[richandfleur]

Has anyone actually confirmed this thing is significantly better at low noise/high ISO shots?

[anu l]

Has anyone actually confirmed this thing is significantly better at low noise/high ISO shots?

have a look here, K-70 at it's highest ISO.... looks good.

VERY low light K-70 pic

[Tau-Ceti]

I did check the first available High ISO Images of the K-70.
I compared them with similiar shots from a K-30.
The differences are a bit strange.
In my opinion, the K-70 clearly rules at ISO 3200 but things change at ISO 12.800.
So, the difference is not that big. Maybe a new firmware might make a change but that is unlikely.
It is still very good, though.

K-70 Samples
Pentax K-70 Review: Preview - Samples

K-30 Samples
Pentax K-30 Review - Thumbnails

[Adam]

Has anyone actually confirmed this thing is significantly better at low noise/high ISO shots?

Hard to tell if there's any improvement compared to the K-3. The two cameras really feel the same in terms of IQ, and I'm not sure why we'd expect it to vary by much. The K-70 does have the latest in-camera processing options from the K-1, though, including 8-level clarity enhancement and skin tone correction.

Adam
PentaxForums.com Webmaster (Site Usage Guide | Site Help | My Photography)



PentaxForums.com server and development costs are user-supported. You can help cover these costs by donating or purchasing one of our Pentax eBooks. Or, buy your photo gear from our affiliates, Adorama, B&H Photo, KEH, or Topaz Labs, and get FREE Marketplace access - click here to see how! Trusted Pentax retailers:

[Kerrowdown]

Gotta install water cooling in my camera!

Another way to do this... you could just remove the bodies WR features.

[Na Horuk]

Hard to tell if there's any improvement compared to the K-3. The two cameras really feel the same in terms of IQ, and I'm not sure why we'd expect it to vary by much.

Yeah? I was hoping the advertised "accelerator" would help noticeably. The high ISO photos seem pretty good. I hope the full review will compare K-70 and K-3 at high ISO (same scene, same lens, same settings, same ISO, then compare the jpeg and raw from both cameras, and maybe video as well?). I think a high-end NR module could be a big step forward, but we have to see how it performs in practice. Its one of the most interesting things about the K-70

[anu l]

Yeah? I was hoping the advertised "accelerator" would help noticeably. The high ISO photos seem pretty good. I hope the full review will compare K-70 and K-3 at high ISO (same scene, same lens, same settings, same ISO, then compare the jpeg and raw from both cameras, and maybe video as well?). I think a high-end NR module could be a big step forward, but we have to see how it performs in practice. Its one of the most interesting things about the K-70

and also with D5500. DPreview is really good at this. Their scene comparison, exposure latitude are an excellent tool for comparing sensor performance.

One criticism which I have about pentaxforums is that in their in depth reviews (which still in itself is one of the best review resource for detailed info), they directly compare their reviewed camera/lens (whenever they do) to only it's predecessor or another Pentax camera/lens made for Pentax.

I think they must compare competing cameras from other companies too, like the K-70 with Canon T6s/i and/or Nikon D5500, say in sensor performance, autofocus performance, handling.

same goes for lenses - sharpness, bokeh, AF performance, color, flare, vignetting etc.

This would show how the camera/lens compares against it's competition from Canon & Nikon etc.

[D1N0]

At imaging resource you can already compare. Imaging Resource "Comparometer" â„¢ Digital Camera Image Comparison Page

This is the K70 (top) vs the K-3 II (bottom) at iso 51.200 The K-70 has a much cleaner look without any loss of detail. (right click and view image for a larger view).






---------- Post added 21-08-16 at 16:34 ----------

Here at iso 25600 compared with the d5500. The K-70 retains much more detail. The D5500 is cooked heavily smudging detail in favour of less noise.



---------- Post added 21-08-16 at 16:41 ----------

The mannequin makes it obvious that this is a big deal.


---------- Post added 21-08-16 at 16:56 ----------

It is approaching the level of detail en cleanness of the K-1. 12.800

[Adam]

Yeah? I was hoping the advertised "accelerator" would help noticeably. The high ISO photos seem pretty good. I hope the full review will compare K-70 and K-3 at high ISO (same scene, same lens, same settings, same ISO, then compare the jpeg and raw from both cameras, and maybe video as well?). I think a high-end NR module could be a big step forward, but we have to see how it performs in practice. Its one of the most interesting things about the K-70

Yes, they will be compared directly.

At the end of the day it's mainly the sensor that determines the image quality. Any other hardware would only improve the in-camera processing, which still isn't going to compare with desktop PP.

[totsmuyco]

Nice results from the K-70. I hope that they'll do much better on the K-3II successor.

[rechmbrs]

Yes, they will be compared directly.

At the end of the day it's mainly the sensor that determines the image quality. Any other hardware would only improve the in-camera processing, which still isn't going to compare with desktop PP.

Isn't the "Accelerator" just a programmable high speed vector processor? Like 8086 and 8087 back in early PC days. If this is the case, I'd expect little or no change in quality but reduced processing time for the JPG output. This would be especially true for introduction of chip. Might mean faster Pixel Shift throughput too??

RONC

[photoptimist]

Isn't the "Accelerator" just a programmable high speed vector processor? Like 8086 and 8087 back in early PC days. If this is the case, I'd expect little or no change in quality but reduced processing time for the JPG output. This would be especially true for introduction of chip. Might mean faster Pixel Shift throughput too??

RONC

One can use an accelerator either to make the old algorithms run faster or to run a new higher-quality algorithm that would have been unacceptably slow on the old processor.

It's not yet clear whether the K-70 images are better because the sensor is better or because the accelerator does a better job of some kind of complex statistical processing that retains/enhances perceived detail while suppressing perceived noise.

What's interesting about IQ assessments is that so much of it is about images that look good to human eyes rather than assessing each pixel's accuracy in measuring photon flux. Perceived IQ certainly should matter more to photographers than photometric IQ but it does imply there's room for new algorithms inside special chips in the camera that can make the sensor data "look" better even if it is not actually better.

[rechmbrs]

One can use an accelerator either to make the old algorithms run faster or to run a new higher-quality algorithm that would have been unacceptably slow on the old processor.

It's not yet clear whether the K-70 images are better because the sensor is better or because the accelerator does a better job of some kind of complex statistical processing that retains/enhances perceived detail while suppressing perceived noise.

What's interesting about IQ assessments is that so much of it is about images that look good to human eyes rather than assessing each pixel's accuracy in measuring photon flux. Perceived IQ certainly should matter more to photographers than photometric IQ but it does imply there's room for new algorithms inside special chips in the camera that can make the sensor data "look" better even if it is not actually better.

I agree with you completely and don't expect any visual difference in the RAW data and little if any in the JPG.

This early in that chip's Pentax K camera life, I'd expect from a practical point of view that there is little new algorithm wise but a lot of getting used to it from system and programming wise. Whatever the chip does has to be programmed with all of the knowledge of timing, sharing memory etc. The hype about it is marketing and not much more.

Does anyone know the make and model of the CPU in the K70? Same as the the K1, I'd expect. The Accelerator is a stablemate to the CPU.

RONC

[anu l]

I agree with you completely and don't expect any visual difference in the RAW data and little if any in the JPG.

This early in that chip's Pentax K camera life, I'd expect from a practical point of view that there is little new algorithm wise but a lot of getting used to it from system and programming wise. Whatever the chip does has to be programmed with all of the knowledge of timing, sharing memory etc. The hype about it is marketing and not much more.

Does anyone know the make and model of the CPU in the K70? Same as the the K1, I'd expect. The Accelerator is a stablemate to the CPU.

RONC

The K-70 has the same older Prime M II processor working with the accelerator while the K-1 has Prime M IV which is 2 generations ahead.

[Kunzite]

Technically speaking, Prime M and Prime (no M) are different lines of processors. The K-1 has PRIME IV, there's no PRIME M IV.