[Serkevan]

I hope too. Even if we got 70 to 90 points with a better and accurate auto-focus would be a huge improvement. I was close to buy a Nikon due to the auto-focus. Thanks to the rumors I saw, I am waiting on the tests in real life like I said from real Pentaxians.

PS: The only flaw I see is the quality of the joystick which may fail in the time.

Honestly, anything over 50 points will probably be good enough for the needs of most people. If the tracking algorithms are good, the performance will be already much better than current bodies. The 7Dii has 65 and the 90D has 45 - both are great. The Nikon D7500 is around 50. The 150 points of the D5/D500 are of course a halo number that would be good, but I do not believe it is necessary to go there as long as the software is up to snuff.

---------- Post added 09-16-20 at 09:08 AM ----------

I have no idea if it is possible to implement a measurement sensor, in combining CDAF (or the variant described in the patent) and the light sensor in that. If the coverage of the AF active area will increase

Theoretically, it should be possible, no? "Just" increasing the resolution of the RGB sensor to the point where it's, as Beholder says, more akin to a smartphone image sensor, could give a DSLR the AF system of a mirrorless camera - the only difference would be moving the light sourcing from the main sensor to a secondary sensor that is already "stealing" light from the viewfinder and recognizing some rudimentary shapes. I doubt we will be quite there with the K-New, but I think it's a logical conclusion of the RGB-for-face-recognition-during-OVF-shooting idea. If a 180K RGB sensor can recognize faces and give the AF module a "use these points" indication, would a, say, 2 MP sensor be capable of managing the entire focusing job?

The main drawback I can think of is stray backlight (through the eyepiece) messing up the focus - it can already throw off the exposure in some circumstances, if they integrate AE/AF in the same chip it stands to reason that both sides would be affected.

The extra excitement part comes from another couple of realizations: No need for submirror means the mirror gets automatically lighter - which might help with mirror slap and/or maximum frame rate. And the free empty space could lead to better arrangement of internal components (is it enough to safely fit a small battery, chargeable by USB?)

It's ironic that contrast detection/on sensor phase detection/dual pixel AF, technologies developed for bringing MILCs up to DSLR level have improved to the point where the DSLR could conceivably reap the benefits back and breathe new life into the OVF way.

---

Last edited by Serkevan; 09-16-2020 at 09:12 AM.

[beholder3]

Honestly, anything over 50 points will probably be good enough for the needs of most people. If the tracking algorithms are good, the performance will be already much better than current bodies. The 7Dii has 65 and the 90D has 45 - both are great. The Nikon D7500 is around 50. The 150 points of the D5/D500 are of course a halo number that would be good, but I do not believe it is necessary to go there as long as the software is up to snuff.

Agree.

The D500 only offers 55 AF Points to select from. The remainder is "support".

So the highest count actually is 65 from the 7d2.

---------- Post added 16th Sep 2020 at 18:12 ----------

That certainly would brighten up the viewfinder image. But to me the K-new cutaway image strongly suggests that the structure is still there for a PDAF assembly, which of course needs that secondary mirror.

It will be very interesting. The patent at least combines AF and meter. So either the metering moves down in the box or the AF moves to the prism.

[swestin]

A new prism made form different materials is actually quite a challenge. The change in refractive index of the glass means the whole shape of the prism had to change too. Directing light is never easy when you have to pass it through a substance.


On top of that each wavelength bends a bit different so they do not all arrive in the same place. As on college professor put it. On one end of the visible spectrum you have the great big red firetruck. On the other end you have the little blue Volkswagen. Now you have to drive it through a sturdy brick wall. Which vehicle stands the best chance of making the straightest path through the wall if they are both going the same velocity?

Actually, no. Where the light passes through a glass/air interface, it is (nominally) normal to the surface, so there should be no refraction. Most of what happens inside a pentaprism is total internal reflection, where only the reflectance varies with angle (how bright), not the geometry. The only way refractive index influence affects where the light goes is where the transitions in and out aren't exactly perpendicular.

[kokol]

Just as a comment from a new moderator: These last few pages are a model of a great speculative discussion. Excellent back and forth, supported by good info and ideas. Might use this in the future to point people to. Thank you all very much.

I agree. I have been lurking here for 13 years and barely ever post, but I have learned a lot about how my trusty-K-5IIs-that-need-to-be-replaced actually works through this thread. Many many thanks to all contributors and moderators!

On the K-New: I have been trying to find references to Ooku who seems to be a Pentax insider. On a Chinese forum I have found the following message, dated Sept 12th:

"Canon’s main market is the mainland. The mainland epidemic has long been under control. Pentax’s main market is Europe, America and Japan. In addition, R5 and R6 production capacity is very low, with a monthly output of more than 10,000 units.
So now R5 and R6 can only be bought with additional money. Pentax factories are in Southeast Asia, and they are also very seriously affected by the epidemic. Now many equipment in production are seriously out of stock"

I hope I am not diverting away too much from the topic. I am just really excited and looking forward to this camera, and can't wait for the specs to be released and the camera to be in my hands. But I hope supply chain issues can be resolved soon and won't delay the release date further. Pentax, please release the thing and take my money already.
Ooku, feel free to chip in

[bencoskater]

The K-5 and K-1 have different conceptions. I drew the path of the light with arrows (red) in the the side view. If the light had to go on the top of the prism, the mirror should be reversed (yellow). Apparently there is an OVF/EVF patent from Ricoh/Pentax: ??2018-151588 | ???????IP Force? (published in sept. 2018. I didn't read it yet) if this is the case, the small components in blue should be the screen for the EVF. A mirrorless in a DSLR? Is it the red light indicating the fockus locked ?



Again, this is just my assumption, maybe I am wrong. We are here to make assumptions and discuss about it.

---------- Post added 09-16-20 at 05:37 PM ----------

Honestly, anything over 50 points will probably be good enough for the needs of most people. If the tracking algorithms are good, the performance will be already much better than current bodies. The 7Dii has 65 and the 90D has 45 - both are great. The Nikon D7500 is around 50. The 150 points of the D5/D500 are of course a halo number that would be good, but I do not believe it is necessary to go there as long as the software is up to snuff.

---------- Post added 09-16-20 at 09:08 AM ----------



Theoretically, it should be possible, no? "Just" increasing the resolution of the RGB sensor to the point where it's, as Beholder says, more akin to a smartphone image sensor, could give a DSLR the AF system of a mirrorless camera - the only difference would be moving the light sourcing from the main sensor to a secondary sensor that is already "stealing" light from the viewfinder and recognizing some rudimentary shapes. I doubt we will be quite there with the K-New, but I think it's a logical conclusion of the RGB-for-face-recognition-during-OVF-shooting idea. If a 180K RGB sensor can recognize faces and give the AF module a "use these points" indication, would a, say, 2 MP sensor be capable of managing the entire focusing job?

The main drawback I can think of is stray backlight (through the eyepiece) messing up the focus - it can already throw off the exposure in some circumstances, if they integrate AE/AF in the same chip it stands to reason that both sides would be affected.

The extra excitement part comes from another couple of realizations: No need for submirror means the mirror gets automatically lighter - which might help with mirror slap and/or maximum frame rate. And the free empty space could lead to better arrangement of internal components (is it enough to safely fit a small battery, chargeable by USB?)

It's ironic that contrast detection/on sensor phase detection/dual pixel AF, technologies developed for bringing MILCs up to DSLR level have improved to the point where the DSLR could conceivably reap the benefits back and breathe new life into the OVF way.

I agree with you. even 50 to 70 points should be enough. Its possible there are selectable points and the rest is used by the cpu for tracking.

[rawr]

The Nikon D7500 is around 50. The 150 points of the D5/D500 are of course a halo number that would be good, but I do not believe it is necessary to go there as long as the software is up to snuff.

And the supporting hardware. A key feature of Nikon's much admired '3-D tracking' AF mode, for example, is the support the RGB sensor provides to the AF by feeding into the AF engine colour telemetry (movement of red shape [football jersey] against green [grass] or white [fence] etc etc).

[repaap]

The K-5 and K-1 have different conceptions. I drew the path of the light with arrows (red) in the the side view. If the light had to go on the top of the prism, the mirror should be reversed (yellow). Apparently there is an OVF/EVF patent from Ricoh/Pentax: ??2018-151588 | ???????IP Force? (published in sept. 2018. I didn't read it yet) if this is the case, the small components in blue should be the screen for the EVF. A mirrorless in a DSLR? Is it the red light indicating the fockus locked ?



instagram view full size

Again, this is just my assumption, maybe I am wrong. We are here to make assumptions and discuss about it.

---------- Post added 09-16-20 at 05:37 PM ----------




I agree with you. even 50 to 70 points should be enough. Its possible there are selectable points and the rest is used by the cpu for tracking.

It would make sense that this item inside of blue circle is new RGB sensor which is actually connected to Safox(ala Nikon version). And that prism being super clear will give it very good reading due new placement of that sensor too. Perhaps it is sensitive so that when there is no eye covering OVF it shuts down something preventing light interrupting readout from ‘behind’

---------- Post added 09-16-20 at 19:57 ----------

And the supporting hardware. A key feature of Nikon's much admired '3-D tracking' AF mode, for example, is the support the RGB sensor provides to the AF by feeding into the AF engine colour telemetry (movement of red shape [football jersey] against green [grass] or white [fence] etc etc).

Indeed. If they have managed to make something in that direction. It would be amazing.

[MMVIII]

It would make sense that this item inside of blue circle is new RGB sensor which is actually connected to Safox(ala Nikon version). And that prism being super clear will give it very good reading due new placement of that sensor too. Perhaps it is sensitive so that when there is no eye covering OVF it shuts down something preventing light interrupting readout from ‘behind’

---------- Post added 09-16-20 at 19:57 ----------



Indeed. If they have managed to make something in that direction. It would be amazing.

I'm sorry, but there is no speculation necessary what the thing in the blue circle is. It might have been lost in the thread by now but I took the drawing I projected into the schematic cut of K-new from the Patent that describes a new viewfinder. It is clearly labelled as the "illumination light source" and is for the superimposing optical system (information display device).

Read more at: Patents possibly linked to the APS-C flagship - Page 3 - PentaxForums.com

This is not a real Ricoh drawing, I just combined two not necessarily related sources. Just for the purpose of fun and speculation.

But it is not a sensor, but an emitter. This is clearly visible also by looking at the angle of the optical element below it. It directs light coming from it to the eyepiece.

[repaap]

I'm sorry, but there is no speculation necessary what the thing in the blue circle is. It might have been lost in the thread by now but I took the drawing I projected into the schematic cut of K-new from the Patent that describes a new viewfinder. It is clearly labelled as the "illumination light source" and is for the superimposing optical system (information display device).

Read more at: Patents possibly linked to the APS-C flagship - Page 3 - PentaxForums.com

This is not a real Ricoh drawing, I just combined two not necessarily related sources. Just for the purpose of fun and speculation.

But it is not a sensor, but an emitter. This is clearly visible also by looking at the angle of the optical element below it. It directs light coming from it to the eyepiece.

Thanks!

[bencoskater]

I'm sorry, but there is no speculation necessary what the thing in the blue circle is. It might have been lost in the thread by now but I took the drawing I projected into the schematic cut of K-new from the Patent that describes a new viewfinder. It is clearly labelled as the "illumination light source" and is for the superimposing optical system (information display device).

Read more at: Patents possibly linked to the APS-C flagship - Page 3 - PentaxForums.com

This is not a real Ricoh drawing, I just combined two not necessarily related sources. Just for the purpose of fun and speculation.

But it is not a sensor, but an emitter. This is clearly visible also by looking at the angle of the optical element below it. It directs light coming from it to the eyepiece.

I agree @MMVIII, in the view I posted earlier I traced the direction of the mirror if it would be a sensor. I read the following patent (Google translated) ??2018-151588 | ???????IP Force?, if it's related to this camera. Strangely they are talking about displaying information, focus points and they're also talking about a LCD display (like used for Live View) giving not enough light to project the image on the OVF:

The optical finder of a single-lens reflex camera deflects the optical path of the subject image by the photographing lens to form it on the focal plate, and the subject image formed on the focal plate is positive via an image inversion means such as a pentaprism. After making a standing image, it is enlarged and displayed by the loupe optical system so that the observer can observe it (OVF for short).[0003]
On the other hand, in such a single-lens reflex camera, the subject image by the photographing lens is imaged on the image sensor when the quick return mirror is retracted, and the signal (image signal) output from the image sensor is displayed on the rear liquid crystal of the camera. Therefore, it is common to have a live view function for confirming the subject image.[0004]
Further, in a mirrorless interchangeable-lens camera that does not have a quick return mirror or an optical viewfinder, there is a type in which a subject image formed on an image sensor is observed with an electronic viewfinder (EVF for short).[0005]
In the "so-called live view" display using the rear liquid crystal of the camera, it is better to observe with the electronic viewfinder because the brightness of the liquid crystal is not sufficient in a bright place such as outdoors in the daytime, and even in the optical viewfinder, it is on the image sensor. A finder device capable of observing an imaged electronic image is desired.[0006]
On the other hand, a superimpose (abbreviated as SI) function for displaying a relatively coarse point image such as a AF point position at the time of autofocus on an image of an optical finder has been known for a long time. Here, if the quick return mirror can be housed and the subject image formed by the photographing lens formed on the image sensor can be displayed by the superimpose function, the optical viewfinder can function as an electronic viewfinder.[0007]
Patent Document 1 and Patent Document 2 disclose a technique related to a superimpose function for displaying an image of an optical finder by superimposing a focus measuring point at the time of autofocus.
....
[0038]
The conventional superimpose optical system that branches at the reflective surface of an upright member such as a prism is reduced by about 0.5 to 0.7 times with a reduction lens with respect to the real image of the photographing lens (reverse through from the eye point side). It is used in the display size (which is reduced when viewed in the arrangement of). Since the loupe magnification of the optical finder of a general single-lens reflex camera is about 3 to 4 times, the loupe magnification of only about 5 to 8 times can be achieved as a superimpose optical system. However, in order to perform superimpose and EVF display over a relatively wide range (70% of the OVF region in the example) using an inexpensive and compact liquid crystal display system, a high loupe magnification (70% of the OVF region in the embodiment) is achieved only with a conventional reduction lens. In the examples, 19.1 times) cannot be achieved.
[0039]
Therefore, in the present embodiment, conventionally, the image of the small liquid crystal display is re-imaged and projected by using the relay optical system (re-imaging optical system) at the position of the display system arranged in front of the reduction lens. That is, the present embodiment is a method in which the display on the small liquid crystal display unit is enlarged by the re-imaging optical system to form a primary image, and the display is connected to the loupe optical system via the condenser lens system (condenser lens surface).

Are they going to implement an LCD to display only information (to display information it's sure at 100%) or will it be used like a real EVF in Live View mode for a better visibility against day light on the LCD screen?

A lot of questions.

[rawr]

there is no speculation necessary what the thing in the blue circle is

Seems legit. If the past placement of Pentax RGB sensors is any guide, the K-new RGB sensor is likely to be a little bit ahead of the blue circle, towards the front of the camera, atop the viewfinder eye-piece lenses.

But in previous Pentaxes the light source for illuminating the superimposed display elements (AF points, settings display etc) has not been at the top of the pentaprism, rather the front. Looks like they may be seriously re-doing the plumbing up there.

[reh321]

I agree with you. even 50 to 70 points should be enough. Its possible there are selectable points and the rest is used by the cpu for tracking.

Most important of all is processor power. I often reduce AF to one point in my KP because I find that works best.

[FozzFoster]

would be ballin' to have focus peaking in the viewfinder.
Or if it displayed a histogram in the bottom corner.

[house]

would be ballin' to have focus peaking in the viewfinder.
Or if it displayed a histogram in the bottom corner.

Ages ago when hybrid viewfinders were discussed I also thought that stuff would be great. Then I realised that I've tried cockpit hud like mirrorless viewfinders and don't like it at all. My K-1 is setup with no grids or af coverage markings in the viewfinder. (i found I needed that central circle to easily find my central af point).

There is of course the possibility that Pentax would come up with a great way of doing these things. One that doesn't ruin your view. A focus patch ala GR could work.

[repaap]

Ages ago when hybrid viewfinders were discussed I also thought that stuff would be great. Then I realised that I've tried cockpit hud like mirrorless viewfinders and don't like it at all. My K-1 is setup with no grids or af coverage markings in the viewfinder. (i found I needed that central circle to easily find my central af point).

There is of course the possibility that Pentax would come up with a great way of doing these things. One that doesn't ruin your view. A focus patch ala GR could work.

I actually agree. I Have focuspoints at there thou, but I want to look through my OVF with just the info I need at the bottom, as usual. If i'll use LCD, then it is different thing all together. I really really hope this new prism thing will be in next K-1 too.