LumoLabs: Pentax K-5 low light focus study is published

[awaldram]

I'm curious, do you think these are both of the same sharpness?

I'm with you on this remark, for me the first one is slightly blurred, but with small pics and high-iso, it's hard to tell...
That's why I advised to use flash and a low iso...

I did that originally and I think it was Walt called me a fool (and worse) for using flash .

Seems I've covered every angle and cannot produce this problem that is so easy to see.

Heres the original flash image taken in around 1 ev , Check the date for my initial Focus test in low EV.

Also, 1000 iso, 1/8, f/2.8 is around EV2.6, and from personal tests, at this level, a contrasted target could still make it... But a more reddish, low contrast one (a face) will have a hard time...

--
My PPG

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=1471087&subSubSection=0&language=EN
My Photo Stream
http://www.flickr.com/photos/awaldram/

 

[awaldram]

I mean 25% inaccuracy should really be easy to spot shouldn't it?

Well you never were one to just go along with everyone else. Had to be different, huh? Had to just go out and have it work right…

Hi Lawson

Long time no hear, how goes it, what you been upto since your 365 finished.?

Yep I just can't get a handle on this one at all, The other P-ttl bugs etc I can reproduce and have reported a P-ttl issue with timer use and wireless (it doesn't ptt at all) to Pentax.

But here I draw a complete blank, though I can create focus shift by using single channel light but I'm so far out of 'normal' use I discount that.

Guess I'll just have to be in Rolands gang

--
----------
Lawson G. Stone
I don't make history, just memories.
http://www.flickr.com/photos/lawsonstone/

--
My PPG

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=1471087&subSubSection=0&language=EN
My Photo Stream
http://www.flickr.com/photos/awaldram/

 

[Ian Stuart Forsyth]

Now You say the exact opposite well as I've covered every aspect (I have some at 400mm if you want) I think I'll drop this subject until it can be decided just what the problem is.

Needless to say with the different views and reports as to what cause the issue and how to reproduce it I'm having a hard time believing there's any systematic problem at all.

--
My PPG

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=1471087&subSubSection=0&language=EN
My Photo Stream
http://www.flickr.com/photos/awaldram/

I have yet to have this problem with the k5 from 15mm- 100mm The only trouble I have seen is that many of my lenses have a shift in the central point of focus as I adjust the F value but I have always seen this sense my 1st slr so it’s not a big surprise

--
A Camera is only a tool, photography is deciding how to use it.
http://IanForsyth.Photography.com

 

[Roland Karlsson]

No I think the phase detect it fractionally sharper, For some reason the Contrast detect is slightly under exposed increasing the contrast level which may fool some.

Hmmmm ... must be some mistake here. The second one is the contrast detect. And it is much sharper ... and somewhat darker.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

 

[Walt_A]

I'm curious, do you think these are both of the same sharpness?

No I think the phase detect it fractionally sharper, For some reason the Contrast detect is slightly under exposed increasing the contrast level which may fool some.

Why what do you think .

That the cdaf version is sharper.

--
Walt

http://picasaweb.google.com/waltdall

--
My PPG

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=1471087&subSubSection=0&language=EN
My Photo Stream
http://www.flickr.com/photos/awaldram/

--
Walt

http://picasaweb.google.com/waltdall

 

[Walt_A]

I'm curious, do you think these are both of the same sharpness?

I'm with you on this remark, for me the first one is slightly blurred, but with small pics and high-iso, it's hard to tell...
That's why I advised to use flash and a low iso...

I did that originally and I think it was Walt called me a fool (and worse) for using flash .

Not me.

--
Walt

http://picasaweb.google.com/waltdall

 

[awaldram]

No I think the phase detect it fractionally sharper, For some reason the Contrast detect is slightly under exposed increasing the contrast level which may fool some.

Hmmmm ... must be some mistake here. The second one is the contrast detect. And it is much sharper ... and somewhat darker.

Well if it under exposed it will be darker.?

Pity I only uploaded straight from camera to gallery so I don't have originals to pixel peek.

But I downloaded the gallery version equalised contrast and exposure and exploded the word set.

Someone in my house must have switched a landing light on the CD image has a shadow slanting from the right.

Overall I think your right the CD might just have the edge but given the slack way I took these and the low shutter speed I'm happy the camera is not at fault.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

--
My PPG

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=1471087&subSubSection=0&language=EN
My Photo Stream
http://www.flickr.com/photos/awaldram/

 

[awaldram]

I'm curious, do you think these are both of the same sharpness?

I'm with you on this remark, for me the first one is slightly blurred, but with small pics and high-iso, it's hard to tell...
That's why I advised to use flash and a low iso...

I did that originally and I think it was Walt called me a fool (and worse) for using flash .

Not me.

Oops sorry.

--
Walt

http://picasaweb.google.com/waltdall

--
My PPG

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=1471087&subSubSection=0&language=EN
My Photo Stream
http://www.flickr.com/photos/awaldram/

 

[falconeyes]

The paper contains an example. IIRC, it says something like 25cm on 1m distance at 31mm f/1.8.

IIRC?? you wrote it yourself....
25cm on a 1 m distance???? That is a lot... sure it is not 25mm????

I think your response isn't polite.
The fact I wrote it doesn't mean I know every sentence by heart.

The focus shift in subject space for a focal length of 31mm and 250µm defocus is easily computed from the thin lens formula (using a pocket calculator).
http://en.wikipedia.org/wiki/Thin_lens

 

[Priscilla Turner]

I am a complete amateur in both mathematics and engineering; but FWIW I have just tried autofocussing my K-x on objects across an incandescent-lit room here at home, using the Pentax DA 18-250 at the long end. The autofocus fixed itself on a point somewhere between 8 and 20 ft. I focussed both with and without an 80A over the lens. Aperture was not altered, the whole rig-out was kept completely stationary. Focus shifted audibly and visibly to a point further off with the 80A on.

Changing WB makes not an atom of difference, but the actual colour temperature makes plenty, at least with my gear.

From a practical point of view, for good neutral well-focussed people-pictures with flash I think that it is clear that with my gear one must cut out the excess yellow-red and set WB on something that produces white whites. This means either a test shot with a gray card or tweaking in post.

--
'To see, not with, but through the eye.' [William Blake]

http://www.flickr.com/photos/22905474@N06/

 

[opiecat]

Imagine the camera Pentax would have if it had Nikons AF, or even the 7D.

Funny d7k AF is even worse

really, oleg?

do you have a d7k or are you relying on internet chatter?

i happen to have one and after using it for 2 months, i can say that the AF is miles better and more consistent than pentax's AF. perfect? no, but more consistent and accurate, a big yes.

--
'when 900 years you reach, look as good you will not'
-- master yoda

http://jordanpaw.zenfolio.com

 

[awaldram]

Imagine the camera Pentax would have if it had Nikons AF, or even the 7D.

Funny d7k AF is even worse

really, oleg?

do you have a d7k or are you relying on internet chatter?

i happen to have one and after using it for 2 months, i can say that the AF is miles better and more consistent than pentax's AF. perfect? no, but more consistent and accurate, a big yes.

You seem to have broadened the statement (pentax's AF) , The allegation was the K5's AF is better.

All reports and tests indicate that the statement is probably true the d7k is yet to outperform the k5 AF wise in any test I've read be it af-s or af-c.

As a a fact your right the d7k has better AF than Pentax's but but Nikons AF is better the the d7k .

I think we need to stick to models here not manufacturers and not mix the two.

--
'when 900 years you reach, look as good you will not'
-- master yoda

http://jordanpaw.zenfolio.com

--
My PPG

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=1471087&subSubSection=0&language=EN
My Photo Stream
http://www.flickr.com/photos/awaldram/

 

[awaldram]

Imagine the camera Pentax would have if it had Nikons AF, or even the 7D.

Funny d7k AF is even worse

really, oleg?

do you have a d7k or are you relying on internet chatter?

i happen to have one and after using it for 2 months, i can say that the AF is miles better and more consistent than pentax's AF. perfect? no, but more consistent and accurate, a big yes.

You seem to have broadened the statement (pentax's AF) , The allegation was the K5's AF is better.

All reports and tests indicate that the statement is probably true the d7k is yet to outperform the k5 AF wise in any test I've read be it af-s or af-c.

As a a fact your right the d7k has better AF than Pentax's but but Nikons AF is better the the d7k .

I think we need to stick to models here not manufacturers and not mix the two.

Oops it was me read it wrong it was already at manufacturer level so ignore all I wrote

--
'when 900 years you reach, look as good you will not'
-- master yoda

http://jordanpaw.zenfolio.com

--
My PPG

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=1471087&subSubSection=0&language=EN
My Photo Stream
http://www.flickr.com/photos/awaldram/

--
My PPG

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=1471087&subSubSection=0&language=EN
My Photo Stream
http://www.flickr.com/photos/awaldram/

 

[brandrx]

Guess I'll just have to be in Rolands gang

Yep! Me too. :-D

Cheers.

Ron

--
Ron - 'We don't have time to go take pics this afternoon Carl.'
Carl - 'What do you mean? It will only take 1/1000s.'

'Keep your eyes looking forward. However, glance back now and then to see where you've come from. It will put a smile on your face.' ~ brandrx

 

[dlacoutu]

Just so my findings are not only for the other forum, here is what I found tonight:

Following Falk's suggestion, I've made some tests involving shining a flashlight into the viewfinder while focusing...

And guess what : using a tungsten flashlight into the viewfinder, my camera focused properly!!! (note that a "daylight" LED flashlight had no effect upon focus).

Test case:

• find a target/light level where you get consistent FF
• M mode, 400 iso, flash on (to be able to better judge focus)
• do pics without the flashlight, pics with it...

I've even found another test, faster and more "obvious"...

"Dynamic" test:

• AF-C (the behavior will be obvious!!!)
• while half-pressing the shutter, intermittently shine the flashlight into the viewfinder.

=> you'll see the camera adjust the focus for tungsten light...

So, under a given light level threshold, the camera will revert to "daylight focus", and not properly compensate for tungsten lighting anymore...

I guess the reports of misfocuses in "low daylight" situations are actually based upon the false assumption that when the sun just went down, the WB is still "daylight", whereas in reality there is an heavy shift toward blue...

 

[Roland Karlsson]

Just so my findings are not only for the other forum, here is what I found tonight:
...

Very interesting findings! Clever idea to shine in the view finder. That shall of course work. At least until you too low contrast on the AF detector.

Now - I am not so sure about your interpretation. I have a K-7 and I have seen no problems with regard to color of the light/object - just problems with the light level.

But - that does not matter. If the camera measures faulty at low light level - then the increased light level will of course cure it.

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

 

[dlacoutu]

Very interesting findings! Clever idea to shine in the view finder. That shall of course work. At least until you too low contrast on the AF detector.

Well, that was Falk's idea, not mine!

Another dumb test : in my "100% FF" scene, I added one small tungsten light source at the edge of the scene, taking care that it did not shine a light upon the AF target.

So, theoretically, it should have no impact whatsoever on focusing, except by giving the colorimetric sensor a good reading...

Result : focus is again spot-on...

Doing the same with a LED light had no effect...

 

[Roland Karlsson]

Very interesting findings! Clever idea to shine in the view finder. That shall of course work. At least until you too low contrast on the AF detector.

Well, that was Falk's idea, not mine!

Another dumb test : in my "100% FF" scene, I added one small tungsten light source at the edge of the scene, taking care that it did not shine a light upon the AF target.

So, theoretically, it should have no impact whatsoever on focusing, except by giving the colorimetric sensor a good reading...

Result : focus is again spot-on...

Doing the same with a LED light had no effect...

How big is the colorimetric sensor?

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

 

[Roland Karlsson]

--
Roland

support http://www.openraw.org/
(Sleeping - so the need to support it is even higher)

X3F tools : http://www.proxel.se/x3f.html

 

[moving_comfort]

.

FYI, 1.03 tests are in.

----> http://falklumo.blogspot.com/2011/03/lumolabs-pentax-k-5-low-light-focus.html

.

Summary:

Basically, I find no significant difference between firmware versions 1.02 and 1.03 when using my testing scenarios. This means that there is no significant improvement if:

The background is white
The target has high contrast
The AF assist light is disabled

Under the above circumstances, the K-5 auto focus starts to lock focus in a false front focus position at tungsten light levels below 0 EV for a fast lens (like f/1.8) or below about 2-4 EV for slower lenses. This did not change at all with release 1.03. Note that the above light level values would read 2-3 steps higher with a target like a Caucasian skin (18% gray).

Moreover, I found that at low enough light levels, the 1.03 firmware seems to rely more than 1.02 on color information outside the direct selected AF spot. With a uniform color distribution, this can lead to somewhat improved results in the vicinity of the focus shift transition. OTOH, the focus system is more easily fooled by features with a singular color. Overall, the advantages may dominate and the effect is small anyway. Version 1.03 still doesn't seem to make use of white balance information, either manually or automatic.

If I believe that Pentax improved the low light focus situation for many users of the 1.03 firmware, then the progress must be bound to any of the following:

AF assist light engages more actively (not tested by myself).
Focus improved in the presence of dark backgrounds (not tested by myself).
Focus improved with low contrast or dim focus features (not tested by myself).

As I don't have a version 1.02 K-5 anymore, I cannot find it out. But any progress brought by the 1.03 firmware upgrade is limited rather than universal.

As much as I applaud Pentax to having addressed the problem, as much I am a bit disappointed they didn't dig deep enough to address the root problem: that the K-5 AF subsystem locks focus even in a situation where the colorimetric sensor(s) fails to determine the light color. It simply shouldn't lock focus at all then. Or ask for user assistance like a priming shot. The light-sensitive AF system of the K-5 has too strong a color dependency to autofocus if the color of the focus feature remains unknown.

A better workaround than 1.03 in firmware is feasible and should stay on Pentax' agenda.

Please, read the full paper linked above if you have further questions and come back here to leave comments or questions. General comments about the issue should still go to the general (earlier) article while firmware release 1.03 comments should go here. Thank You.
POSTED BY FALK LUMO AT 03:53