Panasonic AG-DVX200 Tech Brief - Volume 3 - Page 4

FHD: 1920 x 1080 @ 23.976 to 59.94 Frames Per Second, FHD VFR: 1920 x 1080 @ 2 to 120 Frames Per - used

Page 4 highlights

FHD: 1920 x 1080 @ 23.976 to 59.94 Frames Per Second When set to Full HD mode, the camera delivers the widest eld of view possible (an equivalent to 28.0mm on a photography camera), and it also uses the most pixels to make its image (15.49 megapixels), and it operates at the fastest frame rates, which also means it produces the least rolling-shutter artifacts. How is this all possible? e answer is that the FHD mode uses a technique called "pixel mixing" to read several pixels out together as one large "superpixel". On the sensor itself, the area used to create the FHD image is approximately 5,248 x 2,952 pixels. at many individual pixels are not necessary for making an HD image (which is, of course, 1,920 x 1,080). In fact, it's almost eight times as many pixels as are truly necessary! at's a tremendous amount of data to read, especially at the higher frame rates. To cope with this amount of data, the DVX200 "mixes" each block of 2x2 samecolor pixels together into a "superpixel". In essence, the sensor averages multiple pixels together to create one large pixel that represents generally what the individual pixels had sampled. e camera then only needs to read o the "superpixel" data, rather than the individual pixel data. is reduces the processing load tremendously, allowing the camera to use the full width of the sensor, at the full frame rates, and because the scan happens so quickly, the rolling shutter e ect is greatly reduced. ere are three side e ects to pixel mixing: reduced resolution, reduced noise, and increased aliasing. e e ect of reduced resolution should be obvious: if each group of 2x2 pixels is mixed together to create one super-pixel, then obviously the image won't be as high resolution. e math behind pixel mixing is complex, but it results in su cient resolution for HD footage, and since pixel-mixing is used only in FHD mode, it is a solid and reasonable engineering choice. e second e ect is a notable reduction in image noise. Since four pixels are being averaged together to create each superpixel, the minor uctuations in brightness that are caused by noise are also averaged together, which results in the noise being less noticeable. In general you can consider that a "superpixel" will have approximately twice the signal-to-noise ratio that the individual pixels would have. is makes FHD mode footage noticeably less noisy than UHD or 4K footage. FHD VFR: 1920 x 1080 @ 2 to 120 Frames Per Second When set to record Variable Frame Rates (VFR), the camera may change its scanning mode and perhaps its eld of view, depending on the frame rate chosen. At frame rates between 2 and 60 frames per second, the camera uses the full width of the sensor and the same pixel-mixing technique as described above. Accordingly, images at these frame rates have the same resolution, sharpness, eld of view, and aliasing as standard FHD footage at 23.98, 25.00, 29.97, 50.00 or 59.94 frames per second.

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6

FHD: 1920 x 1080 @ 23.976 to 59.94 Frames Per Second
When set to Full HD mode, the camera delivers the widest field of view pos-
sible (an equivalent to 28.0mm on a photography camera), and it also uses the
most pixels to make its image (15.49 megapixels), and it operates at the fastest
frame rates, which also means it produces the least rolling-shutter artifacts.
How is this all possible?
°e answer is that the FHD mode uses a technique called “pixel mixing” to read several pix-
els out together as one large “superpixel”.
On the sensor itself, the area used to create the FHD
image is approximately 5,248 x 2,952 pixels.
°at many individual pixels are not necessary for
making an HD image (which is, of course, 1,920 x 1,080).
In fact, it’s almost eight times as many
pixels as are truly necessary!
°at’s a tremendous amount of data to read, especially at the high-
er frame rates.
To cope with this amount of data, the DVX200 “mixes” each block of 2x2 same-
color pixels together into a “superpixel”.
In essence, the sensor averages multiple pixels together
to create one large pixel that represents generally what the individual pixels had sampled.
°e
camera then only needs to read off
the “superpixel” data, rather than the individual pixel data.
°is reduces the processing load tremendously, allowing the camera to use the full width of the
sensor, at the full frame rates, and because the scan happens so quickly, the rolling shutter effect
is greatly reduced.
°ere are three side effects to pixel mixing: reduced resolution, reduced noise, and increased
aliasing.
°e effect of reduced resolution should be obvious: if each group of 2x2 pixels is mixed
together to create one super-pixel, then obviously the image won’t be as high resolution.
°e
math behind pixel mixing is complex, but it results in suffi
cient resolution for HD footage, and
since pixel-mixing is used only in FHD mode, it is a solid and reasonable engineering choice.
°e second effect is a notable reduction in image noise.
Since four pixels are being averaged
together to create each superpixel, the minor fluctuations in brightness that are caused by noise
are also averaged together, which results in the noise being less noticeable.
In general you can
consider that a “superpixel” will have approximately twice the signal-to-noise ratio that the indi-
vidual pixels would have.
°is makes FHD mode footage noticeably less noisy than UHD or 4K
footage.
FHD VFR: 1920 x 1080 @ 2 to 120 Frames Per Second
When set to record Variable Frame Rates (VFR), the camera may change its scanning mode and
perhaps its field of view, depending on the frame rate chosen.
At frame rates between 2 and 60 frames per second, the camera uses the full width of the sensor
and the same pixel-mixing technique as described above.
Accordingly, images at these frame
rates have the same resolution, sharpness, field of view, and aliasing as standard FHD footage at
23.98, 25.00, 29.97, 50.00 or 59.94 frames per second.