If you ever wondered how the video stabilization in the latest iPhones compares to what’s in a Nikon or Canon camera, here’s a short video shot simultaneously on a Nikon Z6, a Canon M6, and an iPhone 11, all at about 30mm focal length.
I recently acquired a smartphone again for the first time in about seven years. My last one was the HTC Desire Z: when manufacturers stopped building hardware keyboards into smartphones, I stopped buying them. The camera, particularly the stabilised video in the Apple iPhone 11 Pro looked like something special. As I like big screens, I picked up a Pro Max. The iPhone 11 Pro camera has one other enormous surprise up its sleeve as well.
How does the electronic image stabilisation in modern Canon and Nikon mirrorless cameras stack up against the iPhone 11.
Here are the three cameras configurations.
- iPhone 11 Pro Max, main camera at eqv. 28mm
- Nikon Z6 with 24-70mm f4 kit lens set at 31mm
- Canon M6 Mark I with 22mm pancake lens, eqv. 35mm
The Canon M6 I is at its lightest and most mobile with the 22mm pancake so I tested that. The Nikon Z6 was set at 31mm to be in between the two. The field of view difference was larger than I expected. If I were to do the test again, I’d put on the 15-45mm kit lens at 17.5mm to match exactly the iPhone field of view. The Nikon 24-70mm would then be set at exactly 28mm as well.
Here’s the formats in which I shot:
- iPhone 11 Pro Max: 1080p 50fps
- Nikon Z6 1080p 50fps
- Canon M6 1080p 50fps
Why 1080p? The iPhone 4K is no great shakes, with very pixellated details and the Canon can’t do 4K, plus both Nikon offers far less rolling shutter at 1080p. 1080p offers the most level playing field.
The Nikon Z6 was set with a neutral picture profile, the Canon M6 with a standard picture profile.
Here’s the stabilisation settings:
- iPhone 11 Pro Max: built-in stabilisation, normal
- Nikon Z6: IBIS + electronic stabilisation
- Canon M6: electronic stabilisation only
All three cameras were attached to the same Tilta Cinema Cage with dual wooden hand grips. Here’s what the Tilta Cinema Cage looks like. It’s a reasonably stable platform, designed for hand-holding unstabilised cameras like the original BMCC. The motion is all handheld. There are seven sequences: 360° pan, action pan, walking, action pan ii, walking ii and running.
Video Stabilisation Test: Nikon Z6 vs iPhone 11 Pro vs Canon M6 I
360° Pan with Pause
Both the Z6 and the iPhone are a bit juddery with stop and start, a common issue with stabilisation on pans. The M6 has very smooth start and stop but some issues with up and down motion sneaking into the shot.
All are usable in terms of stabilisation.
Action Pan I
The iPhone 11 Pro is impeccable. The Nikon Z6 is very stable but still suffers from a small amount of horizontal judder. The Canon M6 doesn’t have horizontal judder but the image is not all that stable. Dynamic range on the iPhone 11 and the Nikon Z6 is good but already the M6 suffers from muddy shadows and blown out highlights.
All are usable in terms of stabilisation.
Walking Shot I
This is a slow walk. The iPhone 11 Pro looks like it’s on a gimbal. It’s uncanny.
The Nikon Z6 image is very stable but if you watch closely there’s some jello effect in it. This is the electronic stabilisation. Later tests on a Z50 have persuaded me that it’s usually better to leave Nikon electronic VR turned off. IBIS or optical stabilisation are a good thing but beyond that enhanced stabilisation should be left for post-production.
The M6 is all over the place even during a slow walk. Remember this is the same take, all on the same physical platform. No excuses.
In terms of image quality, this is where the iPhone HDR kicks in. The sky looks gorgeous but the houses and trees all look a bit dark. There’s also some shimmer when the HDR is uncertain how to grade the road. The Nikon Z6 offers a rich, detailed image but would need a grade to bring back some of the sky and deepen the colours.
Alas, the M6 is unable to handle neither sky nor shadows. A blurry, murky image. Canon cripple hammer in full effect, as the M6 still are gorgeous. There’s nothing wrong with the physical sensor.
Action Pan II
This is a more definitive and faster action pan. Both the iPhone 11 and the Nikon Z6 have a very small amount of horizontal judder. The stronger and clearer horizontal move apparently convinced both Nikon and Apple’s motion stabilisation that it’s a horizontal pan and not horizontal wobble. The M6 electronic stabilisation again was faultless in terms of panning. The M6 is marked as orange as the initial shot setup on the dog included a huge amount of vertical wobble.
Walking Shot II
This is a faster walk. The iPhone is again faultless, except for the HDR shimmer on the road which would preclude using this footage in most professional contexts: your colourist would hate you.
The Nikon Z6 is quite stable but does have some micro-judder with the footfalls. It would be easy to clean that up with some post-production stabilisation.
The M6 footage is not stable and one can clearly see every step I take. Not good enough for any real kind of use. The murky shadows look worse than ever. Apparently Canon thinks only people willing to carry at least 2Kg in cinema camera and lens should be eligible to enjoy sharp 1080p, let alone 4K.
Now is the time to separate electronic image stabilisation systems by running after a cyclist and a dog with non-stabilised platform.
The iPhone 11 Pro is incredibly stable. Gimbal substitute for this kind of work. The HDR shimmer is more present than ever, although otherwise the colours and detail are larger than life.
The Nikon Z6 holds up well for the first half of the shot, while I accelerate. At some point, the IBIS gets a bit tired and confused by the larger up and down movement of running and there’s a few big bumps. I mark the Z6 as orange as much of the shot would be usable in casual or documentary footage. In terms of image quality, now that I’m shooting away from the sun, the Z6 offers plenty of dynamic range and offers lots of natural detail. The soft areas are due to depth of field with a relatively wide aperture. The iPhone nominally looks sharper due to (excessive) electronic sharpening and small sensor depth of field.
Canon’s electronic stabilisation holds up surprisingly well against running. It completely falls apart at the end of the sequence but does well for at least the first half. It’s harder to give a fair grade on the stabilisation due to both lack of detail in the picture and the absence of adequate dynamic range.
The iPhone 11 Pro stabilisation is even better than advertised. Well done, Apple. Gimbal makers everywhere must be cursing you. The HDR skies are astonishing but the HDR shimmer precludes much use of iPhone footage as a substitute for an action camera or a real camera on a gimbal. Remember this is best case scenario with 1080p 50fps, the 4K would not have much additional resolved detail and would probably be less stable (tests to come).
Nikon Z6 IBIS is an excellent stabilisation tool. It can handle all kinds of movement well, including slow walking. The picture is detailed with very good dynamic range.
The Canon M6 electronic stabilisation is surprisingly good (particularly in comparison to Nikon electronic VR which created the occasional jello effect). It has its limits which means it’s really for non-moving handheld shots. Canon electronic stabilisation handles panning better than either Apple’s or Nikon’s electronic stabilisation. Canon’s video line skipping and poor video dynamic range (not an issue with stills) makes the M6 a very poor choice of video platform even for 1080p YouTube selfie videos. Poorly done, Canon!
Music courtesy of Carbon Based Lifeforms, track Reaktion. Full discography available on BandCamp.com as special package. Great music to which to work, edit or create.
Alec has been helping businesses succeed online since 2000. Alec is an SEM expert with a background in advertising, as a former Head of Television for Grey Moscow and Senior Television Producer for Bates, Saatchi and Saatchi Russia.
Check out the stabilization in the new GoPro Hero 8, it’s unbelievable, you can literally sprint at full speed and take a selfie video at the same time – never seen this before in my 14 years of working with DSLRs and the like. I’m always shocked at why Apple with a tiny sensor can do better than these other companies just with software – billions of dollars of R&D probably helps. I got the iPhone 7 Plus a few years ago for the optical stabilization but notice how good the software stabilization was in 2x zoom mode too. Of course the Hero is really a toy compared to a proper DSLR.
We’ll definitely checkout the GoPro Hero 8. The issue with GoPro is that their sensors remain tiny and only full daylight capable. I think it’s time for some of these action cameras to include at least a 1 inch if not MFT sensor. Sony might have something like this in their action camera line. I’m very impressed with how the Sony ZV-1 records the gyroscopic data along with the video, allowing videographers to apply accurate digital stabilization in post without warp. I wouldn’t recommend the ZV-1 as a main vlogging camera due to its mediocre image quality at ISO 400 and above. There’s no real substitute for decent sensor size. Even Apple with its all its art creates watercolour images above ISO 80.
Agree with you 100% the Hero8’s video is not brilliant but go to the 1 minute marker of this video and you’ll be blown away by the stabilization – https://www.youtube.com/watch?v=h4FI89G5hU0&t=163s
I tested it too, running as fast as I could with my arms stretched out in front of me – really crazy good! I mainly got it to film my kids biking and swimming etc.
Richard, that GoPro Hero 8 HyperSmooth High footage is impressive. I don’t like the look of the Boost version so much. Still, most of the particular test technique here isn’t really my cup of tea – it’s unreasonable, non-real world demands where the videographer is deliberately giving misleading signals.
It would be very interesting to see the GoPro Here 8 HyperSmooth High vs the iPhone 11 Pro.