Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
06-29-2009, 01:06 PM   #16
Pentaxian




Join Date: Jan 2009
Location: Nevada, USA
Posts: 486
Atomic # 70:

A couple of thoughts:

1.
QuoteQuote:
Manual focus is too hard for my eyes and that damn screen. I cannot even see in viewfinder if the focus was off.
Have you adjusted the eyepiece diopter? Just a thought on the off-chance you haven't checked this already. The degree to which your intended target is out of focus on some of your example shots should easily be seen in the viewfinder, I think. So when the green focus indicator lights up, you should be able to tell if what you want to be in focus is at least close to being in focus.

At least that's how it works for me even with less than perfect eyesight (eyeglass lens 1.2cm thick at edge) and my K200D.

2.
I did an exercise that helped me understand how autofocus behaves when I first got my K200D. View a low contrast object with a small high contrast mark in the middle, like a blank white wall with a black push pin in the middle. Select center focus point and aim away from the center and half press to initiate auto focus. If the object is low enough contrast, it won't be able to find focus. Now, gradually aim closer to the black mark, until autofocus confirms. This gives you the angle of view of the auto focus center. Do this approaching from horizontal and vertical.

I found the effective area of the auto focus sensor was larger - a lot larger - than I thought it would be. The point being, autofocus may be detecting a nearby high contrast feature in your shot, even if that feature is not directly in the center, where the red auto focus selection light blinks.

06-29-2009, 01:21 PM   #17
Veteran Member
ytterbium's Avatar

Join Date: Jan 2008
Posts: 1,076
Original Poster
Im already aware of much larger AF area and know its approximate size.

Btw, i uploaded some more quatative testing. It seems that when you get closer it nails the focus but at certain distance (approx 2m) it starts to get worse.

Those with batteries: Inbox Foto (foto.inbox.lv)

Basically i have no idea what should i do.
06-29-2009, 03:24 PM   #18
Pentaxian
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,686
QuoteOriginally posted by ytterbium Quote
The same sample with bowl:
Inbox Foto (foto.inbox.lv)
Why it seems for you so reasonable and logical?
There is no way a basic front focus issue with the camera would cause a discrepancy as you are showing here. This absolutely has to be something else - either you didn't hold focus locked, or were in AF-C mode, or else this is a crop from a larger shot and the cereal bowl was in fact in range of the AF sensor. Or maybe something else. Without having been there, it's impossible to say which, but again, there is *no way* this is a run of the mill FF problem. Those would be manifested in centimeters, not meters.

You are going to a lot of trouble to set up tests that simply cannot be verified or guaranteed to test what you think they are testing. People create focus test chart and provide good instructions on how to use them for a reason - because it is very easy to mess up a focus test by trying to do it yourself. Put the cereal and batteries and away, wait for the ca to get out of the way, then get a tripod, print a test chart, and follow the instructions carefully. Until then, you aren't performing a focus test at all - you are simply posting pictures in which something went wrong, but we have *no idea* what.

QuoteQuote:
Would you consider it to be logical, if camera focussed on the near end of the guitar in your shot
Well, yes, that would have been within range of the one of AF sensor had I simply pointed and shot. As it is, I was in center point mode, and no, that would not have been in range had I simply pointed in shot. But I didn't do that - I used focus recompose. With the camera pointed at her head, the far of the guitar was not in range, But her elbow was, and so was the microphone, and so was the bassist in the background. To be honest, I don't remember if the camera focused as I expected or not. This was the DA70, which has quick shift, and I use it often when I see in the viewfinder that the camera is not focusing where I want it.

QuoteQuote:
here:
Inbox Foto (foto.inbox.lv)
Where there even isnt anything else in front. The necklace has good contrast, and the dof isn't that shallow for cat to be completely out of focus even if it moved 5..10 centimeters.
Of course it is. That was taken as f/1.8. DOF at that distance is indeed that shallow. It' so shallow that even if the cat didn't move, you'd be unlikely to be able to consistently nail focus

This is all exactly the sort of stuff I mean when I say you aren't testing focus here - you are posting random pictures. Do a *real* focus test and post a single image that clearly shows the results.

QuoteQuote:
Btw, i've observed that you mostly use MF and split prism what eliminates most of those problems. But this does not change the fact that AF should be more or less reliable for those who cannot use other options, or else why include it at all and not sell cameras and lenses for 100..300$ less.
I'm not sure if that was directed at me, but I'd say I'm about 50/50 on manual versus auto focus - and I don't use a split prism for my manual focus. Yes, of course the camera should work as it should. And the photographer should learn how to use it properly. So far, we really have no evidence one way or the other as to what is going on here. A real focus test would eliminate all these questions.
06-30-2009, 02:03 PM   #19
Veteran Member
ytterbium's Avatar

Join Date: Jan 2008
Posts: 1,076
Original Poster
Danke schoon. It is so hot in here .
Anyway, i managed to make myself do something like a real test:
Inbox Foto (foto.inbox.lv) (Sigma, S.closer, DA70).
What do you think? Does this counts as a real test?
If i had only the measured FF (2..10mm's) i'd be a very happy man. Now i still get feets in real life situations and larger subject distances.
But this focus chart test makes me feel like testing brakes of a 200 mph capable motorbike while driving at 3 mph, because i don't shoot macro and i don't get problems at close distances.

The rest is just some philosophical BS:
Just to be on the safe side i also did a test with black line. Centre AF point seems to be much smaller than i expected after said experiences. It barely "touches" the circle showing spot metering area and is perfectly centred - with DA70/2.4, Sigma 30/1.4 and DA 18-55.

Sigma's AF improves a bit when getting closer.
Also it seems that DA70 is doing much better than sigma. I did not do a test to check if it is because of actual lens or 30/1.4 having wider fov (so a test with DA further a way is needed).

After most of the shots i threw the lens OOF, both ways (infinity/closest), on some i left it untouched to confirm that the AF doesn't change its measured distance.

Btw. Why do you keep mentioning all those crazy possibilities like multi-point-auto-AF, cropping, continuous AF and similar basic things. Why would i post picture like that or had any thought of problem under such Dynamic/unpredictable configuration.

There is another question i've always had, but noone has answered clearly. Is simple FF/BF a constant real life distance or a constant lens travel distance?. Because 2mm at 30 cm is somehow bad, but it doesn't matter at 2m if it stays so. 10mm lens turn at 45cm is 10mm, 10mm lens turn at 2m is huge 1m error.


Last edited by ytterbium; 06-30-2009 at 02:35 PM.
06-30-2009, 02:18 PM   #20
Veteran Member
wlachan's Avatar

Join Date: Sep 2006
Location: Canada
Photos: Albums
Posts: 2,626
I took a quick look at your shots and my impression is that the 1st few people shots are sooo FF I feel the problem should be something else. As to the battery tests, the setting is just plain silly cos all Pentax DSLRs have wide area AF sensors. That means each point actually covers quite a larger area than you think and anything being hit by the sensor could be chosen by the camera. The diagram below is my rough finding with my own DS (just to give u an idea how large they are). I am aware tungsten AF issue has been mentioned many times and tried to reproduce the same error w/o success. What I have found is that AF relabililty drops dramatically once below 5EV.

06-30-2009, 02:25 PM   #21
Veteran Member
ytterbium's Avatar

Join Date: Jan 2008
Posts: 1,076
Original Poster
You should check the link in my previous post.
Actually i don't think its a tungsten issue. I've observed it under daylight but not so severe so i thought that it MIGHT be tungsten. Came asking here to find out more.

And the batteries at close range was fine. I admit that shots with batteries at greater distances are useless, because of the AF point size. Just did them just in case.
P.S. I like your pbase photos alot.

QuoteQuote:
Have you adjusted the eyepiece diopter?
Yes, BUT! Against what should i adjust it? I get most sharpest view of various elements (AF/metering brackets, status bar, frame, focusing screen grain) at different slider setting. I believe that AF brackets should be the closest thing to image plane since they are on the FS itself.

Last edited by ytterbium; 06-30-2009 at 02:32 PM.
06-30-2009, 09:48 PM   #22
Pentaxian
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,686
QuoteOriginally posted by ytterbium Quote
Anyway, i managed to make myself do something like a real test:
Inbox Foto (foto.inbox.lv) (Sigma, S.closer, DA70).
What do you think? Does this counts as a real test?
It does, although as I suggested earlier, it would really behoove you to post *one* image that you think demonstrates a problem, rather than expect people to browse through dozens of test chart pictures to see if any show problems. I didn't have the patience for that.

I can say the first couple of pictures were taken from too far away - you need to take the shot from close enough to guarantee that the horizontal line below the focus target is not in range of the AF sensor. As it is, it seems entirely likely you've simply reproduced the same error that you seem to have committed in the previous pictures - providing the camera with something *else* to legitimately focus on.

QuoteQuote:
Btw. Why do you keep mentioning all those crazy possibilities like multi-point-auto-AF, cropping, continuous AF and similar basic things.
Not sure if you're talking to me here, but the possibilities I have mentioned are all things that could cause the effects you showed in the first set of pictures. A front focus problem is *not* something that could cause those effects, so it's got to be something else that caused them. Which isn't to say you don't have a FF problem - you certainly might. Or it might be your camera is perfectly normal - focuses well in well-balanced light, but a little FF in tungsten light, just like virtually every other camera with phase-detect AF ever produced. But that has nothing to do with why you got a relatively in-focus truffle box or cereal bowl when you would have preferred an in-focus picture of someone in the background.

QuoteQuote:
There is another question i've always had, but noone has answered clearly. Is simple FF/BF a constant real life distance or a constant lens travel distance?.
I don't know, but I'd actually suspect neither - I'd guess that it is in some way proportional to focus distance, but not in the way that "constant lens travel distance" would suggest.
06-30-2009, 11:39 PM   #23
Veteran Member
ytterbium's Avatar

Join Date: Jan 2008
Posts: 1,076
Original Poster
You could pick any. They all show the offset. I just upped them all so no one would think that one picture doesn't say anything and it could be just random error. Consistency check.

One picture:
http://foto.inbox.lv/tamrac/testchart/IMGP9436.jpg

But as you said its not the problem, even more if the FF doesn't get much bigger with distance. Since i could clearly determine the AF point size i doubt the camera somehow sees those front subjects or picks other focus points by error (even if prohibited by focus mode).

As for the distance. I was near the limit of Sigmas minimum focusing distance and wanted to leave some 5 cm in front, to the AF travel in both directions relative to subject. Here i illustrate that center AF point is fully within the frame, even with exaggerated point size, compared to actually measured (within spot meter):

Attached Images
View Picture EXIF
PENTAX K200D  Photo 

Last edited by ytterbium; 07-01-2009 at 12:09 AM.
07-01-2009, 08:35 AM   #24
Senior Member




Join Date: Dec 2008
Location: Dawesville,Western Australia.
Photos: Gallery
Posts: 129
I have a k200d and about ten lenses..nothing special..3 cheap DA"s and some old manuals(2 Chinons ). Last week ,in the evening I decided to test them all for focus.Placed a chair in the bedroom, drew a "stick-man" on a piece of paper 1/6th the size of A4. I then placed the paper about 3 metres away on a cupboard. The only lights on were the overhead tungsten and a small sidelamp.
OK.....they all focused immediately without hunting and they were all in focus. I used AF-S and spot focus in AF. I was extremely pleased and rewarded myself with a rare cigarette...... Best of Luck....
07-01-2009, 09:08 AM   #25
Loyal Site Supporter
Loyal Site Supporter
Canada_Rockies's Avatar

Join Date: Nov 2007
Location: Sparwood, BC, Canada
Photos: Gallery | Albums
Posts: 9,554
QuoteOriginally posted by ytterbium Quote
You should check the link in my previous post.
Actually i don't think its a tungsten issue. I've observed it under daylight but not so severe so i thought that it MIGHT be tungsten. Came asking here to find out more.

And the batteries at close range was fine. I admit that shots with batteries at greater distances are useless, because of the AF point size. Just did them just in case.
P.S. I like your pbase photos alot.


Yes, BUT! Against what should i adjust it? I get most sharpest view of various elements (AF/metering brackets, status bar, frame, focusing screen grain) at different slider setting. I believe that AF brackets should be the closest thing to image plane since they are on the FS itself.
Adjust the diopter setting so that the lines on the focus screen are as clear and sharp as you can get them. I use a clear sky to set mine. I alternate between shooting with spectacles and contact lenses and the diopter setting is slightly different for the two. It is a very quick thing to point the camera at something with no objects - a white wall or the sky. I suggest this method on the grounds that the surface of the focusing screen is where you want the best clarity.
07-01-2009, 12:25 PM   #26
Pentaxian
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,686
QuoteOriginally posted by ytterbium Quote
You could pick any. They all show the offset.
The few I spot checked were inconsistent.

Anyhow, the one you've posted here would indeed appear to show some amount of FF. Comparing that result to the same test shot in more balanced light will give you an idea of how much tungsten light really affects things.
07-01-2009, 01:19 PM   #27
Veteran Member
ytterbium's Avatar

Join Date: Jan 2008
Posts: 1,076
Original Poster
Will do.
Just did a shootout with the DA70, Nikon SB-24 flash in auto mode and continious AF with my cat. Most in rather close distances but noone was missed despite complex obstacles.
Interesting . Focusing was done under low yellow light, but flourescent (so no IR component).
I think the sigma must go. I just cannot find a common language with that lens.

[Update:] Another observation i recall was that when pointing the lens at "empty" scene - wide view, like from a hilltop, with any visible objects at far distances it never focussed to infinity but stopped near the 4m mark instead.

Last edited by ytterbium; 07-02-2009 at 01:47 PM.
07-06-2009, 08:17 AM   #28
Veteran Member
ytterbium's Avatar

Join Date: Jan 2008
Posts: 1,076
Original Poster
As a reply to myself. Seems that those 4 meters i observed when focusing at infinity is something Ken already describes:
Sigma 30mm f/1.4

I've red that site, but some time before, when i had k100d and different issues so this must have slipped by. So essentially this means that it is the design of this lens that might be emphasized because of HSM lack.

Since i like wider fov's and greater distances i should switch to something like DA35/2.8 or FA35/2 and maybe the AF speed will benefit as well.

Btw, during the weekend i had more chance to use DA70 and i love the lens. It is snappy, solid and produces wonderful photos.
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!
af, camera, dslr, light, pentax, photography, times, updates
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
K200D low light/high iso performance indytax Pentax DSLR Discussion 22 02-07-2009 07:49 PM
K200d, Low Light, & the Barndoor Das Boot Pentax DSLR Discussion 0 02-05-2009 10:39 PM
k200d consumer reports sees the light dcmsox2004 Pentax News and Rumors 3 11-30-2008 12:22 AM
K100D vs K200D for low light indoor portraits Pierrot Pentax DSLR Discussion 4 04-02-2008 08:25 AM
AF Accuracy's Dependency on Lenses and Yellow Light RiceHigh Pentax DSLR Discussion 10 02-19-2007 06:51 AM



All times are GMT -7. The time now is 11:46 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