Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
02-27-2018, 09:16 AM - 3 Likes   #1
Loyal Site Supporter
Loyal Site Supporter




Join Date: Jun 2009
Location: Tumbleweed, Arizona
Photos: Gallery | Albums
Posts: 5,687
Acceleration Chip (K1mkII, KP and K70) - What's the Difference

This Acceleration Chip that has appeared in the new K1mkII -
  • What is it?
  • What magic will it perform?
  • Does my photography really need it?
... and the current K1 does not have it.When I read about the new K1mkII in the announcement, I had a certain amount of ambivalence. I just acquired the K1, 6 months ago, and just received my new 15-30 lens, and was in no mood to spend any more funds on equipment. This was a camera I was still learning and figuring out. I just want to go out and shoot across a whole bunch of locations that I have been trying to acquire with my K5IIs with some degree of success, but not the images I really wanted. I'm just starting to get the images that I really really wanted.

Back to the "Chip". The Acceleration Chip is in 2 current cameras - the KP and K70 bodies and now in the announced K1mkII. What is it going to do for me? Well - nothing from what I initially saw. Ker-thump, I want to go back making images in my retirement. Done with this nonsense of even thinking about upgrading my new camera.Cutting to the chase here - take a look at the K1mkII feature description (for the Acceleration Chip). Also, the descriptions for the KP and K70 appear to have the Acceleration Chip dealing with the higher ISO values, too. Perhaps, I just read and interpreted it wrong.
QuoteQuote:
To reproduce lively colors and rich gradations close to memory colors in all sensitivity ranges, the PENTAX K-1 Mark II newly incorporates an original accelerator unit, which efficiently processes image signals output by the image sensor before sending them to the imaging engine. This process upgrades both image resolution and color reproduction in a high-sensitivity range, while drastically reducing noise compared to the PENTAX K-1. It also helps bring the camera's top sensitivity up to ISO 819200 (at standard output sensitivity) for high-grade, super-high-sensitivity photography.
Well, the three major points here - in red above, appears to me that right out of the box it's mainly for high ISO values. I trying to remove noise at the source (in camera) at the lowest ISO I'm able to use. Not interested. I want to go back to shooting with my K1, now.

But..... I just happened to see this video (in this post) on the K1mkII that had two sentences, that made me pause and think for a minute.
QuoteQuote:
"The accelerator chip is supposed to be most effective at low and mid ISOs. It retains detail and improves noise and DR."
.... and now the video has gone missing. Pentax, shoot yourself in the foot - yet again. Please, tell me a compelling story about your product!

So, this magical little "Acceleration Chip", has some pixie dust in it, after all. Then there is in this other thread that expresses....
QuoteOriginally posted by adjutant Quote
the new chip is truly the camera's selling point. better noise performance is actually top of my list of reasons to switch from my k-3 ii to kp. i feel like very few people appreciate how big of a deal the new chip is, they're all just saying this is a bizarre non-update to the K-1 and they never even mention the new chip and what big change it is for noise performance. i dont shoot high ISO a lot, but there are times when i've REQUIRED fast shutter speeds with high ISO, but my k-3 ii was just being pushed too much at its limits, even resized for web sizes
This now does create some interest with me - in particular with shooting landscapes in ambient low light (read in the dead of night) with the Milky Way (while I'm freezing my a** off in the middle of the desert - with eyes looking at me from out of the bush in the distance").

I go looking for KP and K70 astro images to see if this Acceleration Chip, does the "star eater" feature that Sony fostered on their customers. The link at...has some excellent imagery and it appears that the stars are nice and clear as well as being sharp. So, it appears that the KP and K70's Acceleration Chip does deliver the images. That's excellent!

OK, Pentax - what is the real story here? Let me rewrite you feature description here, using your own text, but re-working it a bit.

QuoteQuote:
To reproduce lively colors and rich gradations close to memory colors in all sensitivity ranges (most effective at low and mid ISO values), the PENTAX K-1 Mark II newly incorporates an original accelerator unit, which efficiently processes image signals output by the image sensor before sending them to the imaging engine. This process upgrades both image resolution and color reproduction in a high-sensitivity range, while drastically reducing noise, while retaining detail and improves noise and DR, as compared to the PENTAX K-1. It also helps bring the camera's top sensitivity up to ISO 819200 (at standard output sensitivity) for high-grade, super-high-sensitivity photography.
To me, that tells a better story of the "Acceleration Chip". Now, to continue reading about the Prime IV processor...
QuoteQuote:
The PENTAX K-1 Mark II's PRIME IV imaging engine is the culmination of PENTAX's high-speed, high-quality image processing technologies. In addition to highly efficient noise processing, it features Fine Sharpness and Extra Sharpness functions to process the subject's outlines more naturally and more delicately, and the PENTAX Real-time Scene Analysis System, which has adopted a breakthrough deep learning technology. It also effectively compensates for the distortion, vignetting and chromatic aberration caused by specific lens properties, while efficiently correcting the fringe effect.
Putting all of this together - it starts to sound like a pretty reasonable upgrade to the K1, especially when you compare it to the original K1 feature description - Features1 | PENTAX K-1 | RICOH IMAGING

There still is not a whole lot of information on the "Acceleration Chip". By its very nature and implementation, mere mortal photographers are not going to be able to do apple to apple comparison images. There are the KP and K70 bodies out there in the wild, but then you need to double up on the lenses and take side by side shots at essentially the same parameters, but then you have the differences in sensors (I guess a K3 and KP side by side would not be so difficult with two copies of the same lens). The scenes would need to be carefully chosen to potentially highlight the differences.

So, right now - I've went from absolutely not interested in the least, to possibly this may be reasonable for a $550 upgrade. But - Pentax, why are you making it so difficult for me?

Tell me more - the chip has been out there for over a year now - and has generated very little information. How much better is my imagery going to be? I'm willing to at least listen now.




Last edited by interested_observer; 02-27-2018 at 09:50 AM. Reason: Grammer
02-27-2018, 09:40 AM - 1 Like   #2
Administrator
Site Webmaster
Adam's Avatar

Join Date: Sep 2006
Location: Arizona
Photos: Gallery | Albums
Posts: 51,558
We will ask them at CP+ for a clearer technical description of what it does, but I think you've synthesized it pretty well.

My theory is that it simply amplifies the sensor signal before sending it for processing through the PRIME engine, which results in better color accuracy and lower noise. Beauty in simplicity.

We've already seen what it can do first hand:
Pentax KP Review - High ISO Low Light Performance | PentaxForums.com Reviews

Could offer a nice boost to the K-1's image quality in critical use cases. if you look at the high-ISO KP shots above, they do come close to those of the K-1. The update in the K-1 II should once again level the playing field and put the K-1 well ahead.

Now here's the real question- will the 645Z also get this tech, or is it time for Pentax to set it up to a 100Mp sensor? IMO, it will be the latter.

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:
02-27-2018, 09:47 AM - 3 Likes   #3
Loyal Site Supporter
Loyal Site Supporter
monochrome's Avatar

Join Date: Mar 2008
Location: Working From Home
Photos: Gallery | Albums
Posts: 26,276
Let me put it this way.
  • I’ve pre-ordered K-1ll. I’m not doing the upgrade.
  • I’m selling the KP and all my APSc lenses.
  • I’m keeping the K-1 for manual lenses.
  • I’ll figure out which camera is better for what purpose.

The K-1 Next Gen probably won’t be ready until 2020. We’ll see if it is worth upgrading to, but I think my two K-1’s will be enough for probably 5 years from today.
02-27-2018, 10:42 AM - 1 Like   #4
Veteran Member
aleonx3's Avatar

Join Date: Jul 2007
Location: Brampton, Ontario
Photos: Gallery | Albums
Posts: 3,996
QuoteOriginally posted by monochrome Quote
Let me put it this way.
  • I’ve pre-ordered K-1ll. I’m not doing the upgrade.
  • I’m selling the KP and all my APSc lenses.
  • I’m keeping the K-1 for manual lenses.
  • I’ll figure out which camera is better for what purpose.

The K-1 Next Gen probably won’t be ready until 2020. We’ll see if it is worth upgrading to, but I think my two K-1’s will be enough for probably 5 years from today.
I think I will take the same approach since I need to shoot APS-C mode at times when I can not get close to the subject (eg. stage photography). I can mount a DA long lens (eg. 50-135/2.8) on the k-1 and treat it like a k-3. The weight difference between the k-1 and k-3 is not that great.

02-27-2018, 11:27 AM - 3 Likes   #5
Loyal Site Supporter
Loyal Site Supporter




Join Date: Jun 2009
Location: Tumbleweed, Arizona
Photos: Gallery | Albums
Posts: 5,687
Original Poster
QuoteOriginally posted by Adam Quote
We will ask them at CP+ for a clearer technical description of what it does, but I think you've synthesized it pretty well.
Thanks!!! In spite of my snarky comments, I initially took the announcement to be extremely oriented at high ISO, a push on AF a bit, and improved pixel shift. So, to me - there was a lot of non interest. I was just going to punt - as I did not need any of these improved capabilities. However, there was something with the upgrade option that intrigued me. I do experience a level of the "white dot" issue, and was wondering if the Acceleration Chip, would provide a solution (and in my opinion, I think it would as it's position in the right place in the image processing chain). This would be a wonderful gesture on the part of Pentax. That got me started at looking a bit more.

QuoteQuote:
My theory is that it simply amplifies the sensor signal before sending it for processing through the PRIME engine, which results in better color accuracy and lower noise. Beauty in simplicity.
After reading a bit, and peeling the onion on what few comments Pentax has published, that is also my view. The statement - "The accelerator chip is supposed to be most effective at low and mid ISOs. It retains detail and improves noise and DR." was essentially the nail in the coffin, so to speak. It spoke volumes. Cleaning up the incoming signal from the sensor after the D/A conversion - with a very light touch, especially considering the current work in this area (and the new signal processing hardware that is recently available) would certainly provide an elegant solution.

QuoteQuote:
I had started reading this the other day, before my wife re-tasked me - so, I had not gotten to that particular section yet. Right now, it appears to me, that the Acceleration Chip is providing the improvements/solutions to the few issues I am currently having.

QuoteQuote:
Could offer a nice boost to the K-1's image quality in critical use cases. if you look at the high-ISO KP shots above, they do come close to those of the K-1. The update in the K-1 II should once again level the playing field and put the K-1 well ahead.
I'm finding that I'm pushing the envelope in several areas, but with the K1/15-30 getting excellent results. When you are shooting in the dead of night, you are going to get noise - regardless of the ISO you are using. There is just no way around it.

Having said that, I would like to clean up the noise to an extent possible at ISO 800 to 3200, along with eliminate the white dots, while retaining the exceptional detail and rich colors that I have been able to capture (just in the landscape element). Here is a quick test example - I had just received my 15-30 the day before. The K1 with the D FA 15-30 @ f2.8, 30mm, ISO 800, 120 seconds, 4 frame stitched pano. [image 1] Shot at 5am, no moon (just set) before sunrise - actually before astro twilight, I only had 25 minutes, with a fair amount of wind. It's as dark as you can get - 70 miles east of Phoenix, just before Superior, at Picket Post mountain. The lighting on the horizon on the left is Superior and to the right is the Ray open pit copper mine 40 miles away.

On the image below, the detail is excellent - much better than what I expected. The trees are stacked in depth, the details in the rock watering hole are wonderful, and you can even see the strands of the barbed wire. The color tones of greens in the trees, along with their depth are fantastic. Even with the movement from the wind, the details are there. These colors are not available during the day, they are rich and accurate - with excellent contrast. Exposure and contrast are straight out of the camera. Just some black, white, shadow and highlight adjustments, with a bit of clarity and vibrancy. In the full resolution image, there is some noise and a bit of white dots, that are of concern. The sky segment is of no concern as the AstroTrack capability takes care of that, and then it's a matter of compositing the segments together.

QuoteQuote:
Now here's the real question- will the 645Z also get this tech, or is it time for Pentax to set it up to a 100Mp sensor? IMO, it will be the latter.
It will probably be both. I like the 16MP (K5 sensor), along with the K1 (essentially the same pixel pitch of the K5) - which is about the same pixel pitch as the 645Z. I like the large fat light soaking ability of these pixels. Having said that, with out going to 100Mp, it will be viewed as Pentax falling behind. However - if this Acceleration Chip, performs as it appears it does - that would make the results exceptional.

Attached Images
 

Last edited by interested_observer; 02-27-2018 at 11:32 AM.
02-27-2018, 11:31 AM   #6
Site Supporter
Site Supporter
ffking's Avatar

Join Date: Nov 2015
Location: Old South Wales
Posts: 6,026
QuoteOriginally posted by Adam Quote
Now here's the real question- will the 645Z also get this tech, or is it time for Pentax to set it up to a 100Mp sensor? IMO, it will be the latter.
QuoteOriginally posted by interested_observer Quote
It will probably be both
yeah, both, please
02-27-2018, 11:52 AM   #7
Veteran Member
joergens.mi's Avatar

Join Date: Jan 2012
Posts: 408
It may be, that the AC can manipulate the sensor by changing amplification on the senor according to the ISO (- Sony may supply some bits for changing that on the sensor-), and amplifies afterwards in the AC - thereby reducing noise.

What I do not understand, when it works well in low an mid ISO part, why can the max ISO increased?

02-27-2018, 12:13 PM - 2 Likes   #8
Senior Member




Join Date: May 2009
Location: Sweden
Photos: Gallery
Posts: 183
I agree observer, I share many of your thoughts/observations and most of all questions...but there are a couple places I'm not sure I follow.

QuoteOriginally posted by Adam Quote
We will ask them at CP+ for a clearer technical description of what it does, but I think you've synthesized it pretty well.

My theory is that it simply amplifies the sensor signal before sending it for processing through the PRIME engine, which results in better color accuracy and lower noise. Beauty in simplicity.
,,,.
QuoteOriginally posted by interested_observer Quote

After reading a bit, and peeling the onion on what few comments Pentax has published, that is also my view. The statement - "The accelerator chip is supposed to be most effective at low and mid ISOs. It retains detail and improves noise and DR." was essentially the nail in the coffin, so to speak. It spoke volumes. Cleaning up the incoming signal from the sensor after the D/A conversion - with a very light touch, especially considering the current work in this area (and the new signal processing hardware that is recently available) would certainly provide an elegant solution.
Adam, big thanks that you'll ask that!

But regarding your thoughts above, isn't ISO just that, amplification of the signal? I don't see how doing it twice (without help of other processing, thereby risking loosing details) could help, while if done after the A/D conversion, would just be multiplying a digital number, noise and all...just pure amplification shouldn't help one bit from a noise signal processing perspective. And also, if done after the AD-conversion, again, what makes it better than waiting a while more and doing it on the computer in PP? Is the algo/NR better and smarter than others, or is it the same, but just more convenient to have built in in the camera? If other "smarts" are involved in the amplification (if that's what it is) then we're back to the normal pp/noise removal technniques....unless it can do something pp software can't by basing it's smarts on some hardware aspect that pp software wouldn't have access to...

So again..I'm at a loss, the question about how it works on a more technical level would be very appreciated!

But maybe I misunderstand you guys?
02-27-2018, 01:43 PM   #9
Veteran Member
LensBeginner's Avatar

Join Date: Sep 2014
Photos: Albums
Posts: 4,696
QuoteOriginally posted by Igor123 Quote
*snip*

But regarding your thoughts above, isn't ISO just that, amplification of the signal? I don't see how doing it twice (without help of other processing, thereby risking loosing details) could help, while if done after the A/D conversion, would just be multiplying a digital number, noise and all...just pure amplification shouldn't help one bit from a noise signal processing perspective. And also, if done after the AD-conversion, again, what makes it better than waiting a while more and doing it on the computer in PP? Is the algo/NR better and smarter than others, or is it the same, but just more convenient to have built in in the camera? If other "smarts" are involved in the amplification (if that's what it is) then we're back to the normal pp/noise removal technniques....unless it can do something pp software can't by basing it's smarts on some hardware aspect that pp software wouldn't have access to...

So again..I'm at a loss, the question about how it works on a more technical level would be very appreciated!
This is a nice bunch of intelligent questions...
02-27-2018, 02:25 PM   #10
Veteran Member




Join Date: Dec 2013
Location: Buffalo, NY
Posts: 351
QuoteOriginally posted by interested_observer Quote
(while I'm freezing my a** off in the middle of the desert - with eyes looking at me from out of the bush in the distance
Call me silly, but that was my favorite line in your post.

QuoteOriginally posted by Igor123 Quote
But regarding your thoughts above, isn't ISO just that, amplification of the signal?
If it's before the A/D converter it can make a big difference where in the pipeline it's located,* but based on ISO invariance displayed by most Pentax sensors I believe the A/D converter is already on chip (so before the Image Accelerator). It also seems like the image accelerator works magic on JPG processing but not on RAW files. (At least, that's the impression I get from the KP review since they don't show any examples of processed RAW's being better than the K-3 processed RAW's.)

I would say that the main advantage is improved JPEG processing in camera. It reminds me of computers and graphics cards. Sure, some games can be played with the main CPU processing the graphics, but at limited resolution and frame rate. Pass that processing off to a dedicated graphics card instead and you can increase resolution and frame rate substantially while letting the CPU focus on opponent AI, etc. The new image accelerator seems to do the same thing for the camera. Perhaps by offloading some JPEG processing the Prime IV engine might even be able to dedicate more processing power to the auto focus algorithms. (Just thinking out loud here, don't quote me on that!)

*Similar to using an amplifier for a TV antenna 50 ft away from the TV. Putting the amplifier right after the antenna will give the best signal. If you put the amplifier at the TV instead it will have less original signal to work with due to wire resistance, as well as also amplifying any noise generated in the 50 ft of cable.

---------- Post added 02-27-18 at 04:29 PM ----------

While I have a K-3, I don't have a KP (unfortunately!). If someone has both it would be interesting to get their thoughts on processed low light RAW files from the KP vs the K-3. I wouldn't be surprised if the KP is somewhat better just from sensor improvements over time, but if the Image Accelerator is doing it's magic on RAW files then you should see large improvements similar to those shown on JPEG's in the review. (And if it doesn't, I'll feel better about not having a KP yet.)

Last edited by TheOneAndOnlyJH; 02-27-2018 at 02:31 PM.
02-27-2018, 06:06 PM   #11
Veteran Member
dakight's Avatar

Join Date: Sep 2014
Location: Phoenix, AZ
Photos: Albums
Posts: 1,216
QuoteOriginally posted by interested_observer Quote
Thanks!!! In spite of my snarky comments, I initially took the announcement to be extremely oriented at high ISO, a push on AF a bit, and improved pixel shift. So, to me - there was a lot of non interest. I was just going to punt - as I did not need any of these improved capabilities. However, there was something with the upgrade option that intrigued me. I do experience a level of the "white dot" issue, and was wondering if the Acceleration Chip, would provide a solution (and in my opinion, I think it would as it's position in the right place in the image processing chain). This would be a wonderful gesture on the part of Pentax. That got me started at looking a bit more.


After reading a bit, and peeling the onion on what few comments Pentax has published, that is also my view. The statement - "The accelerator chip is supposed to be most effective at low and mid ISOs. It retains detail and improves noise and DR." was essentially the nail in the coffin, so to speak. It spoke volumes. Cleaning up the incoming signal from the sensor after the D/A conversion - with a very light touch, especially considering the current work in this area (and the new signal processing hardware that is recently available) would certainly provide an elegant solution.


I had started reading this the other day, before my wife re-tasked me - so, I had not gotten to that particular section yet. Right now, it appears to me, that the Acceleration Chip is providing the improvements/solutions to the few issues I am currently having.


I'm finding that I'm pushing the envelope in several areas, but with the K1/15-30 getting excellent results. When you are shooting in the dead of night, you are going to get noise - regardless of the ISO you are using. There is just no way around it.

Having said that, I would like to clean up the noise to an extent possible at ISO 800 to 3200, along with eliminate the white dots, while retaining the exceptional detail and rich colors that I have been able to capture (just in the landscape element). Here is a quick test example - I had just received my 15-30 the day before. The K1 with the D FA 15-30 @ f2.8, 30mm, ISO 800, 120 seconds, 4 frame stitched pano. [image 1] Shot at 5am, no moon (just set) before sunrise - actually before astro twilight, I only had 25 minutes, with a fair amount of wind. It's as dark as you can get - 70 miles east of Phoenix, just before Superior, at Picket Post mountain. The lighting on the horizon on the left is Superior and to the right is the Ray open pit copper mine 40 miles away.

On the image below, the detail is excellent - much better than what I expected. The trees are stacked in depth, the details in the rock watering hole are wonderful, and you can even see the strands of the barbed wire. The color tones of greens in the trees, along with their depth are fantastic. Even with the movement from the wind, the details are there. These colors are not available during the day, they are rich and accurate - with excellent contrast. Exposure and contrast are straight out of the camera. Just some black, white, shadow and highlight adjustments, with a bit of clarity and vibrancy. In the full resolution image, there is some noise and a bit of white dots, that are of concern. The sky segment is of no concern as the AstroTrack capability takes care of that, and then it's a matter of compositing the segments together.


It will probably be both. I like the 16MP (K5 sensor), along with the K1 (essentially the same pixel pitch of the K5) - which is about the same pixel pitch as the 645Z. I like the large fat light soaking ability of these pixels. Having said that, with out going to 100Mp, it will be viewed as Pentax falling behind. However - if this Acceleration Chip, performs as it appears it does - that would make the results exceptional.

OK, off topic, but I love that image of the windmill with picket post mountain in the background!
02-27-2018, 07:50 PM   #12
Veteran Member




Join Date: Dec 2016
Location: Southeastern Michigan
Posts: 4,500
Perhaps this Accelerator is a "booster" amplification besides ISO without the usual amplification downside.

The new description being used by Pentax to describe the K-1 Mark II is of course to sell the product. In reality, there's a lot of repetition of that from the original, which for example also has the Prime IV processor. It is always good to see technological progress that has meaning in the visible results gained, but everything said in the new description is not new.

However, I am not one to just dump my highly-valued premium equipment in order to immediately acquire the latest and greatest. If I made a living where the improved results offered would likely produce additional income, I might be thinking otherwise. Only tests and evaluating the impact on one's actual needs will tell. Although I have experienced some definite improvements in results from my KP, like others, I still think very highly of my K-5 IIs, both for its practical advantages, and its still fine IQ. I have no intention of getting rid of it. But the KP has been outstanding in every way for IQ. Both from my own experience, and from all the many outstanding images I have seen from it, including those found in the Astro-bin link above.

This brings me back to the three questions heading this thread. Those questions are the reason I have been rather ambivalent regarding getting a K-1, and also because of the advancements that had appeared in the KP, yet had not yet appeared in the K-1. Now those have appeared in the K-1 Mark II. So once again, those three questions I will have to ask myself to justify the expense, and having to haul around a 2-lb + body and attending lenses, compared to what I am handling now. Someone has rationalized the K-1 as being not much more weight than the K-3 II. In reality, the K-1 or Mark II, loaded and ready, is nearly 8 oz more than the K-3 II! You can add another 4 oz over the KP! Then there's the matter of FF lenses compared with APS-C lenses, and their weight and size. When going by car to take photos, the camera's added weight will matter less, and that is when I may put it to use.

For me, those same three questions must come to mind, with those disadvantages weighed in the mix. And all those superb KP images, like those shown above, speak volumes.

Nonetheless, I still have my own rationale and interest, which relates to my existing lenses. Also, the D-FA 28-105 mm is a very fine zoom lens. And I have not heard of any QC issue having emerged. If it turns out that the K-1 Mark II offers even a near 1-stop additional advantage in low-light shooting, it would be like converting this fine zoom lens to f/2.8-4 capability- a worthwhile gain. I have some very fine FF lenses, even by today's standards. In my case, all I would need is this one excellent, relatively compact new WR zoom lens, and I would have an extensive FF system already. So the enticement is there. But do I actually need it, and how often would I use it?

Many photographers have and use both formats. For me, no doubt I would still use the smaller system most often. And, when it comes to quality telephoto work, it would always be my KP and/or my K-5 IIs, with the lighter weight of the DA* 50-135mm and other excellent telephoto lenses I have. The additional weight and size necessary in going FF to get a comparable image would not make sense for me. So both systems have use advantages, which is why many do own both. The FF system could provide superior edge-to-edge performance with the excellent WR kit zoom, as with other of my best FF lenses, such as the FA 35mm f/2, now a fast moderate wide-angle prime. Then, of course, there's the D-FA 100mm f/2.8 WR macro, the FA 43mm f/1.9 and the 77mm f/1.8 Limiteds. Also the Sigma AF 24-60mm f/2.8 EX DG and Tokina ATX 28-70mm f/2.6-2.8 pro II when need speedier zooms.

How often I would need these advantages, and by how much will they be advantageous- enough to put out another $2,400 are factors to ponder.

Last edited by mikesbike; 03-05-2018 at 01:01 PM.
02-27-2018, 08:08 PM - 3 Likes   #13
Otis Memorial Pentaxian
Otis FanOtis FanOtis FanOtis FanOtis FanOtis Fan
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 16,394
QuoteOriginally posted by mikesbike Quote
Perhaps this Accelerator is a "booster" amplification besides ISO without the usual amplification downside.
Boosting is easy, Mike.

That's what ISO is.

ISO 800 just takes the ISO 100 value and multiplies it - signal and noise - by eight.

To be worthwhile, this thing needs to do something to the noise component, leaving the signal part alone.

This is the challenge to all noise reduction methods - how to not be 'lossy'.

We expect data loss in JPGs, people like myself don't like it in RAW. We do our own selective post-processing in software, and any data reduction beforehand actually limits what we can do. If I see wax-like textures to skin, it's too late for PP, the damage has been done.

The parent company Ricoh has made semiconductor components in the past, so it's possible this thing is real clever, real elegant.

Last edited by clackers; 02-27-2018 at 09:40 PM.
02-27-2018, 09:14 PM   #14
Loyal Site Supporter
Loyal Site Supporter




Join Date: Jun 2009
Location: Tumbleweed, Arizona
Photos: Gallery | Albums
Posts: 5,687
Original Poster
QuoteOriginally posted by joergens.mi Quote
It may be, that the AC can manipulate the sensor by changing amplification on the senor according to the ISO (- Sony may supply some bits for changing that on the sensor-), and amplifies afterwards in the AC - thereby reducing noise.

What I do not understand, when it works well in low an mid ISO part, why can the max ISO increased?
I don't know either, but the KP examples that Adam has shows potential. Does this work on raw as well as JPG?

QuoteOriginally posted by Igor123 Quote
I agree observer, I share many of your thoughts/observations and most of all questions...but there are a couple places I'm not sure I follow.

Adam, big thanks that you'll ask that!

But regarding your thoughts above, isn't ISO just that, amplification of the signal? I don't see how doing it twice (without help of other processing, thereby risking loosing details) could help, while if done after the A/D conversion, would just be multiplying a digital number, noise and all...just pure amplification shouldn't help one bit from a noise signal processing perspective. And also, if done after the AD-conversion, again, what makes it better than waiting a while more and doing it on the computer in PP? Is the algo/NR better and smarter than others, or is it the same, but just more convenient to have built in in the camera? If other "smarts" are involved in the amplification (if that's what it is) then we're back to the normal pp/noise removal technniques....unless it can do something pp software can't by basing it's smarts on some hardware aspect that pp software wouldn't have access to...

So again..I'm at a loss, the question about how it works on a more technical level would be very appreciated!

But maybe I misunderstand you guys?
I haven't got the foggiest idea as to what they are actually doing, nor would I expect Pentax to be forthcoming on that aspect. That said, ....
  • Does it work with RAW as well as JPG?
  • How many stops improvement? 1/3 stop? 1/2 stop, 1 full stop?
  • Does it work across all types of noise, or does it tend to work on a particular type of noise?
  • How much dynamic range might be recoverable?
  • How much additional detail is recovered?
  • etc., etc., etc. ....

QuoteOriginally posted by TheOneAndOnlyJH Quote
Call me silly, but that was my favorite line in your post.
My wife always tells me that the Coyotes are gonna get me. Standing out there for a couple of hours, shooting long exposures, and you can just feel them watching you. Stupid human, why is he just standing there with that infernal contraption? There is a lot of just scrub desert around, but then there are these small areas of lush trees, with water, and some wonderful topography, that really comes together and works well with the Milky Way. This was just a test shot (basically 1 row by 4 frames). Shooting for a 3 row by 7 or 8 frames wide. I wanted to see what would work best - a single 2 minute exposure, several 2 minute exposures stacked, or some Pixel Shift stacked in order to take advantage of the better true color and inherent noise reduction. This was just a set of single 2 minute exposures.

QuoteQuote:
If it's before the A/D converter it can make a big difference where in the pipeline it's located,* but based on ISO invariance displayed by most Pentax sensors I believe the A/D converter is already on chip (so before the Image Accelerator). It also seems like the image accelerator works magic on JPG processing but not on RAW files. (At least, that's the impression I get from the KP review since they don't show any examples of processed RAW's being better than the K-3 processed RAW's.)

I would say that the main advantage is improved JPEG processing in camera. It reminds me of computers and graphics cards. Sure, some games can be played with the main CPU processing the graphics, but at limited resolution and frame rate. Pass that processing off to a dedicated graphics card instead and you can increase resolution and frame rate substantially while letting the CPU focus on opponent AI, etc. The new image accelerator seems to do the same thing for the camera. Perhaps by offloading some JPEG processing the Prime IV engine might even be able to dedicate more processing power to the auto focus algorithms. (Just thinking out loud here, don't quote me on that!)

*Similar to using an amplifier for a TV antenna 50 ft away from the TV. Putting the amplifier right after the antenna will give the best signal. If you put the amplifier at the TV instead it will have less original signal to work with due to wire resistance, as well as also amplifying any noise generated in the 50 ft of cable.

---------- Post added 02-27-18 at 04:29 PM ----------

While I have a K-3, I don't have a KP (unfortunately!). If someone has both it would be interesting to get their thoughts on processed low light RAW files from the KP vs the K-3. I wouldn't be surprised if the KP is somewhat better just from sensor improvements over time, but if the Image Accelerator is doing it's magic on RAW files then you should see large improvements similar to those shown on JPEG's in the review. (And if it doesn't, I'll feel better about not having a KP yet.)
I really don't have any idea as to what they really did. Signal conditioning/filtering up stream before the ADC, worked some magic with the ADCs, maybe the Acceleration Chip is a mixed signal parallel DSP, or combinations of all of these. Good analog engineering is a real black art. Whatever they were able to do, it appears that it does work reasonably well.

Still lots of photography based questions as to ....
  • low ISO? - how low 100 to 1600?
  • JPG or RAW format or both?
  • What types of noise does it work on - all, of better with -----?
  • Richer colors, but any particular color in the RGB set?
Pentax is probably not going to be really forth coming with the engineering details. I went looking for patents filed by Pentax. I really didn't find any that addressed this particular aspect.

Since Pentax made the decision to slide in a whole chip forward of the image processor, they made a conscious decision to spend power, heat and real estate on this, and it appears that the results are there. So, someone really worked with their analog and DSP utilities characterizing these signals, understanding the noise and working some magic. That's a wonderful accomplishment - that has gone pretty much un-noticed by us - the photographers. Not a lot information on this has been posted. Pentax has pretty much soft peddled it also. If they could provide some side by side examples - K1 vs K1mkII would be really helpful, provide a better descriptive write up, this could really help folks (photographers) who shoot in challenging environments.

QuoteOriginally posted by dakight Quote
OK, off topic, but I love that image of the windmill with picket post mountain in the background!
I drove out there for the first time a year ago, one very cold early morning. I had just scouted using google earth, and wound up with a good shot (fingers froze up so much, I could not work the AstroTrace function). I wound up taking a very uninteresting foreground landscape, but it was a test to see how this location would work out. Drove out there again looking for some better vantage points and found this location.

This was a 1 x 4 (1 row, 4 frames). With the Milky Way, and getting some lower/closer foreground, while going wider, will require a 3 x 7 for the final image I really want. I was all set up to do this Feb 12 thru the 26, but it was clouded over and raining these past two weeks. I was able to get one night in - but some clouds were obscuring the Milky Way.

I was able to get close with my K5IIs, but the K1 is doing much better. I'm finding the quality of light (what light there is), really varies - especially with cloud cover. Anyway, it keeps my retirement interesting....


Last edited by interested_observer; 02-27-2018 at 09:21 PM.
03-05-2018, 02:26 AM - 1 Like   #15
Pentaxian
angerdan's Avatar

Join Date: Feb 2015
Photos: Gallery | Albums
Posts: 2,622
QuoteOriginally posted by interested_observer Quote
But..... I just happened to see this video (in this post) on the K1mkII that had two sentences, that made me pause and think for a minute.
.... and now the video has gone missing. Pentax, shoot yourself in the foot - yet again.
Not gone, just revised:
QuoteOriginally posted by pentaxplus:
There was incorrect information shown in the video. The high sensitivity photo comparison between K-1 and K-1 Mark II was incorrect, the video has been revised. We apologize for any inconveniences caused.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
acceleration, accelerator, amplifier, camera, chip, curve, dslr, files, gr, graphics, iii, image, iso, k-1, k-3, k1mkii, k70, kp, lot, noise, pentax, photography, rate, ricoh, tech, tv
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
KP vs K-70 - to KP or not to KP, that is the question OldChE Pentax DSLR Discussion 28 11-02-2019 05:29 AM
Currently have a Kx, looking to move up to a KP, K70, or K3-ii. Recommendations? mojoe_24 Pentax KP 10 11-07-2017 10:05 PM
LR update, May 2017 + GPU Acceleration issue. BruceBanner Digital Processing, Software, and Printing 2 07-29-2017 07:35 PM
Noise Reduction Acceleration Unit reh321 Pentax Q 5 06-15-2016 03:52 AM
Sensor difference due to optics in front of chip rechmbrs Pentax K-5 & K-5 II 5 12-04-2010 05:56 PM



All times are GMT -7. The time now is 07:22 PM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top