Sony PXW-Z450 4K shoulder camcorder

[Glen Vandermolen]

Sony has finally made a 2/3" shoulder camcorder in 4K.

Sony PXW-Z450 Is First 4K XDCAM Shoulder Camcorder With 2/3-type image sensor

[Jack Zhang]

And so the death of 3-chip sensor blocks gradually begins...

I mean, it was very impressive that the HDC-4300 had a 3-chip 4K 2/3'' system, but moving forward we're going to see the death of 3-chip designs for 4K and beyond.

[Christopher Young]

I'm a bit confused? Like I mentioned in the PXW 400 upgrade posts.

If the 450 has a single sensor QFHD/4K conventional type Bayer CMOS sensor once it's de-bayered it's going to have a resolution of around 3.2K... correct?

A conventional 4K grid only has half the Blue and Red sensors so interpolation algorithms are used as part of the de-bayering process to bring up the Red and Blue count.

For example the F65 uses a true 8K diagonal grid to deliver a true non interpolated 4K image. For the 450 to deliver a true 4K from a single Bayer CMOS it's true resolution would have to be well in excess of 3840 x 2160 /4K.

Still confused as to how this camera will deliver true 4K resolution? For true 4K resolution I think they will have to stay with three 2/3" chip RGB style sensors where each chip carries a full compliment of R,G and B in 4K like the HDC series jack mentioned.

Chris Young
CYV Productions
Sydney

[David Heath]

Originally Posted by Jack Zhang

And so the death of 3-chip sensor blocks gradually begins...

I mean, it was very impressive that the HDC-4300 had a 3-chip 4K 2/3'' system, but moving forward we're going to see the death of 3-chip designs for 4K and beyond.

Are we going to see the death of 3 chip designs.....?

Maybe, but for a given size of sensor (say 2/3" in this case) - and with all the implications for lenses that involves - a 3 chip design will always perform better than an equivalent single chip.

With 3 chips and a block, all the light gets used - red, green and blue all get directed to a chip where they get used. That's not true in the case of single chip. For a photosite devoted to red - with a red filter in front - all the green and blue incident upon it gets absorbed, and similarly for green, blue. So it's inherent in single chip design that a proportion of the light doesn't contribute to the picture, or to put it another way, a single chip sensor will always have a lower basic sensitivity, than a 3 chip design of the same size, with all other relevant aspects equal.

Move up to s35, and a 3 chip design becomes impractical, and the much larger size more than makes up for the single versus three nature. But for 2/3", then whilst single chip may be cheaper, it can never have the same sensitivity as 3 chips - expect it to be about 1-1.5 stops worse.

Originally Posted by Christopher Young

I'm a bit confused?

If the 450 has a single sensor QFHD/4K conventional type Bayer CMOS sensor once it's de-bayered it's going to have a resolution of around 3.2K... correct?

Yes - totally correct!

Originally Posted by Christopher Young

Still confused as to how this camera will deliver true 4K resolution?.

If it has a single QFHD Bayer sensor, it won't deliver such resolution - end of story. But... does it matter?

The term 4K seems to be used at the moment to define any camera with a sensor of such dimensions - NOT a camera giving "true 4K resolution". And remember such is the same of many other "4K cameras" - right the way up through such as the FS7, to such as the Varicam and the F55. Yes, it's true that F65 will give better results in that respect, but currently a QFHD Bayer sensor still qualifies for a "4K" badge.

[Christopher Young]

Yes David that's basically my reading of it.

Though I think you will only get the best results out of this type of setup using the latest generation of UHD lenses with the latest ALAC firmware that can talk to the camera's lens file setup to store, register and compensate for any inherent lens aberrations. The smaller the sensor the more acute the angle of convergence and registration of the light path becomes, especially when using lenses designed for three sensor prism registration. Without good communication between lens ALAC and the camera ALAC circuits I would anticipate some not so good results with regard to CA, Coma and edge resolution for starters. Time will tell I guess.

Chris Young
CYV Productions
Sydney

[Anthony Lelli]

[QUOTE
a 3 chip design will always perform better than an equivalent single chip.

With 3 chips and a block, all the light gets used - red, green and blue all get directed to a chip where they get used. That's not true in the case of single chip.


[/QUOTE]

exactly: it comes down to low light performance. I see it on my old pmw-400 , night and day compared to any single chip handycam
But the technical barriers are nothing compared to the marketing barriers. As I (firmly) believe that all the cameras below a certain price point are made darker in purpose.
For that reason I'm going to assume that the low light performance of this single chip 400series may be just fine in low light. But it will need to be seen before buying.
Go to a stadium , shoot 5 minutes of a night football practice from the press box, set the shutter to 1/120 shoot 60p HD and see how the lens wide open produces with standard gain increases. There are no tricks there: either is good and bright in low light with a decent noise level or is not.