G6 resolution difference between 25p and 50p

[Noa Put]

Recently I have been shooting with my gh3 and g6 and noticed a resolution difference, the image looked softer compared to my gh3 when I shot in 50p, I first thougth I might be imagining it untill I just shot a resolution chart (nothing scientific, just printed such a chart and shot it with the gh3 and g6, both with a lumix 12-35mm at 35mm and f2.8 shutter at 1/100 and 800 iso and all shot in the avchd mode and whitebalanced to outdoor (the light I used was 5600k). The image was not completely filled with the chart but that doesn't matter that much as it was only to see if there was any difference.

To my surprise the g6 is softer in 50p mode compared to 25p and what I thought the gh3 is also a tiny bit sharper in 50p mode compared to 25p for the g6. There is no visual difference between 25p and 50p on the gh3. This also explains why I saw a difference in sharpness. Sorry that the image is a bit small but it's easy to see, I"ll try to upload a larger one, what camera and the mode used is in the title of every photo.

Below is a 300% crop from the image so you can see the difference better, is there anyone with a g6 that has the time to confirm this?

[Noa Put]

Here another one, it's a quite large crop but don't know the percentage here, but that doesn't matter as the difference shows, maybe I"m doing something wrong. I find it strange that no-one else has noticed this?

[Noa Put]

I also just notice now that the 50p image is strecthed vertically a bit compared to the 25p one on the g6. The resolution loss also appears to be more vertically then horizontally, lines that run vertical on that chart are somewhat the same, only the ones runining horizontaly show a obvious softness.

[Nigel Barker]

I'm happy to run a test with my G6 but in the interests of comparability I would like to use the same chart. Do you have a link?

I don't have the 12-35mm F/2.8 just the 14mm F/2.5 & the 14-42mm F/3.5-F/5.6 PZ.

[Noa Put]

I used this one: http://koti.welho.com/msalon19/img/R1-ResCH-LG.jpg

Type of lens shouldn't matter, I get the same result with my tamrons.

[Noa Put]

This is also interesting to see how a lens renders an image and how different they can be, below is the lumix 12-35mm on the left en the tamron adaptall II 35-80mm on the right, all camera settings where the same with both lenses, you need to zoom in on the image to see the difference in color and you"ll also notice the blacks "bleed" at the edges on the tamron where the lumix renders it pitchblack all the way.

[Pete Carney]

I have run the test between 30p and 60p and while there is a tiny difference it is nowhere near the difference you are getting between 25 and 50p on your chart.

I did my test with an Olympus OM Zuiko 24mm MC f/2.0 at f/8.0 and iso 400. With a manual lens it removed any influence from camera processing based on lens profile. I can see a difference at 400% crops, but it is indistinguishable at full size even on a 2.5k monitor.

I filmed this morning and switched to 30p just to see, but I forgot that I also went from 1250 to 400 iso at the same time, so of course the 60p at 1250 iso is much noiser and softer :)

I'll try some more real world tests and see if I can reproduce any significant difference. I don't have a 12-35 f/2.8 yet, so I'm not so sure your results aren't from that particular lens and the internal processing being used there.

Cheers,
Pete

[Noa Put]

On my tamron lens there is no communication with the camera either and there I also see the image softening up, it only looks a bit worse on the lumix lens as that one is sharper. I"ll redo the test tomorrow to double check if I did something wrong. Curious to see what Nigel finds.

[Nigel Barker]

I have just shot some very rough charts & I am pleased to find that I don't see the softness with 50p. Here are the complete charts.

14-42mm kit lens at 14mm & F/3.5 ISO 800 1/50 AWB

[Nigel Barker]

Just for completeness sake these are shot with the 14mm lens at 1/100

14mm F/2.5 kit lens at F/3.5 ISO 800 1/100 WB Tungsten

[Nigel Barker]

Finally the same chart shot with the Olympus OM-D E-M5

14mm F/2.5 lens at F/3.5 ISO 800 1/50 AWB 30p

[Pete Carney]

I did another test today and here's a gif animation of an 800x600 crop from 30p and 60p frames.

There definitely is a bit more sharpness with 30p than 60p. I'm going to have to experiment some more with 30p to see if it translates to my final output.

http://www.c3di.com/images/MP4-1080-30p-vs-60p.gif

If you zoom in your browser to 200% and 400% you can really see the difference.

Thanks for bringing this up again Noa. There's more there than I had ever suspected earlier.

Cheers,
Pete

[William Hohauser]

I do see a difference between the 25p and the 50p with both lenses. Look at the vertical rows with the horizontal lines going from 100 lines of resolution on the top of the chart to 1000 lines to the bottom. If you switch between the 25p and 50 while paying attention to the area around 700 lines, you'll see the image go from aliased in 25p to gray mush in 50p The first reason I would ascribe this effect is the compression. The camera is cramming twice as much information into a 50p stream than a 25p stream.This will lead to some sort of image degradation. Interestingly, the vertical lines do not exhibit the same effect at 700 lines.

[Noa Put]

Originally Posted by Nigel Barker

I have just shot some very rough charts & I am pleased to find that I don't see the softness with 50p. Here are the complete charts.

It's weird because I do see a difference between the 2, the 25p chart has some weird kind of artifacts in the image that the 50p image doesn't have which is mostly noticeable on horizontal lines, almost looks like aliasing, something fishy is going on for sure.

I can tell that with my gh3 there is no visible difference whatsoever between 25p and 50p when I shoot that same chart, even when I magnify a lot, they look identical, only not with my g6 where I see a difference. The main question is how much that show on real life footage, this week it's mostly raining where I live so there is not much chance to shoot outside but as soon as I can I"ll make some comparisons, I do have a paid assignment this Thursday where I am shooting with the gh3, might take the g6 along and make a comparison shot when time allows.

You might question if it makes any difference but since I could pick out the g6 shots from that holiday movie I made for me it does matter if I"m planning to combine footage, if 25p proves to be better matching then the g6 will be used in 25p and the gh3 in 50p as that last one is often on the steadicam and I"d liek to slow footage down when I can. But I can easily mix both formats in edius without it showing.

[William Hohauser]

Resolution charts have to be very carefully set up or you well get aliasing and rainbow-like moire in the areas that approach the resolution limit of the camera on the pure horizontal or vertical lines. But even with a chart that is off center and at a slight angle, you can get a lot of useful information about the imaging sensor. How the camera reacts to detail near it's limit speaks to the quality of the design that went into the camera.