Tag Archives: Olympus OM-D EM-1

How much is good enough: Image Quality (IQ) Olympus OM-D E-M1 vs. Nikon D850

This post attemps to question the need to full frame in order to get pleasing prints. We will examine how the image quality of Olympus OM-D E-M1 with a 16 megapixel m43 sensor stack up against a Nikon D850 possessing a full framed 46 megapixel sensor.

Let’s just get one thing out-of-the-way. As pixel peeping will show, the technical image quality of the full-framed Nikon D850 is demonstrably better than that of the 1st generation Olympus OM-D E-M1.This should be no surprise given the D850 has 3 times the number of pixels on a sensor 3.8 times larger.

What about a final print? Will the superior image quality translate to a more pleasing print? Unfortunately, pretty much all the information on the internet focusses around pixel peeping. This makes sense, as one cannot view a paper print on a computer! What is missing is the ability to compare the printed results of the same picture made by two cameras for whatever sizes the photographer typically prints.

Why should this matter? If you are buying into a system, you may feel safer going with a bigger, heavier and more expensive system than you need because of uncertainty in the quality from smaller formats. If you have multiple systems, you may find yourself lugging around very heavy gear on vacations where the results from a smaller camera might provide sufficient results.

In this post, I will first show what everyone knows – the IQ advantages of the D850 over an OM-D E-M1 based on a test photo. But these results will not necessarily translate to a better print. For myself, I will argue my Olympus is totally sufficient and indistinguishable from the D850. But, I am not everyone, and so I offer you the ability to download test photo RAW files so that you can edit, print to whatever dimension and assess the results yourself.

The Test Photo

My colleague (with a D850) and I used a car for a test photo between his and my Olympus. We chose lenses and settings to produce approximately the same angle of view and depth of field:

  • The Nikon was fitted with a Nikkor 24-70mm 2.8 lens, and the Olympus with a M. Zuiko 25mm lens. To achieve roughly the same angle of view, the Nikkor was adjusted to 50mm.
  • To obtain similar depth of field, a m43 lens must be opened up 2 stops compared to a full frame lens with the same angle of view. Therefore, the Nikkor was set to f8.0 and the M. Zuiko to f4.0. For both these lenses, the respective apertures correspond to peak resolution (see the tests at http://www.lenstip.com).
  • Both cameras were set at ISO200.

 

In both cases, we centered and focussed on the license plate of the vehicle. In Lightroom, both photos were sharpened, cropped to 8×10 proportions and adjusted to approximately the same brightness.

Pixel Peep Comparison – D850 is Better

The Headlight

Double clicking on the photos below and adjusting for equal size you will notice a few things:

 

First, the noise from the Olympus is much more noticeable than that of the Nikon. Second, the Olympus shows some pixelation – particularly in the black hood outline. The gradation of the blue bumper is also noticeably smoother with the Nikon, thereby giving a more three-dimensional feel.

The overall image quality of the D850 is superior.

Pine Needles

Towards the top left portion of the photograph are branches with pine needles and cones. Click on the crops below to compare the full-sized images.

 

The D850, with a higher megapixel count, clearly exhibits significantly greater detail. The E-M1 appears washed out in comparison. Of course, the green pine needles are rendered quite differently (which is “better” is entirely personal opinion).

The D850 is clearly able to resolve far greater details.

Pixel Peep Conclusions

In looking at the above comparisons, one can clearly see superior image quality from the D850. Both resolution and gradation are noticeably better when viewed on a monitor.

Print Comparison – It Depends….

In my entire life, I printed 2 photographs larger than 8×10. I like that size, and don’t really wish to make bigger prints. For myself, I might justify going to a larger sensor and/or more pixels if I could see a discernible difference at viewing distances.

I had made 8×10 prints from my Olympus and my friends D850. Here is what I found.

  • Observing the prints at viewing distance (about 1 meter away), myself and friends could not readily distinguish a difference. In isolation, I don’t think one could reliably guess the which camera took the picture.
  • The difference in gradation was much less noticeable in the final prints.
  • Moving a few inches away from the prints, the clarity pine needles clearly identify the picture taken by the D850. The difference, however, appears diminished compared to viewing on a computer monitor.

 

In my mind, the picture needs to look good at viewing distances, and not necessarily a few inches away. As 8×10 is my size, I am happy to say the image quality of my m43 systems is completely sufficient. I am doubly happy because I really don’t have any need to spend huge amounts of money on camera system upgrades, while I have a very compact and light system to carry around.

How About You?

Let’s say 11×14 or 16×20 is your size. Do you need a full frame sensor? You can test for yourself by downloading RAW and full size pictures from this Google Drive location. If a 16 MP m43 sensor is sufficient, then any current m43 or APS camera system should be plenty good. Feel free to edit them in the tool of your choice, and print to whatever size you like. I only ask that you post your results in the comments below!

If you are interested in low light performance of m43 cameras, have a look at my other post:  Low Light Noise Performance – Can m43 Embarrass Full Frame?

Happy shooting 🙂

p.s. In a comment below, Dennis Mook passed on some links to similar tests he did a few years back. Whereas I compared prints at 8×10, he went up to 20×24″! I thought I would reference them here:

“In 2014, roughly carried out the same test and wrote about it as well. I did it mainly to prove to myself that print differences would be minor at best but also to help break down that psychological factor that prejudices photographers against m4/3 as smaller can’t possibly be any good. If you are interested here are the links.

http://www.thewanderinglensman.com/2014/02/the-practical-difference-between-full.html

http://www.thewanderinglensman.com/2014/02/the-practical-difference-part-2-between.html

Thanks Dennis!

Advertisements

A Deep Cooling Effect

Increased differentiation between the subject with the rest of the picture is commonly achieved with narrowed depth of field, or by changing the luminosity of the subject relative to the rest of the photo through post-processing techniques like vignetting or selective brightness manipulations with an adjustment brush.

Making use of colour temperature manipulation is another subject isolation technique I sometimes use. For example, in the photo on the left, I find the squirrel blends in too much with the rest of the picture (evidently, evolution favours squirrels that are camouflaged with the forest). On the right, I used the Lightroom Adjustment Brush to cool the background by (1) Masking everything save for the squirrel and the main branch and (2) Sliding the Adjustment Brush “Temp” slider to -5 (i.e. 5 units cooler).

The slight cooling of the background better isolates the squirrel while providing a bit of depth to the photo.

Some notes on my application of this technique:

  •  I typically go for subtle treatment. Though side by side the manipulation is apparent, the picture on the right would not appear abnormal in isolation.
  • I tend reserve warmth for the subject, while cooling the rest. There is no reason this could not be reversed….

Hopefully this opens up some new ways of working with your photos!

Low Light Noise Performance – Can m43 Embarrass Full Frame?

Conventional wisdom states that in low light situations m43 cameras are inferior to their full frame counterparts because more noise is evident at higher ISO settings. This argument is flawed because, for a given amount of light, it assumes both camera types would need to be set with the same shutter, aperture and (hence) ISO.

For a particular ISO, there is not question that noise tends to increase as the sensor size (and hence pixel dimensions) shrink. Some claim that a m43 sensor is about “2 stops” noisier than the full frame equivalent. As such, a m43 sensor at ISO800 would be expected to produce noise roughly on par with a full frame sensor at ISO3200. This is simply physics and can be accepted as true.

That all being said….

Technologies found in many m43 cameras, and the sensor size itself,  enable significantly lower ISO settings than that on a full frame camera in the same situation. In fact, we will see scenarios where ISO can be reduced 2 stops or more to effectively eliminate the “2 stop noise advantage” offered by full frame sensors.

Case 1 – Technology removes the shakes

Irrespective of sensor size, many lenses and some bodies incorporate image stabilization technologies. Not all technologies are equally effective, however. Anyone doing a bit of research will reach the conclusion that Olympus’ IBIS (in body image stabilization) is leaps ahead of the in-lens stabilization technologies found in (many, but not all) larger sensors lenses (in fact, it is also considered more effective than the Sony A7ii IBIS).

How effective is the Olympus’ stabilization? Take a look at the photo below hand-held at 1/6s. The absence of any sign of camera shake is quite incredible. It is even more amazing because it was taken with a 75mm lens (150mm equivalent angle of view on full frame bodies).

Apartment in KrakowApartment in Krakow – m.Zuiko 75mm 1:1.8 – 1/6s @ 1.8 ISO800 HAND HELD

Two things two consider:

  • With a full frame camera and a 150mm lens would anyone expect to get a decently sharp photo at 1/6s? Even with an in-lens stabilizer, I suspect it would be difficult to  achieve the same results at 1/25s – two stops faster. If the full frame lens has no stabilizer, then the golden rule of thumb dictates 1/150s minimum shutter to avoid camera shake which will require an ISO increase of about 4.5 stops
  • Second, I am not aware of many full frame lenses offering f1.8 at 150mm. If you are lucky, f2.8 is the max. In order to compensate, another 1 stop ISO increase is required.

In summary, the IBIS technology in Olympus cameras is key to sharp photos several stops slower than that of full frame counterparts. To compensate, the full frame camera will require several stops higher ISO. Though a lessor argument, m43 offers some telephoto primes with extremely wide apertures that simply do not exist in larger formats. This further allows lower ISO settings. The full frame noise advantage is essentially lost. 

And, the technology is improving – though perhaps lucky, this  Russian photographer’s blog documents sharp handheld photos at 15s using the new OM-D E-M1ii and 12-100mm f4.0 lens (with its own stabilization).

Case 2 – More depth of field can be good

Razor-thin depth of field lovers correctly state that m43 photos have 2 stops increase in depth of field compared to shooting with a full framed camera at the same aperture for identical angle of views. So, an m43 camera with a 25mm lens at f2.0 will have the same depth of field as a full frame camera with a 50mm lens at f4.0 (2 stops smaller).

Irrespective of whether you like razor-thin depth of field or not (I certainly do not), take a look at the photo below. It simply will not work with less depth of field, and could probably benefit from more.

Apartment in KrakowYoshi – m.Zuiko 12-40mm 1:2.8 – 32mm 1/40s @ f8.0 ISO 1600

Even with my m43 body, a relatively small f8.0 aperture was used at 32mm. On a full frame body, f16 would be required to maintain the same depth of field (i.e. two stops slower). To compensate, either the shutter is reduced to 1/10s or the ISO is increased to 6400. Lowering the shutter speed is not an option. Even if the camera was on a tripod to reduce shake, the photo is of a cat and cats move. Therefore, the ISO on a full frame camera would need to be increased 2 stops to provide the same depth of field necessary for this photo. Again, the full frame noise advantage is essentially lost.

In conclusion…

For a given ISO setting m43 sensors tend to have about 2 stops more noise than full frame sensors. However, technologies found in m43 bodies and the 2 stop increase in depth of field resulting from the reduced sensor size create situation where ISO can be lowered for m43 cameras thereby eliminating much, if not all of the full frame noise advantage.

Do these capabilities carry over in all situations? Certainly not – sports photography, or anything requiring fast shutter speeds to freeze action, will put m43 at a disadvantage. If the subject is relatively stationary, or if it requires significant depth of field then m43 can be a very compelling option for low light photography. Of course, do not forget the m43 size and weight advantage when carrying your camera around during dark nights!

Does m43 really embarrass full frame in low light situations? I wouldn’t go that far, but certainly I would give m43 more consideration than it has thus far received for low light photography.

On the Chase for ‘Image Quality’

Obsessing about technical “image quality” manifests itself in many ways. The drive to acquire, at great expense and physical weight, sharper lenses and cameras possessing extreme megapixel sensors is a prime example. Also, consider the instinctive desire to get the main subject in perfect focus.

Yet, for photographs selling for many hundreds of thousands of dollars, one might  search to find a pot-pourri of technical imperfections including fuzziness, grain (noise) or perhaps less than stellar “dynamic range”.

For example, with all its imperfections, I absolutely love the photo below.  Someone loved it even more to pay $965,000 for it.

The Near One Million Dollar Photo

Alvin Langdon Coburn. ‘Shadows And Reflections, Venice’ – sold for $965000

In the end, a photograph can be great without being perfect  but a perfect “shot” need not be great.

In my photograph below, I accidentally focussed way behind the women. Yet, to my eyes, it somehow works having the sharp mountains and farm in the background in contrast with the soft and somewhat nicely diffuse “main subjects”.

OLYMPUS DIGITAL CAMERA

Near Puno, Peru – Olympus OM-D EM-1, M.Zuiko 25mm 1.8

It is technically imperfect, but I like it almost because of the imperfections.

As far as equipment is concerned, I know my much admired EM-1 cannot possibly match the technical image quality of a full frame 40 MP DSLR. But I don’t care as I believe I will take better pictures with my lighter, more compact camera that just fits me better.

I think I have two points to summarize. First, I believe it is better to make something that “looks right” rather than “is right”. Second, it isn’t necessary to continually seek out new and “better” equipments because, in the end, it won’t make a difference in the value of my photographs.

What are your thoughts on the pursuit of ultimate technical image quality? Feel free to disagree!