Another Olympus E-M1 and SONY a7 walkabout test, high ISO

Started 9 months ago | Discussions thread
blue_skies
Senior MemberPosts: 7,138Gear list
Like?
Re: Another Olympus E-M1 and SONY a7 walkabout test, high ISO
In reply to Daniel Wee, 9 months ago

Daniel Wee wrote:

Recently I wrote up about my comparison between the Olympus E-M1/17mm F1.8 and the SONY a7/35mm F2.8. I figured that many people here might be wondering what the differences may be and how much of a difference one might get going from a micro four-thirds setup to a roughly equivalent full-frame setup. Of course, it was meant to be seen from the point of view of an end-user trying to get roughly the same kind of field of view. I had in mind also to try to show how much the difference in the depth-of-field one could expect with the slightly different setups.

Note: The Olympus 17mm would give a field of view similar to that of 34mm in FF terms, and the DoF afforded by the F1.8 aperture would be what you'd get with F3.6 in FF terms. However, the F1.8 will still be as bright as F1.8 in FF terms, and naturally brighter than F2.8 or F3.6 on FF cameras. I put this here just so that the nitpickers can stop hijacking the thread and turning it into some techno-babble debate.

In the course of that thread, someone asked if I had any high ISO comparisons. I didn't, at that point of time, but today my SONY Carl Zeiss 55mm F1.8 lens arrived. Now, this is one expensive piece of glass, but it is also supposed to be the sharpest glass you can get for the SONY a7/r right now, that offers full AF. If you don't care for AF, you might want to try the Otus if you can afford it.

So, I thought it might be a good idea to compare that with the Panasonic Leica 25mm F1.4 lens that I also had, that was also expensive though not nearly as expensive as the Carl Zeiss. This will give the field of view of a 50mm lens on FF, and the depth of field of a F2.8 FF lens. And yes, F1.4 is as bright as F1.8 on FF, we know that.

My first impressions are that both lenses are incredibly sharp and very high quality. Both are a real pleasure to use. I found that manual focusing (using only the magnification aid and not the peaking function) on the Carl Zeiss was a bit easier than on the Leica. The manual focussing ring on the CZ seemed to be less touchy than on the Leica, notwithstanding the fact that the DoF was shallower. It's not a big difference and certainly not a deal-breaker. In fact, if not for the fact that I was constantly switching between the two cameras, I probably would not have noticed it at all. I mention this here for those who want to know how well manual focusing is implemented on the SONY a7. Answer: Pretty darn good. The E-M1 is no slob either. In one scenario, I found the E-M1's implementation to be particularly better. This is when I am focusing manually and the magnifier is on. If I half-press the shutter release at this point, the magnification is cancelled and I need to move the ring to bring the magnifier back up. On the E-M1, this happens snappily whereas the a7 takes slightly longer to bring back the magnifier.

When it comes to auto-focusing in low-light, the E-M1 feels a lot snappier than the a7. Neither cameras missed their shots but the a7 hunted occasionally whereas with the E-M1, you fire away with a confidence that the camera well deserves. Having said that, I could live with the a7 auto-focusing since I don't really do sports or action photography much, and I don't often shoot in very low light (which is when this problem is more pronounced). When it gets that dark, I prefer to manual focus anyway.

Another thing that I had not written about in my previous comparison thread was the difference between the native aspect ratios. The a7 has a 3:2 aspect ratio while the E-M1 uses a 4:3 aspect ratio. In practice, this means that the a7 produces more elongated frames. I mention this because while trying to get similar framing for the photos from the two cameras, I had to make some adjustments due to the different aspect ratios. I think I could get used to a 3:2 aspect ratio. Yes, I also know that you can select a 3:2 ratio on the E-M1, and yes we can always crop it in post as well. This isn't a criticism of one or the other. Some who are wondering about the a7 may want to take this minor difference into consideration if you strongly favour the 4:3 aspect ratio.
Now, onto the test images. All were shot in RAW, and most were shot wide open, as wide as the lenses will allow - which is F1.4 on the Leica and F1.8 on the Carl Zeiss. I buy these kinds of lenses to use them for their speed so I thought it would be meaningful to test them where they will be used most (by me). In the first test pair, I shot the Leica at F1.8 instead of F1.4 by mistake. In the last test pair, the CZ was shot in F8 while the Leica in F4 on purpose. The field of view of the 55mm is a little narrower than that of the 25mm (FoV of 50mm in FF) so I sometimes had to move a bit to get similar framing. Images were processed in Lightroom 5.3 and luminance noise reduction was set to zero. I notice that the E-M1 tends to expose brighter (using center weighting, sometimes even over-exposing) than the SONY which tends to expose more accurately. As a result, I had to dial in a -0.7EV for the E-M1 on average, sometimes more and sometimes less, to get the correct exposure. I also found the reds quite strong for the E-M1 images (which tends to look warmer) when processed in Lightroom, while the SONY is more muted. I did tweak the colours a bit to get them closer. I do slightly prefer the auto white-balance on the a7 though, although this is easily adjusted in post.
Here we go:-

a7 ISO 1600, 1/1250 sec @F1,8

E-M1 ISO 1600, 1/640 sec @F1.8

a7 ISO 3200, 1/80 sec @F1.8

E-M1 ISO 3200, 1/80 sec @F1.4

a7 ISO 3200, 1/2500 sec @F1.8

E-M1 ISO 3200, 1/4000 sec @F1.4

a7 ISO 3200, 1/800 sec @F1.8

E-M1 ISO 3200, 1/640 @F1.4

a7 ISO 3200, 1/1000 sec @F1.8

E-M1 ISO 3200, 1/1000 sec @F1.4

a7 ISO 1600, 2 sec @F8

E-M1 ISO 1600, 0.6 sec @F4

As you can tell, even up to ISO 3200, both cameras perform admirably and while the a7 has lower noise, as one would expect for the larger sensor, it really isn't radically better than the E-M1. On both cameras, the noise had a very nice and pleasing grain to it and colour noise was not a problem. I'd say that the noise for these cameras at ISO 3200 was similar to what I got at ISO 800 on my Nikon D2H from yesteryear. Details were held well and I wouldn't hesitate to use ISO3200 if I felt I needed to.

As you can probably tell from the test images, both cameras hold their own. The a7 has more resolution and slightly better noise handling but not by an earth-shaking margin. It is what you would expect from a full-frame sensor and a machine/lens combo that costs more. The E-M1 has every reason to hold its head high - we're talking about comparing a m4/3 sensor to a FF one.

In conclusion, I am enjoying both these cameras tremendously and if I were stuck with either one, I'd be just as happy. Each has advantages over the other but neither has an overwhelming advantage. Both the Carl Zeiss and the Leica are really some of the best in their class lenses. You pay for them, but you get what you pay for.
Disclaimer: What is presented above are my observations and opinions. They are NOT intended as "criticism" of either machine, nor presented as gospel truth. The methods used were not intended to be scientific comparisons, but how a typical end user might use both machines to get similar images, so give them some leeway. I own both machines and am brand agnostic. Of course, my own methods are not perfect and there are certainly ways to even things out even more - that is not my intention nor do I have the time for too much elaborate testing. I wish both Olympus and SONY would be spurred by the competition to make even better cameras, because that way we all win.

Thanks for doing this, Daniel, this is very illustrative.

First, a couple of things to consider, that could affect your results, and interpretation, thereof.

Exposure

True and reported ISO are not identical - there is a difference between the two cameras (arguable 0.25 to 0.50 stop) and between the lenses (per the T-value). Moreover, the two cameras were not set up the same way: the A7 tends to protect highlights (=underexpose) and the M1 seems to use average metering (=overexpose). This makes for a very confusing setup. Furthermore, you change the aperture between the two cameras (the F/8 vs. F/4 shot), but this is not compensating for DOF or noise levels.

Flow

As quezra already pointed out - the A7 images appear a lot more noisy that we (A7 users) see ourselves. Was that intentional, or just a result of a simple flow? Showing max noise versus showing best results makes a difference, imho.

Image sizes

If you down-size the A7 images to the 16Mp of the M1, the noise levels drop. This is a very simple noise-mitigation. Comparing them as is, is not showing the true differences.

DOF & focus

Using the M1 at f/1.4 (=f/2.8 FF) versus the A7 at f/1.8 makes a difference in DOF. However, it appears that you did not properly focus the A7, or worse, induced camera shake of sorts.

[EDIT]: from your other post:

"I've not gotten around to trying out EFC on the a7 as yet but that's an interesting suggestion. So far, I've not really had any problems that I am aware of with shutter shock so I turned the anti-shock on the E-M1 off."

You definitely want to use EFC on the A7. In these cases it may have mattered.

[EDIT2]: DxO  lens rating: FE55 = 42, PAN25/1.4 = 23. This is a big difference!

Postprocessing

You said that you moved framing by walking around and use a pp flow to make the pictures similar. But I find them very dissimilar, so I am a bit at a loss what you actually did here.

Also, A7 images need sharpening to be applied - it makes a dramatic difference. The M1 images seem to have sharpening applied. Is this merely camera default, or something that you did?

Results

Below, I have put your images, unprocessed, in LR in a side-by-side view. What I generally see are the following:

ISO3200 is (far?) too high for the M1. ISO1600 is tolerable (true ISO1250?), but noisy.

M1 are overexposed. They are overexposed by themselves, but even more so in comparison to the A7.

ISO3200 on the A7 is actually quite clean - not what I would expect to see, based on personal results, but in comparison to the M1 it is a big difference.

Also

I would almost argue that ISO1250 on the M1 seems to be 'top end', whereas I would put the A7 at ISO5000. Going past that is absolutely possible, but I would not print large any longer. If I downsize the A7 images to M1 image sizes, I would use ISO6400 as the top end.

In other words - I can match M1 noise and DOF levels by shooting the A7 with higher ISO (the images would be a closer match), but then I can trade off noise with DOF: by opening the aperture on the A7, the noise levels drop as do the DOF levels.

In your shoot, you brought in some of the latter - the A7 images are much cleaner as a result, but the images lack sharpness that can be obtained with proper focus and camera technique.

Comparison

In all pictures below, the M1 image is on the left, the A7 image is on the right. These are screenshots taken from LR:

Consider shine and DOF: the M1 image (left) is overexposed, but carries a lot more DOF

Not sure what happened here, the A7 image (right) has shallow DOF, but no real focus. Camera shake?

Most telling image, imho, the M1 image (left) simply 'looses it'

Again, the M1 image (left) is over-exposed, but it works here, and the A7 image is too soft. No sharpening, and possible poor focus and or camera shake? But exposure levels could have been pulled similar (lower M1 and pull up A7) to make them more matching.

Very clear example of the sensor difference: the M1 (left) image has a lot of (processed?) noise, whereas the A7 image is actually rather clean

In this image, the A7 is severely overexposed. It is a pity, because there is much more detail and much less noise in the A7 image. The M1 image seems overprocessed, yet maintains very high noise levels. If you sharpen the A7 image it would have a lot more detail with less noise, but over-exposed as it is, it will be sub-optimal.

-- hide signature --

Cheers,
Henry

 blue_skies's gear list:blue_skies's gear list
Canon PowerShot S95 Sony Alpha NEX-7 Sony Alpha NEX-6 Sony Alpha 7 Sony a6000 +30 more
Reply   Reply with quote   Complain
Post (hide subjects)Posted by
NiceNew
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow