Sony A7 vs. Digital M

[wogg]

I have heard grumbling about the A7's ability to produce good images with ultrawides from a few sources now; presumably this has to do with the sensor's ability to handle light coming in way off axis, which is a special design consideration in the digital Leicas, but is there a substantial improvement with the A7R? Is there a particular consideration for wide angle lenses here, or does it have something to do with the elimination of the anti-aliasing filter???

Cheers,
Dez

It has to do with a combination of the thickness of layers in front of the sensor, including IR filter, as well as the micro lens design. The IR filter is much thicker on both A7 and A7r than M9 or M240. The thicker filter gives the Sony's a better shot at accurate color especially in the reds / near blacks. This design effects the short register lenses a lot, causing moderate to severe vignette, color drift, and loss of resolution toward the edges. At 50mm, this effect starts to taper off, because the angle of the light raking across the sensor is less obtuse. This is why you won't see any Sony native lenses as compact as the M-mount lenses in their future primes-- they'll need longer register for performance reasons. But, they will sure perform. Just will be bigger lenses.

It is very well and very clearly documented on Sean Reid's site, testing a variety of 21, 28, 35, and 50s. For the wider ones, there is resolution loss at edge even at f/8.

 

[CameraQuest]

It has to do with a combination of the thickness of layers in front of the sensor, including IR filter, as well as the micro lens design. The IR filter is much thicker on both A7 and A7r than M9 or M240. The thicker filter gives the Sony's a better shot at accurate color especially in the reds / near blacks. This design effects the short register lenses a lot, causing moderate to severe vignette, color drift, and loss of resolution toward the edges. At 50mm, this effect starts to taper off, because the angle of the light raking across the sensor is less obtuse. This is why you won't see any Sony native lenses as compact as the M-mount lenses in their future primes-- they'll need longer register for performance reasons. But, they will sure perform. Just will be bigger lenses.

It is very well and very clearly documented on Sean Reid's site, testing a variety of 21, 28, 35, and 50s. For the wider ones, there is resolution loss at edge even at f/8.

Its only documented for the A7r. Sean has not tested the A7 yet.

The entire issue is also very much a IS THE GLASS HALF FULL OR HALF EMPTY question.

If you view the A7 as a M9/240 replacement poor full frame edge performance for super wides is an issue.

if you view the A7 as a more advanced and versatile replacement for the Ricoh GR, its not an issue at all. I realized that looking at the comments in this thread, something like "this lens does not work on the A7, but its great on the Ricoh GR." Yet the Ricoh just amounts to a cropped full frame sensor. I am guessing a cropped A7 sensor shooting ultra wides will outperform the Ricoh GR sensor with the Sony's larger files. If you view the A7 as an improved more versatile GR instead of a lacking M9/M240, it starts looking a lot more attractive.

Stephen

 

[Pioneer]

Here Goes Nothing

Here Goes Nothing

1, 2, 6, 7, 8, 9, 10 - Bottom = Leica

3, 4, 5 - Top = Leica

????

😕

 

[Godfrey]

1, 2, 6, 7, 8, 9, 10 - Bottom = Leica
3, 4, 5 - Top = Leica
????
😕

Finally, someone who just tries to pick which are which.
You are correct on four out of ten. 🙂

Let's see if a few other people will venture to choose before I put the legend up.

G

 

[Particular]

Ok, I'll play.

01-A7 is top
02-A7 is bottom
03-A7 is top
04-A7 is bottom
05-A7 is top
06-A7 is bottom
07-A7 is bottom
08-A7 is bottom
09-A7 is top
10-A7 is top

I may have them all backwards, but I thought I could see more darkening in the corners in the A7 ones. It is harder to see on the close up shots. It seems the ones I think are M9 are more saturated, which I first mistook for the bright sony colors I get from my nex5. I can't really judge detail and resolution.
Are we meant to conclude it is the lens that makes an image and not the megapixels? I think my old R-d1 is still available in the classifieds...

 

[uhoh7]

Wow Godfrey, hard to tell them apart in your images.

But I'm looking at a7 and m9 results everyday, and you rarely can't tell which is which Raw as they come into lightroom. The whole look is very different with my cameras and editing system.

I rarely shoot over iso 800 and usually am close to 160 with the M9.

Right away, raw the M9 has exaggerated contrast. However raising the exposure adds little noise in the Leica Raws, which I find much easier to manipulate in LR, than the sony raws, which you have to very careful with or noise ensues.

Bokeh is radically different with the same lens, the Sony makes creamy bokeh with everything, whereas the M9 does not forgive.

Detail: it's not that every single M9 shot is wicked sharp. I miss a number, often for reasons I'm not yet totally clear about. But when you "hit" it, the M9 is alot sharper than the A7. It's jaw dropping.

So much so, that even though the Sony is easier to use and focus, in general, I now mostly take the M9. It's unnerving to have no idea what you really have till you get home. But the Sony can be misleading also, since the excellent LCD often makes it seem like you got a real sharp shot that's not. The zoom will tell you.

For that matter, as everyone knows, the M240 shots are very distinct from the M9 on Raw import.

As a professional tool each has it's strong points. Most real world jobs do not need M9 results, and an out of focus M9 shot is useless. Once their is a decent AF lens selection for the Sony, it's the the more practical day to day tool.

But as a tool for "fine art"---what I'm trying to acomplish in my "own" shots, the M9 is way better on my screen. I rarely make fine art, but the camera can deliver basically unbeatable results in early 2014.

And the thing is, we are no longer talking about a 6k camera. 3200 is plenty to find a very nice M9.

It's a golden age 🙂

M9 + CV 75/2.5 cropped:

 

[EastNeuk]

I haven't owned either camera, so I won't try to guess which is which. Instead I've tried to guess which photographs came from the same camera (whichever one that happens to be).

1. Top A, Bottom B
2. Top A, Bottom B
3. Top B, Bottom A
4. Top B, Bottom A
5. Top A, Bottom B
6. Top B, Bottom A
7. Top B, Bottom A
8. Top A, Bottom B
9. Top A, Bottom B
10. Top A, Bottom B

My guesses are mainly based on the rendering of the greens.

 

[YYV_146]

I have heard grumbling about the A7's ability to produce good images with ultrawides from a few sources now; presumably this has to do with the sensor's ability to handle light coming in way off axis, which is a special design consideration in the digital Leicas, but is there a substantial improvement with the A7R? Is there a particular consideration for wide angle lenses here, or does it have something to do with the elimination of the anti-aliasing filter???

Cheers,
Dez

The A7r fares worse on most Leica M lenses, the result is pretty much unacceptable unless you limit yourself to lenses longer or equal to 50mm.

The A7 can produce great images with some Leica ultra-wides. The 16-18-21mm immediately comes to mind. My 21mm Summilux is corner-to-corner sharp at f2.8, and at f1.4 corner smearing is not an issue because of the extremely small DOF. A variety of factors contribute to performance of M lenses on the A7, including the placement of the rear element, aperture size and the design of the lens.

My conclusion is that most modern M lenses should work on the A7 and achieve usable results, if not wide open then stopped down by 1-2 stops. Some lenses can do better, and some (especially those with older designs) will do worse.

 

[YYV_146]

When I consider cameras I consider what kind of images could be made with them.

The shot below was made at iso 5000 with a 35mm Summilux wide open and pushed about a stop in post. With the M9 this photo would be impossible to make - pushing two stops over the M9's max iso does not deliver usable files, that much I know. Of course there is the M type 240, which should be usable at such iso levels.

Another example. I used a macro adapter to focus the 21mm Summilux down to 0.4m. Again not possible with any M body. A 50mm or 75mm might have made a better macro lens, but my point stands.

 

[hepcat]

When I consider cameras I consider what kind of images could be made with them.


Another example. I used a macro adapter to focus the 21mm Summilux down to 0.4m. Again not possible with any M body. A 50mm or 75mm might have made a better macro lens, but my point stands.

Actually, Victor, it IS possible, and can be done with the M body quite easily; perhaps just not as easily or quickly as with live-view. I use the Visoflex III with standard focusing mounts (rather than Visoflex mounts) regularly as an extension tube. It works well, and has worked well for shooters of M bodies for at least 75 years. Last week, though, I used my Panny GX-1 with the kit lens and live view for macro in an industrial setting as I just didn't want to drag the M9 and Visoflex III out.

 

[Roger Hicks]

No, not at all. make a few guesses, Roger. It's for fun.

G

Dear Godfrey,

Why would I WANT to guess which pictures were made with which camera, based on a computer screen, when I have no real interest in an A7 anyway? For that matter, how much do I normally care what camera was used to take ANY picture? You said it was for fun. Sorry: can't see how it could begin to qualify.

Cheers,

R.

 

[Godfrey]

Any one else care to offer their opinions? I'll post the key a bit later today.

thx! 🙂
G

 

[dcsang]

I have been extremely happy with the A7. Sure it's got it's "cons" (start up time, black out after a shot and maybe a couple others) but every camera has "cons". I have yet to buy one of the Sony AF lenses for it. The camera, for me, has been used strictly as a platform for the m-mount (and LTM mount with adapter) lenses.

I have yet to be concerned about really blurrry corners - especially with ultra wides. I don't own a 28mm lens but do have a 21mm and a 12mm which, to me, perform just fine.

This shot with the 12mm worked out just fine and it was easy to get a decent exposure.


I had an opportunity to get a digital M - heck, I had ordered one back in April of 2013 and it took till December of the same year to have someone say "it's ready"; but by that point, I had the A7 in hand, already was taking great photos with it, and other than an actual rangefinder, I didn't see a "need" to drop the camera for the Leica.

The files are what I like, the size of the camera is just about perfect, it takes the same great lenses I can use on my M7 and it didn't "break the bank" per se. No, it's not a Leica nor is it even a rangefinder but it has a lot of potential and is small enough and versatile enough that I have rarely let it leave my bag since I bought it.

Cheers,
Dave

 

[f16sunshine]

Hard to tell from detail and sharpness.
There does seem to be a bit of additional IR sensitivity showing in a couple images. (yellower than normal leaves/grass etc..).
I would guess those are the M9 images.
Either would be a good tool.

Thanks for the comparison G 🙂 I appreciate your sharing your useful position of having both cameras on hand.
I'm not too grumpy to post my guesses rather just too lazy (maybe later at my computer rather than phone).

Victor makes an excellent point in his post. The a7 is capable of coming home with some images the M9 could not. Maybe the M240 could also (at 5 times the price!)

 

[Particular]

Hard to tell from detail and sharpness.
There does seem to be a bit of additional IR sensitivity showing in a couple images. (yellower than normal leaves/grass etc..).
I would guess those are the M9 images.
Either would be a good tool.

Thanks for the comparison G 🙂 I appreciate your sharing your useful position of having both cameras on hand.
I'm not too grumpy to post my guesses rather just too lazy (maybe later at my computer rather than phone).

Victor makes an excellent point in his post. The a7 is capable of coming home with some images the M9 could not. Maybe the M240 could also (at 5 times the price!)

If you are going by the saturation of the greenery, you can copy/paste most of my answers.

 

[cosmonaut]

The Sony A7 is replacing my Leica M9. I have a broader range of top notch lenses for it (Leica R from 19 to 180mm, a few select Nikkors, and the M-Rokkor 40 and 90 that I usually use on my CL). It produces results I like more, and I like shooting with it more.

Nothing wrong with the M9, I like it a lot. But the Sony A7 does it better for me.

G

Have you warmed up to the ergonomics of the a7? I know at first you didn't like it. Also just what do you think should be changed? I know you probably never shoot Jpeg but you should try using the B&W setting.

 

[gdi]

While I got rid of my M9 when I did controlled testing that showed the X-Pro1 matching its results, I dusted off my M8 this AM for a quick, down and dirty test.

As most know there are many who will vehemently defend the M8 as the equal to the M9 in overall results - and surpassing it in sharpness. Their claim is that it is just a cropped M9 with less filtration. Personally, I find the M9 photos to be nicer than the M8's, and I don't recall seeing an advantage going to the M8 in sharpness, but they may be close enough to make a transitive judgment versus the A7, if all you have is an M8 at hand.

That said, the results of my comparing the M8 and A7 shows that the A7 displays significantly more detail and sharpness in the center, than the M8. This was using the same high quality 50mm lens stopped down a stop and a half. When adding the A7r to the mix the difference is much more evident.

What do the results mean to me? Not much, just satisfies my curiosity and confirms my logical preconceptions. The Sonys are very advanced and high quality cameras, but they are certainly not fabulous M mount platforms....

 

[Godfrey]

Have you warmed up to the ergonomics of the a7? I know at first you didn't like it. Also just what do you think should be changed? I know you probably never shoot Jpeg but you should try using the B&W setting.

I think it's kind of clunky in most ways, too many things I'd change (including the basic button control layout) to detail. However, it has just enough customization capability that I can make it work in a useful way most of the time.

For instance, I just returned from a photo walk with the 135mm lens fitted. The light was quite variable due to a certain amount of high-altitude patchy clouds; when from full bright sun to moderate shade in an instant as I walked. The A7's Av mode with AutoISO works well, but you can't set a floor value to keep the exposure time fast enough for a 135mm lens to keep the camera still. So I went to the excellent Manual mode with AutoISO and fixed 1/200 second ... And then the ability to map the Exposure Compensation control to the front dial where I find it most useful doesn't work, I have to remember to use the awkward right-rear corner dial to control EV compensation, which when I switch back to Av mode overrides the front dial customization rather than working with it.

Stupid, clunky design. If it didn't have a good sensor that worked well with my R lenses and a good viewfinder, I'd never be interested in it. The Olympus E-M1 is a FAR more sophisticated camera in every way; these things work seamlessly across all the E-M1's multiplicity of customization options.

G

 

[cosmonaut]

Well I agree with you the right rear dial is a little awkward. As much as I shoot I still have to feel for it.

I think it's kind of clunky in most ways, too many things I'd change (including the basic button control layout) to detail. However, it has just enough customization capability that I can make it work in a useful way most of the time.

For instance, I just returned from a photo walk with the 135mm lens fitted. The light was quite variable due to a certain amount of high-altitude patchy clouds; when from full bright sun to moderate shade in an instant as I walked. The A7's Av mode with AutoISO works well, but you can't set a floor value to keep the exposure time fast enough for a 135mm lens to keep the camera still. So I went to the excellent Manual mode with AutoISO and fixed 1/200 second ... And then the ability to map the Exposure Compensation control to the front dial where I find it most useful doesn't work, I have to remember to use the awkward right-rear corner dial to control EV compensation, which when I switch back to Av mode overrides the front dial customization rather than working with it.

Stupid, clunky design. If it didn't have a good sensor that worked well with my R lenses and a good viewfinder, I'd never be interested in it. The Olympus E-M1 is a FAR more sophisticated camera in every way; these things work seamlessly across all the E-M1's multiplicity of customization options.

G

 

[Godfrey]

Ok, I guess that's enough fun .. 🙂

The comparison photos pairs line up like this:

The raw files coming out of the two cameras are very different processed to Lightroom's defaults, both in white balance (AWB) and overall contrast, as you can (hopefully) see from the unprocessed raw files in these pair strips:

Top row is Leica M9, bottom is Sony A7, in both sets of five frames.​

That would have made it too easy to guess which camera made which image for anyone who was familiar with processing raw files from either. So I took a few minutes time to adjust the white balance, contrast, and exposure so that they had a similar feel.

What I didn't do was any changes to lens characteristics, other than turning on LaCA removal for all exposures from both cameras. The thing to look for (as one person commented, you're the winner of the guessing game, CNNY! 🙂 was that the M-Rokkor 40 at f/4.7 on the A7 would produce slightly darkened corners compared to the M9 at the same lens opening. The other indicator (which several people commented on) was that the M9 produces very vivid greens which are difficult to match with the standard camera calibration profiles provided by Lightroom for the two cameras.

Note:
If I create a custom camera profile for both with Xrite Passport software and a color checker, the resulting colors are virtually indistinguishable.. This is why I've always maintained that using the colors produced by the CCD sensor in the Leica as a proof of its so-called superiority is hogwash. The color spec is far more influenced by calibration and image processing than by the technology of the sensor. ​

Sharpness was not a good discriminator. Even if you had the full resolution images to work with, all of these photos were made hand-held. It would be impossible to say with any credibility that one camera out-performed the other based upon resolution potential given the same lens and this test.

So which camera wins for me?

Well, it's a mixed bag. Neither camera is perfect, each has advantages and disadvantages. And I like shooting with both of them. I can afford to keep both if I want to. But I know that I'll get more use out of the Sony due to its viewfinder, greater sensitivity, and lens versatility. And I know that I will likely not be able to afford buying comparable quality lenses for the M9 across the board like I have for the A7. I feel the overall imaging performance is close enough to a wash that that's not a consideration.

A fun comparison. 🙂

Godfrey