See post 1892.
"Of course you can deinterlace 60i but you will never get real 60p out of that."
And i can get SD out of HD material but do I want that?
Anyway you are right, we should stop here.
GH2 official discussion thread
- Thread starter Vitaliy Kiselev
- Start date
"Of course you can deinterlace 60i but you will never get real 60p out of that."
Anyway you are right, we should stop here.
It can be argued that in the realm of digital anyone can really create anything. Want to turn that old clunky VHS tape into "FullHD" footage? No problem, there is a process for that. Anything is possible and anyone (granted access to the right tools) can be the Creator. The only question remaining is how much aliasing is introduced in the process? How close will end result be to the original?
In case of 60i-to-60p process, the end result will very much look like the 60i it was "created" from. How close will it resemble the original 60p sampled from the sensor however? And that is the question that is harder to answer, because there are so many variables in the equation -- complexity of detail and motion characteristics. One thing is known with certainty however -- SD-level vertical resolution in the worst case scenario. For some people this is unacceptable.
Very sharp with a not too large DOF, but the typical 25p stuttery motion, even though the kid is moving slow.
I do not like that.
I prefer 720/50p because it is smoother.
It's still different to "every other interlaced camcorder" since you never get a true progressive picture out of them when deinterlacing the video (aliasing). With the Gh2 you can as I see it since the picture is progressive from the beginning. So it means that we'll get true 25p/30p. With a hack we'd get true 50p/60p @ 1920x1080
Last edited:
Xian, when you sample a progressive source there are two ways you can go about constructing interlaced output of the same frequency. There is enough evidence out there that Panasonic made the choice described by lpowell. You are right however about what is possible with the hack.
Ok, I haven't read anything about it. I just thought it makes sense but I guess there is always a way to make things worse than you have to. It's funny that Panasonic want's the 50p/60p sensor output to be pointless when it could have been awesome. I guess it's too early to give it to us 2011. Hopefully someone else will 
Last edited:
Ok, I haven't read anything about it. I just thought it makes sense but I guess there is always a way to make things worse than you have to. It's funny that Panasonic want's the 50p/60p sensor output to be pointless when it could have been awesome. I guess it's too early to give it to us 2011. Hopefully someone else will
If they gave us 1080p @ 60 FPS in the GH2 then what would they give us in the GH3? There is only so much you can do when you are limited to a 4/3” sized sensor. Eventually, you are going to max out the sensor technology and you are left with only software upgrades.
Rick Burnett
Section Moderator
They are far from maxing out the 4/3" sensor size. I'd say it's not the size of the sensor that matters as much as the production methodology. I work in CMOS design, and most of the limitations on CMOS technology start and end with the fab process. As those technologies get better (and they do ALL the time) they can make sensors that don't have as much variance across them, which is one of the biggest problems in CMOS technology). The bigger a sensor, the more variance across them. The more voltage drop to the center of the sensor for switching circuitry. The more delay of signal data to get out of the farthest pixels. Given the RED sensor can handle 120fps at 5k resolution, imagine a 2k sensor and how fast that could be made! Given the m4/3 is not that much smaller than the RED MX sensor, clearly there is room for improvement in the future.
IBIS, Global shutter, 4k resolution, improved DR and noise, HDR, improved pre-amps, built-in variable ND, GPS, WiFi-n, 3G/4G/HSDPA modem, a life time supply of pizza and beer, and a £500 voucher for the local knocking shop
Matthew P
Well-known member
Vegas should be able to do it just fine. Just set the framerate to double that of the interlaced footage, and set the 'deinterlacing method' to 'none'.
It must be exact though, and you'll need to pay extra attention to NTSC framerates as they're unhelpfully fractional.
Why would I turn deinterlacing off? I want to deinterlace the content. And doubling the 59.94 frame rate would be 119.88. That would just be wasting bandwidth.
This just in! Here's some GH2 MTS footage, scroll down to the bottom and right-click and save.
http://translate.google.com/transla...p/docs/review/newproduct/20101117_407184.html
http://translate.google.com/transla...p/docs/review/newproduct/20101117_407184.html
Well it looks like I am wrong then. Thanks for the post. lot's of good info there and I hope they do continue to push the limits of the 4/3s sensor.
We have had that footage for days now. In all of its interlaced Glory/Gory.
haha, i'm late to the game.
Why is it so interlaced? I take it has to be converted in post?
Here's a much better set of GH2 files
http://www.mediafire.com/?pcdr5yo287wnd
source: http://vimeo.com/16844081
http://www.mediafire.com/?pcdr5yo287wnd
source: http://vimeo.com/16844081
Matthew P
Well-known member
You have to turn deinterlacing off because you're doubling the framerate. Vegas will interpolate the footage into progressive...
And it wouldn't be 119.88, as that would be double 60p, not double 30i/60i (it may be 60 fields a second, but it's still 30 interlaced frames...).
Try rendering with half the vertical resolution (540) with a pixel aspect ratio of .5, if you don't want to waste bandwidth, and if your output format supports it.