[FozzFoster]

Otherwise it will be a nightmare.

I agree. Sorry all - I had no bases to state my assumption and it was off topic for the thread anyway

[bencoskater]

No worries, it was not for you especially. It's a general reminder to prevent a closing action of the thread from an administrator.

[Rondec]

Ahk I should have said "I personally wouldn't expect 'decent' 4K".
As in, "good enough 4K to appease the critics".

I imagine it depends on what image processing engine they've chosen to replace PRIME. Certainly Pentax isn't a video company -- they've made that clear -- but I would imagine that this sensor can do 4K read out speeds and so it is more a matter of them enabling it and using a decent codec for it. That will be more about licensing someone else's software like they did with Milbeaut, but I don't know much about this sort of thing.

[repaap]

I imagine it depends on what image processing engine they've chosen to replace PRIME. Certainly Pentax isn't a video company -- they've made that clear -- but I would imagine that this sensor can do 4K read out speeds and so it is more a matter of them enabling it and using a decent codec for it. That will be more about licensing someone else's software like they did with Milbeaut, but I don't know much about this sort of thing.

Edit: they don’t seem to have 4K at GR it is only theta, which is different all together. Dunno, it could be just fullHD. Just like it is in older cameras. They do have now new processor thou. And sensor. So, I would not bury the hope for basic 4 K 8 bit or so. Who knows.

---

Last edited by repaap; 09-19-2020 at 05:28 AM.

[angerdan]

My guess is that Pentax may emulate [fellow Milbeaut image engine user] Nikon in terms of their AF evolution, and also maybe a few other camera functions.

The D6 does all this with a relatively old 180k pixel RGB metering sensor (introduced with the D5 in 2015, I think), working alongside the AF system, so it wouldn't be a big step for Pentax to also upgrade their 86k imaging sensor to something that could work the way the D6 works.

Using the metering sensor to use face detection in PDAF like with Nikons 3D-Tracking Face-Detection would be a great feature:
a4: 3D-Tracking Face-Detection

[gatorguy]

Using the metering sensor to use face detection in PDAF like with Nikons 3D-Tracking Face-Detection would be a great feature:
a4: 3D-Tracking Face-Detection

The Nikon D500 seems to do pretty well on that front, at least according to those owning it, and also making use of the RGB metering sensor just as the new K3III may use.
FWIW Pentax is 306k and the D500 is 180k so the hope would be it's as effective as the Nikon if I understand all this.

Is that the same as the Nikon 3D tracking you've mentioned a couple of times?

[MMVIII]

The Nikon D500 seems to do pretty well on that front, at least according to those owning it, and also making use of the RGB metering sensor just as the new K3III may use.
FWIW Pentax is 306k and the D500 is 180k so the hope would be it's as effective as the Nikon if I understand all this.

Is that the same as the Nikon 3D tracking you've mentioned a couple of times?

306k would be very close to VGA resolution with 640x480. It is described as a RGB Ir sensor, I have no idea what the arrangement of the pixels is/how the color filter is arranged.
But if the luminance of all pixels can be used it is actually quite a fine grained picture to work with for pattern recognition. And the color information for tracking algorithms.

[angerdan]

Is that the same as the Nikon 3D tracking you've mentioned a couple of times?

Yes, i think so.

[Kunzite]

This patent seems related to the new RGBIr metering sensor:
??2021-5762 | ???????IP Force?

Not very interesting I guess, at a glance they were trying to reduce the cost for such a sensor.

[MMVIII]

This patent seems related to the new RGBIr metering sensor:
??2021-5762 | ???????IP Force?

Not very interesting I guess, at a glance they were trying to reduce the cost for such a sensor.

Thank you! Could be. My Japanese reading skills are zero, my patent reading skills are almost zero, Google translate translation skills are so so. However, they describe a new way to apply color filters to photosensorarrays. Thus using photolithography for a R G and B filter layer stack and as all of them transmit IR a combination of two shields all visible light and can be used as for IR pixels.
They speak of a Bayer filter array where onebof the green pixels is substituted by IR.
This would effectively mean the 306k sensor to resolve 320x240 pixels after interpolation if my math is correct?
And there is the IR image. They say "In addition to acquiring an infrared image, the IR pixel can also be used as a dark correction pixel that corrects a pixel signal in the visible region. The IR pixel can also be used as a pixel for infrared removal correction that removes an infrared component mixed in the visible region.".

This is an image with 300x240 as a comparison.

[fs999]

Damned a dog on a bicycle can't be missed now !

[repaap]

Damned a dog on a bicycle can't be missed now !

Perhaps:
A damned dog on a bicycle? Coming slowly towards with a mask. It will be a challenge.

[MMVIII]

Damned a dog on a bicycle can't be missed now !

Right! One would think so!

[house]

I was thinking the horror of Cujo learning to cycle : 0

[fs999]

Right! One would think so!

Oh you're cheating ! It's a tricycle !