Nailing (auto-)focus with the Panasonic 100-300?

I think the problem here is the nature of CDAF auto-focus combined with the relatively slow focus performance of the 100-300 lens.

CDAF works by checking the contrast in the focus area, moving the lens a bit, and checking again. If it has moved in the right direction, the image will be sharper (more contrast). This repeats until the AF converges on the best focus.

The critical part of this process working is the converges part. i.e. the camera keeps moving in the same direction until it goes too far and then backs up a bit. This is fine if the samples in the focus area are of exactly the same part of the subject (like an eye for example) each time the AF checks for contrast. This is fairly easy to achieve with a static subject and a lens that moves reasonably quickly to focus and IS probably helps as well.

However, when the subject is moving, it is very, very difficult to ensure that the same part of the subject is under the focus box each time the image is sampled. This could lead the AF to assume it has found the best focus because the image gets less contrasty between two steps when actually, the bit of the image being sampled is different and either has different contrast or is at a different distance.

There is not much that you can do about this with any particular lens other than use a smaller focus box and try to track the subject more accurately. Lenses that focus faster and resolve more fine detail work better.

The "answer" is PDAF but this too isn't perfect on the E-M1, especially with older, slower lenses.

My challenge in that first 100/6.7 photo is that I do not see _anything_ in "proper" focus, although the camera claims in the EXIF data that focus is on the shoulder (S-AF focus priority release).

I agree, it looks simply out of focus. However I doubt if the above is a correct interpretation of the EXIF data. It tells what your chosen AF area was, but probably not if the camera actually reached focus there (for that, see Focus Distance - but that value won't be very helpful if you don't know your actual distance to the subject). Or are you confident that the camera confirmed focus (and you didn't have Release Priority on)?

From there, you can figure out how to achieve it in practice.

That's what I am trying to get to right now

Well, I think you'll need to do some actual experimenting with focus point size, autofocussing twice (or prefocussing), continuous-AF, et cetera to find the method that works best. With my 50-200mm Fourthirds lens, C-AF and focussing twice often helped with this kind of event photography, but that was using phase detect AF which isn't really comparable.

You must have a different batch than I do.  My focus ring is not sticky and I find it very easy to use.  Now that the lens spends most of it's time on either the GX8 or GH4 (sometimes, the GM5), I find that auto focus works a lot better.  I also use the pinpoint (cross hairs) focus target.

shoffmeister wrote:

LMNCT wrote:

My cure on the GX1 was to focus manually and that worked quite well.

The focus ring on my 100-300 is so stubborn and sticky and cranky - how are (were?) you able to operate that?

I'd love to shoot "S-AF + manual", because turning the focus ring will enable manual focus assist magnification automatically. For that to work, it seems as if I need to keep the shutter button half-pressed with the right hand while using thumb and index finger of the left hand to pinch and operate the focus ring. I seem to be lacking the right technique here. (And, e.g., a monopod firmly stuck into the ground to provide leverage against the left-handed turning of the ring)

on the 100-300 the more you squeeze the focus ring the stiffer/sticker it becomes. While by no means an "answer", i try to squeeze as little as possible.

Focus ring might be an option but i dont think it will solve the stickiness.

I've *considered* trying to get some corn starch or graphite inside the ring, but havent actually done so.

I've *considered* trying to get some corn starch or graphite inside the ring, but havent actually done so.

Just a couple of days ago I read an account from someone who said that her new 100-400mm became much smoother after she'd used it in the rain, and the extended part of the lens had gotten wet (after which she retracted it, probably pulling a tiny bit of moisture into the lens). I don't recall if she was talking about the zoom ring or the focus ring, but it might be worth a try. I'd hesitate to use graphite (let alone starch!) because that would probably affect the sealing.

Danielvr wrote:

I've *considered* trying to get some corn starch or graphite inside the ring, but havent actually done so.

Just a couple of days ago I read an account from someone who said that her new 100-400mm became much smoother after she'd used it in the rain, and the extended part of the lens had gotten wet (after which she retracted it, probably pulling a tiny bit of moisture into the lens). I don't recall if she was talking about the zoom ring or the focus ring, but it might be worth a try. I'd hesitate to use graphite (let alone starch!) because that would probably affect the sealing.

The 100-300 is not weather sealed; that still might not justify trying any powder based lubricant.

I read that thread as well but didnt really know what to make of it. Moisture would dry out, and perhaps leave some residue; hard to see why that should help despite the fact that it did.

The 100-300 is one of the oldest M43 lens out there. Besides its focus accuracy, the burst speed / (re)focus speed is not very good (i cant recall the reasons but it relates to the lens, not the body, not the photographer).

The 100-400 should be better on speed based on this , specifically this line:

High speed digital signal exchange at 240 fps

The image sharpness one can reasonably infer from other people's image but the ease of use, focus speed, focus accuracy are much harder till you have one in your hands. (And even then it may be body specific as the GX8 would give the "dual IS" while others wont).

The 100-300 is not weather sealed;

I know, but I think that you'll find a seal around the extending part of the barrel even so (I have no 100-300 to verify this, but it seems to me that it would be too vulnerable without one there).

I read that thread as well but didnt really know what to make of it.

Do you happen to recall where it was? I meant to revisit that thread because of some excellent images from the 100-400mm. I'm a bit of Zuikoholic myself, but to be honest I think the Panaleica renders textures much nicer, more 'analog', than the 300mm f4 does. Its colors are beautiful as well.

The 100-400 zoom that got smoother after getting wet: here

Thanks much!

Just look at those bird photos she took from her windowsill.. beautifully rendered textures, excellent colors and incredible pop. I know this is the wrong thread, but anyone looking for 100-400mm bird samples should check them out:

https://www.mu-43.com/threads/pana-leica-100-400mm-f-4-6-3-first-impressions.83425/page-9#post-881469

shoffmeister wrote:

Danielvr wrote:

The computed DOF of the 100/6.7 shot should have been 90cm (online calculator and EXIF data agree). _If_ focus as claimed by the camera had _really_ been on the shoulder, then a whole lot more of the arm, the helmet, and the face would have been in focus.

The sharpnesss standard underlying 'depth of field' is rather arbitrary

... and contains a perceptual component derived by Zeiss some fourty years ago, taking into account display (i.e.printing) technology of the times as well viewing distance of the output.

My challenge in that first 100/6.7 photo is that I do not see _anything_ in "proper" focus, although the camera claims in the EXIF data that focus is on the shoulder (S-AF focus priority release). At 1/1000 exposure time and with three axis IBIS (Olympus E-PL7, overrides OIS) and a mostly static subject (Danny McAskill was standing still on the bike at the time), I would tend to rule out subject movement or equipment movement in this specific case (in the general case, sure, those would be major factors)

I suspect the actual culprit here is that the part of the bikers jersey that you were trying to focus on (the shoulder) has almost no contrast, just gray and some darker gray. Since the focusing of the camera is reliant on the contrast of the selected focus area there just is not enough contrast there for it to work properly. I have the same problem with foliage. Couple that with the slow focusing motors of the 100-300 and it can be frustrating.

Maybe you shoot start by establishing the sharpness of your lens by shooting something static at similar distance, aperture, etc, using a tripod and electronic shutter. Then you'll know what kind of sharpness is possible.

My copy of the Panasonic 100-300 has decent sharpness. It is not an Olympus 300/4 or a Panasonic-Leica 100-400, but even at the far, far (bad, bad) 300mm end it can produce the sharpness of the second photo (300/5.6 at 1/750). If I am not mistaken, that specific photo is not even the best the lens can produce at that focal length - it is good enough for me, though, because down-sampling to sizes I use removes the softness. At 100mm the Panasonic tends to be much easier to shoot and has significantly better optical quality.

From there, you can figure out how to achieve it in practice.

That's what I am trying to get to right now

Danielvr wrote:

The 100-300 is not weather sealed;

I know, but I think that you'll find a seal around the extending part of the barrel even so (I have no 100-300 to verify this, but it seems to me that it would be too vulnerable without one there).

I read that thread as well but didnt really know what to make of it.

Do you happen to recall where it was? I meant to revisit that thread because of some excellent images from the 100-400mm. I'm a bit of Zuikoholic myself, but to be honest I think the Panaleica renders textures much nicer, more 'analog', than the 300mm f4 does. Its colors are beautiful as well.

From this thread: it seems the interplay of burst rate and diapraghm.  IIRC, the two slower burst rates (on Panny bodies) *do* autofocus between shots, the two fastest do not

Are you shooting using C-AF or S-AF?

The 100-300mm problem occurs when shooting at rates over 2-4 fps using C-AF. The type of diaphragm used in the older Panasonic lenses cannot open and close fast enough for high speed C-AF.

My 100-300mm could only shoot at 2-3 fps in C-AF mode which is why I changed it for the Olympus 75-300mm.

Hope that helps.  For me the 'focus hunting' at the long end is the biggest drawback - but I'm usually shooting airplanes, not birds.

LMNCT wrote:

The 100-300 is not "garbage". The trick to getting sharp results at the long end is to use an aperture like 7.1.

I've tried from f/5.6 to f/11 at 300mm and I don't get sharpness I'd consider acceptable at any aperture. It's fine from 100-200mm, acceptable up to 250mm or so when stopped down, but zoomed all the way to 300mm I think I'm justified in calling it garbage. Of course I'm just talking about my copy of the lens - it's possible I don't have a particularly good one.

I just acquired the p100-300 and was also surprised at OOF shots. I think, however as one poster said, at the long end of telephoto with a subject moving, and at slowish shutter speeds, you are more than likely going to get soft, unless you get lucky. I do think this lens can be sharp, however, as i caught this squirrel pausing for a moment and shot him. Out of 7 s-af frames at 300mm only when he paused did it nail the focus: (heavily cropped to 1100-1600 as well, his paws are a little OOF because he was busy with them but i think the rest of him is in focus)

not sure if exif will show but it's 100 iso at 5.6 1/400 at 300mm

I think, however as one poster said, at the long end of telephoto with a subject moving, and at slowish shutter speeds, you are more than likely going to get soft, unless you get lucky.

Telephoto lenses that still produce a sharp image under those circumstances are exceedingly rare.