Reading mode: Light Dark

New Pentax - maybe Nikon show the way forward..

Status
Not open for further replies.
Go to last post
Mark Ransom said:
Alex Sarbu said:
xmeda said:
You know what I mean. It was already too late 3 years ago. Many more guys left K mount since then.
Click to expand...
How about those who didn't?
Click to expand...
I'm sure you will both be very happy.

Of course I exaggerate, but you only have to look at the forum traffic to know how things have changed.
Click to expand...
And why exactly are you exaggerating, in such a disparaging way?

I'm seeing in this forum/site how quite a few of us bought K-1s and D FA*s and other products which are upper level from what Pentax had in their digital past. There must be a few hundred FF-using Pentaxians around.
Mark Ransom said:
Mark Ransom said:
Mark Ransom said:
But system needs to show some continuity instead being dead :)
Click to expand...
Except we know they're making the K-new, and they launch the occasional lens - we know it's not dead.
Click to expand...
Reminds me of a certain Monty Python sketch involving a parrot.
Click to expand...
It reminds me of another certain Monty Python sketch, the "bring out your dead" one. Particularly when that dude whacked the "I'm not dead!" guy in the head - problem solved.

There are people with clubs looking to whack Pentax in the head once again. And again. And again.
Mark Ransom said:
Mark Ransom said:
Things are moving slowly, way too slowly by my liking but exaggerating and finger pointing won't make them any faster.

Neither would killing the FF, effectively abandoning their best-paying customers. That would stop things... forever.
Click to expand...
Here I must agree. The biggest problem as I see it is that they're not big enough to support both FF and APS-C simultaneously.
Click to expand...
As I see it, they're too small to support only one of FF or APS-C.
Mark Ransom said:
Mark Ransom said:
Mark Ransom said:
Look at good old times.. Ist Ds, K10D, K20D, K7, K5... every 2 years new beast. Although with incremental upgrades. Then K5II distorted it a bit.. and since K3 things went down the hill.
Click to expand...
Hmm... "good times"?

The *istDS was my first DSLR; yet at that time Pentax was unable to do more than rehashing the *istD.
Click to expand...
Mine too. The big innovation there was taking the *ist-D and making it affordable.
Click to expand...
Everyone did that though - making affordable DSLRs.
Mark Ransom said:
Mark Ransom said:
The K10D was an amazing camera, but it had some nasty issues.

The K-5, when they presented it at Photokina they didn't even know how large is its buffer (that part of the firmware was work in progress). Then, sensor spots and AF issues in low light.

They launched more cameras, but it wasn't all rosy.
Mark Ransom said:
Same for low-end. Ist Dl, K100D, K200D, Km, Kx, Kr, K30, K50/500, (+KS1,KS2) and then only K70 and big gap since 2016 till today :(

And several lenses every year too.
Click to expand...
The lenses dropped dramatically after 2008 - when Hoya launched what Pentax Corporation was working on. In 2009 they made 3 WR versions of existing lenses and the remaining DA Limited; in 2010, just 2 lenses. In 2011, no K-mount lens was introduced.
Click to expand...
There's not much point in assigning blame any more. The past is what it is. It's much more instructive to look at the recent history.
Click to expand...
Isn't that also assigning blame, only exclusively to Ricoh?

Let's keep in mind, Ricoh only bought what was left of Pentax, and the market is in sharp decline. And they had their own problems.

I'm afraid the only company for which Pentax Imaging Systems was a core business was Asahi Optical/Pentax Corporation. This is the reality we have to deal with... no, Ricoh won't pull an Olympus losing large sums of money just to stay in the spotlight.

But they're continuing Pentax.
Mark Ransom said:
Mark Ransom said:
Mark Ransom said:
I'd love to get K mount 60-600 or the 50-100/1.8 to complement my 18-35/1.8 for lowlight/ shallow DOF oriented photos... or the 24/1.4 and 105/1.4.. :(

These are nowhere near in Ricoh future...
Click to expand...
Right, Pentax is no longer an option if you want 3rd-party lenses.
Click to expand...
The availability of third party lenses was a big part of what convinced me to go with Pentax in the first place. My two favorite lenses for the *ist-DS were Tamron. When Sigma announced that they weren't doing any new K-mount lenses I knew that things would never be the same.
Click to expand...
That's why I said what I said - 3rd-party lenses aren't important to me, but they are to others.

It's unfortunate Sigma dropped support (right when they could step forward and fill the D FA line gaps). It is also because we didn't buy enough Sigma lenses (and likely, they established we're too price driven to buy their expensive Art series).
Mark Ransom said:
Mark Ransom said:
Too bad they're so slow. I very much prefer the D FA* 50mm f/1.4 to the Sigma 50 and 40mm, and I'm quite sure I would prefer the upcoming D FA* 85mm f/1.4 instead of the 105.

And that (hopefully) reasonable sized ultra wide angle on the roadmap?
Click to expand...
Alex, I truly hope your your loyalty is someday rewarded.
Click to expand...
I'm not expecting any reward, except the occasional joy in using Pentax products.

I wouldn't say no to a factory or R&D center tour, though ;-)

Alex
 
DAVID MANZE said:
bob5050 said:
Can anyone say that they think Canon's or Nikon's model-to-model evolution is as significant as it once was?

bob5050
Click to expand...
Yes , I can ! ....... the last significant Nikon cameras were the D810 ...... D500 ....... D850 ......... Z6 ...... Z7 ...... (with the D850 thought to be the pinnacle of the DSLR)
Click to expand...
Is this a list of recent mid-range Nikon models? :-p

The D810, not that much difference from the D800(e). Both the Z6 and Z7? Then I'd better count the K-1 and K-1 II in the Pentax list.
DAVID MANZE said:
All of those cameras were huge model to model evolutions ..... especially when they played the ML card.
Click to expand...
No, not all of them.

Alex
 
DAVID MANZE said:
bob5050 said:
Can anyone say that they think Canon's or Nikon's model-to-model evolution is as significant as it once was?

bob5050
Click to expand...
Yes , I can ! ....... the last significant Nikon cameras were the D810 ...... D500 ....... D850 ......... Z6 ...... Z7 ...... (with the D850 thought to be the pinnacle of the DSLR)

All of those cameras were huge model to model evolutions ..... especially when they played the ML card.
Click to expand...
The D800 -> D810 update was similarily impactful as K-1 -> K-1 II. The D850 was a significant upgrade of this line in all respects, but I'd say, that it took around 5 years, dating back from D800.

D300S -> D500 evolution was absolutely significant, but then again, that took 7 years.

Mirrorless Zs are first Nikon cameras truly good for video, but that's about it. Nothing huge here, IMO.
 
DAVID MANZE said:
lowerCamelCase said:
DAVID MANZE said:
iso rivolta said:
ceterum said:
iso rivolta said:
Contrast detection with a secondary sensor was first introduced by Canon T80 in 1986. I'm not sure if that would help much. Alignment problems would make this again prone to misfocus. Would be better to have CDAF as a final step on the main sensor.
Click to expand...
Pentax could really deliver something useful here by offering an automatic AF-calibration (comparing the image-sensor-AF-result to the PD-sensor-AF-result, ideally for different focal lengths and distances).
Click to expand...
Nikon already has something similar in D850.

https://nps.nikonimaging.com/technical_solutions/d850_tips/useful/auto_af_fine-tuning/
Click to expand...
It may sound like a wonderful feature ... auto AF fine tune .... also in the D500 ....... it's just that it is hit and miss with wide shot to shot variations ....... to drive you to distraction ......

...... forget that feature ....... it will have you running round the houses !!
Click to expand...
This can be easily improved by software changes taking a bigger number of shots and simply looking at the convergence of the standard deviation.
Click to expand...
Exactly, you take a whole load of shots ..... too many, as the repeatability is poor with results like -4 to +6 for example ....... we know that theoretically LV CD AF is more accurate and that OVF PDAF is slightly less reliable ........ but averaging out accumulated errors from two AF systems is just adding errors not eliminating them!!

..... and if you have a range of lenses to calibrate, you can be there for hours as the whole process is based on averaging outliers .......

Dave's clichés
Click to expand...
Not quite, averaging statistical errors is not accumulating them. Averaging systematical errors is. This is why you look at the standard derivation and how it changes over the measurements and in the end you will get a better result.

The whole procedure of course assumes that the LV AF is without significant errors.

I don't think this will take hours. With modern cameras you can shoot 100 pictures in 10 seconds.
 
Heliar said:
The D800 -> D810 update was similarily impactful as K-1 -> K-1 II. The D850 was a significant upgrade of this line in all respects, but I'd say, that it took around 5 years, dating back from D800.

D300S -> D500 evolution was absolutely significant, but then again, that took 7 years.
Click to expand...
And that's essentially what my point was: that current technological advances only justify a new camera model maybe every 5-7 years. And getting one maybe about that often IF the areas of advancement happen to line up with your needs.

But, for the moment, the era of constant replacement is over.

bob5050
 
bob5050 said:
Heliar said:
The D800 -> D810 update was similarily impactful as K-1 -> K-1 II. The D850 was a significant upgrade of this line in all respects, but I'd say, that it took around 5 years, dating back from D800.

D300S -> D500 evolution was absolutely significant, but then again, that took 7 years.
Click to expand...
And that's essentially what my point was: that current technological advances only justify a new camera model maybe every 5-7 years. And getting one maybe about that often IF the areas of advancement happen to line up with your needs.

But, for the moment, the era of constant replacement is over.

bob5050
Click to expand...
A bright side of this is that we have more sources to buy lenses :)
 
bob5050 said:
Heliar said:
The D800 -> D810 update was similarily impactful as K-1 -> K-1 II. The D850 was a significant upgrade of this line in all respects, but I'd say, that it took around 5 years, dating back from D800.

D300S -> D500 evolution was absolutely significant, but then again, that took 7 years.
Click to expand...
And that's essentially what my point was: that current technological advances only justify a new camera model maybe every 5-7 years. And getting one maybe about that often IF the areas of advancement happen to line up with your needs.

But, for the moment, the era of constant replacement is over.

bob5050
Click to expand...
All the more reasons to finally bring something new, because if you are right Pentax has outdated cameras in a field where no significant progress is made. To me that makes it even worse.

Not that the K1 or K3 are bad cameras, but others can do better now, at least in the technology part.
 
lowerCamelCase said:
All the more reasons to finally bring something new, because if you are right Pentax has outdated cameras in a field where no significant progress is made. To me that makes it even worse.
Click to expand...
Or better, depending on your perspective. Sales for the K-new should be reasonably robust precisely because Pentax APS-C shooters are actually ready for an upgrade, and the technology justifies the buy. It won't be greeted as a ho-hum event.
bob5050
 
bob5050 said:
lowerCamelCase said:
All the more reasons to finally bring something new, because if you are right Pentax has outdated cameras in a field where no significant progress is made. To me that makes it even worse.
Click to expand...
Or better, depending on your perspective. Sales for the K-new should be reasonably robust precisely because Pentax APS-C shooters are actually ready for an upgrade, and the technology justifies the buy. It won't be greeted as a ho-hum event.
Click to expand...
Or maybe not. I was ready for an upgrade 2 years ago, but Pentax didn't have what I wanted so I went elsewhere. I'll be checking out the K-new just out of curiosity, but I won't be buying it.

Their lack of retail presence means that upgraders are pretty much their only market. Even if the K-new is the greatest camera ever invented, it's unlikely to bring newcomers to the brand.
 
Alex Sarbu said:
Mark Ransom said:
Alex Sarbu said:
xmeda said:
You know what I mean. It was already too late 3 years ago. Many more guys left K mount since then.
Click to expand...
How about those who didn't?
Click to expand...
I'm sure you will both be very happy.

Of course I exaggerate, but you only have to look at the forum traffic to know how things have changed.
Click to expand...
And why exactly are you exaggerating, in such a disparaging way?
Click to expand...
I guess I'm still a little bitter that I wasn't able to get what I wanted from them, when I wanted it. The observation about forum activity is almost certainly true, which means I'm not alone.
Alex Sarbu said:
I'm seeing in this forum/site how quite a few of us bought K-1s and D FA*s and other products which are upper level from what Pentax had in their digital past. There must be a few hundred FF-using Pentaxians around.
Click to expand...
Of course, which is why I freely admitted I was exaggerating.
Alex Sarbu said:
Alex Sarbu said:
Alex Sarbu said:
Alex Sarbu said:
But system needs to show some continuity instead being dead :)
Click to expand...
Except we know they're making the K-new, and they launch the occasional lens - we know it's not dead.
Click to expand...
Reminds me of a certain Monty Python sketch involving a parrot.
Click to expand...
It reminds me of another certain Monty Python sketch, the "bring out your dead" one. Particularly when that dude whacked the "I'm not dead!" guy in the head - problem solved.

There are people with clubs looking to whack Pentax in the head once again. And again. And again.
Click to expand...
Touché.
Alex Sarbu said:
Alex Sarbu said:
Alex Sarbu said:
Things are moving slowly, way too slowly by my liking but exaggerating and finger pointing won't make them any faster.

Neither would killing the FF, effectively abandoning their best-paying customers. That would stop things... forever.
Click to expand...
Here I must agree. The biggest problem as I see it is that they're not big enough to support both FF and APS-C simultaneously.
Click to expand...
As I see it, they're too small to support only one of FF or APS-C.
Click to expand...
You might be right about that. Damned if they do, damned if they don't.
Alex Sarbu said:
Alex Sarbu said:
Alex Sarbu said:
Alex Sarbu said:
Look at good old times.. Ist Ds, K10D, K20D, K7, K5... every 2 years new beast. Although with incremental upgrades. Then K5II distorted it a bit.. and since K3 things went down the hill.
Click to expand...
Hmm... "good times"?

The *istDS was my first DSLR; yet at that time Pentax was unable to do more than rehashing the *istD.
Click to expand...
Mine too. The big innovation there was taking the *ist-D and making it affordable.
Click to expand...
Everyone did that though - making affordable DSLRs.
Click to expand...
That's what led to the golden age of DSLRs, and Pentax was there in the thick of it.
Alex Sarbu said:
Alex Sarbu said:
Alex Sarbu said:
The K10D was an amazing camera, but it had some nasty issues.

The K-5, when they presented it at Photokina they didn't even know how large is its buffer (that part of the firmware was work in progress). Then, sensor spots and AF issues in low light.

They launched more cameras, but it wasn't all rosy.
Alex Sarbu said:
Same for low-end. Ist Dl, K100D, K200D, Km, Kx, Kr, K30, K50/500, (+KS1,KS2) and then only K70 and big gap since 2016 till today :(

And several lenses every year too.
Click to expand...
The lenses dropped dramatically after 2008 - when Hoya launched what Pentax Corporation was working on. In 2009 they made 3 WR versions of existing lenses and the remaining DA Limited; in 2010, just 2 lenses. In 2011, no K-mount lens was introduced.
Click to expand...
There's not much point in assigning blame any more. The past is what it is. It's much more instructive to look at the recent history.
Click to expand...
Isn't that also assigning blame, only exclusively to Ricoh?

Let's keep in mind, Ricoh only bought what was left of Pentax, and the market is in sharp decline. And they had their own problems.

I'm afraid the only company for which Pentax Imaging Systems was a core business was Asahi Optical/Pentax Corporation. This is the reality we have to deal with... no, Ricoh won't pull an Olympus losing large sums of money just to stay in the spotlight.

But they're continuing Pentax.
Click to expand...
My point was that looking at ancient history is not informative to how things are going to go in the future since so much has changed.
Alex Sarbu said:
Alex Sarbu said:
Alex Sarbu said:
Alex Sarbu said:
I'd love to get K mount 60-600 or the 50-100/1.8 to complement my 18-35/1.8 for lowlight/ shallow DOF oriented photos... or the 24/1.4 and 105/1.4.. :(

These are nowhere near in Ricoh future...
Click to expand...
Right, Pentax is no longer an option if you want 3rd-party lenses.
Click to expand...
The availability of third party lenses was a big part of what convinced me to go with Pentax in the first place. My two favorite lenses for the *ist-DS were Tamron. When Sigma announced that they weren't doing any new K-mount lenses I knew that things would never be the same.
Click to expand...
That's why I said what I said - 3rd-party lenses aren't important to me, but they are to others.

It's unfortunate Sigma dropped support (right when they could step forward and fill the D FA line gaps). It is also because we didn't buy enough Sigma lenses (and likely, they established we're too price driven to buy their expensive Art series).
Click to expand...
Even if you never buy a third party lens, it helps assure you that the system is a safe choice. It was a big part of my decision to buy my first Pentax DSLR, even though I didn't know at the time which lenses I would be getting.

I don't blame Sigma or anybody else for their withdrawal from K-mount support, I'm sure it makes perfect sense from their point of view. Unfortunate is indeed the right word.
Alex Sarbu said:
Alex Sarbu said:
Alex Sarbu said:
Too bad they're so slow. I very much prefer the D FA* 50mm f/1.4 to the Sigma 50 and 40mm, and I'm quite sure I would prefer the upcoming D FA* 85mm f/1.4 instead of the 105.

And that (hopefully) reasonable sized ultra wide angle on the roadmap?
Click to expand...
Alex, I truly hope your your loyalty is someday rewarded.
Click to expand...
I'm not expecting any reward, except the occasional joy in using Pentax products.

I wouldn't say no to a factory or R&D center tour, though ;-)
Click to expand...
Occasional joy would be reward enough, I think.
 
Mark Ransom said:
bob5050 said:
lowerCamelCase said:
All the more reasons to finally bring something new, because if you are right Pentax has outdated cameras in a field where no significant progress is made. To me that makes it even worse.
Click to expand...
Or better, depending on your perspective. Sales for the K-new should be reasonably robust precisely because Pentax APS-C shooters are actually ready for an upgrade, and the technology justifies the buy. It won't be greeted as a ho-hum event.
Click to expand...
Or maybe not. I was ready for an upgrade 2 years ago, but Pentax didn't have what I wanted so I went elsewhere. I'll be checking out the K-new just out of curiosity, but I won't be buying it.

Their lack of retail presence means that upgraders are pretty much their only market. Even if the K-new is the greatest camera ever invented, it's unlikely to bring newcomers to the brand.
Click to expand...
Actually, if Covid-19 did not happen, which cost me enough money to reconsider hobby investments (I am in the lucky group of people however who will financially recover and are not in serious issues, which I am thankful for), I would have switched last month, as the new Pentax FF is not even rumored yet.

I now decided to wait for the k_new and see what I has. Otherwise I will just keep the FA77 and K1-II, get a FA31 again and sell everything else but my analogue bodies and seriously invest into another brand. Which would be a shame, because I do like Pentax a lot in many ways. I must say however, I am not a fan of the changes from K3 to K1 in body design, with the smaller top display and neither like the kp body much, so maybe time for me is over anyway. No bad feelings, just that I hate to leave the brand behind I liked so much. Independently from my personal situation I do hope they make the right decisions to survive. Everything else would be tragic.
 
Mark Ransom said:
Or maybe not. I was ready for an upgrade 2 years ago,
Click to expand...
so was I
Mark Ransom said:
but Pentax didn't have what I wanted so I went elsewhere.
Click to expand...
And I didn't see anything I wanted bad enough to justify starting over. Pentax leaving money on the table doesn't automatically mean I'm willing to send it (and a lot more) to someone else. The K-3 still works fine, so I can wait on the K-next.
Mark Ransom said:
Their lack of retail presence
Click to expand...
I've always been a bit skeptical of this argument. We haven't had a camera store around here in maybe 5-6 years or longer. One used to be able to buy a Canon, Nikon, or Sony at Best Buy or CostCo, but not the good ones, just the low end. In both, the selection has paired back dramatically over the last two years to nearly nothing. So effectively, nobody has on-the-shelf presence any more.

bob5050
 
Last edited:
lowerCamelCase said:
DAVID MANZE said:
lowerCamelCase said:
DAVID MANZE said:
iso rivolta said:
ceterum said:
iso rivolta said:
Contrast detection with a secondary sensor was first introduced by Canon T80 in 1986. I'm not sure if that would help much. Alignment problems would make this again prone to misfocus. Would be better to have CDAF as a final step on the main sensor.
Click to expand...
Pentax could really deliver something useful here by offering an automatic AF-calibration (comparing the image-sensor-AF-result to the PD-sensor-AF-result, ideally for different focal lengths and distances).
Click to expand...
Nikon already has something similar in D850.

https://nps.nikonimaging.com/technical_solutions/d850_tips/useful/auto_af_fine-tuning/
Click to expand...
It may sound like a wonderful feature ... auto AF fine tune .... also in the D500 ....... it's just that it is hit and miss with wide shot to shot variations ....... to drive you to distraction ......

...... forget that feature ....... it will have you running round the houses !!
Click to expand...
This can be easily improved by software changes taking a bigger number of shots and simply looking at the convergence of the standard deviation.
Click to expand...
Exactly, you take a whole load of shots ..... too many, as the repeatability is poor with results like -4 to +6 for example ....... we know that theoretically LV CD AF is more accurate and that OVF PDAF is slightly less reliable ........ but averaging out accumulated errors from two AF systems is just adding errors not eliminating them!!

..... and if you have a range of lenses to calibrate, you can be there for hours as the whole process is based on averaging outliers .......

Dave's clichés
Click to expand...
Not quite, averaging statistical errors is not accumulating them.
Click to expand...
In effect you are averaging errors ..... too many spurious results.
lowerCamelCase said:
Averaging systematical errors is. This is why you look at the standard derivation and how it changes over the measurements and in the end you will get a better result.

The whole procedure of course assumes that the LV AF is without significant errors.

I don't think this will take hours. With modern cameras you can shoot 100 pictures in 10 seconds.
Click to expand...
The process is a slow and clunky one requiring tripods and targets, many button presses as well as waiting for the result, which "in theory was a definitive AFFT" ..... (at least 20 shots to get this average) ..... in reality it turned out not to be ..... and this meant this averaging workaround ..... pen and paper ...... and then the time spent taking a few test shots .......

...... if you have 15 lenses ....... ugh!

--
Dave's clichés
 
Last edited:
Heliar said:
DAVID MANZE said:
bob5050 said:
Can anyone say that they think Canon's or Nikon's model-to-model evolution is as significant as it once was?

bob5050
Click to expand...
Yes , I can ! ....... the last significant Nikon cameras were the D810 ...... D500 ....... D850 ......... Z6 ...... Z7 ...... (with the D850 thought to be the pinnacle of the DSLR)

All of those cameras were huge model to model evolutions ..... especially when they played the ML card.
Click to expand...
The D800 -> D810 update was similarily impactful as K-1 -> K-1 II. The D850 was a significant upgrade of this line in all respects, but I'd say, that it took around 5 years, dating back from D800.
Click to expand...
Five years is not a long time for that series ...... and don't forget the D7200/D7500 ...... the D5 ..... and now the D780 ....... how many cameras does it take to impress you?
Member said:
D300S -> D500 evolution was absolutely significant, but then again, that took 7 years.
Click to expand...
That was mainly stagnation not evolution and wasn't well received ...... once Nikon got on the case though ...... they nailed it !
Member said:
Mirrorless Zs are first Nikon cameras truly good for video, but that's about it. Nothing huge here, IMO.
Click to expand...
--
Dave's clichés
 
Last edited:
DAVID MANZE said:
Heliar said:
DAVID MANZE said:
bob5050 said:
Can anyone say that they think Canon's or Nikon's model-to-model evolution is as significant as it once was?

bob5050
Click to expand...
Yes , I can ! ....... the last significant Nikon cameras were the D810 ...... D500 ....... D850 ......... Z6 ...... Z7 ...... (with the D850 thought to be the pinnacle of the DSLR)

All of those cameras were huge model to model evolutions ..... especially when they played the ML card.
Click to expand...
The D800 -> D810 update was similarily impactful as K-1 -> K-1 II. The D850 was a significant upgrade of this line in all respects, but I'd say, that it took around 5 years, dating back from D800.
Click to expand...
Five years is not a long time for that series ...... and don't forget the D7200/D7500 ...... the D5 ..... and now the D780 ....... how many cameras does it take to impress you
Click to expand...
When I was considering my first camera, it was a tie between D7200 and Pentax K-3 II. In the end I went for Pentax. The D7500 wasn't well received afaik.

Don't get me wrong, I am impressed by what Nikon did. D850 is still the greatest FF camera on the market IMO, and it's almost 3 years old. Same goes for D500, ruling the APS segment 4 years after its release.

I just wanted to emphasize the time it took them to do all that great stuff, and how Ricoh Pentax, being much, much smaller, deserves at least some patience from the community.

Nikonians waited 7 years for crop sensor flagship update. Canoneers are still waiting and it will probably never come. I went for K-1 few years ago and sold my crop glass, but I'm excited to see what's coming.
 
DAVID MANZE said:
lowerCamelCase said:
DAVID MANZE said:
lowerCamelCase said:
DAVID MANZE said:
iso rivolta said:
ceterum said:
iso rivolta said:
Contrast detection with a secondary sensor was first introduced by Canon T80 in 1986. I'm not sure if that would help much. Alignment problems would make this again prone to misfocus. Would be better to have CDAF as a final step on the main sensor.
Click to expand...
Pentax could really deliver something useful here by offering an automatic AF-calibration (comparing the image-sensor-AF-result to the PD-sensor-AF-result, ideally for different focal lengths and distances).
Click to expand...
Nikon already has something similar in D850.

https://nps.nikonimaging.com/technical_solutions/d850_tips/useful/auto_af_fine-tuning/
Click to expand...
It may sound like a wonderful feature ... auto AF fine tune .... also in the D500 ....... it's just that it is hit and miss with wide shot to shot variations ....... to drive you to distraction ......

...... forget that feature ....... it will have you running round the houses !!
Click to expand...
This can be easily improved by software changes taking a bigger number of shots and simply looking at the convergence of the standard deviation.
Click to expand...
Exactly, you take a whole load of shots ..... too many, as the repeatability is poor with results like -4 to +6 for example ....... we know that theoretically LV CD AF is more accurate and that OVF PDAF is slightly less reliable ........ but averaging out accumulated errors from two AF systems is just adding errors not eliminating them!!

..... and if you have a range of lenses to calibrate, you can be there for hours as the whole process is based on averaging outliers .......

Dave's clichés
Click to expand...
Not quite, averaging statistical errors is not accumulating them.
Click to expand...
In effect you are averaging errors ..... too many spurious results.
Click to expand...
Let's do some very basic statistics: Every measurement has errors. Errors are basically of two kinds, statistical errors, those who follow some probability distribution and systematical errors, for example an offset.

In our case it is save to assume that both ways of af have a statistical error which is more or less a gaussian distribution if you are not too close to infinity or the nearest focus plane possible. If this is the case, the gaussian distribution, it is quite reliable to determine "real" value within a maybe 200 measurements. Let's assume one measurement has 2sigma=5 correction steps. That means, with the probability of roughly 96% your measurement does not be further off the real value more than 5 points of correction.

With 200 measurements 2sigma of the average value is roughly 0.3cs. So with the probability of 96% our value is not further off than 0.3cs. In this case, as our corrections smallest delta is 1cs (correction step), our measurement is better than we can make use of. This is the simplest of all ways, using weights and such we can get a even way better result with the same measurements done.

Averaging errors, when in an gaussian distribution, means eliminating them.

We of course assume, that there is no systematical error to the lv af, otherwise its useless.
DAVID MANZE said:
DAVID MANZE said:
Averaging systematical errors is. This is why you look at the standard derivation and how it changes over the measurements and in the end you will get a better result.

The whole procedure of course assumes that the LV AF is without significant errors.

I don't think this will take hours. With modern cameras you can shoot 100 pictures in 10 seconds.
Click to expand...
The process is a slow and clunky one requiring tripods and targets, many button presses as well as waiting for the result, which "in theory was a definitive AFFT" ..... (at least 20 shots to get this average) ..... in reality it turned out not to be ..... and this meant this averaging workaround ..... pen and paper ...... and then the time spent taking a few test shots .......

...... if you have 15 lenses ....... ugh!
Click to expand...
 
...as the rate of derogatory personal attacks increased
 
Status
Not open for further replies.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top