Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
03-20-2016, 01:58 PM   #136
Site Supporter




Join Date: Mar 2016
Posts: 155
QuoteOriginally posted by falconeye Quote
The original source is about "substrate" (基板, where also mentioned 基板は means board). From the context of the full article, it is clear they talk about the circuit layout of the sensor daughterboard. This is the stock D800 sensor from Sony and there is no way to alter its physical properties as its readout is all digital anyway. I said this elsewhere already, this is a misleading part of the interview, probably from over-enthusiastic marketing


This is true. The D810 sensor is more expensive to manufacture and has physical properties which have been measured to be different. The K-1 has a stock sensor (the same as D800/D800E/A7r).

Hi,

I had not interpreted Pentax's site Challengers | PENTAX K-1 Special site | RICOH IMAGING as such originally but I think that you are right - on second reading it does sound like they were referring to the daughterboard .

I was at a Canadian demo recently for the K1 and I got to play with the K1 for about an hour or so. The rep presenting was very knowledgeable and quite adamant when I pushed him that the sensor in the K1 is not a 1:1 sensor for the D800 variant. Instead he insisted that it is an evolution of the D800/D810 family of sensor that has been built to Pentax specifications. If I'm honest, this doesn't give me any more confidence one way or the other that the K1 sensor is different to the D800E (as reps will always try to differentiate their product).

That being said, the entire imaging pipeline is more than just a sensor, so it's quite reasonable that Pentax's packaging of the daughter card plays it's part in overall SnR somewhat. However I would imagine that the leaked high ISO images vis-a-vis a Nikons D800/D810 class appear to show that the K1 has better SnR at high ISO.

With respect to the theory that the K1 sensor is a 1:1 for the D800E (design, manufacture, packaging) - until we see actual tear downs or an official statement from Pentax it is all conjecture at this point based on reasonable assumption and rumor - but assumption nonetheless.

Pentax K-1 Review: Hands-On Preview - Specifications
Nikon D800E Review - Specifications
Nikon D810 Review - Specifications

Another possibility is that the K1 sensor is the same design (not a bad thing) as the D800E but that in the intervening years Sony has perfected the manufacturing side and that the 36mp variant in the K1 makes use of newer manufacturing processes to yield a better SnR. I'm looking forward to seeing/hearing about a teardown of the K1.

03-20-2016, 03:08 PM - 1 Like   #137
Loyal Site Supporter
monochrome's Avatar

Join Date: Mar 2008
Location: Kirkwood (St. Louis) MO
Photos: Gallery | Albums
Posts: 20,190
A modest but genuine question.

QuoteOriginally posted by falconeye Quote
The original source is about "substrate" (基板, where also mentioned 基板は means board). From the context of the full article, it is clear they talk about the circuit layout of the sensor daughterboard. This is the stock D800 sensor from Sony and there is no way to alter its physical properties as its readout is all digital anyway. I said this elsewhere already, this is a misleading part of the interview, probably from over-enthusiastic marketing


This is true. The D810 sensor is more expensive to manufacture and has physical properties which have been measured to be different. The K-1 has a stock sensor (the same as D800/D800E/A7r).
QuoteOriginally posted by tomO2013 Quote
Hi,

I had not interpreted Pentax's site Challengers | PENTAX K-1 Special site | RICOH IMAGING as such originally but I think that you are right - on second reading it does sound like they were referring to the daughterboard .

I was at a Canadian demo recently for the K1 and I got to play with the K1 for about an hour or so. The rep presenting was very knowledgeable and quite adamant when I pushed him that the sensor in the K1 is not a 1:1 sensor for the D800 variant. Instead he insisted that it is an evolution of the D800/D810 family of sensor that has been built to Pentax specifications. If I'm honest, this doesn't give me any more confidence one way or the other that the K1 sensor is different to the D800E (as reps will always try to differentiate their product).

That being said, the entire imaging pipeline is more than just a sensor, so it's quite reasonable that Pentax's packaging of the daughter card plays it's part in overall SnR somewhat. However I would imagine that the leaked high ISO images vis-a-vis a Nikons D800/D810 class appear to show that the K1 has better SnR at high ISO.

With respect to the theory that the K1 sensor is a 1:1 for the D800E (design, manufacture, packaging) - until we see actual tear downs or an official statement from Pentax it is all conjecture at this point based on reasonable assumption and rumor - but assumption nonetheless.

Pentax K-1 Review: Hands-On Preview - Specifications
Nikon D800E Review - Specifications
Nikon D810 Review - Specifications

Another possibility is that the K1 sensor is the same design (not a bad thing) as the D800E but that in the intervening years Sony has perfected the manufacturing side and that the 36mp variant in the K1 makes use of newer manufacturing processes to yield a better SnR. I'm looking forward to seeing/hearing about a teardown of the K1.
What is the tipping point that makes our resident experts (and I mean that with respect), and those who have conversations with Pentax employees who adamantly state the sensor is made to Pentax spec. - what is the tipping point at which we believe the words written on their home page, and that come out of their mouths? When do we stop doubting, because doubting is what we do?
03-20-2016, 03:22 PM - 1 Like   #138
Pentaxian




Join Date: Jun 2009
Location: GMT +10
Photos: Albums
Posts: 10,751
QuoteOriginally posted by tomO2013 Quote
in the intervening years Sony has perfected the manufacturing side and that the 36mp variant in the K1 makes use of newer manufacturing processes to yield a better SnR
I think that is highly likely. It's not as if Sony chip engineers in the IMX094 production line just put down their CAD tools the day the first 36MP chip of that design rolled off the assembly line in late 2011.

Further, note that the suffix of the chip name (eg IMX094AQP for the D800) is said to reflect:
- revision number (A, B, C etc), then
- the mono/chroma for the bayer plane (usually Q is the color bayer and L is monochromatic), and
- the last letter is the chip package or 'substrate?' (Q for plastic QFP [Quad Flat Package], E/P for ceramic LCC [Leadless Chip Carrier], L/R for ceramic LGA [Land Grid Array package], Z for QFN [Quad Flat No-leads]).

The Sony 36MP A7R chip is the IMX094AQR, indicating package variation from the D800 (LGA vs LCC), so it's certainly possible that the K-1 may also have package variation from the D800 or A7R. I think some of these package variations (eg D800 plastic vs A7R ceramic) may certainly have different thermal or other properties, as suggested in the Pentax engineer interview, which might play a role in chip performance.

Someone needs to donate their K-1 to iFixit or Chipworks to stop this speculation.

Last edited by rawr; 03-20-2016 at 03:30 PM.
03-21-2016, 03:10 AM   #139
Senior Member




Join Date: Jan 2013
Posts: 242
QuoteOriginally posted by rawr Quote
I think that is highly likely. It's not as if Sony chip engineers in the IMX094 production line just put down their CAD tools the day the first 36MP chip of that design rolled off the assembly line in late 2011.

Further, note that the suffix of the chip name (eg IMX094AQP for the D800) is said to reflect:
- revision number (A, B, C etc), then
- the mono/chroma for the bayer plane (usually Q is the color bayer and L is monochromatic), and
- the last letter is the chip package or 'substrate?' (Q for plastic QFP [Quad Flat Package], E/P for ceramic LCC [Leadless Chip Carrier], L/R for ceramic LGA [Land Grid Array package], Z for QFN [Quad Flat No-leads]).

The Sony 36MP A7R chip is the IMX094AQR, indicating package variation from the D800 (LGA vs LCC), so it's certainly possible that the K-1 may also have package variation from the D800 or A7R. I think some of these package variations (eg D800 plastic vs A7R ceramic) may certainly have different thermal or other properties, as suggested in the Pentax engineer interview, which might play a role in chip performance.

Someone needs to donate their K-1 to iFixit or Chipworks to stop this speculation.
It may not be that important to know what sensor specifications are from Pentax' input. There is still the sensor stack, that plays an important role in producing a typical Pentax IQ. Sony experimented with/improved the sensor microlens array for its own A7r.

What is pretty much beyond doubt though, is that the Pentax K1 Sony 36 mp sensor is part of the same family as the Sony A7r sensor and the Nikon D800(e) sensor, whereas the Nikon D810 sensor is the odd one out, featuring a more expensive design introduced by Nikon for maximum detail, smoothness and dynamic range in bright light and high contrast scenarios. E.g. daylight landscapes, full daylight portaiture.
The Pentax K1 on the other hand, is boasting maximum performance at high iso levels, where the more expensive design of a Nikon D810 sensor is not required, but other factors like clever noise control are far more important.

I have not seen a single add or advertising snippet where Pentax emphasised superior IQ in bright light scenarios. They are fully going for the clean high iso thrill. I would have liked a superior bright light camera, but that would not have been possible at the K1 price. Still, if you are paying 1900,- for a camera body, why not pay 200,- more for a true iso 64 body with superior smoothness and dynamic range in bright scenarios with high contrast. I don't do much night photography myself....

Chris

03-21-2016, 03:29 AM   #140
Loyal Site Supporter




Join Date: Mar 2009
Location: Gladys, Virginia
Photos: Gallery
Posts: 15,744
QuoteOriginally posted by Chris Mak Quote
It may not be that important to know what sensor specifications are from Pentax' input. There is still the sensor stack, that plays an important role in producing a typical Pentax IQ. Sony experimented with/improved the sensor microlens array for its own A7r.

What is pretty much beyond doubt though, is that the Pentax K1 Sony 36 mp sensor is part of the same family as the Sony A7r sensor and the Nikon D800(e) sensor, whereas the Nikon D810 sensor is the odd one out, featuring a more expensive design introduced by Nikon for maximum detail, smoothness and dynamic range in bright light and high contrast scenarios. E.g. daylight landscapes, full daylight portaiture.
The Pentax K1 on the other hand, is boasting maximum performance at high iso levels, where the more expensive design of a Nikon D810 sensor is not required, but other factors like clever noise control are far more important.

I have not seen a single add or advertising snippet where Pentax emphasised superior IQ in bright light scenarios. They are fully going for the clean high iso thrill. I would have liked a superior bright light camera, but that would not have been possible at the K1 price. Still, if you are paying 1900,- for a camera body, why not pay 200,- more for a true iso 64 body with superior smoothness and dynamic range in bright scenarios with high contrast. I don't do much night photography myself....

Chris
I would just say that most places where you will be able to use iso 64 you could also use pixel shift, which should get you a bump in dynamic range, resolution and color depth. Obviously if there is much of a breeze going on or something like that, pixel shift isn't going to give as much benefit -- my understanding is that artifacted areas will just be replaced by the first image capture -- but in those situations, iso 64 probably won't be an option either due to slow shutter speed.
03-21-2016, 07:46 AM   #141
Pentaxian




Join Date: Jun 2009
Location: GMT +10
Photos: Albums
Posts: 10,751
QuoteOriginally posted by Chris Mak Quote
They are fully going for the clean high iso thrill. I would have liked a superior bright light camera, but that would not have been possible at the K1 price. Still, if you are paying 1900,- for a camera body, why not pay 200,- more for a true iso 64 body with superior smoothness and dynamic range in bright scenarios with high contrast.
???
You are wildly speculating here, both about the tuning of the D810, and design bias of the K-1 towards 'the clean high iso thrill'.

It had never even occured to me, in all that I have seen, that Pentax were making high-ISO IQ their top priority with the K-1. I don't know where you are getting your ideas from.

Let's just see how K-1 v1.0 tests out once it is released. The K-1 and D810 may very well prove to be twins.
03-21-2016, 08:52 AM   #142
Site Supporter




Join Date: Mar 2016
Posts: 155
QuoteOriginally posted by rawr Quote
???
You are wildly speculating here, both about the tuning of the D810, and design bias of the K-1 towards 'the clean high iso thrill'.

It had never even occured to me, in all that I have seen, that Pentax were making high-ISO IQ their top priority with the K-1. I don't know where you are getting your ideas from.

Let's just see how K-1 v1.0 tests out once it is released. The K-1 and D810 may very well prove to be twins.
I completely agree with you

The big changes from the D800E to the D810 were the revised shutter mechanism to reduce vibration, lower base ISO 64 and the physical removal of the anti aliasing filter (disabled on the D800E and absent on the D810).
In terms of sensor performance the difference between the D810 and D800E is fairly negligible. The D810 tends to hold dynamic range a little better at higher ISO's but is noisier than the D800E which tends to have a little better noise handling but also noticeably better color handling at high ISO. pro's and cons to both. To see any difference you really need to be pixel peeping between both.

In terms of tonality at ISO 100 the D800E and D810 across the majority of the ISO range are a wash.
Nikon D810 versus Nikon D800 versus Nikon D800E: Outstanding performance - DxOMark

I have not seen any evidence to suggest that the D810 sensor is a more expensive to produce sensor based on any technical advantage over the D800E, merely that it is an evolution of the same sensor family.
Even if it was technically superior, personally I would take a sensor with in-body image-stabilization any day of the week over one without.
In practical real world shooting conditions, this feature alone will allow you to explore more of your cameras theoretical image quality, more of the time.

As for sensor shift - where it can be leveraged it should provide significantly higher colour accuity, colour accuracy, tonality, lower noise and detail as it effectively cancels out the effect of the bayer array. To put this in perspective it's like getting a FF foveon sensor or even (for black and whites ) could give us a glimpse of what a theoretical 36mp leica monochrom could achieve as you are getting true pixel luminance values at each pixel!

To get better detail/tonality than a sensor shift K1 Lloyd Chambers suggested that you may need to look at medium format north of 50mp.
diglloyd: Pentax K1: Super Resolution Mode Appears to be Best DSLR Quality Yet
03-21-2016, 09:04 AM   #143
Senior Member




Join Date: Jan 2013
Posts: 242
QuoteOriginally posted by rawr Quote
???
You are wildly speculating here, both about the tuning of the D810, and design bias of the K-1 towards 'the clean high iso thrill'.

It had never even occured to me, in all that I have seen, that Pentax were making high-ISO IQ their top priority with the K-1. I don't know where you are getting your ideas from.

Let's just see how K-1 v1.0 tests out once it is released. The K-1 and D810 may very well prove to be twins.
There is no "design bias" with the K1 sensor. Pentax simply adopted the most common and more affordable standard version of the Sony 36mp sensor. And there is nothing wrong with that, I use the Sony A7r myself, and it is a great sensor.
What I meant to say, is that all the claims (from Pentax themselves and others) about the K1 sensor being "better" or "better tweaked" than the Nikon D800(E) sensor are revolving around iso 51000 or somewhat being cleaner than on the Nikon.
I don't understand the drive to prove that the K1 sensor is "not the old Nikon D800 sensor" or "better than the Nikon D800 sensor" to begin with, it is simply the going Sony 36mp sensor.
But it starts to get shady when even Pentax rep's or even allegedly, engineers, start coming up with all sorts of misinformation merely to play on people's irrational wishes for the K1 to beat the Nikon bodies.

I am NOT wildly speculating on the Nikon D810 sensor being a different, more expensive version of the standard Sony 36mp sensor. The specs, the lower true base iso should be enough for anybody with an unbiased mindset and some sound technical insight.

Chris

03-21-2016, 09:17 AM   #144
Loyal Site Supporter
monochrome's Avatar

Join Date: Mar 2008
Location: Kirkwood (St. Louis) MO
Photos: Gallery | Albums
Posts: 20,190
Do you know that for a fact or are you also speculating? I'm just curious - though it sounds snarky it isn't meant that way.
03-21-2016, 03:50 PM   #145
Loyal Site Supporter




Join Date: Apr 2010
Location: Tennessee
Posts: 5,708
Challengers | PENTAX K-1 Special site | RICOH IMAGING
03-21-2016, 04:06 PM - 3 Likes   #146
Pentaxian
noelpolar's Avatar

Join Date: Oct 2013
Location: Goolwa, SA
Posts: 3,065
Sometimes I go to type stuff and then think..... why bother...... how many more sleeps before mine arrives.....
03-21-2016, 05:02 PM   #147
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,863
QuoteOriginally posted by Rondec Quote
I would just say that most places where you will be able to use iso 64 you could also use pixel shift, which should get you a bump in dynamic range, resolution and color depth.
I really like and admire the pixel shift mode Pentax and Olympus have pioneered. Esp. as Pentax may now offer the effective highest resolution short of medium format.

But comments like the one quoted IMHO carry it too far.

It is ALWAYS an option with any camera to apply stacking technology. Either in post production, or in camera as offered by Sony as their "Smooth Reflection App" (a much underrated feature which creates a RAW from not 4, but up to 256 exposures in burst).

One should NOT confuse native camera performance with results obtained by stacking software. I understand that Pentax pixel shift does better than just stack, as it gets rid of the demosaicing step. But that may reduce artefacts and improve resolution, it just equals other stacking results in dynamic range or color depth.

Moreover and this is my biggest criticism to date, there doesn't seem to be an unbiased analysis of pixel-shift results vs. stacking the same #frames, e.g., using PhotoAcute and allowing for minor random variations between frames (as in monopod landscape bursts) to achieve super-resolution. Most comparisons are on/off comparisons with out-of-camera JPGs which are more a test of the firmware JPG engine (i.e., its demosaicer) than anything else.

I really like the pixel shift feature. But IMHO, it is mentioned FAR too often in current discussions on pentax forums (and Pentax forums only).
03-21-2016, 06:08 PM   #148
Pentaxian




Join Date: Jun 2009
Location: GMT +10
Photos: Albums
Posts: 10,751
'PENTAX has developed a solid policy for defining the image quality of a new SLR camera by optimizing the balance between image resolution and sensitivity.'

That whole article works through how they struck their balance. It doesn't read, to me, like they were hell-bent on making the K-1 a dedicated high-ISO machine.

Even the high-ISO K-1 samples Ricoh provide in the 'Expression' page of the sub-site max out at ISO 25600. Probably as high as anyone would want to practically go with a K-1/D810/A7R, or even A7RII.

Pentax did the same thing with the K-5 vs the D7000, even though they used an identical sensor - Pentax claimed K-5 max ISO was ISO 51200, Nikon said the D7000's ISO ceiling was ISO 25600. All mere marketing.
03-21-2016, 06:35 PM   #149
Loyal Site Supporter




Join Date: Mar 2009
Location: Gladys, Virginia
Photos: Gallery
Posts: 15,744
QuoteOriginally posted by falconeye Quote
I really like and admire the pixel shift mode Pentax and Olympus have pioneered. Esp. as Pentax may now offer the effective highest resolution short of medium format.

But comments like the one quoted IMHO carry it too far.

It is ALWAYS an option with any camera to apply stacking technology. Either in post production, or in camera as offered by Sony as their "Smooth Reflection App" (a much underrated feature which creates a RAW from not 4, but up to 256 exposures in burst).

One should NOT confuse native camera performance with results obtained by stacking software. I understand that Pentax pixel shift does better than just stack, as it gets rid of the demosaicing step. But that may reduce artefacts and improve resolution, it just equals other stacking results in dynamic range or color depth.

Moreover and this is my biggest criticism to date, there doesn't seem to be an unbiased analysis of pixel-shift results vs. stacking the same #frames, e.g., using PhotoAcute and allowing for minor random variations between frames (as in monopod landscape bursts) to achieve super-resolution. Most comparisons are on/off comparisons with out-of-camera JPGs which are more a test of the firmware JPG engine (i.e., its demosaicer) than anything else.

I really like the pixel shift feature. But IMHO, it is mentioned FAR too often in current discussions on pentax forums (and Pentax forums only).
Thanks for the put-down.

My point was merely that both iso 64 and pixel shift are limited in usefulness in that best results will come using a tripod. For many people, hand holding is what they rely on and if that is the case, then both of these features will have limited value, as iso 64 is going to be limited to very good light.

Once you decide you will use a tripod, you can shoot whatever iso you want or as many exposures as you want and combine them in post. That said, I do think the pixel shift feels like an easier way to get better image quality, but I will see when I get a K-1. I guess my hope wasn't for "super-resolution," but for improvement in color depth and dynamic range, that can aid ultimately in improved image processing afterwards.
03-21-2016, 06:36 PM   #150
Loyal Site Supporter




Join Date: Apr 2010
Location: Tennessee
Posts: 5,708
QuoteOriginally posted by rawr Quote
'PENTAX has developed a solid policy for defining the image quality of a new SLR camera by optimizing the balance between image resolution and sensitivity.'

That whole article works through how they struck their balance. It doesn't read, to me, like they were hell-bent on making the K-1 a dedicated high-ISO machine.

Even the high-ISO K-1 samples Ricoh provide in the 'Expression' page of the sub-site max out at ISO 25600. Probably as high as anyone would want to practically go with a K-1/D810/A7R, or even A7RII.

Pentax did the same thing with the K-5 vs the D7000, even though they used an identical sensor - Pentax claimed K-5 max ISO was ISO 51200, Nikon said the D7000's ISO ceiling was ISO 25600. All mere marketing.
I think they needed to hit the 36MP mark so that APS-C crop mode would give them a good number of MPs. With all of the existing APS-C users who will be moving over to the K-1 that was a very important consideration.

The sensor is cut from the same wafer as the Nikon D810, but that doesn't mean it has the same micro lenses or color array. Ricoh says they made their own substrate for the sensor. What that means is beyond my limited knowledge of sensor production so it may just be marketing. It appear Ricoh has a very custom version of the D810 sensor.

ďItís not possible to achieve high sensitivity simply by reducing sensor noise, because the circuit configuration of the substrate to which the sensor is mounted greatly affects the type of generated noise. Thatís why we designed an original substrate for the PENTAX K-1 by making the best use of our know-how.Ē
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!
camera, cameras, chip, d800, dslr, firmware, full frame, full-frame, high-iso, iso, jpeg, k-1, k1, louis, nikon, pentax, pentax k-1, photo, rest, sensor, settings, sony, sr, yr
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Hands on with the Pentax K-1 Mousehill Pentax Full Frame 25 03-05-2016 01:03 PM
Details and Shots on the upcoming FF infinity Pentax Full Frame 45 12-01-2015 10:14 AM
DA 50mm 1.8 - Brief hands-on impressions and samples JPT Pentax SLR Lens Discussion 190 09-12-2014 10:17 AM
Pentax K-3 Hands-on and Sample Photos frank Pentax K-3 116 10-12-2013 03:57 AM
Hands-On Photos of the K-01 from Pentax Singapore User Group creampuff Pentax K-01 299 12-03-2012 08:01 AM



All times are GMT -7. The time now is 02:37 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