arbitrage wrote:
I know Canon and Nikon have left AFMA in their MILCs which seems like it shouldn't need to be there but it seems this lens would do well with a -10 adjustment on the R4.
Sony has AFMA in the menu but it only works when using a lens adapter with its own focus sensor.
I did not know Canon and Nikon MILCs have AFMA. What is it used for? How is it calibrated?
dclark wrote:
Sony has AFMA in the menu but it only works when using a lens adapter with its own focus sensor.
I did not know Canon and Nikon MILCs have AFMA. What is it used for? How is it calibrated?
dclark wrote:
Sony has AFMA in the menu but it only works when using a lens adapter with its own focus sensor.
I did not know Canon and Nikon MILCs have AFMA. What is it used for? How is it calibrated?
Dave
Wouldn't it be great, as fast as these bodies focus, to be able to focus bracket with BIF?
There quick shots in succession. Post software could handle the slight movement between the shots. Just dreaming!
kimknapp wrote:
Wouldn't it be great, as fast as these bodies focus, to be able to focus bracket with BIF?
There quick shots in succession. Post software could handle the slight movement between the shots. Just dreaming!
Great idea! The problem with getting that implemented is none of the major mfg's ever seem to admit any of their offerings have AF issues!
kimknapp wrote:
Wouldn't it be great, as fast as these bodies focus, to be able to focus bracket with BIF?
There quick shots in succession. Post software could handle the slight movement between the shots. Just dreaming!
Canon's dual pixel raw allows some shift of focus.
Pls. check this out and also the comments in the link:
&t=148s
arbitrage wrote:
Yes so frustrating. I should now amend my post from May 14th....when I actually got those images back to the computer (from the 2-6 with 1.4TC) the critical hit rate was actually pretty bad. Every miss was back focus. Not a single front focused shot. And all just slightly back focused where the perch was almost always perfect and back of the bird. But eyes/head were always out. I guess stopping down for more DOF would help some??
The more and more I try this combo, the more I confirm that it is best in cloudy light. In bright light it usually back focuses by just enough to ruin most shots. TC makes this "feature" worse
Thinking about it, this combo would be the perfect thing for an AFMA feature. I know Canon and Nikon have left AFMA in their MILCs which seems like it shouldn't need to be there but it seems this lens would do well with a -10 adjustment on the R4....Show more →
Thanks for posting that gannis. While the images He posted leave a lot to be desired and may give the false impression of user error. Stuart is witnessing the problems we’re all experiencing. I think Geoff captured it in comparing the af system to the old 7dmkii. Similar inconsistencies with af. I’ve tried tweaking every setting from af priority to Af sensitivity and nothing seems to affect the keeper rate. Perhaps the most perplexing aspect that Stuart observes as do I and many others, it seems that the af is tracking the subject as evidenced by the little green squares lighting up in the evf over the target, but when reviewing the image the actual af point it frequently is way off target.
Sony hasn’t been much help other than telling me to send the lens in. It works fine on the a9 so that’s not the problem, it simply doesn’t play nice withe the a7riv.
I want to thank all in this thread. Very enlightening. I shoot the A9/7R3/200-600 and have had many similar focus issues with the 3 that the 4 is having. It to seems to perform better in overcast conditions. I''ve only been serious about BIF for a couple years and feel I'm still learning. Only had the Sony gear about a year after more than a decade with Nikon shooting mostly sports.
I have a question about minimum focus settings. I don't ever see them discussed. Can I assume it is always better to be in 10 meters to Infiniti setting when shooting BIF with the 200-600 ?
Bryan, generally it is assumed that the lens will be more accurate when it doesn't have to "rack" the full focus distance. Thus if one were shooting objects > 10m then one should select 10m - inf.
Back to the conversation regarding the AF issues with the a7riv + 200-600 combination. As Mr. James observed in his video, the AF points don't seem to correlate with what the photographer is seeing in the EVF. It's as though there is some sort of disconnect between what is being displayed and what the camera is selecting for AF.
Whereas the other afternoon, the combination seemed to be working fine, another trip to the beach following a sandpiper in the surf was simply exasperating. Working in "wide", the small green boxes were lighting up and following the sandpiper as it was slowly moving near the surf. One would expect that all of the images would be in "adequate" focus. Upon review in camera, the af points were no where near the subject in 50% of this sequence. The first two grabbed the sandpiper's legs and the other two were grabbing space. This was confirmed when viewed in A7info and I've included the four image sequence below...
Interesting! Let us hope Sony fixes this via a firmware update. I know a lot of folks use A7R4 for static shoot and an A9/A9II for action. I cannot justify owning 2 Sony cameras as i am heavily invested in Nikon gear. I was planning to sell the R4 and buy an A92 instead but the resolution on R4 is simply superb so, hoping Sony would fix this in the next FW update.
@armd - What is the software that you use to get the AF point info on the images?
There are several ways to determine the a7riv AF point. First, if you look at the image in camera and press on the 100% mag it supposedly magnifies to the af point. This correlates with the program that I used for the images, A7info written by Soens (www.soens.de/download). If you download it, please send a few $ his way. Other programs will accomplish this as well.
armd wrote:
As Mr. James observed in his video, the AF points don't seem to correlate with what the photographer is seeing in the EVF. It's as though there is some sort of disconnect between what is being displayed and what the camera is selecting for AF.
I'm using the same combo (R4 & 200-600) and I have investigated this effect in a few of my own bursts. My impression is that a lot of said "disconnect" is lag. Many shown focus areas correspond more to where the bird was in the previous image of a sequence.
My take on how a a burst on "Hi" works, is a sequence roughly like follows:
- take a preview image (for "Hi+" the last taken image is used instead)
- analyze the image to track the subject
- display the image together with the green rectangle signifying the found subject
- read the PDAF pixels with the area of the subject
- calculate the required focus compensation
- tell the AF motors to do their thing
- capture RAW image
- repeat, the sequence takes 1/8s
So there are significant delays between 1) taking the preview, 2) reading PDAF and 3) taking the image. When the subject is stationary within the view (so with perfect panning and IS disabled), the displayed focus area matches where the bird is ending up. But when the subject moves too much between two frames, the camera may try to focus where the bird no longer is. Either way, even if the camera finds the moved bird in the PDAF data and focuses perfectly, the area where it focused will be somewhat different from where it is in the image taken a tad later.
Of course this is only one of the effects that can soften the resulting image, but keeping the subject as stationary in the view as possible did seem to help. Just keeping the bird somewhere in view isn't good enough. When I had my own go at the beach, I noticed that my panning was particularly bad when trying to go low to the ground.
Some of us have had the same severe problems getting a bird in focus that is not moving quickly. I remember trying to photograph a Harlequin Duck moving slowly through the frame - but also waves (slight up and down) were at work too. If I recall I shot about 20 images in a burst and two were in focus...So my thinking is the background (or foreground) can affect focus on this combo - especially if there are horizontal lines (waves or branches) in the image. Others have reported that the Sony sensor gets "fooled" by vertical lines...
But then the same R4 works well with other lenses, including 100-400. What is strange is even adapted canon glass seems to be doing better than Sony 200-600.
Daran wrote:
I'm using the same combo (R4 & 200-600) and I have investigated this effect in a few of my own bursts. My impression is that a lot of said "disconnect" is lag. Many shown focus areas correspond more to where the bird was in the previous image of a sequence.
My take on how a a burst on "Hi" works, is a sequence roughly like follows:
- take a preview image (for "Hi+" the last taken image is used instead)
- analyze the image to track the subject
- display the image together with the green rectangle signifying the found subject
- read the PDAF pixels with the area of the subject
- calculate the required focus compensation
- tell the AF motors to do their thing
- capture RAW image
- repeat, the sequence takes 1/8s
So there are significant delays between 1) taking the preview, 2) reading PDAF and 3) taking the image. When the subject is stationary within the view (so with perfect panning and IS disabled), the displayed focus area matches where the bird is ending up. But when the subject moves too much between two frames, the camera may try to focus where the bird no longer is. Either way, even if the camera finds the moved bird in the PDAF data and focuses perfectly, the area where it focused will be somewhat different from where it is in the image taken a tad later.
Of course this is only one of the effects that can soften the resulting image, but keeping the subject as stationary in the view as possible did seem to help. Just keeping the bird somewhere in view isn't good enough. When I had my own go at the beach, I noticed that my panning was particularly bad when trying to go low to the ground.
rdcny wrote:
Some of us have had the same severe problems getting a bird in focus that is not moving quickly. I remember trying to photograph a Harlequin Duck moving slowly through the frame - but also waves (slight up and down) were at work too. If I recall I shot about 20 images in a burst and two were in focus...So my thinking is the background (or foreground) can affect focus on this combo - especially if there are horizontal lines (waves or branches) in the image. Others have reported that the Sony sensor gets "fooled" by vertical lines...
Yes...in brighter light the combo has back focus issues even on a perched bird. Yet the 600GM has no issues at all on the R4.
gannis wrote:
But then the same R4 works well with other lenses, including 100-400. What is strange is even adapted canon glass seems to be doing better than Sony 200-600.
My best guess is it is the combination of the slower focus motor in the 200-600 combined with only 20 AF calculations per second of the A7RIV.
My 600GM with its dual-linear motors may be making up for "only" 20 calculations. Alternatively the 60 AF calculations of the A9 may be making up for the slower focus motor in the 200-600.
But that said, I still see issues with contrasty, bright light and practically still subjects. And why changing the OSS mode off of 1 makes a difference is beyond me. Unless that is just more taxing on the computer system and slowing down AF also??
gannis wrote:
But then the same R4 works well with other lenses, including 100-400. What is strange is even adapted canon glass seems to be doing better than Sony 200-600.
I am not so sure about that. I have photos taken with that combo with small flexible spot that is entirely within the bird and the background ended up being in perfect focus.
Also, in harsh light, I have had problems focusing the 600GM occasionally.
gannis wrote:
But then the same R4 works well with other lenses, including 100-400. What is strange is even adapted canon glass seems to be doing better than Sony 200-600.
Therein lies the rub. If the issue is related to the relative speed of the 200-600’s af motors, then it shouldn’t perform well on the a9, and yet it does. If the af challenges were related to limitations in af calculations on the a7riv then why would adapted lenses such as a Canon 500 f/4 is ii perform better with or without a tc and not manifest the degree of misfocus/back focus or tracking problems? To me, this suggests some control or algorithm issues rather than primary lens/camera deficiencies.
kimknapp wrote:
I am not so sure about that. I have photos taken with that combo with small flexible spot that is entirely within the bird and the background ended up being in perfect focus.
Also, in harsh light, I have had problems focusing the 600GM occasionally.
Focusing on the background is a problem, but it is a different problem. What we are discussing is the focusing on nothing problem. It is focusing not on target or background and is just slightly OOF.
As to the problem you have mentioned, I suspect the shot occurred during the delay between having acquired the background and locked in on it and reacquiring the bird which is now under the AF point. Was the next frame in focus?
billsnature wrote:
Focusing on the background is a problem, but it is a different problem. What we are discussing is the focusing on nothing problem. It is focusing not on target or background and is just slightly OOF.
As to the problem you have mentioned, I suspect the shot occurred during the delay between having acquired the background and locked in on it and reacquiring the bird which is now under the AF point. Was the next frame in focus?
Bill
Actually, everything was stationary.
I am camping right now, so I can't put up the image, but like you said, it may not relate to the problems discussed on this thread.
