arbitrage wrote:
I also wanted to add that when I've done these types of shots I use Wide (no tracking). Of course I do 90% of my BIF in Wide (no tracking) and the other 10% in Zone (no tracking).
These examples were all coming off a perch in a messy Arbutus tree...
Daran wrote:
Short answer: no. BEAF (bird eye AF) is a fantastic feature when it works, alas it works only if the eye is reasonable in size. For estimating the eye position without having a decently sized eye you'd need the new A7RV or A9III with the AI chip. Of course the A7RV is too slow for this and the A9III lacks resolution. Besides, BEAF works like tracking and hence incurs a small delay. Which is already a no-go for what you attempt to do. I'm shooting the A1 with the 600GM these days. Even that lens without TC can't usually keep up with a song bird jumping into the camera.
There a lots of small things that can help. I'd certainly loose the TC. Disable tracking, too, if the background allows it. Maybe close the aperture a bit, increasing the DOF. Just a little though, as the DOF increases only linearly while the exposure time increases squared. Cleaner background obviously helps, but there will always be the branch the bird sits on. Leaving more distance (or zooming out) helps with the AF speed and DOF, but also makes it less likely that the camera still considers the increasingly smaller blob as the actual subject.
What really helps is: if the bird sits on the tip of a soft branch and launches your way, the acceleration is *much* smaller as the branch is pushed away, sometimes allowing the lens to follow the bird.
arbitrage wrote:
I also wanted to add that when I've done these types of shots I use Wide (no tracking). Of course I do 90% of my BIF in Wide (no tracking) and the other 10% in Zone (no tracking).
These examples were all coming off a perch in a messy Arbutus tree...
arbitrage wrote:
May I ask why is tracking not used?
I also wanted to add that when I've done these types of shots I use Wide (no tracking). Of course I do 90% of my BIF in Wide (no tracking) and the other 10% in Zone (no tracking).
These examples were all coming off a perch in a messy Arbutus tree...
Based on personal experience with the A1 and 400 and 600 GMs - the chance to get the second and third photos in focus is low. Of course it is possible but quite hard. The reason is the bird's acceleration is the highest at the moment of takeoff. I tried with tracking, with no tracking, different levels of sensitivity - all makes a little difference. At the end it all depends on the initial acceleration. If it's "reasonable" - there's a chance of taking a sharp photo. I hope the A9 III can improve things a little. We'll see in February...
docusync wrote:
Based on personal experience with the A1 and 400 and 600 GMs - the chance to get the second and third photos in focus is low. Of course it is possible but quite hard. The reason is the bird's acceleration is the highest at the moment of takeoff. I tried with tracking, with no tracking, different levels of sensitivity - all makes a little difference. At the end it all depends on the initial acceleration. If it's "reasonable" - there's a chance of taking a sharp photo. I hope the A9 III can improve things a little. We'll see in February......Show more →
Actually pre-capture is what I really want, and have for years. I must have half a terabyte of useless sequences of 100 identical shots where I thought the bird might do something interesting but just sat there, or flew off 2 seconds after I gave up. If they never put it in the A1 then actually I am on the fence about an upgrade path... Precapture vs pixels is not an obvious choice for me.
If you mean A1, not A1 II, then sorry to break it to you but Sony thinks you would not buy A1 II if pre-capture was added to the original A1 . photonoclast wrote:
Actually pre-capture is what I really want, and have for years. I must have half a terabyte of useless sequences of 100 identical shots where I thought the bird might do something interesting but just sat there, or flew off 2 seconds after I gave up. If they never put it in the A1 then actually I am on the fence about an upgrade path... Precapture vs pixels is not an obvious choice for me.
Are you me? Lol the exact same thing.. course I don't have a good birding lens anyway but still. It's really depressing holding the shutter to only stop and then it does something amazing. Just can't help but sigh..
Having used both the A9 and now the A1 to photograph small birds like this I've found there is virtually no scenario when the camera will accurately track a bird flying towards you if (as in this scenario) the bird is a similar color to the background, the background is very busy, and the lens is not fast enough to be able to blur most of that background (or if camera movement does not blur the background via panning).
In other words, I'm not surprised the OP was not able to get this shot. Heck, my A1 sometimes struggles to even focus on a stationary bird against a busy, similarly-colored background! The only solution is often to get skilled enough to be able to keep the small spot focus area on the moving bird, although even then I've occasionally seen focus jump somewhere else.
Maybe the A7R5's AI autofocus can do better. Isn't that what AI is supposed to be for?!
photonoclast wrote:
I have no doubt there's plenty of operator error! The thing is, I do have thousands of successful images against clean backgrounds. That's a piece of cake with the A9.
But what I'm trying to do now is move more into showing the bird as part of its habitat, maybe even moving in its habitat. Isolated "portraits" against clean backgrounds are getting a bit boring to me - I think I've got at least a few dozen clean portraits of every species that comes through my area, and I'm trying go beyond that. I agree the images I posted are complete rubbish, but I posted them to show the challenge.
Your suggestion of removing the TC is exactly what I'm wondering. If I were to upgrade to the A1 and take off the TC, after cropping I would get about the same number of pixels per bird, as I've seen Arbitrage put it here, so that would be OK, but then the question is how much AF capability I gain with (1) the extra light available since the TC is off and (2) the increased AF speed of the A1....Show more →
I certainly agree having habitat or even a hint of it, when tastefully added to the frame gives it life. but as you point out this example is not the correct starting point. I suspect once you have a beautiful background dialed in (with elements of habitat but not so much clutter) the problems with the AF will reduce.
The camera is looking to first discern the subject from the BG and then track focus on it, when there is too much clutter the initial discern process make take just a fraction longer. with a tiny fast bird darting towards you, the bird will become totally OOF within a split second so the camera will just ignore it as some blurry blob and continue holding AF on the clutter
The A1 AF is certainly more capable than the A9, in your scenario though the major factors are 1. the BG and 2. the TC added to a zoom lens that's not exactly the fastest in terms of AF. the A1 will solve the 2nd issue as you can take off the TC to speed the AF up a bit. regarding the cluttered BG issue, it is more of a fundamental problem and there is no magical solution to it. but once you find a pretty BG to illuminate this issue both cameras will perform to your satisfaction I believe, at least after several attempts. The A1 will be better for sure, but the A9 will give you some frames to work with as well, maybe with more attempts. Also I do find the A1 with the bare zoom lens sharper than the A9 with the zoom and the TC, especially for dynamic frames that need some cropping. to be honest while I like 2-6 very much, I am not a fan of it with the TC..
I also agree with the consensus here that the "non-tracking" AF is way to go for this type of scenario. Where possible a faster lens example f/4 will also help in that the shallower DOF will blur the clutter more and thus reduce the distraction for the AF to some degree.
photonoclast wrote:
I’d like to get some input from people who have experience with the A1 and small birds in flight.
I have the A9, which I use with the 200-600 + 1.4TC. Subject is 95% small birds (migratory warblers, vireos, etc). Mostly it works great and gets the images I want. There is one scenario, though, that arises frequently and the autofocus just can’t keep up. Here are three consecutive images in a sequence at 20fps of a bluebird launching. In this case I had the focus on the perched bird and the focus setting was Lock-on AF expand Flexible Spot.
Obviously the focus remained nailed to the branch. It doesn’t seem to matter what AF setting I use (wide, tracking or no tracking, etc.), it never follows the bird since they move so fast. The only time I’m successful with these types of images is if the bird flies parallel, i.e. stays in the plane of focus, and not even slightly towards or away from me.
I’m just curious if the A1 (+ the 200-600, with or without the TC; if it takes the 600/4 that's not an option) can handle this situation. Large birds are easy – the issue is these little ones when the background is close and busy (these birds try hard to stay in the brush for safety!).
I’ve seen some great images here and can manage ok myself when the bird is flying against a distant background. The specific issue is when they are among branches or launching, and the AF stays nailed to some static feature (usually the branch it was perched on). So, I’m curious if the A1 eye detection plus tracking solves the problem, or not.
It's not a lens problem - all the lenses, with the exception of some of the macro lenses, can rack focus much much faster than any living thing can move.
You need the A9iii - I will let you know how that works in Feb - pre-capture is one thing but is it fast enough to follow them as they launch ?
Well if the a6700 is anything to go it should be. Keep an eye out for JP commentary in this regard - by all accounts he filled a few cards using pre-capture on wildlife recently.
photonoclast wrote:
Actually pre-capture is what I really want, and have for years. I must have half a terabyte of useless sequences of 100 identical shots where I thought the bird might do something interesting but just sat there, or flew off 2 seconds after I gave up.
Lol this is EXACTLY my experience!!!
photonoclast wrote:
If they never put it in the A1 then actually I am on the fence about an upgrade path... Precapture vs pixels is not an obvious choice for me.
There is a very low chance Sony may still add pre-capture to the A1. They already announced a firmware update (March 2024?) and it's just some random BS nobody cares about.
My concern that pre-capture alone may not be enough. It's mostly the sensor read out speed + AF/prediction algorithms (as Duncan also mentioned). Besides my Sony gear I have the Canon R7 (pre-capture!) + the 100-500L lens. While in theory it sounds like a nice lightweight system - IRL it's pretty much unusable. Yes, I never missed a taking off bird but I never got a usable shot either The R7's sensor is super slow, probably the same readout speed as the Sony A7R5. The A9 III should be in a different league, so there's hope... While 24Mpx doesn't sound like much IMHO it's better to have a low res sharp photo than a high res but blurry and unusable.
shac wrote:
Geoff - what sensitivity level are you using?
Thanks
David
I typically leave it set at 1. Although I can't be certain these particular shots were at 1 as I have fiddled with it from time to time but good chance they were at 1.
I've just found that Tracking results in more misses and slower recovery to the actual subject. If tracking starts tracking a water ripple or stick it won't let go. Wide or Zone without tracking may focus on the water or stick but will usually correct to the moving subject after a couple frames. The camera gets way more distracted in Tracking Wide and Tracking Zone in my experience with BIF.
docusync wrote:
Based on personal experience with the A1 and 400 and 600 GMs - the chance to get the second and third photos in focus is low. Of course it is possible but quite hard. The reason is the bird's acceleration is the highest at the moment of takeoff. I tried with tracking, with no tracking, different levels of sensitivity - all makes a little difference. At the end it all depends on the initial acceleration. If it's "reasonable" - there's a chance of taking a sharp photo. I hope the A9 III can improve things a little. We'll see in February......Show more →
True, the second or third are rarely sharp. In the examples I posted those were probably fifth or later once the focus recovered to the moving bird. The first few frames are always back focused.
arbitrage wrote:
In the examples I posted those were probably fifth or later once the focus recovered to the moving bird.
5-6th frame means 0.2s @ 30s. Likely by this time the bird’s trajectory becomes clear and acceleration decreases, so the camera can catch up…
The A1 is supposed to perform 120 AF calculations per second. It’s the same number as the A9 III, but hopefully the latter will utilize smarter AF algorithms that would allow it to predict the trajectory faster and more reliably. I’m actually surprised they didn’t bump the AE/AF calcs to 240 per second if the sensor readout is so crazy fast.
I typically leave it set at 1. Although I can't be certain these particular shots were at 1 as I have fiddled with it from time to time but good chance they were at 1.
I hate tele convertors. They all adversely affect focus speed and accuracy and it's just how much of that you are fine with. With the 50mp of the a1, you don't need it nearly as much.
timgangloff wrote:
I hate tele convertors. They all adversely affect focus speed and accuracy and it's just how much of that you are fine with. With the 50mp of the a1, you don't need it nearly as much.
This seems to be the general take-away of this thread. There seems to be a pretty strong consensus that the A1+ bare telephoto has somewhat better resolution than the A9 + TC. So, the A1 would let me take off the TC and that will give a better hit rate for these types of challenges. (I also have a lot of BIF shots with the bird at the very edge of the frame due to the difficulty of following at speed; now more of these would be crop-able into useful images.)
Light loss and slowdown in AF speed associated with teleconverter is always a big issue, and makes it a net negative in a lot of situations. But I don't think A1+200-600 natively will solve your problem. The combo still doesn't respond fast enough. It may also have something to do with the algorithm. The way I see it, it's like a stock price is flat for a long time. Then suddenly it has an uptick. Is it real or just noise? By the time the camera concludes it's real, there is a lot of catching up to do already. And that lens is not known for acceleration.
One approach that semi works in this situation is the age old method of shooting in manual mode (I have the lens button programmed to toggle between AF/MF easily), and pre focus slightly in front of the bird. I say semi works because you can end up with one frame where the bird is still approaching the focus plane while the next frame it's already past. But fps helps in this situation, and it's a probabilistic thing. So you should end up with good ones given enough tries.
If you stick with AF, you can also pre focus in front of the bird, and only hit the focus/shutter when it launches. In this case, the lens and the bird are moving in opposite directions meeting each other. So slowness in the lens is less a concern. Also since the bird will be closer to the focus plane than the branches behind, it's more likely for the camera to focus on the bird than the branches behind. However, if there are busy branches in front, most likely the camera will still end up focusing to those instead. It is indeed annoying many times the camera doesn't even seem to recognize it's a bird and thus the object when it's a little blurred. Maybe with the AI chip in next generation, it can work much better.
.
The R7's sensor is super slow, probably the same readout speed as the Sony A7R5. The A9 III should be in a different league, so there's hope... While 24Mpx doesn't sound like much IMHO it's better to have a low res sharp photo than a high res but blurry and unusable.