FZ1000 and D800E - a surprising comparison

[luxborealis]

In an earlier post, I described how I was looking for a travel camera with decent image quality, a decent lens and raw capability to extract every bit of quality I could - and I seem to have found it in the Panasonic Lumix FZ1000 with its 1" sensor and Leica lens. Now that I have one in hand, I thought about a simple test to compare it to what I am currently using - a Nikon D800E.

I've tried to replicate what I would typically do in travel and walkabout photography: hand-held, aperture priority using a small-enough aperture to give suitable dof at the lowest ISO possible that would allow hand-held shooting. I also set the lens to 25mm focal length (35mm equiv for the FZ1000) and chose a lens for the D800E I would often use, an 18-35mm. To be honest, I first used my 24-85mm but the corners were significantly softer than the FZ1000, so I switched to the 18-35. Chalk one up for the Leica lens!

I shot the same scene with both cameras, attempting to align the photos. It's a view of my backyard with snow - appropriate, because for me, maintaining highlight detail is important and I had read the FZ1000 had some trouble with that. Granted, the shadows are not being put to the test in this scene on a flatly-lit day like today.

Both cameras were set to their respective base ISOs: 125 for the FZ1000 and 100 for the D800E. The FZ1000 was set for ƒ5.6 while the D800E was set for ƒ8 - a closer approximation of equivalent dof than using the same aperture for both. Surprisingly, the shutter speed was identical despite a full stop difference in aperture and a 1/3-stop difference in ISO. Hmmmm.

After shooting, the raw files were imported and processed in LR6.1.1 and I attempted to match the FZ1000 colour and sharpening to the D800E files. What is surprising is they are not significantly different as you can see from the Develop Panels included below. Both underwent Lens Correction, automatically for the FZ1000 and with the correct profile for the 18-35 Nikkor. The only additional lens correction I made was to apply 4 points of Purple Defringing to the FZ1000 files, but that's minimal.

For the 100% screenshots, I exported the raw files as jpegs that matched the smaller size of the FZ1000 sensor, 5472x3648 (vs 7360x4912 for D800E). Given that both files would be printed to the same size (10x13", for example), this is not an unrealistic comparison.

I think I can successfully report the FZ1000 has stood up to the test. No, this isn't a perfect test with the best Nikkor optics available for the D800E, rather, this is a realistic test based on my usage. I find it encouraging that the quality of output is more similar than the vast price difference would predict. There is slightly more "grain" (see the steel fence posts) and not quite the acutance, however I am cautiously optimistic that it will perform well enough to leave the D800E behind on walkabouts. There are other subtle differences that I could spend an afternoon correcting, but they are minimal and do not warrant the time.

Further testing is needed in the golden hours, in particular to see how much the shadows can be pulled.

Actually, when I stop and think about it, I'm rather pleased with and excited by the quality of the FZ1000. There isn't quite the same acutance along the edges, but for being 1/4 the cost, I'm amazed. Factor in the additional weight of a FF kit to match even part of the range of the 25-400 ƒ2.8-4 Leica lens on the FZ1000 and you're looking at a very quality- and cost-efficient camera.

[Slobodan Blagojevic]

...The FZ1000 was set for ƒ5.6 while the D800E was set for ƒ8 - a closer approximation of equivalent dof than using the same aperture for both...

Thanks for the test, interesting.

As for the f/stop... wouldn't f/15 be an equivalent for Nikon, rather than f/8? This would also introduce more diffraction for Nikon, making it closer to Panasonic in resolution?

My first impression with your provided pictures is that Panasonic details lack depth and tonal variation, look flat and lack 3-D in other words, something we often hear when medium format is compared to 35mm as well.

[luxborealis]

Thanks for the test, interesting.

As for the f/stop... wouldn't f/15 be an equivalent for Nikon, rather than f/8? This would also introduce more diffraction for Nikon, making it closer to Panasonic in resolution?

My first impression with your provided pictures is that Panasonic details lack depth and tonal variation, look flat and lack 3-D in other words, something we often hear when medium format is compared to 35mm as well.

Thanks Slobodan, you are correct, and, perhaps a test using ƒ11 or ƒ16 would be more appropriate. It would surely give an additional edge to the FZ1000 with the increase in ISO and/or decrease in shutter speed needed by the D800E to match exposure.

Certainly depth and tonal variation is slightly impaired with the FZ1000, but I have yet to print it to see how much difference there is. As well, I might be able to bring some of that back with some additional processing. This is a "first" test under only one condition, so time will tell.

I'm realistic enough to understand the FZ1000 will not be able to replicate the output from a D800E. I'm more curious about how much the differences really are in my style of shooting and final output - for these shots, web, books and the odd print.

[Slobodan Blagojevic]

...I'm more curious about how much the differences really are in my style of shooting and final output - for these shots, web, books and the odd print.

That is critical, of course (personal use). I tend to print bigger and bigger these days (for art fairs), and am about to do a 36"x48". My biggest so far were 24"x36" canvas from an 8 Mpx file (Canon 20D, APS-C sensor) and 30"x40" canvas from a 15 Mpx file (Canon G10, jpeg, 1/1.7" sensor).

In other words, I have no doubt that your new camera will deliver great images, even if printed big. The only time you would be able to tell a difference is in a direct comparison like this. As viewers or buyers will never see the comparison, they'd fall in love with the subject matter and your interpretation, not technical details. If someone like me comes into your booth and starts complaining about the "lack of depth and tonal variation," just kick them out, they are tire-kickers, not real buyers anyway

[armand]

With the right scenario you can get good enough shots. It doesn't have much room for compensations though.

Here are 2 shots where it worked and tolerated postprocessing quite well. I have them printed on canvas at ~ 20"x13", and there isn't much fine detail to start with. My D750 probably wouldn't have been significantly better.

[armand]

And here is a shot where it wasn't good enough and where a better sensor camera would have been helpful.

[armand]

I'm uploading some files from FZ1000 and I suddenly recall another problem with it: the LR profile sucks! Big time. It loads as a lifeless photo, particularly the colors. I get a glimpse of the jpeg colors (from the embedded one) while the photo is loading but then everything goes away. Yes it can be tweaked but it's not easy, takes a lot of time and many times the colors are still not "right".
Did you find a decent profile anywhere out there?

[b_rubenstein]

I have a GX8 and LX100 and there are no camera specific profiles for them in LR either. I bought a XRite Color Checker Passport that comes with software to create LR profiles via a plug-in. The generated profile gives a much better starting point to work from.

[luxborealis]

I'm uploading some files from FZ1000 and I suddenly recall another problem with it: the LR profile sucks! Big time. It loads as a lifeless photo, particularly the colors. I get a glimpse of the jpeg colors (from the embedded one) while the photo is loading but then everything goes away. Yes it can be tweaked but it's not easy, takes a lot of time and many times the colors are still not "right".
Did you find a decent profile anywhere out there?

I must agree. The colours, even from the JPEGs, are "off". Sometimes they are quite good (outside with and without fill flash), but indoors, they are a dog's breakfast. It's even difficult to get a white balance that seems pleasing as often the shadows seem to have a different cast than the highlights, even when lit by the same indoor source.