(Re)Learning Autofocus on Olympus E-M1 Mark I and my wildlife photography experience

L0n3Gr3yW0lf

Wall-Eeeeeeeeeeeeee
Joined
Jul 31, 2013
Messages
1,645
Location
UK
Real Name
Ovidiu
This is going to be a very long (and may get sidetracked too) threads to be able to put all my thoughts down on a contained explanation and topic. Going to use lots of pictures too (all of the images have been edited extensively to beat of my skills and I preferred look to them) but going to put them behind spoilers (though data is a lot more generous available this days). Also thank you very much if you can manage to stay with me to the end of this, I really appreciate it.

So I got my Olympus E-M1 Mark I almost a year ago, mainly because I had bought the Olympus 14-54mm Mark I (still have it and it’s my main landscape lens) and the Olympus 5o-2oomm Mark I (my favourite lens ever, well until a month and a half ago, also it died at the begging Of this year). I knew about the limitations of AF of Four Thirds lenses with Micro Four Thirds cameras, but I was quite content with AF-S as I started learning to do Wildlife photography and very limited for Birds. It was quite a refresh from using Manual Focus only with my Olympus E-M5 Mark II I used for a few months before, being able to focus (no pun intended) on composition, action, exposure. I have tried C-AF once in blue moon but the results were pretty disappointing so I just stuck with S-AF. But then my Olympus 5o-2oomm died (mechanical failure) and after a very long time of trying to convince myself I got the Panasonic Leica 5o-2oomm to replace it. And what a beast.

So for the last 2 months I have been relearning and discovering C-AF and Tracking AF on the Olympus E-M1 Mark I. And here is what I learn: I need more learning :p

So here we go (if anyone has tips and advices feel free to drop them and I will do my best to used them and get better at it). First, things that are not related to Continuous Autofocus but that can help get more out of it is finding wildlife, tracking their movement and be as less intrusive as possible.

Finding or spotting wildlife: sounds easy, not so easy to do actually. From thick foliage to ver good camouflage it’s not always easy to spot animals in nature (unless they are really loud, like birds):

B6CF2298-E195-4D12-93EF-3184506CFF8F.jpeg
Subscribe to see EXIF info for this image (if available)
From a distance the chicks were very hard to spot because they blended very well with the tree bark, even making the pictures it was needed some post editing to bring a bit more contrast and separation to see them well.

FC4297F4-DFE0-4DCB-AAD4-38D12D25BE91.jpeg
Subscribe to see EXIF info for this image (if available)
I change upon a wild fox in the same place where I usually observe wild bunnies. Just for a split second I saw in my peripheral vision the foxes head before I stopped and slowly tried to get a better angle to make the picture, this is as close as I physically was able to get.

After finding usual spots for specific types of animals I slow down and pay close attention to movement as well as colour shifts, like the one above. The wildlife menu I have quick access (as in my daily walk to work) to are wild bunnies, (very rare) foxes, (even more rare) deer, a plethoras of small birds as well as crows, buzzers, pigeons. One aspect I am discovering is spotting by sound (kind of duh, I know). I have permanent ear damage from a very old untreated ear infection so I have a permanent (and sometimes mentally overwhelming) ringing in my ear so most of the time I have some kind of sound playing to balance or cover the ringing (music on headphones or YouTube Podcasts or videos playing at home). But I have started to not use my headphones when I finish work and have a day off and stay around the fields and the edge of the town.

12FA49E0-1D96-4626-B184-4402E3ECD4FC.jpeg
Subscribe to see EXIF info for this image (if available)
I would have not been able to notice this guy if not for the loud sounds he was making out there.

Tracking animals on how they move and predict where they might be or heading can be a very though challenge, especially if you don’t know the animals behaviour. Some can be predictive:

219C3F51-CDD5-4D9F-9DE3-075B757DD764.jpeg
Subscribe to see EXIF info for this image (if available)
This little chicks stayed clumped on the same spot for the entire hour that I was stalking them (I think to make their parents job easier to find them when bringing the food). After 10 minutes of observation I realised that they are not going anywhere as long as I don’t get to close and/or threaten them in any way.

AB88FAF4-64B2-46DF-AE84-3604F7404EB1.jpeg
Subscribe to see EXIF info for this image (if available)
Or very unpredictable behaviour, like squirrels, as they go up and down, left and right in a matter of seconds.

Observation and patience is key ... and learning from books or other sources about different animals behaviour and instincts.

Getting as close as possible is a beast in itself as a skill. ALWAYS be careful and do not approach dangerous animals (duh?!?). Less harmful animals is a different matter. Because most of my early photography I did not own long lenses and manual focus only most of the time I learn early with lots of practice to move very slowly when approaching wildlife (mainly butterflies at that time). But it works with bigger animals too:

138BB443-FE4F-485C-B9D9-346F6DCAE6A6.jpeg
Subscribe to see EXIF info for this image (if available)

252CEB91-6EDB-469C-9706-DC1D39EDA818.jpeg
Subscribe to see EXIF info for this image (if available)

705B0F5C-0893-4568-8F13-441BAB5EB123.jpeg
Subscribe to see EXIF info for this image (if available)

Walk very slowly with slow footsteps with soles first and heels last, do not change body posture, size or shape (like raising arms to the side), keep camera close to the chest and the face so you can slowly react and make pictures. Because lenses are cylinders for animals it basically looks like a big eye is watching them, to be expected they would feel scared of being watched by a predator with 67mm eye.

What may help with stability when shooting such long (or close) distances is not just the body stance/posture (don’t worry about the dirt on the knees, do go closer to the ground) is breathing. Snipers in the military learn to control well their breathing before and during taking a shot. It applies for wildlife too. Take deep and constant breaths when racking and the viewfinder will be more stable. Just before taking an image or a burst hold your breath and you can get stable enough for either slow shutter speeds or less camera shake. (I guess all those 10 years of playing Counter Strike shooter when I was a teen finally paid off :p , I didn’t do military service).

Another point of interest is camera angle and height. Especially for birds. While birds will tend to stay up to the tree top, if you are patiently waiting for more then a few minutes the birds will go lower on the tree. The closer the birds are to your hight the more natural it feels (and less of making picture of birds butts).

A721C4E2-F552-4E66-9E76-CD3FF9FE9987.jpeg
Subscribe to see EXIF info for this image (if available)

Same with ground level animals, the closer you are to their level the more the images look and feel like it’s a window into their world and not a look as they are being observed in an overlord fashion (am I going to overboard with metaphoric speak?). Simply put I find the closer you are to the subjects level to more engaging the picture feels.

F04B5418-D98F-454C-8F4C-206D6BD1BD04.jpeg
Subscribe to see EXIF info for this image (if available)

Now, what does all of this shinanigins have to do with Autofocus? Well without a good base of those above no Autofocus performance will save you when the animals are nowhere to be seen or constantly run away from you. But if you do have a good use of them you can get more out of the Autofocus performance, especially C-AF.

How does Panasonic Leica 5o-2oomm and Olympus E-M1 Mark I perform? Well, confusingly. While trying to get used to C-AF I have been noticing inconsistence performance. For about 2 months I have been focusing mainly on C-AF with Tracking to see how well it works, or not, in a variety of light conditions and subjects.
I will start with the most exciting opportunity and it starts with a story (yes, I am wondering how much longer does this thread drag on too):

One morning, after work, I decided to stay around the grass field nearby and see what might happen (encouraged by some nice picture opportunities I had a few hours earlier). And I noticed a big bird in the sky, surveying the area like a drone:

44B460B2-F6D4-4F49-AF98-9D7DE485145B.jpeg
Subscribe to see EXIF info for this image (if available)

It’s a buzzard (or at least I think it is, I’m defined no biologist). So I decided to stay and watch its behaviour. So it swoops high and low until one moment it stops .... and like a helicopter it hovers, beating its big wings and keeping itself, incredibly, in one spot and was looking down, scanning very intensely bellow (never seen a bird that big do that before, and I used to watch a lot of National Geographic documentaries as a kid):

1B7F9501-6051-4617-971A-41F66496BEBA.jpeg
Subscribe to see EXIF info for this image (if available)

Back to Autofocus, I had the camera set to C-AF with Tracking, 5 FPS anti-shock mechanical burst, 9 square AF area and centre weight metering. I keep the shutter speed as high as I could, usually1/1000 to 1/4000 of a second. It seems to catch focus on the first image of the burst but then it slowly loses focus and can’t recover unless I stop the burst and refocus:

9F2971EB-0776-49DD-AE00-3EB8747677D2.jpeg
Subscribe to see EXIF info for this image (if available)

While I shot almost 2000 pictures that morning, the happy rate for me was around 25%. Most of the time the background was blue sky or grey clouds, while the subject was just the size of a single AF point. Longer lens might have helped, but a focus limited would have made it a lot quicker for the AF to recover.

So when the bird was keeping a keen eye on its pray bellow it would make small dives to inch itself closer to the grass field:

CF6FFDFF-033E-4628-91D5-2D797BE3AA29.jpeg
Subscribe to see EXIF info for this image (if available)

I wish I had use spit metering because on most of the images the bird was underexposed and even at ISO 100 shadow recovery is not looking good.

EC41C9C7-98EB-4A61-B2C7-646024EFDF43.jpeg
Subscribe to see EXIF info for this image (if available)

When it has finally got a full lock on its pray bellow, like a guided missile it plunged so fast to the ground level at even at 10 FPS I got only one picture in the entire burst as it dived. Now comes the most disappointing part of that morning:

0E2D78D8-6502-4D47-AC6D-0F35175659F4.jpeg
Subscribe to see EXIF info for this image (if available)

This is the best that I could do to make a picture of where it flew off with its meal. The Tracking AF could not keep up with the bird no matter how hard I tried. While the ISO sky rocketed because of the dark trees behind and because clouds were rolling in and changing the brightness dramatically quick. This is the only picture where it was in focus enough (but it’s still to out of focus) to see the field mouse in its talons. The AF was going in and out, from minimum to infinity and while I could track the bird by shape the camera could not put the focus on the bird at all. I had tried 5 times, the bird had 5 dives about every 20 to 30 minutes (I stayed there for about 2 hours and a half).

I hope I can try again soon, maybe get better luck next time.

I have tried Tracking AF with still subjects and slow subjects to see how it behaves. While it does get focus on the first frame, it immediately starts flickering back and forth, up and down, left and right even if the subject doesn’t move (that was to be expected) or starts to move (meaning it didn’t track where it was moving, just where it was before).

5ABBE0A1-F032-4C89-8B54-BA46C3963362.jpeg
Subscribe to see EXIF info for this image (if available)

One such example, where the left bird was sitting alone on the branch, and a male showed up out of nowhere from the left side, the camera was late to “see” the new comer, it stuck with 2 frames on the new bird and it completely lost track into the background while the male and female bird were still there and on the same focal plane as before. As the male bird left the camera tried to keep the tracking on it it couldn’t keep up:

981AFD29-6093-42F5-8ACB-D07DD92F385C.jpeg
Subscribe to see EXIF info for this image (if available)

AAC799F2-2BB4-4818-B820-52F97AECE6A3.jpeg
Subscribe to see EXIF info for this image (if available)

I have tried to get closer to make the subject bigger in frame to the AF points have more to work with but it didn’t change anything:

BE9E5B43-6A39-4C8E-8994-C1586EE044C5.jpeg
Subscribe to see EXIF info for this image (if available)

Most of the time I have tried using the single tiny AF point size because I have noticed the camera had a tendency to grab anything it wanted that was closer to me wishing the AF points, from tree branches to blades of grass.

I will try to use AF-C only (no tracking) for the next month or so and see how that one behaves. I am sure there is a lot more I need to learn. I have not dived to much into some of the deeper menus, like the AF-C Lock settings, I left it set on Off.
 

L0n3Gr3yW0lf

Wall-Eeeeeeeeeeeeee
Joined
Jul 31, 2013
Messages
1,645
Location
UK
Real Name
Ovidiu
I have tried different flight path scenarios, like flying away from me:

54A9E933-B4EF-4F18-8227-C9F55697B65C.jpeg
Subscribe to see EXIF info for this image (if available)

1F3A8710-1610-434D-B4A8-F597B982F99E.jpeg
Subscribe to see EXIF info for this image (if available)

It started the tracking well but after a few frames it got lost again.

Same with this bird:

29D21493-44D7-4BD3-9307-D47CC6DEEDAA.jpeg
Subscribe to see EXIF info for this image (if available)

FBD630DB-D2AC-434C-BD73-60F0B063EB90.jpeg
Subscribe to see EXIF info for this image (if available)

It started well and even though there was significant contrast between the black bird and the green grass it still lost the subject midway to the end.

I have tried flight path from left to right and vice versa and the Tracking AF was bleeding into the background.

0161E3A5-0E44-4788-9EB9-368C37F5B811.jpeg
Subscribe to see EXIF info for this image (if available)

Out of 20 images about 4 came out sharp, like the one above.

I have tried with dogs (only one quick test, dog owners seem to be weary about this bad weather) and out of a 30 sequence I got 2 sharp images, the dog was coming towards me:

9B0300C3-DD04-45FF-97FC-518F9D25C535.jpeg
Subscribe to see EXIF info for this image (if available)

C3B489F3-E102-42EF-92CB-9AA32266D04B.jpeg
Subscribe to see EXIF info for this image (if available)

I do wonder if the camera is holding back the lens, if the Panasonic Leica 5o-2oomm would not have such issues with Panasonic G9 or Olympus E-M1 Mark II or E-M1X.

I think a focus limited would have helped this lens a lot in any situation because it limits how much it needs to hunt and improve the acquisition speed and recovery on both Panasonic and Olympus (but even more so with Panasonic because CD-AF only). A custom setting would be an awesome feature, set it for 2 to 5 or 7 meters for small birds, 15 to infinity for big birds or long range tracking.
 

L0n3Gr3yW0lf

Wall-Eeeeeeeeeeeeee
Joined
Jul 31, 2013
Messages
1,645
Location
UK
Real Name
Ovidiu
More samples made with this combo:

39344316-BC98-47B7-80BA-C22444D11054.jpeg
Subscribe to see EXIF info for this image (if available)
AC74BAD5-FD7C-427C-A63F-33ECA6365A2E.jpeg
Subscribe to see EXIF info for this image (if available)
9C3B2C0E-6926-4C5F-826D-1C1AD52D1A26.jpeg
Subscribe to see EXIF info for this image (if available)
C9C932C6-CFB2-4C8A-B041-E5ACB66ED2A8.jpeg
Subscribe to see EXIF info for this image (if available)
24E19720-0F17-493B-BEB6-446377CF52D7.jpeg
Subscribe to see EXIF info for this image (if available)
8FD993CF-A55C-4048-B1A9-51CEC8E2A0EB.jpeg
Subscribe to see EXIF info for this image (if available)
511AB2F6-FC64-45F8-8D8F-63E653165399.jpeg
Subscribe to see EXIF info for this image (if available)
42077F78-4EC5-4D29-9E9E-0C2DBBE78BE5.jpeg
Subscribe to see EXIF info for this image (if available)
BBD17E0E-D002-4BF6-9DEF-66F9249E422B.jpeg
Subscribe to see EXIF info for this image (if available)
9319FD0A-7311-48A4-B59C-1ABC8AAD355E.jpeg
Subscribe to see EXIF info for this image (if available)
845BFF27-1827-4C29-B158-56494D954ECC.jpeg
Subscribe to see EXIF info for this image (if available)
BDAE29FA-D5B7-47DC-B7FC-37D2491F0514.jpeg
Subscribe to see EXIF info for this image (if available)
034FAEA5-711C-46D3-93D4-6EE6D34AF42E.jpeg
Subscribe to see EXIF info for this image (if available)
CD75A22E-B997-4827-99B8-54381900BDD6.jpeg
Subscribe to see EXIF info for this image (if available)
 

ac12

Mu-43 Legend
Joined
Apr 24, 2018
Messages
5,259
Location
SF Bay Area, California, USA
Have you updated to the current Firmware version?
I recall that the AF was addressed in some of the FW updates.

A problem that I have always had with area/zone AF with Canon and Nikon is that the camera does not know what MY subject is. And I have no control over what it will choose to focus on. Standard logic with Canon is "closest subject," so if ANYTHING within the focus area is closer to me, that is what the camera will focus on. As a result of this, I do NOT use area AF for the sports that I shoot, because 80% of the time I will have another player in the AF area, with a good chance of that player being closer to me than my subject.
I am guessing that this was the problem when the AF was going in and out. It may have been focusing on the grass in front of you.

Likewise, tracking with a busy background has given me problems on Nikon. The camera will sometimes loose focus of the subject and focus on something in the background.

C-AF on the EM1 for me has be iffy. I was shooting a hurdler coming straight to me, and the camera would sometimes loose focus of the hurdler and focus on the background, then it would come back to the hurdler. Quite irritating, as I rarely had that problem with my Nikon.
 

L0n3Gr3yW0lf

Wall-Eeeeeeeeeeeeee
Joined
Jul 31, 2013
Messages
1,645
Location
UK
Real Name
Ovidiu
Have you updated to the current Firmware version?
I recall that the AF was addressed in some of the FW updates.

A problem that I have always had with area/zone AF with Canon and Nikon is that the camera does not know what MY subject is. And I have no control over what it will choose to focus on. Standard logic with Canon is "closest subject," so if ANYTHING within the focus area is closer to me, that is what the camera will focus on. As a result of this, I do NOT use area AF for the sports that I shoot, because 80% of the time I will have another player in the AF area, with a good chance of that player being closer to me than my subject.
I am guessing that this was the problem when the AF was going in and out. It may have been focusing on the grass in front of you.

Likewise, tracking with a busy background has given me problems on Nikon. The camera will sometimes loose focus of the subject and focus on something in the background.

C-AF on the EM1 for me has be iffy. I was shooting a hurdler coming straight to me, and the camera would sometimes loose focus of the hurdler and focus on the background, then it would come back to the hurdler. Quite irritating, as I rarely had that problem with my Nikon.

Yes, both camera and lens have the latest firmware updates. While I do not use zone AF with foreground elements that can and will confuse the AF, I do wish it had a custom setting to make your own AF area size and shape (I do think Panasonic G9 lets you do that but don't quote me on that). I use zone AF only for big birds in flight, like the buzzard, where the sky or the clouds are the only background elements. Even in cases like that, I have seen the Olympus E-M1 Mark I lose its stickiness and drift out of focus. I do somewhat understand the less than great AF experience because it was Olympus's first PDAF in a sensor implementation. I do believe that the Panasonic Leica 5o-2oomm is NOT a good fit for the E-M1 Mark I, the camera will drag down the lense's greatness, but since I do not have a better/newer camera to test it with I can't guarantee that (though I would love to test that myself).


AF close focus limit is in the mk2, not the mk1.

The AF limiter I mentioned was for Panasonic Leica 5o-2oomm. Not the old Olympus 5o-2oomm. Not sure how expensive they might be to add to a telephoto lens BUT since Panasonic is so adamant about CDAF and DFD I would see a focus limiter nothing but a huge advantage to their system. Limiting any set of range in the AF length would improve the speed of acquisition of the subject, the speed of the recovery in case of a missed focus as well as tracking performance, stickiness, when there is sufficient distance between the subject and the background/foreground elements. Software Focus Range limiter would be another great addition, where you could put in custom values on active focus distance you might be looking for (example: 2 meters to 10 meters for close-up birds, 10 meters to 50 meters for big birds or birds in flight). Top that off with a Focus Distance Re-engaging, when you have a Focus Lock in Continous or Tracking AF, to specify how many meters should the AF look for forward or backwards to re-engage AF if the focus gets lost (perfect for the birds in flights with a path from left to right and right to left).

This made me think of an idea for AI focus tracking and detection. When the AI finds it's subject it keeps track of its location by shape, colour, etc. Why not have a Focus Limiter option in the AI where you can tell: track me only THIS type of animal. And have a set of options on what kind of animal you want, it doesn't have to be on very specific species but on animal type: birds, small birds, large birds. Like, give us a rotary menu setting (like the PASM) where we can choose different types of animals after we selected them from an alphabetic genera type (Aquatic, Birds, Mammals, Reptiles, etc). And the AI AF can then lock its algorithm to ignore everything within the frame except for the selected Animal Focus Lock Limiter. Have more than one bird in the shot: track all of them and let us select (with joystick or arrow pad or dials) which ones to prioritise.
 

Mack

Mu-43 Hall of Famer
Joined
Jan 14, 2018
Messages
3,123
I do find the in-body auto focus limiter to be a great asset on the E-M1 Mark II and the E-M1X. I have a tree stump where the birds land that is 33 feet distant. I set the first of three Auto-Focus Limiter selections to 25 to 40 feet and the lens responds much faster. I set up a 3x3 or 9x9 target over the top of the stump where the body of the bird will be and so far so good once I get a green square and the focus beep from the body. The second AF Limiter range I've picked is 30-220 feet to keep the thing from going off into infinity (or Olympus calls it 999.9 feet in their AF Limiter.). A normal-sized yard bird over 185 feet distant is pretty much useless for me in feather detail resolution with the 420mm setup.

Something else I've noticed is even though I have worked through the LensAlign and FocusTune software for the 300mm+1.4 teleconverter combo (i.e. 420mm), it seems to help to add about a 1-2 points plus to what the software says. Basically, I'm making the lens backfocus a bit on the bird. Could be the AF is locking on a front plane of the bird where the rest of the body and tail is further back so shifting the lens into a forced backfocus seems to help with the overall bird's sharpness.

I've also found stabbing the E-M1 Mark II shutter will cancel out my AF Tuned number and camera will go to a zero or default state. My E-M1.2 shutter button "stabbing" is very slow to get my AF tuned number into the mix verses the faster E-M1X. I have to slow down and be cognizant of pressing the button slowly and fully and not going "Look! A Bird! Must be fast!" and then stabbing the shutter too quick. The FocusTune software alerted me to that matter with some crazy numbers when I was thinking "This camera needs to go in for service," but it was me stabbing the release button too fast rather than squeezing it off slowly.
 

L0n3Gr3yW0lf

Wall-Eeeeeeeeeeeeee
Joined
Jul 31, 2013
Messages
1,645
Location
UK
Real Name
Ovidiu
I do find the in-body auto focus limiter to be a great asset on the E-M1 Mark II and the E-M1X. I have a tree stump where the birds land that is 33 feet distant. I set the first of three Auto-Focus Limiter selections to 25 to 40 feet and the lens responds much faster. I set up a 3x3 or 9x9 target over the top of the stump where the body of the bird will be and so far so good once I get a green square and the focus beep from the body. The second AF Limiter range I've picked is 30-220 feet to keep the thing from going off into infinity (or Olympus calls it 999.9 feet in their AF Limiter.). A normal-sized yard bird over 185 feet distant is pretty much useless for me in feather detail resolution with the 420mm setup.

Something else I've noticed is even though I have worked through the LensAlign and FocusTune software for the 300mm+1.4 teleconverter combo (i.e. 420mm), it seems to help to add about a 1-2 points plus to what the software says. Basically, I'm making the lens backfocus a bit on the bird. Could be the AF is locking on a front plane of the bird where the rest of the body and tail is further back so shifting the lens into a forced backfocus seems to help with the overall bird's sharpness.

I've also found stabbing the E-M1 Mark II shutter will cancel out my AF Tuned number and camera will go to a zero or default state. My E-M1.2 shutter button "stabbing" is very slow to get my AF tuned number into the mix verses the faster E-M1X. I have to slow down and be cognizant of pressing the button slowly and fully and not going "Look! A Bird! Must be fast!" and then stabbing the shutter too quick. The FocusTune software alerted me to that matter with some crazy numbers when I was thinking "This camera needs to go in for service," but it was me stabbing the release button too fast rather than squeezing it off slowly.
That sounds very interesting. I wasn’t sure if Olympus was the one who had in camera focus limited (I thought one of the brands had it but wasn’t sure which). I remember that Sony have a feature where it shows you the distance between you and the subject in the viewfinder interface, a good tool to better understand and keep track of your subjects position relative to you.

Unless the bird is just in cruise mode I tend to burst my burst shots. I go for around 5 to 15 images if I’m just waiting for the bird to do something interesting, while in let the buffer clear and the AF to recover and continue to observe the behaviour of the subject. I have the AF engage on half press shutter, I have tried to use back button AF on the AF-L button but the camera is to small for my hand, it makes my hand shift uncomfortable to use that button all the time (it makes my hand separate from the camera grip on the lower right side of the grip, making the camera feel less sturdy in my hand from less contact with my palm.

I am excited about Olympus E-M1 Mark III, if they can squeeze or even improve the performance of E-M1X inside a smaller body. Maybe invest in a more extensive AI tracking, like animal eye or shape tracking.

And, please, a better and more comprehensive Chromatic Aberration correction with Panasonic lenses. While, funnily enough, Chromatic Aberration show up so intense in out of focus images, sometimes even on slight front or back focusing, that one can even judge the sharpness and focus accuracy by the, less, amount of blue/purple and yellow colour shift around the subject I would rather not have to correct for it.
Blue is hard to correct if the background is blue as well, as it becomes a grey hallow. Yellow is even more annoying because correcting it can drift or destroy the entire yellow pallet in the image (and there’s a lot of yellow in green/living foliage).
 
Last edited:

ac12

Mu-43 Legend
Joined
Apr 24, 2018
Messages
5,259
Location
SF Bay Area, California, USA
The EM1-mk3 is not going to be as BIG as the EM1X, so there isn't the space for all the computing circuits and heat dissipation.
So don't expect to get all the AI and other stuff that is in the X in the mk3.
All we can hope for is a reasonable subset of what is in the X in the mk3.
 

L0n3Gr3yW0lf

Wall-Eeeeeeeeeeeeee
Joined
Jul 31, 2013
Messages
1,645
Location
UK
Real Name
Ovidiu
The EM1-mk3 is not going to be as BIG as the EM1X, so there isn't the space for all the computing circuits and heat dissipation.
So don't expect to get all the AI and other stuff that is in the X in the mk3.
All we can hope for is a reasonable subset of what is in the X in the mk3.
While compute chips are shrinking (in size), getting faster and more energy efficient each year cycle (give or take, given the state of Moore's Law is reaching its limit with the limits of the laws of physics on the current architectural design of 2D electronic that has been used for around 60 years now). Cameras should be able to benefit from this if not for the economic state of the camera industry (past and present). Computers and smartphones (and other devices) far outsell any brand or the entire camera industry combined, making their components in high demand and lower price. This makes camera chips and technology on the higher side and less on fast evolution, except for Sony.

No camera brand or manufacturer have their own electronic chip manufacturing capability (think of their main processors like Canon's Digic, Nikon's Expeed, Olympus's True Pic, etc), except for Samsung, so they are more on the side of what it's available then what's the best (also the bulk they can buy doesn't even shine a light to the bulk that Apple buys in terms of chips for example). Think how much Sony has to handle in a 24 MP sensor 20 FPS with A9 and that's around the same space as an Olympus E-M1 Mark II (or even heavier data processing of Sony A7 Mark II, 42 MP, 10 FPS maximum, at 82 MB per RAW file that's 820 MB per second at minimum to be processed). It can be done but the problem is the kind of technology they can get their hands on and can afford to put in in an affordable price (given the E-M1X's launch price I wouldn't expect it to be as fast or half the price).

And there's also the software side, optimization of AI coding and processing can get you with slower and/or older chips, but you needs lots of software engineers to do a lot of crunching to get more out of what they have now, given Olympus's tiny size as a (photography division) company compared to the likes of Sony or Canon they wouldn't be able to bring out more updates with faster processing (unless they have incredibly talented coders there).

(On a side note, this reminds me how big of a loss Samsung really was when they silently killed their mirrorless camera manufacturing. While many would not give half a penny on them, I think they would have brought the power and innovation that Sony can only do right now. They had very good sensors [the 28MP one in NX1] and with the capability of sourcing their own chips [right now Samsung is the #2 in chip manufacturing and selling] they could have challenged any camera for performance that we have right now)
 

AussiePhil

Mu-43 Hall of Famer
Joined
Jun 1, 2014
Messages
3,397
Location
Canberra, ACT, Aust
Real Name
Phil
I'll preface this with the following information
I have used the EM1.1, em1.2 and 1x now quite extensively with the 40-150Pro and the PL100-400 but not as a birding person but the thoughts should apply regardless of the subject matter.
Let's discuss the EM1.1
I first started with the 12-60 and 50-200swd 4/3rds lenses and some hopes for CAF but tempered by the various reviews. The results were less than stellar and bad enough to drive me back to SAF, considering the years of MF with film and SAF on other 4/3rds cameras i was only back to where I started.
When i got the 40-150Pro we tried again with some limited success but again for most of what i was shooting at the time SAF still did the job for me.
Then we got the PL100-400 so we thought it was time for another test however this didn't even get out of the office and leads to an observation and thought/theory on using CAF with effectively stationary objects that are not changing the distance in any meaningful way between the camera and subject.

What had happened was i was taking some test shots at various focal lengths of a simple but large black and white barcode on the side of a box, the barcode would have been 3x5 inches in size, you know those big ones on boxes.
in SAF focus was nailed fast and accurately

Now read this sentence carefully.....
In CAF the focus was nailed fast and accurate but would then drift in and out of focus. the CAF algorithm seemed to be expecting and wanting the barcode to be moving and the focus point was shifting just a little and then it would snap back into focus. Remember this is a high contrast barcode.
Due to it being close, DOF was thin enough to see this happening.
i have seen this repeated all the time with various lenses on the em1.1 and to a lesser extent on the em1.2... untested on the 1X
The working theory is that the CAF code is expecting the object to move and is shifting the focus fractionally to test this, with greater distances the action may not be visible due to available DOF but still result in a slightly soft photo.
You can also observe this when tracking a subject in a constant direction that stops changing distance from the camera, focus will overshoot during a burst.

All this was enough, for what i was shooting, at the time to push me back to SAF.

With the mk2 and EM1X i have adopted BBF and with CAF and static distance subjects will focus and then stop.
 

L0n3Gr3yW0lf

Wall-Eeeeeeeeeeeeee
Joined
Jul 31, 2013
Messages
1,645
Location
UK
Real Name
Ovidiu
I'll preface this with the following information
I have used the EM1.1, em1.2 and 1x now quite extensively with the 40-150Pro and the PL100-400 but not as a birding person but the thoughts should apply regardless of the subject matter.
Let's discuss the EM1.1
I first started with the 12-60 and 50-200swd 4/3rds lenses and some hopes for CAF but tempered by the various reviews. The results were less than stellar and bad enough to drive me back to SAF, considering the years of MF with film and SAF on other 4/3rds cameras i was only back to where I started.
When i got the 40-150Pro we tried again with some limited success but again for most of what i was shooting at the time SAF still did the job for me.
Then we got the PL100-400 so we thought it was time for another test however this didn't even get out of the office and leads to an observation and thought/theory on using CAF with effectively stationary objects that are not changing the distance in any meaningful way between the camera and subject.

What had happened was i was taking some test shots at various focal lengths of a simple but large black and white barcode on the side of a box, the barcode would have been 3x5 inches in size, you know those big ones on boxes.
in SAF focus was nailed fast and accurately

Now read this sentence carefully.....
In CAF the focus was nailed fast and accurate but would then drift in and out of focus. the CAF algorithm seemed to be expecting and wanting the barcode to be moving and the focus point was shifting just a little and then it would snap back into focus. Remember this is a high contrast barcode.
Due to it being close, DOF was thin enough to see this happening.
i have seen this repeated all the time with various lenses on the em1.1 and to a lesser extent on the em1.2... untested on the 1X
The working theory is that the CAF code is expecting the object to move and is shifting the focus fractionally to test this, with greater distances the action may not be visible due to available DOF but still result in a slightly soft photo.
You can also observe this when tracking a subject in a constant direction that stops changing distance from the camera, focus will overshoot during a burst.

All this was enough, for what i was shooting, at the time to push me back to SAF.

With the mk2 and EM1X i have adopted BBF and with CAF and static distance subjects will focus and then stop.

I noticed the same thing. Makes me wonder if because of the nature of hour Contrast Detect AF works even if the subject is moving only lateral (with little to no distance change between you and it) the camera would still go in and out of focus when tracking, even correctly, in one direction. I wonder how well would Panasonic's AFF, Flexible, would work with a subject with very varied type and speed of movement.

I am starting to think that the best way to use C-AF and Burst shooting is in short and small control bursts to let the AF regain its control for the subject. I might be going back, like you, to S-AF as my default option for AF.
 

AussiePhil

Mu-43 Hall of Famer
Joined
Jun 1, 2014
Messages
3,397
Location
Canberra, ACT, Aust
Real Name
Phil
I noticed the same thing. Makes me wonder if because of the nature of hour Contrast Detect AF works even if the subject is moving only lateral (with little to no distance change between you and it) the camera would still go in and out of focus when tracking, even correctly, in one direction. I wonder how well would Panasonic's AFF, Flexible, would work with a subject with very varied type and speed of movement.

I am starting to think that the best way to use C-AF and Burst shooting is in short and small control bursts to let the AF regain its control for the subject. I might be going back, like you, to S-AF as my default option for AF.

Continue to experiment with CAF as you may well find very useful scenarios to use it... it just didn't suit what i was doing at the time and i wasn't prepared to put more time/effort into it.
I haven't directly used my em1.1 for a number of years now as it's on loan to my son along with the 12-60, 50-200 combo.

The mk2 was a significant step up in CAF but still when dealing with what i call static distance even if moving subjects SAF is my preferred option but my mk2 tends to live in CAF and i just use BBF to stop the focusing action. SAF is still better for some situations though.

The EM1X changes the game a bit more with MF being available whilst in CAF mode so it's practical to leave the camera in CAF/TR+MF mode more often and just use BBF and MF override as required.
 

Latest threads

Top Bottom