Disorganised thoughts on the Sony a7R II

  • It has some outright bugs around handling of SD cards.  For example, it will not record high quality video to a SanDisk Extreme Pro card.  It falsely claims that the card is not UHS-I compliant, nor fast enough – both utter garbage.
  • “Silent” mode (electronic front & rear curtain shutters) makes a significant difference to the sharpness of the shot, at pretty much all shutter speeds.  However, it does result in a surprisingly pronounced jelly effect (due to the rolling shutter) which makes even mostly stationary subjects, like a flower barely moving in a light breeze, very distorted in some shots.  This is not necessarily self-evident when looking at a single shot, but is readily apparent when comparing several shots of the same subject.  And you can generally forget about using this mode for genuinely moving subjects – I already posted an example of what happens then.
  • The higher resolution (vs the 24 Mpix I’m used to) is noticeable, and in particular makes a big difference in cropping flexibility.  I was very happy with how aggressively I could crop and still get very usable images.
  • With the Sony 90mm f2.8 macro lens, the system is really slow to autofocus.  Glacial.  I was very surprised by how much slower it is than the Sigma 105/2.8 on my D7100.
  • Speaking of autofocus… I was very disappointed.  This is likely in no small part limited by the lens, but autofocus in general just didn’t work for most subjects – certainly not macro subjects, for which the aforementioned Sigma 105/2.8 & D7100 have almost no trouble at all.  I ultimately shot manual-focus pretty much 100% of the time, in the end.
    • And by “just didn’t work” I mean the focus system just hunts back and forth throughout the entire range, sometimes repeatedly, other times just halting at some random spot, often completely out of focus.  This happens even if the subject is already in focus when you hit the focus button.  I was amazed at how bad this is.  I hope the a7R II that I was using was just outright broken in some fashion – if this is normal, then I will never buy an a7R II.
  • The Sony 90mm f2.8 is optically fantastic, but has some surprising and kind of stupid usability issues.  It utilises a push-pull focus ring mechanism to shift between autofocus and manual.  In autofocus mode, the ring doesn’t move, and the focus scale associated with it is meaningless.  But when you pop it into manual focus, it then does follow the focus scale, and oftentimes the point the ring happened to be set at is way off from what autofocus had.  So making manual adjustments to autofocus is impossible.  I’m just dumbfounded by this.  It’s a terrible design.
  • Aperture (and any other Mac OS X photo editors that use the Apple-provided RAW interpreter) is glacial with the a7R II’s files.  And I believe this is compounded by the nasty regression in El Capitan by which RAWs take much longer than they used to, in Yosemite and earlier, to render.  i.e. in Yosemite it’d take a second or two to render a D7100 NEF.  Not ideal but usable.  In El Capitan it often takes ten to twenty seconds.  A huge pain.  But the a7R II’s files can take minutes.  Multiple minutes.  It’s insane.  As always, Apple doesn’t test their shit.  Or care.
  • The Sony’s RAWs are significantly less malleable than the D7100’s.  Posterisation kicks in surprisingly quickly when doing even basic edits, like levels or curves.  I assume this is because of the Sony’s daft lossy compression of the RAWs, and will consequently be fixed when they finally release the pre-announced firmware update that fixes that design flaw.  But that’s still not actually out at time of writing.
  • Capture One – which, pleasantly, is freely available for Sony cameras like the a7R II – does a noticeably better job interpreting the RAWs than Apple’s RAW engine.  Which isn’t surprising – Apple’s is not very good and Capture One’s is broadly held to be the best.  Capture One also does a significantly better job with D7100 RAWs, too.
  • Photo review on the a7R II itself isn’t great either.  The camera is generally a bit sluggish, but especially when you tell it to zoom in – e.g. to check sharpness & focus – for which it can take five or more seconds just to respond.
  • Adjusting the active autofocus point is a bit awkward.  You have to hit a button or two, then use the control wheels on front & back to move vertical and horizontal, a little like an Etch-a-Sketch.  It’s slow to the point that I typically just used manual focus instead.
  • Focus peaking is very useful, but not very reliable.  For some scenes it claims nothing is in focus, when it clearly is.  For others, it claims everything’s in focus, but is strongly exaggerating.  It’s also not very good at indicating focus for moving subjects.  In general it’s significantly better than not having it, but you still sometimes have to do a fair bit of focus bracketing and spray-and-pray in order to get the shot.
  • Noise performance is exceptional.  It’s easily two stops better than the D7100.  One of those stops is more or less expected because of the larger sensor, but the other is an awesome bonus.  I was surprised to find that using ISOs of up to 10,000 still yielded photos I liked, whereas on my D7100 it’s rare that I’m happy with ISOs above 1,600 or so.
  • Its exposure warnings, particularly for blown highlights, are really inaccurate.  It’ll say a highlight is blown – totally past 100% – when in fact it typically isn’t.  It seems to claim highlights are blown at least a stop, maybe two whole stops, from when they really are.  That was frustrating – I ended up with a lot of unnecessarily under-exposed shots early on, until I gradually came to accept that all the zebra stripes were just lying.
  • I hadn’t really acknowledged, in my own mind, the fact that you can’t get as close to your subjects with a full frame camera.  Meaning, on a 1:1 macro lens, you get fewer pixels on your subject, because your pixel density is lower, despite the higher overall pixel count.  And your preview in the viewfinder is smaller, because it’s “zoomed out” vs a crop sensor, which makes focus a bit trickier.  Granted having more cropping and framing options later is an upside of this, but… I was a bit disappointed by the lack of magnification possible with a regular macro lens.

Despite the number of critiques listed above, overall I did enjoy using the Sony a7R II.  I rented it somewhat on a whim, as a precursor to buying it – I almost just bought one sight-unseen.  I’m glad I didn’t – my particular attraction to it was for macro work, where on paper it’s just superb, but because of many of these design flaws and limitations, it’s not nearly as good as it should be.  Still a decent camera.  I just won’t be buying one.  Which makes me sad – Nikon’s clearly not interested in competing, especially in the crop-sensor realm.

Leave a Reply