When you shoot test charts, sometimes you learn things that aren't directly related to the charts.
Here's one thing I learned today: when you shoot charts, you're supposed to expose for the brightest chip set to clip (overexpose). I always thought you exposed the middle-gray chip (the one in the middle) to be set at 55 IRE on a waveform monitor. After all, this is how you properly expose a gray card. Apparently this isn't how it's done, as the DSC Labs instructions say otherwise. So, it turns out I've been doing it wrong for a long time.
Another thing I learned about the GH2 doing this test. The "Multiple" metering mode can bite you in the a** when shooting charts, and "Center-Weighted" is preferable. I didn't always fill the frame with the chart, and left a little black around the edges. Well, that little black around the edges "tricks" Multiple Meter mode into telling you to underexpose by 1/3 stop. For my own personal workflow, I'll use "Center-Weighted" metering from now on.
Anyway, I have some stills to post from footage I shot with the GH1 and GH2. These are just a small portion of many tests that I did. I may post those other results later. But here's what I found in a nutshell:
The 24mb 1080 24p is sharper than the 17mb 1080i. This may seem obvious, but the difference is striking. It looks like the GH2's 1080 24p is noticeably sharper than the GH1.
Also, I compared the Lumix 20mm f1.7 to the 14-140mm kit lens at the same aperture (f6.3) and found the 20mm to be sharper. Again, not surprising. What's interesting is that the kit lens seems to be 1/3 stop slower than the 20mm when set to the same aperture.
All shots are on Smooth setting with Sharpness at minus 2. Contrast, Color and NR are at 0. ISO 400.
GH2 Lumix 20mm @f6.3
http://farm6.static.flickr.com/5009/5315168682_8d3a91bdb5_o.jpg
GH2 Lumix 14-140mm, set to 20mm @f6.3
http://farm6.static.flickr.com/5123/5314572209_729cf2216c_o.jpg
GH2 1080i
http://farm6.static.flickr.com/5087/5314573703_15c7d40325_o.jpg
GH1 1080i
http://farm6.static.flickr.com/5245/5314575311_34622edced_o.jpg
Lastly, just to illustrate how important the glass is when judging image quality, contrast and color fidelity, look at the quad photo using the 20mm and three different 50mm lenses...a Canon FD f1.2, an Olympus Zuiko OM f1.8, and a c-mount Cosmicar f1.9. BTW, the Zuiko looks soft because it's wide open at f1.8. Also, if you look closely, you'll see the Canon has red/blue chromatic aberration. But the point is they all have different color casts even though they were all shot with the same white balance under the same lighting. I guess this is why it's important to have a matching set of lenses when shooting a film.
And a bonus photo: this is shot with the "Intelligent Resolution" set to "High". Notice the added white outlines around the text on the chart. That's not there in real life. It basically sharpens the image, which probably isn't good if you want a cinematic look.
http://farm6.static.flickr.com/5247/5315170140_83d1995bef_o.jpg
I did some tests regarding white balance settings and the dynamic range of different film modes, but I'll post them later. I may reshoot the charts and shoot some real-world footage. Hope these photos are informative.
Here's one thing I learned today: when you shoot charts, you're supposed to expose for the brightest chip set to clip (overexpose). I always thought you exposed the middle-gray chip (the one in the middle) to be set at 55 IRE on a waveform monitor. After all, this is how you properly expose a gray card. Apparently this isn't how it's done, as the DSC Labs instructions say otherwise. So, it turns out I've been doing it wrong for a long time.
Another thing I learned about the GH2 doing this test. The "Multiple" metering mode can bite you in the a** when shooting charts, and "Center-Weighted" is preferable. I didn't always fill the frame with the chart, and left a little black around the edges. Well, that little black around the edges "tricks" Multiple Meter mode into telling you to underexpose by 1/3 stop. For my own personal workflow, I'll use "Center-Weighted" metering from now on.
Anyway, I have some stills to post from footage I shot with the GH1 and GH2. These are just a small portion of many tests that I did. I may post those other results later. But here's what I found in a nutshell:
The 24mb 1080 24p is sharper than the 17mb 1080i. This may seem obvious, but the difference is striking. It looks like the GH2's 1080 24p is noticeably sharper than the GH1.
Also, I compared the Lumix 20mm f1.7 to the 14-140mm kit lens at the same aperture (f6.3) and found the 20mm to be sharper. Again, not surprising. What's interesting is that the kit lens seems to be 1/3 stop slower than the 20mm when set to the same aperture.
All shots are on Smooth setting with Sharpness at minus 2. Contrast, Color and NR are at 0. ISO 400.
GH2 Lumix 20mm @f6.3
http://farm6.static.flickr.com/5009/5315168682_8d3a91bdb5_o.jpg
GH2 Lumix 14-140mm, set to 20mm @f6.3
http://farm6.static.flickr.com/5123/5314572209_729cf2216c_o.jpg
GH2 1080i
http://farm6.static.flickr.com/5087/5314573703_15c7d40325_o.jpg
GH1 1080i
http://farm6.static.flickr.com/5245/5314575311_34622edced_o.jpg
Lastly, just to illustrate how important the glass is when judging image quality, contrast and color fidelity, look at the quad photo using the 20mm and three different 50mm lenses...a Canon FD f1.2, an Olympus Zuiko OM f1.8, and a c-mount Cosmicar f1.9. BTW, the Zuiko looks soft because it's wide open at f1.8. Also, if you look closely, you'll see the Canon has red/blue chromatic aberration. But the point is they all have different color casts even though they were all shot with the same white balance under the same lighting. I guess this is why it's important to have a matching set of lenses when shooting a film.
And a bonus photo: this is shot with the "Intelligent Resolution" set to "High". Notice the added white outlines around the text on the chart. That's not there in real life. It basically sharpens the image, which probably isn't good if you want a cinematic look.
http://farm6.static.flickr.com/5247/5315170140_83d1995bef_o.jpg
I did some tests regarding white balance settings and the dynamic range of different film modes, but I'll post them later. I may reshoot the charts and shoot some real-world footage. Hope these photos are informative.
Last edited: