Fully mirrorless, six months in

This post is by Ming Thein from Ming Thein | Photographer

Click here to view on the original site: Original Post

_Z730319 copy

At the end of last year, I sold my last DSLRs. In a way, they represented the apex of development in the smaller format: the D850, with high frame rates, resolution, high ISO capability, color accuracy, AF tracking and a great viewfinder – if you must still have an optical finder, and unless you need much lighter weight or crazy frame rates, this is probably as good as a DSLR is going to get. The D8xx line proved so good that the D3X high resolution pro body never even got a successor – there was simply no need. It challenged

WhatsApp Image 2019-12-24 at 19.52.53

medium format cameras of its day, and arguably still continues to do so at the 50MP 44x33mm end, especially if you need AF tracking, frame rates, or do a lot of low light work that needs fast lenses. Sitting at the other end of the spectrum, I also bid farewell to the D3500 – the synthesis of consumer manufacturing efficiency; complete with a decently performing, stabilised lens and state of the art sensor at a price less than most entry level mirrorless – or even a decent 1″ compact. Cheaper even, than a spare battery for some cameras. Yet with all of this, you get performance and image quality pros would have done highly immoral things for not that long ago. Despite my various hardware experimentations and diversions, I’ve always kept a DSLR of some sort somewhere in the lineup – even if not primary body. Since then, I’ve been living a mirrorless life – to make up for it, today I offer some reflections on the topic.

1. EVFs are really good enough, and in many ways, better than the optical finders in small format cameras.
The things they couldn’t do – dynamic range, detail, refresh rate – have become a non-issue with the latest generation of panels. Some are so good you can’t even make our individual pixels anymore, no matter how hard you look. It’s worth noting that early restrictions were not down to the availability of panel hardware, but readout rates of the sensor. Sensors can read faster at lower resolutions, either by binning or skipping pixels; to have sufficient pixels to map 1:1 to a 4K panel and sufficient fluidity requires 4K at 60fps readout – which is not a trivial technical undertaking. It’s the reason why a lot of cameras using earlier sensors have such lousy live view; we finally have nuance. And this doesn’t count the things we can now do effortlessly: focus on the sensor plane for both better accuracy and the ability to see DOF effects and focal plane movements as they will be captured; zoom to any point of the image to check focus or details; not go blind when shooting into the sun or other bright point sources. And for some, overlay so many icons you can no longer see your subject. They also make reliability binary: a DSLR’s mirror mechanism is a complex piece of engineering with a lot of moving pieces, and can go out of alignment or not operate properly, causing focusing issues since the AF sub mirror is also attached to the main one, but an EVF either works or doesn’t. In fact, reliability as a whole has become a lot more binary – which is both good and bad.

2. A combination of factors means effectively better optics.
We have a nice convergence here that results in a meaningful bump in image quality even for the same given resolution: new optical technology mainly around computation of more complex and better-corrected optical formulae; new mounts which allow for larger exit pupils, more even frame coverage, new electronic communication pathways and faster apertures; and the continual iterative improvements in manufacturing technology. We’ve actually been seeing some of this for the last few years in regular DSLR lenses, but a lot of these things were restricted by the diameters of older mounts. And we haven’t even started talking about faster AF motors, different AF systems (CDAF requires a different way to move lens elements than PDAF, which in turn requires different movement in the same amount of time; hybrid is something else entirely) and in the case of the Nikon Z, I suspect better assembly precision now that the VR elements have been left out. The risk of decentering is greatly reduced since there isn’t an element that has to be able to move out of the optical axis in the first place.

3. Stabilisation opens up new opportunities and expands the shooting envelope dramatically
Expansion of shooting envelope requires no explanation – not only is every lens stabilised, but stabilised well. It allows for longer exposures and more precise composition and lower ISOs (and thus better image quality). It makes it easier to focus manual lenses since the zoomed view is stable. I was previously of the opinion that optical stabilisation was more effective than sensor shift; I think with improvements in sensor shift tech, it’s better for wider perspectives because of precision of motion, but only about equal at the telephoto end because of limitations in range of motion (and image circle of the lenses). However, it remains true that larger sensors are still more difficult to stabilise; there’s more mass that has to be moved at both higher speeds (to cover/compensate the same effective portion of the frame) and higher precision (since resolution typically also increases with sensor size) – which makes the latest generation of cameras all the more effective. Not only have I not used a tripod in recent memory outside high precision studio work, but I’ve also found myself shooting less and less in the silly ISO realms – even though I can.

4. Battery life is no longer an issue
At least with the Z7 – 1500+ shots per charge shooting with medium format-style shot discipline (i.e. single shot) is common for me. Much higher if I’m shooting bursts. In practice, this is more than a heavy day’s shooting on a documentary job and almost what I got out of the D850, which was the best of the non-gripped DSLRs. I just carry one spare per camera; if I need more either I have bigger problems or seriously need to think about my curation. Gone are the days when I needed eight to ten batteries for the A7RII to do the same…

WhatsApp Image 2019-12-24 at 19.52.53
Let me first apologise for the crappy image – it was the only one I could find in my archives with this relative comparison.

5. The size and weight improvements are meaningful, but with caveats
Whilst ‘best of’ class lenses exist for mirrorless cameras with the expected performance, it’s also very easy to get carried away and land making your kit both heavier and less ergonomic (smaller bodies, smaller grips) than a comparable DLSR setup. The real strength is just how small you can go whilst maintaining image quality, and allowing other factors like stabilisers to compensate for slower apertures. The accompanying image says it all: consumer level DSLR with kit lens vs pro-grade mirrorless with pancake zoom; similar effective practical resolution (Z7 in DX mode delivers a very high quality 20MP, and better acuity than the D3500’s 24MP and not as good kit lens) and a wider shooting envelope due to sensor and stabiliser differences. The Z7 doesn’t feel like it has control or ergonomic limitations, I don’t have to choose what I allocate to the single command dial, and it’s smaller to pack. Note: I could just as easily make this unwieldy by using the 50/0.95 Noct, which looks optically superb but completely unnecessary over the very impressive 50/1.8 S if you don’t shoot it wide open.

6. We are only just beginning to see the beginnings of system maturity
It takes time to design, produce and release lenses. The dust has settled and we’re seeing a usable lineup from all manufacturers – no, it’s not complete, but perhaps we also need to accept that some of the lenses that were considered necessary or impossible previously are now less critical and imminently doable. Don’t be surprised if we actually see a superzoom that’s surprisingly good in the near future – I’ll be the first in line, as I now know that perspective versatility trumps speed under most of the situations I typically shoot in.

7. I still don’t know why low end mirrorless costs more than a DSLR
Maybe it’s the tail end of R&D amortisation, but surely – something with no moving parts (if you have an e-shutter only), no in-body optics (most entry level stuff doesn’t even have an EVF) and a much simpler body design should be cheaper to make. It isn’t cheaper to buy. Pricing this way is not the way to attract new users nor is it the way to save the dying entry level product lines that aren’t getting internal attention anyway (just look at the age of the sensors in some of Canon’s offerings, and lack of difference to earlier ones other than shuffling some buttons around and upping the model number).

8. Whilst control schemes have improved, they’ve still got a long way to go
I’m going to end on this one: digital started by porting over the controls of film to a series of form factors that changed and matured and eventually found consistency. Most modern digitals have more or less settled on a common way of labelling features or settings. Yet few, if any, take advantage of UI improvements learned from mass adoption of other technology; yes, we still need buttons so we can shoot with gloves on and get tactile feedback, but we don’t need never-ending menus. And those buttons can do other things – Canon’s trackpad thingy in the 1DxIII is particularly clever, and should also work with a glove, though the duplication of controls in the 8-way hat switch seems a bit pointless. If Nikon did it, you could add a D-pad and touchscreen to that, leaving just too many ways to accidentally set something incorrectly. Apple got their early phone UI right, bloated a bit with iOS 12 with too many scrolling settings, and now bloated even further with submenus – whilst I understand the desire from some quarters for this, you still don’t have enough control to make it truly user-driven, but you have enough confusion to get in the way. I’m not even sure future computational cameras can have that much control – a lot of the benefits like stacking and exposure fusion or noise reduction are probably most effective when left to AI. They’re certainly not settings I’d want to change on the fly.

But all in all, and in the context of what and how I shoot? I think we’re there, and then some. And the proof is the images that I’m making now that I couldn’t have done before. MT


Visit the Teaching Store to up your photographic game – including workshop videos, and the individual Email School of Photography. You can also support the site by purchasing from B&H and Amazon – thanks!

We are also on Facebook and there is a curated reader Flickr pool.

Images and content copyright Ming Thein | mingthein.com 2012 onwards unless otherwise stated. All rights reserved

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.