GH2 Tele Conversion Function

That clip is beautiful, whomever said that EX tele would be one of the most enviable features of the GH2 from those of other camps was dead on.

Strangely I get a real Macross Plus vibe from that video.
 
sasi_yaad said:
Crop Factor

35mm = 1
super 16 = 2.97
2/3" = 3.93
GH2 1080 = 1.8 x 2.6 = 4.68
1/2" = 5.41
GH2 720 = 1.8 x 3.9 = 7.02
1/3" = 7.21

Theoretically 2/3" lens will not vignette but 1/2" lens will vignette on the GH2 1080 mode,
try to use those numbers to figure if a certain lens will vignette on the GH2.
Click to expand...

I've found this graphic on http://www.dpreview.com/galleries/8409887347/photos/558543/gh2-sensor-aspect-ratioswhere where is clearly visible that the sensor diagonal size does not change when the aspect ratio is switched. Even in 16:9 mode, the diagonal is 21.6mm with a 2X crop factor, not a 1.8X.
 
Yes, I've just explained it in the main gh2 thread, the GH2 has a horizontal FOV of 1.8 in 16:9 aspect, video and stills, this is not the crop factor which is dictated so theory goes by the dia circle, however, I argued that that tthe camera and crop now think the dia circle is wider, given that this was originally calculated for a 3:2 aspect(FULL FRAME) and doesnt apply to a 16:9 aspect. In any event it is what it is, but it aint no x2 crop in 16:9 ratio. Just to prove the situation, someone recently compared a Canon with a 400mm lense in video with the GH2 and the 300mm. I think they were expecting to see similar results, in fact what they saw was an effective 640 FL vs 540 FL, which proved without doubt that the GH2 in 16:9 in 1.8 fov.
 
Adventsam said:
Yes, I've just explained it in the main gh2 thread, the GH2 has a horizontal FOV of 1.8 in 16:9 aspect, video and stills, this is not the crop factor which is dictated so theory goes by the dia circle, however, I argued that that tthe camera and crop now think the dia circle is wider, given that this was originally calculated for a 3:2 aspect(FULL FRAME) and doesnt apply to a 16:9 aspect. In any event it is what it is, but it aint no x2 crop in 16:9 ratio. Just to prove the situation, someone recently compared a Canon with a 400mm lense in video with the GH2 and the 300mm. I think they were expecting to see similar results, in fact what they saw was an effective 640 FL vs 540 FL, which proved without doubt that the GH2 in 16:9 in 1.8 fov.
Click to expand...

Did they measure the effective focal lengths accurately or did they just make an assumption based on the difference they saw? If they just assumed that the canon would be 640mm equivalent based on a 1.6 crop factor they probably made a mistake, because in 16:9 ratio the canon would have to crop part of it's image and therefore reduce it's angle of view. So it would be the Canon's crop factor becoming larger not the Panasonic's becoming smaller.

The Gh2 never uses the entire sensor so the diagonal of the whole sensor doesn't matter. It just crops it differently based on the chosen aspect ratio. The max diagonal is determined by the image circle and it should be the same for all aspect ratios. Of course, probably no camera uses the whole image circle, because the image quality deteriorates towards the edges. It could be GH1/2 are a little less conservative and go a little closer to the edges of the image circle, but I'd like to see some actual examples of this.

Edit: Nevermind. I guess you were just talking about horizontal FOV.
 
Last edited:
Any differences between ETC mode and full-sensor mode in terms of color sampling?

Any differences between ETC mode and full-sensor mode in terms of color sampling?

It's clear that one of the advantages of the Extra Tele Conversion mode (ETC) is that you can avoid line skipping, and binning. But because, in this mode, the ratio between photosites and pixels is going to be 1:1 (and assuming a kind of Bayer color filter pattern), well, I would predict some drop in color resolution. This, because having more photosites relative to the number of output pixels should allow the system to interpolate better the 'real' color that is reaching that particular area that the pixel is representing.

As far as I know, this is the strategy that you find in video cameras like Arri Alexa and the new Sony F3. Alexa sensor is 3.5k (horizontal res) and F3 sensor is 3.43k , so there is really not a 1:1 correspondence between photosites and pixels there (e.g. Alexa has a real resolution of 3392 x 2200 photosites which is then downsized to 1920x1080 pixels). The approach here is trying to find a good compromise between maximizing photosite size (bringing photosite count close to the HD output of 2MP) and maximizing photosite count (to get a good color sampling). Also, increasing photosensor size also has the positive effect of improving sensitivity, and signal to noise ratio.

On the other hand, the ETC mode is more like getting the image quality in still mode, which we know is better than the full sensor video image. But don't forget that because we are only capturing the center of the image, a proper comparison with the still mode will consider magnifying the ETC image by a factor of 2.4 or so, and this will obviously decrease the resolution of the image compared to an equivalent FoV in still mode. However, ETC mode should affect resolution (sharpness) compared to still mode, but not necessarily color sampling, which should stay the same.

So far so good comparing ETC mode with still mode. My concern here is what happens to color resolution when you compare ETC mode with full-sensor video mode. Shouldn't we expect some drop in color resolution in the ETC mode compared the full-sensor video mode? This because having more photosites relative to the number of output pixels (which is the case in full-sensor mode) should allow the system to interpolate better the 'real' color that is reaching that particular area that the pixel is representing. Any comments on this?
 
SteB said:
One question I have wondered about is whether with extra teleconversion function, the rolling shutter effect will be less of a problem or the same? What I mean by this is whether the read out from the sensor is faster because it is only using the central portion. My only reason for asking this is one of the main reasons I will be getting a GH2 is for some macro work and long lens stuff, both of which obviously exaggerate the jello effect. It just occurred to me that it might be different in the cropped mode. So I'd be interested to hear anyone's experiences when they get to try out their GH2's as to if they find the extra teleconversion function improves the rolling shutter effect or doesn't make any difference. I appreciate that if you simply use this feature at the same focal length, that the movement would be exaggerated because you have narrowed down the field of view, essentially giving yourself a longer lens. So I mean at roughly the same angle of view.
Click to expand...

Some testing I did shows rolling shutter to be less in ETC mode at an equivalent field of view, not surprising since only a small part of the sensor has to be read out instead of the whole thing
 
richtrav said:
Some testing I did shows rolling shutter to be less in ETC mode at an equivalent field of view, not surprising since only a small part of the sensor has to be read out instead of the whole thing
Click to expand...

Thanks for the insight. I did wonder if that maybe the case as the readout would be from a smaller proportion of the sensor. What I didn't know was if the sensor would be read at the same speed in crop mode.
 
Why is everybody rushing into wide C-mount lenses?

Surely a Lumix 7-14mm would do a better job...and be more versatile ie switching from full frame to crop. I don't have a GH2 yet...Australia comes just before Outer Mongolia for delivery priority. Has anyone tried crop mode with a 7-14mm?
 
isnt the 7-14mm a little expensive to just throw away alot of that beautiful wide angle imagery with ETC? i plan on using that lens in cinema mode and get as much of that awesome wide image as i can
 
Skeptikal said:
Why is everybody rushing into wide C-mount lenses?

Surely a Lumix 7-14mm would do a better job...and be more versatile ie switching from full frame to crop. I don't have a GH2 yet...Australia comes just before Outer Mongolia for delivery priority. Has anyone tried crop mode with a 7-14mm?
Click to expand...

You can get insanely faster lens in c-mount
 
Jeremy_Sawatzky said:
isnt the 7-14mm a little expensive to just throw away alot of that beautiful wide angle imagery with ETC? i plan on using that lens in cinema mode and get as much of that awesome wide image as i can
Click to expand...

The 7-14mm is superb for underwater work...the big potential advantage of the crop mode is that it will hopefully turn the 7-14mm into a macro lens as well. An incredible advantage for dual-use during the same dive...without needing to surface and change lenses and ports.

I'm just hoping crop-mode works well with the 7-14mm...in principle it should.
 
You must log in or register to reply here.
Back
Top