Doxa750
Veteran Member
I forgot to mentioned that his friend is now looking at OM1.2 as well as we were all out shooting early this morning. He is using R5 and 100-500mm lens.
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Continuation: OM-1: Imprecision of AF Target?
- Thread starter EZGritz
- Start date
David5833
Veteran Member
Do you often shoot small critters among trees and bushes, like the real-world example with the woodpecker on a distant tree that Mait posted in the previous thread? Your "tests" aren't at all relevant to that situation. Nor are they relevant to macro situations where imprecise focus points can be an issue.
Many times you just can't move around to get a clear shot and a good composition of a bird on a tree trunk or among branches and other trunks, and in those cases you might reasonably expect that the camera will focus on what is in the focus box.
tomhongkong
Veteran Member
Exactly. Especially if you are using a long lens and stretching things so far as light is concerned.
If EZGritz could show us how to get reliable focus for birds in branches in these conditions I would be very grateful
tom
Shooting with OM1 mk i and 300/4 in a low light interior, smallest spot target in centre of screen, SAF and CAF. Both firmwares up to date. Focus lock confirmed - AF priority. Not using tracking or subject detection.
Screen grabs from Workspace.
SAF
CAF
Reproducible results across 2 days.
While the CAF image is sharply focussed on the card, it's less clear where the SAF image focussed - perils of smallest target!
SAF image processed in C1 with default settings
In general, my experience is that CAF will focus in lower light than SAF but SAF is more accurate. I've not seen this in general shooting and it's clearly a bug. If I moved the target point away from the card about twice as far, CAF would focus on the giraffe reasonably reliably.
I think that matches OP's experience in the previous thread. Not had this as a problem personally.
Andrew
--
Infinite are the arguments of mages. Truth is a jewel with many facets. Ursula K LeGuin
Please feel free to edit any images that I post
Screen grabs from Workspace.
SAF
CAF
Reproducible results across 2 days.
While the CAF image is sharply focussed on the card, it's less clear where the SAF image focussed - perils of smallest target!
SAF image processed in C1 with default settings
In general, my experience is that CAF will focus in lower light than SAF but SAF is more accurate. I've not seen this in general shooting and it's clearly a bug. If I moved the target point away from the card about twice as far, CAF would focus on the giraffe reasonably reliably.
I think that matches OP's experience in the previous thread. Not had this as a problem personally.
Andrew
--
Infinite are the arguments of mages. Truth is a jewel with many facets. Ursula K LeGuin
Please feel free to edit any images that I post
Last edited:
Neilatolds
Forum Enthusiast
- Messages
- 253
- Reaction score
- 1,282
I replied to the previous post and will again. Studio setups trying to make the camera fail are irrelavent, I care about in the field. To address the woodpecker senario we need to do a little math. I captured this image of a hard to get bird in Argentina, sharp and lots of detail.
Yellow-billed Tit-Tyrant Argentina
However this is a crop from this image
I cropped this tightly in photoshop and measured the size of the bird.
What this means is that this bird is actually hitting less than 4% of the pixels on the sensor yet the focus was instantaneous and acurate. I was using bird detection with all points active. The woodpecker example shown on a previous post was even further so probably 1% of the pixels. Surprise it doesn't work well. If you want good pictures you do have to move to a better location or accept the results.
As for how to get a good image in a busy background you need to use single point and keep it as far from the distractions as you can and it does work. In this case on the head of the bird. Single point with bird detection on.
Marico Sunbird Kenya
And how does it do in low light. This image is ISO 12800 late in the day. Autofocus is still very acurate and most importantly very fast which is essential for hummingbirds. Again bird detection on all points active.
Snowy bellied hummingbird Panama
Finally I will say I use the OM1 with the 300F4 ususally with the 1.4TC, a killer combination. This combination has helped me get over 250 bird species in Kenya, 250 birds species in Panama and 220 bird species in Argentina and we only count birds we have an image of. If I miss a shot I do not blame the camera.
Yellow-billed Tit-Tyrant Argentina
However this is a crop from this image
I cropped this tightly in photoshop and measured the size of the bird.
What this means is that this bird is actually hitting less than 4% of the pixels on the sensor yet the focus was instantaneous and acurate. I was using bird detection with all points active. The woodpecker example shown on a previous post was even further so probably 1% of the pixels. Surprise it doesn't work well. If you want good pictures you do have to move to a better location or accept the results.
As for how to get a good image in a busy background you need to use single point and keep it as far from the distractions as you can and it does work. In this case on the head of the bird. Single point with bird detection on.
Marico Sunbird Kenya
And how does it do in low light. This image is ISO 12800 late in the day. Autofocus is still very acurate and most importantly very fast which is essential for hummingbirds. Again bird detection on all points active.
Snowy bellied hummingbird Panama
Finally I will say I use the OM1 with the 300F4 ususally with the 1.4TC, a killer combination. This combination has helped me get over 250 bird species in Kenya, 250 birds species in Panama and 220 bird species in Argentina and we only count birds we have an image of. If I miss a shot I do not blame the camera.
Yes. So that raises the questions if focused on the wrong part or if it did not achieve focus at all? And is that due to a bug, due to low light, or some other factor, e.g. lack of contrast in the small AF Target area?
This suggests that C-AF uses a larger AF target than S-AF, but that seems unlikely.
I really wonder what would have happened if you had used two identical objects, in identical light (e.g. two chess pieces). That way, the AF system should have no preference for the brightness or contrast of one over the other.
I think so too, but it still leaves some questions unanswered. Thanks for running this test.
tomhongkong
Veteran Member
Those are nice photos, but do not have too many distractions round the subject. That is where the problem starts with birds in bushesI replied to the previous post and will again. Studio setups trying to make the camera fail are irrelavent, I care about in the field. To address the woodpecker senario we need to do a little math. I captured this image of a hard to get bird in Argentina, sharp and lots of detail.
Yellow-billed Tit-Tyrant Argentina
However this is a crop from this image
I cropped this tightly in photoshop and measured the size of the bird.
What this means is that this bird is actually hitting less than 4% of the pixels on the sensor yet the focus was instantaneous and acurate. I was using bird detection with all points active. The woodpecker example shown on a previous post was even further so probably 1% of the pixels. Surprise it doesn't work well. If you want good pictures you do have to move to a better location or accept the results.
As for how to get a good image in a busy background you need to use single point and keep it as far from the distractions as you can and it does work. In this case on the head of the bird. Single point with bird detection on.
Marico Sunbird Kenya
And how does it do in low light. This image is ISO 12800 late in the day. Autofocus is still very acurate and most importantly very fast which is essential for hummingbirds. Again bird detection on all points active.
Snowy bellied hummingbird Panama
Finally I will say I use the OM1 with the 300F4 ususally with the 1.4TC, a killer combination. This combination has helped me get over 250 bird species in Kenya, 250 birds species in Panama and 220 bird species in Argentina and we only count birds we have an image of. If I miss a shot I do not blame the camera.
I agree that staged photos with Giraffes or bottles do not tell the whole picture but may give clues as to what is happening. I can reliably reproduce those results.
It's unfortunate that OMS (or previously Olympus) feels unable to support its customers by providing a clear description of how they mean AF to work.
Perhaps they don't know?
tom
PonyOfHope
Active member
Dear friends,i see that some of us here are die-hard olympus fans and try to convince others that they are imagining there is a weird behavior of the af in some situations.
Others show that indeed there is a focus inaccuracy by posting photos that prove this behavior.
Well imo the camera should focus in the given box or not focus at all if it can't"see" the thing in the focus box a person tries to photograph (i am not talking about subject detection btw, just focus box, 1x1). If it behaved like that people would be more pleased instead the camera picking something close to the focus box.
What's the point of the focus box if the camera ignores it ? Why is there a single area focus box at all if the camera grabs focus to objects near the box ? C'mon,if i want the camera to grab focus something near the tiny subject i would use larger box,wouldn't you? If someone uses the tiniest "box", he/she/it probably has got a reason to use it......hmmm....maybe he/she/it needs a pinpoint accuracy in this situation ?
Others show that indeed there is a focus inaccuracy by posting photos that prove this behavior.
Well imo the camera should focus in the given box or not focus at all if it can't"see" the thing in the focus box a person tries to photograph (i am not talking about subject detection btw, just focus box, 1x1). If it behaved like that people would be more pleased instead the camera picking something close to the focus box.
What's the point of the focus box if the camera ignores it ? Why is there a single area focus box at all if the camera grabs focus to objects near the box ? C'mon,if i want the camera to grab focus something near the tiny subject i would use larger box,wouldn't you? If someone uses the tiniest "box", he/she/it probably has got a reason to use it......hmmm....maybe he/she/it needs a pinpoint accuracy in this situation ?
MEDISN
Veteran Member
Focus point "reaching" is an issue on every MILC I have tested. Some are worse than others but it happens.
IsaacFocus point "reaching" is an issue on every MILC I have tested. Some are worse than others but it happens.
I’ve never actually had a problem with it in use. It’s reassuring that SAF doesn’t do it any more, as it did in EM1.n bodies. The AF system on the OM1 is very different from the OM5.
I only posted because there seemed to be disbelief with the original thread.
Andrew
--
Infinite are the arguments of mages. Truth is a jewel with many facets. Ursula K LeGuin
Please feel free to edit any images that I post
HicHic
Veteran Member
- Messages
- 2,662
- Reaction score
- 3,554
That is a very cute bearI rarely find focus problems with a distinctive image with detail on the target.
I do find that the OM1.1 sometimes indicates that the camera is focused on the target when it is actually focused on something else. I know when that is likely to happen and modify my approach to minimize the problem.
A good example where this is likely to be a problem is photographing Black Bear (small focus point, Subject AI). They are dark and have little detail. In addition, Black Bear do not like high outside temperatures, so they tend to be out in the late afternoon/evening, rarely in good light, so illumination will be low and you will tend to have a high ISO.
If the bear is stationary, it is easy to simply check focus before exposure. However, if the bear is moving, this is far more difficult to confirm when the EVF indicates focus on the target. My solution is to shoot short bursts unless the head/eye is clearly visible. While it is difficult to see if the bear is clearly in focus, I can often see if the background is in focus and stop the burst. If the background is in focus, I stop and shoot another short burst. If the bear turns so that the head/eye is turned toward me, then I simply continue the burst because I find that the camera is likely to correct focus.
The first image (#1) below is the fifth (last) image in a short burst, all images where focus but the eye is not visible, so I stopped and shot a new short burst. The next image (#2 next burst) is out of focus and all 4 images in the burst are out of focus even though the camera indicated the bear was in focus. However, the background is in focus, so I stopped the burst.
The next burst starts out of focus, but the camera starts to correct focus as the bear turns its head toward me on the fourth image of the burst (image #3 below), so I continue the burst since the bears head is now facing me. By the fifth image (#4 below) the camera has returned focus to the bear and all images after this image were in focus until I stopped the burst after 34 images.
All cameras sometimes fail to provide the image you want, you need do understand the situations where potential failures are likely and adapt your techniques to maximize probability of success.
#1 end of burst with in focus bear
#2 first image of new burst with all four images of burst out of focus
#3 fourth image of new burst as bear turns toward camera and camera begins focus correction
#4 fifth image of burst with all images through 34th image of burst in focus
MEDISN
Veteran Member
Yeah it never affected me in the field either. I liked your example. I did not read the original thread but this issue has come up before. Since then I have tested this on a dozen cameras and they all "reach" beyond what the focus box appears to cover.IsaacFocus point "reaching" is an issue on every MILC I have tested. Some are worse than others but it happens.
I’ve never actually had a problem with it in use. It’s reassuring that SAF doesn’t do it any more, as it did in EM1.n bodies. The AF system on the OM1 is very different from the OM5.
I only posted because there seemed to be disbelief with the original thread.
Andrew
gary0319
Veteran Member
Olympus/ OM has two options for releasing the shutter. 1) release the shutter only when focus is achieved, and 2) release the shutter, even if focus in not achieved. It's at the desecration of the user.
Last edited:
Neilatolds
Forum Enthusiast
- Messages
- 253
- Reaction score
- 1,282
This is where opinions of distrations is different. There is a branch right in front of the marico sunbird. In fact for me this is an ID only shot and I did move and got a clear shot. I will not show or post a shot where the bird is covered by branches as for me it is not good enough.
As you do not think this is enough distractions how about this one. We walked all around this tree trying to get a clear shot of these owls and this is the best we could do.
and the crop.
This is using the 300F4 with the MC20 from a great distance. Can't use subject detection in this senario but single point does what I want it to.
To be clear the camera is not perfect and I can make it misfocus if I want to but I am more interested in making it work right not fail. If I miss a shot it is almost always human error and not the camera.
As you do not think this is enough distractions how about this one. We walked all around this tree trying to get a clear shot of these owls and this is the best we could do.
and the crop.
This is using the 300F4 with the MC20 from a great distance. Can't use subject detection in this senario but single point does what I want it to.
To be clear the camera is not perfect and I can make it misfocus if I want to but I am more interested in making it work right not fail. If I miss a shot it is almost always human error and not the camera.
PonyOfHope
Active member
Of course it has those options,but this has nothing to do with focusing outside the box,those who shared examples used "1) release the shutter only when focus is achieved"
I got nothing against OM System,just funny to see how some dear users here try to strongly persuade others that the camera doesn't behave like examples show,yet it does grab focus outside the tiny box sometimes as some users have shown clearly,with good examples.
Now,does it mean it is a bad camera,well,hell no. It is just one thing you need to be aware of,that's it.
EZGritz
Veteran Member
When it does this it thinks it is focused, just not where you ask it to. It can do that. All cameras can. None of them are perfect.
EZGritz
Veteran Member
On single point focus I don't see my OMS or OLY cameras doing this on SAF or CAF. If the subject is moving around you can accidentley fire the shutter after it moves in which case the camera will quickly focus somewhere else.
With subject detect, the OM-1 can lose the subject when it's moving around and you can fire the shutter before it refocuses. Most of the time it focuses on the subject inside the large box. It can miss the eye with active animals like dogs but still focus in the larger box. Close enough to call it in focus if the DOF is large enough. Not always critical focus where you want it. Sometimes the DOF is too shallow where you place the focus point or where subject detect is trying to find the eye in which case it can choose something in front or behind the subject. That's been my experience. If I try too shallow a DOF for the camera to focus where I place the focus point and the subject moves that point out of the focal plane, if I'm trying to use ISO 200 for example to make the cleanest image where I want it but the DOF is too shallow it will show the focus point where I want it but the focus somewhere else. It might be close enough for the image to be in general focus but not critical focus.
--
Author of "The Pelican Squadron" - Harvey Gene Sherman
https://www.amazon.com/Pelican-Squadron-Tale-Internet-Bubble-ebook/dp/B08FCY6V7Y
"A great photo can have mediocre IQ, but a mediocre photo cannot be saved by great IQ"
With subject detect, the OM-1 can lose the subject when it's moving around and you can fire the shutter before it refocuses. Most of the time it focuses on the subject inside the large box. It can miss the eye with active animals like dogs but still focus in the larger box. Close enough to call it in focus if the DOF is large enough. Not always critical focus where you want it. Sometimes the DOF is too shallow where you place the focus point or where subject detect is trying to find the eye in which case it can choose something in front or behind the subject. That's been my experience. If I try too shallow a DOF for the camera to focus where I place the focus point and the subject moves that point out of the focal plane, if I'm trying to use ISO 200 for example to make the cleanest image where I want it but the DOF is too shallow it will show the focus point where I want it but the focus somewhere else. It might be close enough for the image to be in general focus but not critical focus.
--
Author of "The Pelican Squadron" - Harvey Gene Sherman
https://www.amazon.com/Pelican-Squadron-Tale-Internet-Bubble-ebook/dp/B08FCY6V7Y
"A great photo can have mediocre IQ, but a mediocre photo cannot be saved by great IQ"
Last edited:
EZGritz
Veteran Member
Thomas Eisel says CAF is faster but SAF is more accurate. For moving subjects I do better with CAF because it focuses before the moving subject moves out of focus. I have no problem with CAF or SAF and static subjects in any light. The illuminator seems to work well when the camera needs it. I leave it set on automatic.
EZGritz
Veteran Member
I wonder if it has something to do with DOF?
EZGritz
Veteran Member
It can miss and focus on a branch. I can usually see when it's focusing in the wrong place. I wait until I can see the bird is in focus. Sometimes I move the camera around. Sometimes the bird moves and identifies itself as the subject. Sometimes I recompose until it focuses on the bird. It isn't perfect and neither am I. I do the same with older bodies but the OM-1 is much more accurate and easier to find focus where I want it most of the time if not all the time.
It will follow a small bird camaflaged to the background hoping around in low light but not all the time. Sometimes I have to recompose. It doesn't prevent me from getting the photo but it isn't much of a photo because the bird looks so much like the leaves it's hoping on it's hard to see the bird in the image after taking it. If I can barely see the bird it isn't much of a photo and I don't care if the AF system can't identify it either.
It will follow a small bird camaflaged to the background hoping around in low light but not all the time. Sometimes I have to recompose. It doesn't prevent me from getting the photo but it isn't much of a photo because the bird looks so much like the leaves it's hoping on it's hard to see the bird in the image after taking it. If I can barely see the bird it isn't much of a photo and I don't care if the AF system can't identify it either.
Keyboard shortcuts
- f
- Forum