Liza Witz 4 years ago Here's what I'd like to see: -- Use AVC-Intra 50 instead of AVCHD, or have it as an option. This is a panasonic created format that Final Cut Pro edits natively. Its also 50Mbits/sec and has a deeper color space and more 10 bits per pixel. This makes the camera "pro" answers the naysayers who bitch about bitrate and means there's no log and transfer process for anyone using FCP. -- Allow pre-set focus positions with a toggle between them and a user configured rate of adjusting focus. EG: a software defined focus puller! This would be so useful and is so cheap, doesn't even need to be a custom button, just a mode on an existing button. -- CMOS Global Shutter- Would eliminate the issues of rolling shutter. Or if global shutter isn't practical, then a distributed regional shutter. (eg: a bit of circuitry on the chip that will read out a region of pixels very fast, and store the values locally so that the data can be read off later. RAM is super cheap, RAM can be built on a cmos process, so why not integrate RAM on the sensor? Seperate out the process of *reading* the frame from the chip from the process of *capturing* the frame. (for those who don't know what I'm talking about the hard part is pulling a 8megapixel data stream off of the sensor 60 times a second. So what happens is there is no buffering and the camera just reads the values off line by line, while the image is still being captured. This is what causes the jello effect.) -- Dithering the full resolution image down to HD resolution rather than dropping pixels. All these DSLRs have to convert the full resolution image to HD resolution before encoding it..... the sensor is producing 8 megapixels (or 4,000 pixel across) images when in 16:9 mode, but HD is only 1920 pixels across. Rather than just dropping every other pixel, it would be nice if the camera could take every four pixels and average them down to one, to turn the 4000x3000 sensor image into a 1920x1080 video frame (about three to one vertically, and 2:1 horizontally) This would immediately boost high ISO image quality by 2-4 stops, making the GH2 outperform anything out there right now. It would also make the image quality of every frame in every situation much better.... and we'd be really using all the data the sensor the sensor is gathering. This would double the lines horizontally and triple vertically, producing a dramatic improvement. However, it would also mean a fast method of reading data off the sensor, or some processing on the sensor (possibly combined with the regional shutter idea I mentioned earlier) to make it work. ------ What I think we might see: -- Higher dynamic range sensor We know panasonic is working on this so its probably to be expected. -- Some more lenses, maybe more HD type lenses. And that's about all I can think of that we should expect. I think the odds are probably very good that there is a GH2 and that it will be announced in the first half of 2010, probably in the first couple months of 2010. But I have no idea what the obvious next step is for for panasonic--- unless they're working on something like the features on my wishlist. AVCIntra is technically possible but they may be reserving it for "pro" products. My sensor design ideas may not be technically possible, though from what I know of HVLSI design they are probably feasable (though they may be too expensive) ..... but they seem like the direction that we'd want them to go. The key issues facing these kinds of cameras are: image quality, low light sensitivity and aliasing and rolling shutter artifacts. So I have to figure this is where panasonic would be spending time. These issues affect all cameras that use 4/3 sensors for Video-- but then, is this format really for video, or is video just an add on? If the latter, then they will likely be pursuing other features.