GH2 XF300 vs GH2 Resolution Test

[Isaac_Brody]

Has anyone ever posted resolution charts comparing the GH1, AF100 and GH2? Seems kinda like an obvious thing people would do but I can't find anything...

Bruce Allen
www.boacinema.com

Have you shot with a hacked GH2? It is perhaps unfortunately for the AF100 in a different league now. I have shot with both, and the difference between them in sharpness and IQ is night and day. I would say don't knock it till you try it. I was completely surprised by the results.

 

[Kholi]

Has anyone ever posted resolution charts comparing the GH1, AF100 and GH2? Seems kinda like the obvious thing people would do...

Bruce Allen
www.boacinema.com

There actually might be resolution charts on Personal-View with post hack stuff. I wouldn't guess so, since the hack is evolving constantly. Hell I just patched V2 and now they're up to 5...

But Blooms little shootout thing is the closest comparison to AF100 and hacked GH2, or newest.

 

[cowpunk52]

The colored moire is def more pronounced at the FD lens - however, they had slightly different frames and it looks like the frequency through the FD lens was hitting the moire sweet spot.

One thing about the FD's that I noticed doing this test is that they have significantly lower contrast than the m4/3 lenses. I really had to do a number to the curve in order to make the images come close to matching. The intensely low local contrast on Canon FD's do make them appear to be much softer - before I applied contrast I really thought the m4/3 was resolving a lot more detail. But when I made the FD lens punchier, it became clear that there wasn't more resolved detail in the m4/3 version, it was just an illusion.

I think that low contrast look gives the FD lenses it's cinematic mojo - and I think that's the same way the Voightlander's and SLR Magic lenses get their cinema mojo. The m4/3 lenses do appear sharper, but I attribute that mostly to their higher contrast (thanks to better coatings), but especially in-cam processor sharpening, which in reality is just increased local contrast - and unfortunatly I'm not sure that can be turned off.

 

[DBP]

I just applied the Aquamotion V2 hack the other day. It's 100mbs.

My cat has been my main test subject with GH2 experimental vids, and the biggest difference I notice between the hack and non-hack is that small details in her fur are preserved much better. With the stock codec, some areas turn a bit mushy.

 

[Isaac_Brody]

Try Quantum V2 or the latest V5, it will blow your mind with the details it's retaining. Also, I recommend running your footage through 5DtoRGB before editing.

 

[cowpunk52]

Try Quantum V2 or the latest V5, it will blow your mind with the details it's retaining. Also, I recommend running your footage through 5DtoRGB before editing.

Yes, 5DtoRGB is really a fantastic tool. I use it pretty much exclusively now, and the pro batch version is worth the $50 - especially since that cost helps out developers who are helping us out.

 

[DBP]

Yes, 5DtoRGB is really a fantastic tool. I use it pretty much exclusively now, and the pro batch version is worth the $50 - especially since that cost helps out developers who are helping us out.

I've tried the free version, which I like. Does anyone know how it compares to Cineform Neoscene?

 

[Kholi]

Yes, 5DtoRGB is really a fantastic tool. I use it pretty much exclusively now, and the pro batch version is worth the $50 - especially since that cost helps out developers who are helping us out.

Quite seriously, it should be a stapled part of the workflow. You have to transcode anyway, the results are obvious enough to warrant the purchase.

 

[Ryan Farnes]

So does 5DtoRGB actually extract a little more detail from raw MTS files? (or however the video comes out of the camera)

 

[cowpunk52]

So does 5DtoRGB actually extract a little more detail from raw MTS files? (or however the video comes out of the camera)

I smell another test coming on...

But in all seriousness - yes, it does. Go to their website at rarevision.com and read about how it works and what exactly it does. If you want to squeeze every last drop of quality out of your juicy juicy DSLR, it will help you do it.

 

[Danielvilliers]

Totally agree with these findings.

The GH1, GH2 and AF100 all only seem to give about a little higher than 720p resolution to me.

I said that on Barry's AF100 vs FS100 thread (very few people agreed with me - probably because they were only watching the 720p Vimeo file and couldn't see the difference then?)...

I'm not sure why this isn't obvious to everyone else.

Personally, it looks like I did a CG render with too low anti-aliasing settings and slightly too low resolution, then did a blow-up and do a combination of blurring / sharpening to make it "acceptable" 1080p.

Of course, crisp 800p or whatever might be absolutely fine for most folks' purposes.

Bruce Allen
www.boacinema.com

If you look at the latest Philip Bloom shootout, and download the 1080p images embeded into his page you will see that the hacked gh2 is smidge lower than the C300. I mean I am not even sure because of the different FOV and moving target. It is definitely higher than the other DSLR and at least the same than the F3. The C300/f3/gh2 are really a tie in sharpness. If the c300 is supposed to resolve 1000 TVL then the gh2 is very very very close to that. The af100 example is just terrible, but the subject was perhaps moving.