Image quality

Out-of-camera JPEG using the Standard profile.
Panasonic Leica 10-25mm F1.7 @ 21mm | ISO 200 | 1/800 sec | F4.5
Photo by Jeff Keller

The DC-G95 uses the same (or very similar) 20MP Four Thirds sensor that we've seen on other Panasonic models, such as the DC-GX9. It performs well, though its Raw files aren't as malleable as those from cameras with larger sensors.

Key takeaways:

  • Image quality is strong at low to middle ISOs, with pleasing colors and good detail capture
  • At higher ISOs, noise in Raw images is more noticeable than on cameras with larger sensors, like the Sony a6400
  • Noise reduction is light at low ISOs for JPEGs, though it smothers fine detail at high sensitivities
  • Raw files can be brightened in post-processing letting you reduce exposure and better preserve highlights, with a fairly low increase in noise

Our test scene is designed to simulate a variety of textures, colors and detail types you'll encounter in the real world. It also has two illumination modes to see the effect of different lighting conditions.

At low ISOs, the G95 captures a reasonable amount of fine detail, and is competitive with the Sony a6400 and Panasonic's sports-oriented G9 (which we believe uses a newer sensor). At midrange ISOs, the G95 shows more noise, with the difference becoming rather glaring as the sensitivity increases a bit more.

Switching to JPEG: We think that the G95's color saturation is just about right, though it's not quite as punchy as the Fujifilm X-T30. Sharpening is solid, but not as sophisticated as Sony's implementation, which brings out more fine detail. Stair-stepping artifacts, present in JPEGs from most Panasonic cameras, shows up as jagged edges in curved and diagonal lines. Three stops above its base ISO of 200, the G95's noise reduction starts to take its toll on fine detail. At high ISOs, noise reduction leaves behind artifacts, and takes a toll on fine detail.

Raw dynamic range

The G95 drops to 10-bit mode in bracketing and continuous drive mode, lowering dynamic range. These tests were unintentionally shot in one of these modes, meaning they do not represent the camera's full capability. We will re-shoot and re-assess this test and the score as soon as we can.

Our ISO invariance test is designed to show how much noise the camera is adding, and hence how much it's minimized as amplification increases. We do this by using the same aperture and shutter speed at every ISO, thus removing shot noise from the equation. When amplification levels are low (ie: base ISO), the camera is able to capture more highlight tones. If you crank up the ISO in order to brighten the image, you lose those extra highlights.

In the case of the G95, there is a slight noise penalty to be had if you keep ISO down, then brighten the image by three stops, but in exchange you're getting 3 stops of highlight detail. Thus, if you're in a situation where you need to capture more highlight tones, and don't mind a little extra noise, then shooting at base ISO (but underexposing your image) then brightening in Adobe Camera Raw (or similar) might be worth considering.

In our exposure latitude test, we use the camera's base ISO and try brightening increasingly dark exposures, as you would do if you pull up the shadows from images you'd exposed to protect highlights. In the example above, you can see that when shutter speeds are matched, the Sony a6400 has less noise, thanks to its larger sensor and lower base ISO. The G95's sensor performance falls behind that of other four-thirds cameras like the G9: when given the same exposure and pushed the same amount, more noise is evident.

When you brighten the image even more, you can really see the difference in noise between the two cameras, illustrating how the a6400 allows for further brightening of shadows, without as big a hit in noise as the Panasonic. Not every photograph will require shadow-lifting (and not every photographer will require the ability to do it) but it's worth knowing how well preserved the dynamic range is when this is done.