[Rondec]

Lets have a look at the current models with bigger viewfinder
Canon EOS-1D X Mark III 100% 0.76×
Canon EOS 5D Mark IV 100% 0.71×
Nikon D6 100% 0.72×
Nikon D850 100% 0.75×

the same only 2 models
Canon EOS 6D Mark II 98% 0.71×
Nikon D780 100% 0.7×

As for EVF - the magnification of FF mirrorless Sony, Panasonic, Canon and Nikon is higher...

Others have addressed it, but this thread is about an APS-C camera that is going to have a larger viewfinder. The K-1 has an average full frame viewfinder. It is nice, but nothing particularly stellar. That is going to change with the K-new which will be further advanced. As for the EVFs, most of us here aren't big EVF fans, hence why we are still shooting with a brand that offers nice OVFs.

---------- Post added 08-30-20 at 05:47 AM ----------

I dispute the idea that image processing can improve the image quality of a camera.

It comes down to what you define as "image quality".

One definition, which I submit makes the most sense in the context of talking about cameras, is

"1. Level of fidelity with respect to the captured scene."

Another definition is

"2. Level of absence of undesirable image properties, for some definition of 'undesirable'."

This second definition clearly needs the subjective qualification I added since some people may want grain, incorrect colours, exaggerated contrast, etc. in their images. Some people don't find these aberrations to be "undesirable", on the contrary they would perceive a loss of "image quality" if skies didn't render with artificially blue colours or artificially low levels of noise, for instance.

Note that scenes with very low light levels will inevitably result in noisy images due to the stochastic nature of light itself. According the the first definition, the noise that was in the scene should be captured in the RAW data.

According to the second definition, that noise should be smoothed out (causing as little as possible damage to actual scene detail).

In my view, the role of a camera should be to record the scene as it is. Any choices regard aesthetics, e.g., what the appropriate level of noise is, should be left to JPEG processing in-camera or to post-processing. I don't object to an option of having some of those decisions made in-camera already, for those who appreciate not having to do the processing of their own. However, this latter option, that may improve "image quality" according to definition 2. above, should never ever be confused with increasing "image quality" according to definition 1. above.

To preempt the argument that the "accelerator unit" somehow manages to improve IQ according to definition "1." above: This is almost certainly impossible. Almost certainly the "accelerator unit" is in the same position as any RAW converter that has to process the data without any knowledge about the original scene or the artefacts introduced by the sensor during the capture process. In other words, it is very likely that neither of them have any additional information that would help them to reliably increase fidelity.

FWIW, if the "accelerator unit" used some additional information (such as black frames) that would not be available to a RAW converter then that same information could be made available to a RAW converter by storing the respective data in the RAW file.

We know this because the Sony sensors Ricoh use only provide a digital output stream, i.e., there is no way any "magic sauce" could be introduces by Ricoh to actually increase the quality of the data they are getting out of the sensor.

In summary, of the two definitions of "image quality" above, only the first one is independent of any subjective evaluations and is arguably the most natural when it comes to defining the "image quality" in the context of cameras. One can entertain thoughts about what processing steps increase a subjectively defined notion of "image quality" of final products, such as prints, but such considerations have nothing to do with the quality of the capturing process.

I can forgive Pentax marketing for trying to present their image processing efforts (they openly admit to "image processing") as improvements to "image quality", but I maintain that we must not confuse this with improvements that are the result of sensor-based innovations such as increased dynamic range, lowered noise floors, etc.

I have no problem with the accelerator, but to try to appease folks who do, I would suggest that it kick in at a point where it is agreed that image quality is starting to decline -- say, after iso 800 on APS-C and iso 1600 on a K-1. It is unlikely to me that it is adjustable, that is, I don't think you7 can switch it on or off, otherwise that might be a solution as well.

It does seem as though this is a big thing as Panasonic and Canon both have been shown to do some smoothing to their RAWs by Photons to Photos.

[howieb101]

RP is a very good camera for slow stills and 1080p video,IQ is nice.

IF,a new crop Pentax is better will be a surprise.However it wont be priced @ U$ 899!

Depends. An average image sensor. More a means by Canon to get them into RF mount.

[MMVIII]

I dispute the idea that image processing can improve the image quality of a camera.

It comes down to what you define as "image quality".

One definition, which I submit makes the most sense in the context of talking about cameras, is

"1. Level of fidelity with respect to the captured scene."

Another definition is

"2. Level of absence of undesirable image properties, for some definition of 'undesirable'."

This second definition clearly needs the subjective qualification I added since some people may want grain, incorrect colours, exaggerated contrast, etc. in their images. Some people don't find these aberrations to be "undesirable", on the contrary they would perceive a loss of "image quality" if skies didn't render with artificially blue colours or artificially low levels of noise, for instance.

Note that scenes with very low light levels will inevitably result in noisy images due to the stochastic nature of light itself. According the the first definition, the noise that was in the scene should be captured in the RAW data.

According to the second definition, that noise should be smoothed out (causing as little as possible damage to actual scene detail).

In my view, the role of a camera should be to record the scene as it is. Any choices regard aesthetics, e.g., what the appropriate level of noise is, should be left to JPEG processing in-camera or to post-processing. I don't object to an option of having some of those decisions made in-camera already, for those who appreciate not having to do the processing of their own. However, this latter option, that may improve "image quality" according to definition 2. above, should never ever be confused with increasing "image quality" according to definition 1. above.

To preempt the argument that the "accelerator unit" somehow manages to improve IQ according to definition "1." above: This is almost certainly impossible. Almost certainly the "accelerator unit" is in the same position as any RAW converter that has to process the data without any knowledge about the original scene or the artefacts introduced by the sensor during the capture process. In other words, it is very likely that neither of them have any additional information that would help them to reliably increase fidelity.

FWIW, if the "accelerator unit" used some additional information (such as black frames) that would not be available to a RAW converter then that same information could be made available to a RAW converter by storing the respective data in the RAW file.

We know this because the Sony sensors Ricoh use only provide a digital output stream, i.e., there is no way any "magic sauce" could be introduces by Ricoh to actually increase the quality of the data they are getting out of the sensor.

In summary, of the two definitions of "image quality" above, only the first one is independent of any subjective evaluations and is arguably the most natural when it comes to defining the "image quality" in the context of cameras. One can entertain thoughts about what processing steps increase a subjectively defined notion of "image quality" of final products, such as prints, but such considerations have nothing to do with the quality of the capturing process.

I can forgive Pentax marketing for trying to present their image processing efforts (they openly admit to "image processing") as improvements to "image quality", but I maintain that we must not confuse this with improvements that are the result of sensor-based innovations such as increased dynamic range, lowered noise floors, etc.

Isn't that funny. You claim to keep this topic alive by perpetuating an extreme theoretical position, because you are afraid that others, with this extreme theoretical position will bash a new camera that does not fulfill the generic rules of this extreme theoretical position. By this you program yourself for disappointment, because you proclaim a solution that did not get a single positive clue to be fulfilled by the manufacturer in its next camera. To the contrary, the pre-raw processing pipeline might include hardware already known and used by this manufacturer and they seem to be quite confident that they can deliver flawless image quality. Your position would not allow that in any case, even if you have absolutely no practical evidence that this decreases the image quality the slightest.
Good luck with that, I don't see any exit scenario for you if you continue your way.

[alfa75ts]

See the middle slot-like notch flushed with the camera back surface?

https://www.pentaxforums.com/forums/16-pentax-news-rumors/411737-another-sni...ml#post5063295

[reh321]

I dispute the idea that image processing can improve the image quality of a camera.

You may dispute it all you want to. That does not make your opinion true, any more than those opinions held 120 years ago that man can not fly or that carriages could not travel without horses were true.

Pentax has made few statements about the "K-3ii replacement"; we shall see what Pentax delivers.

[Class A]

I don't understand this purism in a system where all pictures are an interpretation of light as 1s and 0s that hit a sensor.

This "interpretation" is based on solid reconstruction techniques.

Yes, the Bayer CFA bandwidth characteristics do not coincide with that of a human standard observer which is why cameras have trouble getting certain hues right.
Yes, only one colour is captured per pixel position and the full RGB image must be interpolated through a process called demosaicing which can be performed in many different ways, with every approach yielding a slightly different reconstruction. The list goes on.

However, engineers know what to do in order to obtain a faithful as possible reconstruction because they know what information has been lost in which ways.

The fact that a perfect reconstruction is not possible does not imply that interpretations become arbitrary.

Of course no manufacturer is obliged to provide the most faithful reconstruction and many manufacturers deliberately do not do that in order to give you nicer "memory colours", sunsets, etc. However, again, this does not detract from the fact there is an objective notion of "image quality" and another, subjective one, that would be better referred to as "image appeal".

All I wanted to achieve is to point out the difference between these two notions.

[Class A]

I have no problem with the accelerator, but to try to appease folks who do, I would suggest that it kick in at a point where it is agreed that image quality is starting to decline -- say, after iso 800 on APS-C and iso 1600 on a K-1.

You use "image quality" in the sense of "image appeal", which is OK.

Just note that if the "accelerator unit" is a signal processor -- and there is no evidence to suggest anything else -- it can per definition not improve "image quality" in the sense of "image fidelity".

It is unlikely to me that it is adjustable, that is, I don't think you7 can switch it on or off, otherwise that might be a solution as well.

The "accelerator unit" is already "switched off" below ISO 640 in the K-1 II. It does not kick in gradually.
A signal processor can be designed to have an effect or not have an effect, depending on a single bit of input.
I understand that some don't want to believe this, but it remains a fact.

It does seem as though this is a big thing as Panasonic and Canon both have been shown to do some smoothing to their RAWs by Photons to Photos.

Why would an issue be bigger or smaller depending on whether other manufacturers apply similar technique in some ISO ranges or apply single pixel denoising, etc.?

There appears to be a misunderstanding that I wanted to "rant" against Pentax when I all I was doing was to point out that there are two sensible ways to interpret "image quality".

[Class A]

That does not make your opinion true, any more than those opinions held 120 years ago that man can not fly or that carriages could not travel without horses were true.

Did you read my post?

I did not offer an "opinion" on any technology limitations.

I defined a certain way of reading "image quality" that disallows, per definition, any reliable improvements after the fact.
In the future certain AI techniques may be able to increase image fidelity in many cases, but it -- per definition -- will always remain guesswork and hence sometimes fail.

Your characterisation of my post as an "opinion" and your comparison to ill-advised technology predictions do not make any sense because my argument is based on information preservation. Information that has been lost during capture cannot be recreated, the best you can do is to take a gamble on making the information look more complete in the knowledge that you'll be wrong many times.

Let me restate my position as follows: Put marbles with engravings on them into a cement mixer with, turn the latter on and let it run until all engravings have been polished away.
When I now claim that you won't be able to restore the marbles into their original states without ever having seen the original engravings, will you still make comparisons to airplanes and cars not having been predicted by some people? Will you still claim that some day there might be some technology that can restore the engravings?

[reh321]

Did you read my post?

I did not offer an "opinion" on any technology limitations.

I defined a certain way of reading "image quality" that disallows, per definition, any reliable improvements after the fact.
In the future certain AI techniques may be able to increase image fidelity in many cases, but it -- per definition -- will always remain guesswork and hence sometimes fail.

Your characterisation of my post as an "opinion" and your comparison to ill-advised technology predictions do not make any sense because my argument is based on information preservation. Information that has been lost during capture cannot be recreated, the best you can do is to take a gamble on making the information look more complete in the k nowledge that you'll be wrong many times.

Let me restate my position as follows: Put marbles with engravings on them into a cement mixer with, turn the latter on and let it run until all engravings have been polished away.
When I now claim that you won't be able to restore the marbles into their original states without ever having seen the original engravings, will you still make comparisons to airplanes and cars not having been predicted by some people? Will you still claim that some day there might be some technology that can restore the engravings?

Yes, I completely read through it several times. I believe you have used your interpretation of physics to substantiate opinions you already had; my KP gets enough information that your latest strawman is also irrelevant. As a user of a KP - often at "high ISO values" - my opinion is that first you need to use a modern Pentax camera. My experience does not agree with your theoretical view at all. You are welcome to use a K-1 or K-3 the rest of your life and hand-craft 'raw' files all you want to to produce your interpretation of the scene, but I am very happy with what is being delivered to me.

[Not a Number]

This thread has gone way beyond News and Rumor and has wandered off in many directions which often circle back to themselves. In addition the arguments are tending towards being personal.

The thread has outlived its useful life.

Closed for the duration.

[Parallax]

Another one bites the dust
Another one bites the dust
And another one gone, and another one gone
Another one bites the dust