1/3" 10-bit 4:2:2 vs. 1/2" 8-bit 4:2:0 for Dynamic Range ?

doondoon

doondoon

Well-known member
I'm wondering how the HPX300 will hold up against the solid 10 stops of the EX1.

EX1 has less pixel dense larger sensor (but has an 8-bit bandwidth limited codec).

VS.

The HPX-300 which has a smaller sensor with more pixels packed into it (but has a 10-bit 4:2:2 signal @ 100Mbps).

I'm hoping the HPX300 can come through with a solid 10 stops as well. Have there been any tests done on dynamic range?
 
The supposed boost in DR from the larger ex1 sensor doesn't make a difference when bit depth is concerned. It's still 8 bit, so you'll still have a higher chance of seeing banding than on the 10 bit footage. The 10 bit footage will be more tonally accurate, and the HPX300 quantizes the brightness from the sensor at 20bits intially which I believe is higher than the EX1's processing.
 
CovenantPictures said:
One advantage of the larger chips is less noise in the shadows by providing more signal.
Click to expand...
Yes, but -- not always. It's more about relative pixel size than it is about chip size.

By extension, it can be about chip size, because a 1920x1080 1/3" chip is going to have smaller pixels than a 1920x1080 2/3" chip, so therefore, all other things being equal, the 2/3" chip will have a lot less noise.

But, if you were to compare an SD 720x480 1/2" cam against an HD 1920x1080 2/3" cam (of comparable chip technology) it's possible the 1/2" cam would have less noise, because the relative pixel size would be so much bigger.

In general, like for like, yes a bigger sensor will have correspondingly bigger pixels, and an attendant reduction in noise because of it.
 
CovenantPictures said:
The red has been reported to get I believe 11 or 12 stops
Click to expand...
Red's testing on a Stouffer chart showed 11.3 stops of total dynamic range. Many people on RedUser have been saying that it has maybe 9 stops of "usable" range, which I guess is a subjective opinion. But chart-wise, it's 11.3.
 
Dynamic range is a measurement of contrast.

Where do you lose contrast? Everywere, the lens, the sensor, the processing, and the recording. The result is the product of all of them:

MTF(lens) X MTF(sensor) x MTF(processing) x MTF (compression)

Let's look at them individually.

Lens: roughly the same. Jan, will the HPX300 be available with the Canon in the kit? I think the brochure just mentions the Fujinon.

Sensor: One is 18 months newer design, the other is bigger. Same resolution. You would need to test to determine if one was better.

Processing: Testing would be necessary, but again the HPX300 is a newer design and technology runs, it doesn't walk.

Compression: Sony has a slightly faster HDV; 40% in 30p/60pi, 63% in 24p HQ. Panasonic has the latest H.264 technology recording at over twice the bit rate with more chroma depth (10 vs 8) and more sampling (4:2:2 vs 4:2:0). Sony brought a gun to a knife fight. There are ten pages back and forth in another thread comparing it to 100Mb/s XDCAM long GOP or 160Mb/s I=frame, not 35Mb/s.

If you are recording the HD-SDI output, then the recording advantage goes away. In that case the output from both is uncompressed 10-bit 4:2:2.

Last, your clients or workflow may favor one or the other in post. If your primary client requires XDCAM RAW footage, your options are limited. Or if they require DVCProHD or AVC Intra.

Net, both great cameras, with the Panasonic having a huge edge in on-camera recording resulting in a better on camera recorded image. If you recorded both through the HD-SDI port, I image operator skill differences would be bigger than camera differences.
 
I asked Jan about the lens as well. She said that only the Fuji lens would be provided with the camera. And as of now, anyway, Panasonic will not offer this package as a "Body only option".
 
You must log in or register to reply here.
Back
Top