and bite to them, whereas the Sony looks like there was a sheet of plastic in between the camera and the scene.

Is this due to the larger sensor size, ie, larger pixel size (from the MF camera) driving better acuity? Or is the 16 bit files lending more texture to the image? Or is diffraction also coming into play earlier for the Sony by cramming a huge number of pixels into a very small space? Maybe all ...?

 

[Kolor-Pikker]

I agree, and the differences look very obvious to me too - in favour of the Pentax. IMHO, details from the Pentax have more texture and bite to them, whereas the Sony looks like there was a sheet of plastic in between the camera and the scene.

Is this due to the larger sensor size, ie, larger pixel size (from the MF camera) driving better acuity? Or is the 16 bit files lending more texture to the image? Or is diffraction also coming into play earlier for the Sony by cramming a huge number of pixels into a very small space? Maybe all ...?

A similar difference was noted in the IQ180 vs 645Z thread, with the larger sensor having an even higher increase in the qualities you've mentioned. It seems to be the nature of going up in sensor size.

 

[jagsiva]

A similar difference was noted in the IQ180 vs 645Z thread, with the larger sensor having an even higher increase in the qualities you've mentioned. It seems to be the nature of going up in sensor size.

I saw that test as well, and was surprised at how close the 645Z got to the IQ180. The IQ180 file was downsized, but the margin was much narrower than I would have guessed, especially considering the IQ was using the new 35LS.

I am really quite impressed here by the Pentax.

 

[rhern213]

I made another comparison now that my 120 macro got in.

Same process, RAW conversion in LR, no other adjustments output to JPEG.

One Note: Man am I annoyed by the speed difference between LR and C1. I'm going to need to send some hate mail to C1 to get their S#%$ straight with Pentax, and do some research on how to get the 645Z files into C1.

Here are the specs for this test:
A7R2 - C/Y 100-300 @ 100mm, F8
645Z - 120 A Macro, F8

With the A7R2 and C/Y lens, there was no difference between F5.6-F8 for diffraction, they were identical.

A7R2
_DSC5485 by Richel Hernandez, on Flickr

645Z
IMGP0032 by Richel Hernandez, on Flickr

A7R2 Center
_DSC5485_C by Richel Hernandez, on Flickr

645Z Center
IMGP0032_C by Richel Hernandez, on Flickr

A7R2 Bottom
_DSC5485_B by Richel Hernandez, on Flickr

645Z Bottom
IMGP0032_B by Richel Hernandez, on Flickr

 

[torger]

In this comparison the A7R-II looks to perform even a little bit better than 645z (sharper, better local contrast), so lens has definitely a lot to do with it.

 

[Kolor-Pikker]

In this comparison the A7R-II looks to perform even a little bit better than 645z (sharper, better local contrast), so lens has definitely a lot to do with it.

Exposure and processing seem to different, so it's hard to say, the shadow levels on the Z shot are much darker than the A7 that's for sure. For some reason the A7 shot actually looks somewhat processed while the Z example is more of what I'd expect SOOC. It probably has to do with the defaults that Lightroom applies for each camera.

 

[tsjanik]

Torger, I'm not convinced this result is attributable to the lenses used. Based on my experience with the Pentax lenses, the 120mm should be even better than the 75mm used in the first test. I don't know Sony lenses, but I would expect a 55mm prime to out perform the zoom used in the second test; so the results are opposite what I would expect based on the lenses used. I wonder if this is a focus issue. Looking at the Flicker images, the bushes in the lower right (closer to the camera I assume) are sharper in the Pentax shot, whereas the cars in the distance appear softer, perhaps indicating a closer focus with the Pentax.

Tom

PS Thanks for the profiles you've posted.

 

[rhern213]

I'm also surprised given the results of the pentax 75 vs sony 55. I thought the 120 macro would clearly outperform anything I put on the A7R2. As per many reviews the 120 macro should be even sharper than the 75. Perhaps my copy of the 120 isn't quite up to snuff? Although it's pretty uniform across the image, I've never had a bad lens copy that behaved like that so idk. Maybe I'll try to get an AF version and see how that compares, even thought they should be identical.

For reference, the focus was on the center of the first electricity pole. The trees on the sides are just literally 1-2 ft. in front of the pole and I'm over 100ft away. So at F8 the pole and the trees should be in the same plane of focus.

I took several shots in live view zoom to be sure the focus was as accurate as possible. So if there is something off, it may be due to possible shutter shake? Perhaps wind?

Regardless on the other hand, man the Contax 100-300 is spectacular, I don't know why it isn't a more popular lens for the A7 cameras. In the 35mm world I've never had a telephoto lens of any kind that's as sharp as this, and now combined with IBIS it's even better.

Torger, I'm not convinced this result is attributable to the lenses used. Based on my experience with the Pentax lenses, the 120mm should be even better than the 75mm used in the first test. I don't know Sony lenses, but I would expect a 55mm prime to out perform the zoom used in the second test; so the results are opposite what I would expect based on the lenses used. I wonder if this is a focus issue. Looking at the Flicker images, the bushes in the lower right (closer to the camera I assume) are sharper in the Pentax shot, whereas the cars in the distance appear softer, perhaps indicating a closer focus with the Pentax.

Tom

PS Thanks for the profiles you've posted.

 

[torger]

Yes both vibration and slightly out of focus could cause the effect if the 645z lens should be better. Or a faulty lens...

 

[hcubell]

I'm also surprised given the results of the pentax 75 vs sony 55. I thought the 120 macro would clearly outperform anything I put on the A7R2. As per many reviews the 120 macro should be even sharper than the 75. Perhaps my copy of the 120 isn't quite up to snuff? Although it's pretty uniform across the image, I've never had a bad lens copy that behaved like that so idk. Maybe I'll try to get an AF version and see how that compares, even thought they should be identical.

For reference, the focus was on the center of the first electricity pole. The trees on the sides are just literally 1-2 ft. in front of the pole and I'm over 100ft away. So at F8 the pole and the trees should be in the same plane of focus.

I took several shots in live view zoom to be sure the focus was as accurate as possible. So if there is something off, it may be due to possible shutter shake? Perhaps wind?

Regardless on the other hand, man the Contax 100-300 is spectacular, I don't know why it isn't a more popular lens for the A7 cameras. In the 35mm world I've never had a telephoto lens of any kind that's as sharp as this, and now combined with IBIS it's even better.

I accept that the zoom is of exceptional quality. Still, we are talking about a telephoto zoom, with an adapter no less, compared to what should be an exceptionally sharp lens. Is it possible that, as a macro lens, it is not optimized for focusing and resolving at infinity?

 

[rhern213]

I have heard some claims on forums that this particular lens may not be optimal at infinity, yet others say it's a myth. I'm gonna try to take some closer focusing shots and try to compare it that way.

I have a Mamiya 120 macro adapted to the A7R2 for comparison, such my luck.

I accept that the zoom is of exceptional quality. Still, we are talking about a telephoto zoom, with an adapter no less, compared to what should be an exceptionally sharp lens. Is it possible that, as a macro lens, it is not optimized for focusing and resolving at infinity?

 

[Kolor-Pikker]

I have heard some claims on forums that this particular lens may not be optimal at infinity, yet others say it's a myth. I'm gonna try to take some closer focusing shots and try to compare it that way.

I have a Mamiya 120 macro adapted to the A7R2 for comparison, such my luck.

FWIW, my 120mm FA is pretty bad compared to the 150mm at infinity focus, at any aperture. The 150mm doesn't do macro, so there's nothing to compare against. The 120mm is still vastly sharper than a macro on my old Canon.

 

[hcubell]

Yes both vibration and slightly out of focus could cause the effect if the 645z lens should be better. Or a faulty lens...

Of course, if vibration and/or difficulty focusing are part of the package with how the lens performs out in the wild, then that's part of how we evaluate performance.

 

[rhern213]

So here's another sample:

A7R2 - Mamiya 120mm Macro @ F8
645Z - 120mm Macro @ F11

A7R2
_DSC5492 by Richel Hernandez, on Flickr

645Z
_IMG0046 by Richel Hernandez, on Flickr

A7R2
_DSC5492_C by Richel Hernandez, on Flickr

645Z
_IMG0046_C by Richel Hernandez, on Flickr

 

[tsjanik]

Thanks for the post. To my eye, sharpness is a wash, but I prefer the contrast in the Pentax image.

 

[ErikKaffehr]

Hi,

The Mamiya lens is a floating point design corrected for both near focus and infinity and so is then Sony 90/2.8G.

Exposures and processing are different, and I know from experience that it is very difficult to match processing between different systems. Having a studio setup and including a colour checker helps.

If we check out the watch samples, the view angle is a bit different between the two shots and that would affect the rendition of shiny areas.

What I would say is important for me is that both systems make a very good job. If you would make a wall side print and display on two different walls you wouldn't see the difference. But, you may still prefer one over the other. All small changes put together make for a difference. A bit higher contrast here, a bit of better shadow detail there, a slightly better composition. In all probability, the first exposure would be the best one. Because you build the scene and the image. When you take the second shot you essentially replace the camera and take another shot.

Best regards
Erik


Best regards
Erik

I have heard some claims on forums that this particular lens may not be optimal at infinity, yet others say it's a myth. I'm gonna try to take some closer focusing shots and try to compare it that way.

I have a Mamiya 120 macro adapted to the A7R2 for comparison, such my luck.

 

[justalexander]

This thread is an incredibly interesting and revealing read! Thanks for all the previous posts so far. I know mine is VERY lengthy but I hope some of you will gain some insights from my 10 year professional career...

I'm looking for a new camera system and considering the 645Z. I shoot interiors and architecture. Time is money for me and my requirements are (in no particular order):

• Lens profiles which perfectly correct lens distortion;
• Sharp files, corner to corner, minimum CA or fringing;
• Efficient workflow (stable tethered connection/fast processing/good profiles that require minimum colour correction in post);
• Great handling camera. I'm looking to do more personal work so weather sealing would be a bonus (solid build quality/efficient operation).

PREVIOUS SYSTEM
H3DII 22MP

• Very solid. Always shoot on tripod so size was not a problem.
• Phocus software corrected all distortion in a single click. Colour profile seemed to require less post production than any other system I've used since.
• Film like quality to files. Not soft, but more natural than every other digital system I've used.
• The sensor died and its not worth spending the $5000 required to replace the power board.

CURRENT SYSTEM/S
CANON 5DIII

• Solid build but the USB connection is not as solid as the firewire connection on the Hasselblad leading to frequent drop outs. Canons 1DX and Nikons D4 have ethernet connections which are secure but both cameras are targeted at sports shooters with low resolutions and high frame rates.
• Better dynamic range than H3D.
• Discovered 17 and 24mm TSE lenses which are stellar performers for their price. Unfortunately when shifted you need to resize the canvas correctly around the image before applying CA/fringing corrections for them to work correctly. This requires me to remember the exact amount of shift used but unfortunately I'm often so rushed that I forget to do this.

SONY A7R

• Built like a plastic toy. The mount would flex leading to images that had a tilt effect.
• Stellar dynamic range.
• Worked well with 17 and 24mm TSE via Metabones but every other Canon lens looked terrible. Much softer than on the Canon. Returned the Canon 24-70 2.8 MKII because it was mushy and smeared.
• Sony's 24-70 F4 is a joke and not worthy of the Zeiss label. Returned it also.
• USB port even worse than Canon. Embarrassing to tell a client 24 times a day to hold on a second because I need to jiggle the cord to reconnect.
• EVF is a joke! I know you get things like focus peaking and exposure zebras but I cant possibly see how any of this makes up for the low quality of an EVF.
• Fly by wire focus is a DOG! I can focus a traditional lens by instinct, by using my eyes, or if necessary by looking at the distance on the focus ring. It's also easy to calculate the hyperlocal distance looking at the distance scale. Sony's Fly by wire focus systems either require you to autofocus (which requires you to pick 1 of 4 different focus options, then 1 of 12 different focus modes, THEN move your focus selection point around the screen; or focus manually for which the EVF is not sharp enough requiring you to turn on focus peaking which is not always accurate, and then to turn it off again so you can view the image without it being covered with red pixels.
• I have to admit though that the EYE AF is great for the occasional portrait, but in my opinion/field not worth the trade off.
• Batis and Loxia lenses holds some promise but still no/very limited availability in Australia and gaps in the lens lineup without having to mix the two different ranges!

SONY A7RII

• Mount is much better than A7R but camera still feels like a toy and still has all the problems associated with EVF and fly by wire focus.
• Too many features hidden under too many menus. I'm a professional; time is money; I don't need 99% of the cameras features but the features I do need are sometimes a pain to activate.
• 17mm performs well but the 24mm TSE doesn't perform as well as it did on the A7R!? Anyone else have the same experience? I even updated to the latest Metabones and no change. The drop in quality is so significant that I shooting with the 17mm TSE and cropping results in a sharper file than using the 24mm!

RECENTLY TESTED
LEICA S006

• Beautiful build quality.
• Just the right feature set for my needs.
• Lemo connector is secure. Not a single dropped connection when shooting tethered.
• Like the idea that it's weather sealed.
• Generally STELLAR lens performance except for the following which is unfortunately a deal breaker:
• The lens profiles do not adequately correct lens distortion based on focus distance. I have seen a number of reports to this effect since the introduction of the S system but Leica have so far failed to resolve this fairly simple problem. The attached shot of my kitchen shows moustache distortion along the lower edge of the bench even with the lens correction profile applied. There is no way I could give this image to a client!
• I cant understand why Leica leave file conversion to Adobe. Would you buy a Bentley and take it to the guy around the corner for a service? I believe that if Leica developed their own convertor with 1 click corrections (like Phocus) they could extract even higher quality from their files as well as attract even more customers.
• Leica have only introduced the 120mm Tilt/Shift Schneider into their system but what about the 28, 50 and 90. Why and just who is the 120mm intended for?

H5D50C

• Loved the HTS, gave me the tilt/shift I'd come to love from the Canons but because Phocus does all corrections automatically, not having to resize the canvas around the shifted image to apply corrections was a real time saver.
• Unfortunately with 1.5 crop factor on top of the 1.3 of the CMOS sensor I would no longer have a true wide angle lens for architecture.
• Everything I loved about my old H3DII 22MP but the lenses simply aren't good enough for 50MP. I'd wondered why Hasselblad didnt have an 80MP back like Phase One and now I know!
• Unfortunately a real disappointment because I thought this would be the answer.

PHASE ONE XF IQ260

• Stunning camera.
• Tested the 35mmLS which had STELLAR performance. The 28mm had mushy edges and even my 24TSE on 5D MKIII outperformed it!
• Files look a little over processed/digital without a bit of tweaking.
• Capture One is a nice piece of software, at least in terms of extracting maximum quality from their backs.
• PROHIBITIVE price. The XF system and lenses would cost $75,000 here in Australia and I would then need to add an Alpa FPS (another $20,000 without any lenses) in order to get some shift action!!!

SO...

• They were just some thoughts on my experience using different systems. Everyones needs differ but as a professional interiors and architecture photographer the 2 systems that I've tried and liked are either unsuitable (Leica S lens distortion) or require me to sell my soul to the devil (Phase One).
• From everything I've read the 645Z could be perfect. Unfortunately I don't know anyone who owns one; the dealers down here have the bodies but no lenses in stock; AND they won't even let me test the camera in-store because to do so would make the camera a demo unit which would require selling at a discount!
• So until I manage to convince them to get a lens in stock and allow me to test the camera, can anyone/everyone please tell me whether the 25mm and 28-45mm profiles completely remove distortion? I often shoot buildings/interiors front on and any amount of distortion is really obvious.
• I'm also a little concerned about the lack of decent mid range lens. From all reports the 55mm is pretty average.

Thanks again and appreciate any feedback...

 

[GrahamWelland]

Have you looked at the Actus DB and Rodenstock digital lenses? All movements for significantly less than the cost of the Alpa FPS and add a CMOS MFDB for quality and live view. It's an outstanding system and cheaper than other tech solutions with their helicoid taxes.

 

[rhern213]

From the examples I've seen the 55 is a dud in the corners.

However I have the 75 2.8, and it's great edge to edge at optimum apertures, you won't be disappointed with that one.

I'm now waiting to receive the 45-85 zoom to see how that is, the reviews have been good.

I have the new 35-D on pre-order, so hopefully soon I can compare it to the older 35-A I have now.

Unfortunately I can't afford to spend the loot on the 28-45, and likewise the 90mm macro which is also supposed to be spectacular.

I'm also a little concerned about the lack of decent mid range lens. From all reports the 55mm is pretty average.

 

[steve_cor]

Although the Leaf Credo 50 is $27,000, you can get it with the Cambo Actus DB Plus and a lens, for only $20,000. They are trying to compete with the Hasselblad CFV-50c, so you have to get it in Hasselblad V-mount.

https://www.digitalback.com/product/credo-50-cambo-actus-db-bundle/


--Steve.