Parariss wrote:
By "shutter slap," I was referring to the vibration in the camera (and therefore possible blurring of the image) caused by shutter actuation, not the noise.
Fred Miranda wrote:
Just came back from a children's party. It was in a gym with equipment that reminds of a circus. Pretty cool!
The place was very dim (as you can see from the ISO setting) and also had artificial lighting (fluorescent). There were windows around the gym which made exposure a nightmare.
My settings were:
A7R III with FE 55/1.8 ZA,
Metering "Highlight", Exposure comp +2 EV (+1EV internally and +1 on the Dial),
Continuous mode,
Compressed RAW,
"Hi" frames per second (8fps),
Lock-on: Flexible spot: "M" with a recall button to toggle it to Lock-on "Wide".
Electronic shutter for all images.
Custom button (C4) for APS-C crop on the fly. (In a few images)
I had a great rate of keepers (about 90%) which was great considering the lighting and venue. Here are a few shots:...Show more →
Excellent family photos., Fred. Still playing with the settings on
my new A7riii. I am assuming you chose for your auto-iso the max setting of 6400?
Al
The A7RIII and Voigtlander 40/1.2 arrived yesterday. So far, I've only had time for a little basic exploration, with initial setup and a few quick test shots, but I'm favorably impressed with both. More to follow after get into detailed settings, etc.
At the moment, I've set the camera for Fred's "automatic ETTR" process, but I'm getting uncorrectable burned-out highlights with 255 in all 3 channels. Seems that something isn't right, but I thought I've made the correct settings:
I notice that exposure changes drastically if a bright light source is in the center of the frame vs near the edge. Some of this would be due to increasing lens vignetting as the light source nears the frame edge, but not all of it (it's at least 3 stops different).
DannyBurkPhoto wrote:
The A7RIII and Voigtlander 40/1.2 arrived yesterday. So far, I've only had time for a little basic exploration, with initial setup and a few quick test shots, but I'm favorably impressed with both. More to follow after get into detailed settings, etc.
At the moment, I've set the camera for Fred's "automatic ETTR" process, but I'm getting uncorrectable burned-out highlights with 255 in all 3 channels. Seems that something isn't right, but I thought I've made the correct settings:
I notice that exposure changes drastically if a bright light source is in the center of the frame vs near the edge. Some of this would be due to increasing lens vignetting as the light source nears the frame edge, but not all of it (it's at least 3 stops different).
Parariss wrote:
Shutter speed max'd? The 40/1.2 can run into that in bright days if used wide.
Also, I assume you're measuring 255 off a RAW.
1) No, shutter speed was moderate (indoor lighting only).
2) Yes, measured from RAW in ACR.
By far the worst instance was shooting a predominantly dark-midtone range, but with some bright highlights. It seems that the camera based its metering on the dark areas and exposed them far to the right, but ignored the actual highlights and grossly overexposed them.
DannyBurkPhoto wrote:
By far the worst instance was shooting a predominantly dark-midtone range, but with some bright highlights. It seems that the camera based its metering on the dark areas and exposed them far to the right, but ignored the actual highlights and grossly overexposed them.
I've seen same thing happen with certain light situations where even with only +1,5 EV positive correction with highlight priority. Most of the time I can get very very nice/accurate ETTR with highlight priority with +2 EV correction, but with certain smallish highlights Rawdigger can show overexposure even with +1,5 EV.
Just want to give a quick shout out to the battery life on these things. Shot an event with strobes/tethered for 5 hours last night and I didn't even go through one battery. Such a great feeling after having the mark ii for so long!
tn1krr wrote:
I've seen same thing happen with certain light situations where even with only +1,5 EV positive correction with highlight priority. Most of the time I can get very very nice/accurate ETTR with highlight priority with +2 EV correction, but with certain smallish highlights Rawdigger can show overexposure even with +1,5 EV.
I've seen this as well. When the highlights are not predominant on a scene, I usually dial compensation to +1 to be on the safe side. However for most of my shooting, +1.7 or +2 has worked great.
Fred Miranda wrote:
I've seen this as well. When the highlights are not predominant on a scene, I usually dial compensation to +1 to be on the safe side. However for most of my shooting, +1.7 or +2 has worked great.
Your method should be perfect, Fred, if the camera actually meters strictly on the highlights...but it apparently doesn't! Maybe it's "priority" unless it feels like doing something differently.
DannyBurkPhoto wrote:
Your method should be perfect, Fred, if the camera actually meters strictly on the highlights...but it apparently doesn't! Maybe it's "priority" unless it feels like doing something differently.
Some days ago I did a interesting test in a sunny and contrasty day to compare the 3 main metering methods usefulness in that particular environment (very common in Spain :-). During a city walk, I took the main shoot carefully using "Entire Screen Average" metering and the zebras for the perfect ETTR exposure, in Av mode (so the EC goes registered in the EXIF data). Obviously, some minor specular highlights were sacrifized here and there, but the main intent was conservative. After this "artistic" shot I took a couple of additional quick shots (but still trying to frame the exact same scene) using Highlight priority and Multi metering modes, both with EC 0 so the camera metered scene was also registered in the EXIF data. I was patient enough to do this with all the shots (35 photos).
At home I compared the perfect ETTR exposure (achieved with certain manual EC in Average metering) with the other metering modes, calculating in a spreadsheet the EC they'd require to achieve the same exposure (same absolute exposure value). The results show that no metering mode is fine. One usually needs to dial EC to optimize the Av autoexposure. In fact, the Sunny 16 rule in manual mode is not only more accurate than the camera autoexposure modes, but saves you from dialing EC :-). Welcome again to the sad fact that not a single Digital SLR manufacturer has implemented to date a proper automated RAW ETTR exposure mode... and ONLY Sony gives the photographers a configurable manual RAW ETTR exposure mode (the zebras) albeit not configured from factory...
In summary, the Average metering required in average a +0.2 EC, while the Highlight priority needed +1.25EV. Only the Multi metering required an average of +0EV... but only a third of its shots were fine at around 0 EC (a 43% required a positive EC and other 23% a negative EC). The same random results applies to the other modes (except the Highlight priority, which never required negative EC).
Here is a capture of the data. Each row represents the same photo (my Average metering 1/3 and 2/3 EC values rounded to 0.3 and 0.7, but the other metering deducted values may end rounded to other close frontiers):
For using the zebras as a guide, I setup a custom zebra level of 108+ at the camera bare defaults. At home, all pictures were ETTR exposed as intended (by looking at the true RAW data). I usually have a picture profile enabled with HLG gamma, but for this test I disabled it, because otherwise every metering mode becomes about +0.5EV more conservative, and I wanted to generate "default" results here. And by the way, why the heck the entire camera metering depends on a active "picture profile" settings?. Any camera metering in a given mode should only be related to the scene lighting, or am I missing something?
Anyway I'm happy with the A7R3, thanks to the zebras fine control when you have time to meter (which I usually have).
With a Min EC of zero, it looks like Highlight mode is doing exactly what it's supposed to when it comes to protecting highlights. But, it's too conservative sometimes. Interesting analysis, thank you.
Parariss wrote:
With a Min EC of zero, it looks like Highlight mode is doing exactly what it's supposed to when it comes to protecting highlights. But, it's too conservative sometimes. Interesting analysis, thank you.
Because curiosity I just have studied the outlier shot requiring 0 EC in Highlight metering mode. It is the following one. At the left with such conservative exposure at 1/80 sec, and at the right the Multi metering variant, which yielded 1/30 sec at the same aperture and ISO (also at 0 EC):
So it was a very contrasty image with a huge area in the shadow, which fooled the Multi metering and also required -1/3 EC in Average metering. The shot was taken about 2 hours before sunset. This is the actual extent of the clipping in both RAW images (yes, even the Highlight metering with 0 EC has some minor clipping):
The RAW converter (Capture One) was unable to properly restore detail in the overexposed areas. And except for the sky, the colors are also severely affected in most of the burned area, which is not always the case, but is always on risk whenever there is clipping. Besides the exposure correction in the second image, to be comparable I applied a 75% highlights recovery on both conversions (I usually don't go beyond 50% to avoid unnatural looking). Here are 100% crops:
So even a Highlight Priority +1.3 EC is not a good ETTR exposure for this image. The conservative exposure (left image) already included a 0.1% clipped area:
How a +2 EC on Highlight Priority would have impacted this very same image?. We can simulate such setting with the tool I posted a few weeks ago, by moving the white point two stops to the left, manually selecting a white point of 4480 (for 14 bit RAW) using the -w switch while generating the clipping tiff. At the left EC+1.3 and at the right the EC+2.0 simulated from the conservative exposure:
Even the red channel would have clipped in the building wall (yellow = green + red) so the bricks would be almost totally blown to white there. Taking into account how good this sensor is recovering the shadows (except if there are skin tones on them) many times it has no sense to overexpose, even under high contrast. The full image here is pretty good in the shadows, despite the considerable recovery done. This is a crop from the conservative exposure, with the noise reduction set to 15/15 (instead of the C1 aggresive default 50/50):
In summary many times one can not rely on any camera metering mode (nor should use systematically Highlight metering + 0EC, which on average "underexposes" between 1-2 EV). Please Sony (and Canikon!) bring us a true ETTR metering mode!. Mirrorless has a full real time capture of the entire scene!. As users we shouldn't agree with the current situation if we want to change things (in fact it took 7 years to convince Canon to take care about dynamic range).
Fred Miranda wrote:
I've seen this as well. When the highlights are not predominant on a scene, I usually dial compensation to +1 to be on the safe side. However for most of my shooting, +1.7 or +2 has worked great.
I guess I'm lost! I thought you reduced your exposure to preserve highlights, then lifted shadow areas. How do you keep from clipping highlights when you increase exposure nearly two stops?
With a7R and a7RII, I reduce exposure between 0 and -1EV, and still find some instances where I've clipped the highlights in side lit or partially back lit scenes. Is the a7RIII so different in handling highlights/shadows?
I thought you reduced your exposure to preserve highlights, then lifted shadow areas. How do you keep from clipping highlights when you increase exposure nearly two stops?