Spot Metering changes ISO in a burst on Z9

If you would like to post, you'll need to register. Note that if you have a BCG store account, you'll need a new, separate account here (we keep the two sites separate for security purposes).

D

Dinusaur

Active member
Member's gallery
I am a bird photographer and I have been using Spot Metering on Nikon DSLRs (D810, D5, D6) and now on Z9 for many years now with success. The other day I was shooting a male Ruby-throated Hummingbird on a flowering plant with green background. In a burst of 10 shots I noticed that my ISO changed from 640 to 720 to 1400 and to 2500 even though the relative position of the bird (and hence the focus point that was on the bird) didn't change in any of the frames. I read Steve's crash course on Spot Metering and understand that since it is bigger than the size of the AF point, it is likely underexpose or overexpose depending on where it lands. However, when the subject position is relatively constant one should expect the exposure remaining the same over a burst of frames. By the way I was using 3D AF mode with 4.0 FW. Here are screen captures of two full frame photos (7th and 10th frame) from the burst, the focal point is shown as red rectangle. Any suggestions/comments?
ISO 640.JPG
You can only see EXIF info for this image if you are logged in.

ISO 2500.JPG
You can only see EXIF info for this image if you are logged in.
 
Member's gallery
Palouse said:
The humming birds wing beats are supper fast. Could it have out paced the camera's ability to keep up, since the wing has different tones on the top vs the bottom surface?
Click to expand...
Good suggestion; however, that is very unlikely as similar wing position a few frames earlier had ISO 640 and 1400. Also, why didn't the camera got back to 640 when the wing position changed, instead it progressively got worse.
 
Member's gallery
there is an option (I think) to have the camera use the initial exposure for the entire burst. Or go to manual. Suspect that this may be due to beating of the wings and cause area on the hummer to look darker than it is.
 
Member's gallery
Palouse said:
The humming birds wing beats are supper fast. Could it have out paced the camera's ability to keep up, since the wing has different tones on the top vs the bottom surface?
Click to expand...
I was thinking the same thing. Notice, between these two shots the darker image is when the wings are up and the hummer’s white breast is exposed to the spot metering. Hard to say without seeing a lot more shots but I expect the beating wings are having a big impact on spot metering.

P.S. seeing the reply to your suggestion, it’s hard to say what’s going on but personally I don’t use spot metering for any fast moving subjects as it’s so easy to slip a tiny bit and dramatically change the exposure. It’s great for static scenes or for metering when running full manual with manual ISO but I don’t use it for dynamic subjects in an auto exposure mode.
 
Member's gallery
I think it might be behaving like that because the area that the AF point is on is relatively dark (especially if the head of the bird is close enough to the AF point to be factored into the metering calculation). It appears that in the second shot the camera has lightened the body of the bird more towards middle gray.
 
Member's gallery
Dinusaur said:
Any suggestions/comments?
Click to expand...
One thought after looking at these again:

Assuming you had Subject Detection enabled it appears as though the 3D AF recognized the hummingbird as a subject but did not achieve eye detection in these images. If that's true and knowing how 3D AF with Subject Detection continually tries to reach the next level (e.g. first tries to detect subject, then head, then eyes) it's possible the AF algorithms were continuing to search and as it did the center of the spot metering point was continuing to jump between shots in the sequence. The fast beating wings likely made eye detection even harder.

Yeah, a lot of assumption buried in that thought based only on a couple of static images but we've all seen the green eye detect boxes jump around on small or low contrast targets as subject detection tries to find the eyes. knowing the spot metering follows the active AF point, that metering spot may have been jumping around between individual shots even though the overall subject detection box stayed in roughly the same place.

Maybe you didn't have Subject Detection enabled and something completely different was happening but if indeed the active AF point was searching as the camera attempted to detect a head or eye then I'd add that as one more reason to avoid spot metering on live action subjects when using Subject Detection as in that scenario the actual spot metering point may not be right where we think it is at a given moment.

Just speculating here...
 
Member's gallery
Thanks everyone for your suggestions and comments. I know it is not an exact science; however, I'll have another opportunity this week to photograph the hummingbird at the same location, I'll try out some of the suggestions and see if it makes any difference. In the meantime here are some more details about this shooting sequence that may help narrow the reason.
I started shooting when the bird appeared near the flower pot and stopped when it left the scene. First frame in the sequence was #4110, the time reads 11:40:23.27. Last frame in the sequence 4202 with time reading 11:40:29.30 - the whole sequence lasted 6 seconds. At Z9 frame rates I should've 120 frames; however, I got 92. What gives? There was 1 sec pause between frames 4150 and 4151 and half second pause between frames 4184 and 4185; possibly for recomposing. After accounting for this pause the math works out ok for 30 or so missing frames. Now here is the kicker. Between frames 4110 (first in the sequence) and 4192, ISO remained close to 640, with very little variation. Then the ISO started drifting from the next frame on (6th second into shooting). Here's how it reads:

Frame #4110 - #4192: ISO 640 with maximum hitting 800 in some frames.
Frame #4193 - #4194: ISO 1400
Frame #4195- #4197: ISO 2500
Frame #4198: ISO 3600
Frame #4199 - #4200: ISO 4500
Frame #4201 - #4202: ISO 5600

If anyone wants to look into last 20 frames of the sequence here's the link to my OneDrive with the photos.

Microsoft OneDrive

1drv.ms 1drv.ms

Wonder if 6 seconds of continuous shooting has anything to do with it.
 
Member's gallery
Dinusaur said:
I'll have another opportunity this week to photograph the hummingbird at the same location, I'll try out some of the suggestions and see if it makes any difference.
Click to expand...
FWIW, it's hard to do deep forensics in a relatively uncontrolled environment like shooting small fast moving subjects under natural light. So you may not pin down exactly what's happening without carefully controlled testing.

That said, if you'll shoot in similar circumstances and have multiple shooting opportunities I'd recommend shooting a sequence with Matrix Metering and perhaps a sequence with Center Weighted Metering. I suspect some or all of the variation in auto exposure is due to the metering mode choice. Even if you'll always use Spot Metering for your future work it may help you narrow down what's happening when you see dramatic exposure changes and help to understand things better.
 
Member's gallery
Dinusaur said:
Thanks everyone for your suggestions and comments. I know it is not an exact science; however, I'll have another opportunity this week to photograph the hummingbird at the same location, I'll try out some of the suggestions and see if it makes any difference. In the meantime here are some more details about this shooting sequence that may help narrow the reason.
I started shooting when the bird appeared near the flower pot and stopped when it left the scene. First frame in the sequence was #4110, the time reads 11:40:23.27. Last frame in the sequence 4202 with time reading 11:40:29.30 - the whole sequence lasted 6 seconds. At Z9 frame rates I should've 120 frames; however, I got 92. What gives? There was 1 sec pause between frames 4150 and 4151 and half second pause between frames 4184 and 4185; possibly for recomposing. After accounting for this pause the math works out ok for 30 or so missing frames. Now here is the kicker. Between frames 4110 (first in the sequence) and 4192, ISO remained close to 640, with very little variation. Then the ISO started drifting from the next frame on (6th second into shooting). Here's how it reads:

Frame #4110 - #4192: ISO 640 with maximum hitting 800 in some frames.
Frame #4193 - #4194: ISO 1400
Frame #4195- #4197: ISO 2500
Frame #4198: ISO 3600
Frame #4199 - #4200: ISO 4500
Frame #4201 - #4202: ISO 5600

If anyone wants to look into last 20 frames of the sequence here's the link to my OneDrive with the photos.

Microsoft OneDrive

1drv.ms 1drv.ms

Wonder if 6 seconds of continuous shooting has anything to do with it.
Click to expand...
Did this happen on multiple bursts. Was the trend always an increasing ISO or did the ISO fluctuate?
 
Member's gallery
The reflected light values under the spot are constantly changing as wing positions change and small changes occur in the bird's position relative to the spot. The camera, IIRC, is sampling the scene 120 times a second and is doing exactly what you have set it up to do.
 
Member's gallery
Shoot in full manual and you don't have to worry about stupid metering modes. So much easier. Humans are smart....cameras are dumb...why make your life more difficult relying on dumb camera decision making? Especially with MILC's WYSIWYG in the EVF it just is so much easier to do it yourself.
 
Member's gallery
I'd think about reverting to full manual. Matrix metering uses scene recognition for both exposure and also to support subject recognition.

Spot metering for a moving subject is going to have a lot of variation depending on where the spot metering is in the frame and what is under that area. If the bird is temporarily out of the AF box selected, your exposure will be wrong. The second exposure in the first post- while overly bright - seems to be a correct exposure of the bird.
 
Member's gallery
I vaguely remember reading somewhere a suggestion to turn off auto ISO when shooting bursts. That probably had to do with shooting manual, but it could be related.
 
Member's gallery
SandyW said:
I vaguely remember reading somewhere a suggestion to turn off auto ISO when shooting bursts. That probably had to do with shooting manual, but it could be related.
Click to expand...
i use auto iso and shoot bursts almost exclusively and haven't had difficulties and i typically do this on purpose because auto ISO provides me some benefits i prefer. HOWEVER, i use matrix, not spot so i'm metering to the scene, not what's under the af point.
 
Member's gallery
Dinusaur said:
I am a bird photographer and I have been using Spot Metering on Nikon DSLRs (D810, D5, D6) and now on Z9 for many years now with success. The other day I was shooting a male Ruby-throated Hummingbird on a flowering plant with green background. In a burst of 10 shots I noticed that my ISO changed from 640 to 720 to 1400 and to 2500 even though the relative position of the bird (and hence the focus point that was on the bird) didn't change in any of the frames. I read Steve's crash course on Spot Metering and understand that since it is bigger than the size of the AF point, it is likely underexpose or overexpose depending on where it lands. However, when the subject position is relatively constant one should expect the exposure remaining the same over a burst of frames. By the way I was using 3D AF mode with 4.0 FW. Here are screen captures of two full frame photos (7th and 10th frame) from the burst, the focal point is shown as red rectangle. Any suggestions/comments?
View attachment 64886
View attachment 64887
Click to expand...
Hi there.

I looked at the EXIFs of the photos again, and here is my un-scientific guess: the 2nd photo ( much brighter, +4EV) is using the metered luminance of the 1st photo IF these two are consecutive shots.

Kind regards

Oliver
 
Member's gallery
Quite truthfully I use auto ISO the majority of the time whether shooting bursts or whatever. That said there are times I turn it off, for example panos. If I was seeing something that I questioned I would include it in my "things to try to see if it makes a difference."
 
Member's gallery
Dinusaur said:
I am a bird photographer and I have been using Spot Metering on Nikon DSLRs (D810, D5, D6) and now on Z9 for many years now with success. The other day I was shooting a male Ruby-throated Hummingbird on a flowering plant with green background. In a burst of 10 shots I noticed that my ISO changed from 640 to 720 to 1400 and to 2500 even though the relative position of the bird (and hence the focus point that was on the bird) didn't change in any of the frames. I read Steve's crash course on Spot Metering and understand that since it is bigger than the size of the AF point, it is likely underexpose or overexpose depending on where it lands. However, when the subject position is relatively constant one should expect the exposure remaining the same over a burst of frames. By the way I was using 3D AF mode with 4.0 FW. Here are screen captures of two full frame photos (7th and 10th frame) from the burst, the focal point is shown as red rectangle. Any suggestions/comments?
View attachment 64886
View attachment 64887
Click to expand...
The first thing I do when shooting multiple shots of anything or focus stacking or Pano is to go fully manual ... 🦘
 
Member's gallery
John Navitsky said:
i use auto iso and shoot bursts almost exclusively and haven't had difficulties and i typically do this on purpose because auto ISO provides me some benefits i prefer. HOWEVER, i use matrix, not spot so i'm metering to the scene, not what's under the af point.
Click to expand...
Since I do mostly bird photography, what's under AF point is typically the bird and I would rather have the metering on the bird correct than the entire environment; that's the main reason I use spot metering. Why would the exposure value drift so much even when the AF point is mostly on the bird, I don't know - that's the mystery. I wished there's a way to see the area used for spot metering similar to the way we see focus area.
 
Member's gallery
Butlerkid said:
If the appropriate metering option is chosen, there is no reason (other than personal preference) to not use auto ISO when shooting bursts. All my wildlife shooting is done in bursts, in manual with auto ISO.
Click to expand...
unless the subject or background change. In this case the beating of the wings may give a different exposure. Or the background could change when as the bird moves around the scene (probably not here)
 
  • Like
Reactions: Roy
You must log in or register to reply here.
Back
Top