How does it compare to the GH5S?

It's easy to assume that a large sensor camera simply offers better image quality, which is mostly recognizable through less noise in the shadows, but also tends to mean cleaner tones across the whole image. But depending on which of the camera's modes you use, this may not be the difference you experience.

Key Takeaways

  • S1H can offer better quality than the GH5S, in situations that tolerate shallower depth-of-field or in bright light if you can exploit its wider operating envelope
  • Some of the camera's Photo Styles trade this direct tonal quality benefit for additional highlight capture

Benefits of a larger sensor

When talking about stills, we usually say that large sensor cameras have a benefit over smaller sensor cameras. This benefit tends to come in two situations: in low light, where a large sensor can receive more light at the same exposure settings, if paired with a lens giving the same angle of view. This benefit is only realized if you can tolerate the shallower depth-of-field that comes with it.

The other benefit comes in bright light where, if you match depth-of-field, the large sensor camera will use a bigger F-number. This would require a higher ISO but would give comparable image quality, but there may be an option to extend the shutter speed and lower the ISO to gain an advantage.

For example, when shooting stills in the following scenario:

GH5S S1H
EV 13 F4.0 | 1/800 sec | ISO 160 F7.1 | 1/800 sec | ISO 500

These setting would be expected to give comparable images. However there's also the option to extend the S1H's shutter speed to 1/400 and reduce the ISO to 250, giving you a one EV benefit. If your subject's movement allowed it, you might even have the option to drop to 1/160 sec and ISO 100 to get even more of an image quality benefit.

You don't have this option in video though, since you have less flexibility over the shutter speed:

GH5S S1H
EV 9 F4.0 | 1/50 sec | ISO 160 F7.1 | 1/50 sec | ISO 500

Even if you're happy to let your shutter diverge from the 180 degree shutter rule, you could make the same adjustment on both cameras, so to get more light to the S1H, you'd need to open up the aperture and deal with the shallower depth-of-field.

However, this doesn't close-off all the bright-light benefits of the larger sensor. In any situation brighter than the GH5S can cope with (i.e.: when you're having to add neutral density filters to the GH5S), the S1H will offer a slightly wider operational envelope: tolerating more light before you need to start using NDs.

GH5S S1H
EV 11.3 F4.0 | 1/50 sec | ISO 160
+2.3EV ND
F7.1 | 1/50 sec | ISO 100

However, these direct comparisons assume both cameras are using directly comparable tone/response/gamma curves.

Panasonic has done something interesting, though: it's adjusted several of the Photo Styles, which combine the tone response curve (gamma) and color response (gamut), to make the S1H's capabilities more accessible.

We're going to have a look at the 'Like 709' profile, which is standard ready-for-broadcast response, the more flexible 'Cinelike D' profile, which gives a near-finished look, erring on the side of capturing a wider dynamic range, and V-Log, which is a response designed to provide maximum flexibility during the color-grading process that it demands.


Like 709

In the output-ready 'Like 709' color mode, both cameras behave as you'd expect: matched exposure values give matched output brightness.

Panasonic S1H 'Like 709'
ISO 100
F4.0
Panasonic GH5S 'Like 709'
ISO 160
F5.0

As implied by its lower base ISO setting in this color mode, the S1H requires 2/3EV more exposure than the GH5S. Or, looking at it the other way around, can tolerate 2/3EV more light than the GH5S. Because this Photo Style is essentially matched across both cameras*, the S1H's larger sensor will simply deliver better tonal quality, so long as you're able to accommodate the shallower depth-of-field that comes with it.

The tone curve here is presenting just over 10 stops of DR, with 7 of these above a value of 64 (a very dark tone roughly comparable to IRE 0: the lowest value you can safely assume people will be able to view).

Compared shadow regions of the two cameras: S1H top, GH5S lower half

Cinelike D2

The 'Cinelike' responses are different on the two cameras, though:

The S1H offers the 'Cinelike D2' curve, rather than the Cinelike D mode of the GH5S. Cinelike D2 adjusts the camera's minimum ISO from 100 to 200, encouraging 1EV less exposure than Like 709 mode. There is no such change on the GH5S.

Panasonic S1H 'Cinelike D2'
ISO 200
F2.8
Panasonic GH5S 'Cinelike D'
ISO 160
F2.2

Sure enough, if we compare them side-by-side, we can see the Cinelike D2 tone curve is accommodating an additional 1EV of highlights, giving a smoother roll-off compared with the GH5S. Examining the footage shows that the S1H gives up some of its tonal benefit in order to capture this additional highlight information.

This tone curve incorporates around 12.3EV of DR, with the 9th stop starting at a value of around 64, which could make those last few tones difficult to lift. The GH5S is showing nearer 10.7EV of DR with tone 8.3EV down from clipping hitting 64, and a higher noise floor making the last few tones difficult to distinguish between.

With its flatter Cinelike D2 curve, and lower exposure, the S1H (Top) doesn't show as much of a noise benefit, but it retains more highlight information than the GH5S (Bottom).

But the different exposures required gives you some changes to close down the aperture, so the depth-of-field difference need not be so great (obviously you could apply ND filters, rather than stopping-down if you want the shallower depth-of-field).

V-Log

Finally, we look at the cameras' log modes. The GH5S uses a truncated version of the V-Log tone curve, called V-Log L. This is a clipped version of the curve with around 4.6EV of headroom above middle grey. The S1H uses a version of the curve with approximately 6.3EV of headroom above middle grey.

Panasonic S1H 'V-Log'
ISO 640
F2.8
Panasonic GH5S 'V-LogL'
ISO 320
F1.6

Again, this is an interesting trade-off. Essentially the S1H will capture around 1 2/3EV of additional highlights, compared with the GH5S. But in doing so, it delivers similar noise levels to the GH5S. In other words, you're gaining the camera's large sensor capability in the form of additional highlights, rather than tonal quality improvements.

The S1H is showing all 13 stops of our test wedge but the last step is almost indistinguishable from the noise floor. The GH5S is showing around 11EV of DR.

The S1H (Top) captures more highlight information (not illustrated) than the GH5S (bottom) but across the tones that both cameras capture, there's no quality benefit to the S1H.

The nature of Log is that it's pretty flexible when you edit it, so there is some scope to intentionally over-expose your footage if you want to strike a different headroom/tonal quality balance. Interestingly, between the raised ISO rating and the inherent low sensitivity (greater light tolerance) of the S1H's sensor, you're likely to end up shooting V-Log at around 1 2/3EV bigger F-number, essentially cancelling out any depth-of-field difference between the two cameras.

What does this mean?

The upshot of this is that the S1H can offer better image quality than the GH5S, but not necessarily in the way you'd expect. Depending on which gamma/gamut mode you choose to shoot you essentially have choices between the following extremes:

  • A similar dynamic range with better tonal quality but shallower depth-of-field (or less ND in bright light)
  • Nearly 2EV of additional dynamic range but with comparable tonal quality and depth-of-field to a GH5S

This suggests you should start by exposing for 'Like 709' (so long as you aren't losing significant highlights) as this will show the most obvious tonal quality benefit, especially in low light. V-Log provides the scope to capture much more highlight information, and maximizes the gradeability of your footage, but comes with a noise cost, so should only be used when you need it (in bright, high-contrast scenes where there's enough light to minimize the visual impact of the noise). Cinelike D2 offers a useful balance between these two extremes.

Ultimately, the difference between full frame and a four thirds sensor isn't a simple case of 'better quality,' it's more of an expansion in the operational envelope of the camera, and a broadening of creative options.


*When Knee is set to 'Off'