You probably don't know what ISO means

Hi,

It is very simple:

Maximising exposure give least noise in midtones/highlights.
  • Increasing ISO throws away midtone signal noise ratio and throws away highlight data but may give a small improvement in the darkest detail. With modern CMOS, that improvement is small.
  • Reducing exposure has exactly the same effect on midtones as raising ISO, but it offers additional protection for highlights.
Exposing to the right maximizes signal/noise ratio. But, it is limited by the need of cropping.

Raw converters do play a lot of foul games. Capture One shows highlights far to bright, while Lightroom can apply highlight reconstruction without giving any information about it.

If you want to understand what your camera/sensor is doing, the best way is to use a tool like RawDigger that shows actual raw data.

It is absolutely OK to ignore all this. The folks writing camera software and raw converters are no fools. It is quiet probable that the camera makes a decent choice for you!

If you want to understand the basics, this article is probably the most approachable one regarding that issue: http://www.photonstophotos.net/Emil Martinec/noise-p3.html

Best regards

Erik
https://www.dpreview.com/forums/create?type=discussion&forum=1067

This DPR article suggests that different cameras handle ISO differently. I think I remember it being said the GFX is ISO invariant (I don't think the 645z is ISO invariant) but I don't know how to translate that information into finding the least noisy settings as the article suggests doing.

Any thoughts or help for the befuddled?
If you want some perspective, you can't do any better than ETTR at base ISO. All this other stuff is about what to do when that's impractical.
Essentially, this is what I am doing via bracketing but this passage in the article had me doubting this:

there are times that exposing-to-the-right will result in noisier midtones than you want. In these situations, you have to let the highlights go. However, fixating on JPEG midtones isn't helpful.

This brings us to the biggest problem with using a clumsy metaphor for film sensitivity as the way of setting image brightness in digital: it means we aren’t given the tools to optimally expose our sensors.

ISO ends up conflating the effects of amplification and of tone curve, meaning you have to do your own research to find out what your camera’s doing behind the scenes, and what the best way to expose it is.


This post is my attempt at research and/or obtaining advice. Upon a second reading of the article, it isn't a "how to" article. It's a "beware of this" article.
 
Last edited:
The exposure the camera suggests is based on an average within 1/3rd stop of 18% gray in the image, or better yet an arbitrary (as the standard puts it) good exposure, for an image with color space assigned. Since raw files don't have color space, the exposure value provided by the camera is based on a processed file. I believe it's almost universally a jpeg, and probably srgb by default, but maybe tiff or otherwise? So your EC from a "proper" exposure, or your highlight blinkies, are based on a jpeg with baked in adjustments. It isn't representing the actual headroom you have in the raw file.

The other thing the excerpt seems to be saying really ignores iso itself. For a given scene with relatively bright highlights, if you expose for the mids, the amount of light the aperture & shutter settings allows through to the sensor leads to blown highlights. If instead you expose for the highlights, they will not be blown, but then the amount of light let through isn't sufficient for the signal in the mids to be sufficient, so they're noisier.
 
Hi,

It is very simple:

Maximising exposure give least noise in midtones/highlights.
  • Increasing ISO throws away midtone signal noise ratio and throws away highlight data but may give a small improvement in the darkest detail. With modern CMOS, that improvement is small.
  • Reducing exposure has exactly the same effect on midtones as raising ISO, but it offers additional protection for highlights.
Exposing to the right maximizes signal/noise ratio. But, it is limited by the need of cropping.

Raw converters do play a lot of foul games. Capture One shows highlights far to bright, while Lightroom can apply highlight reconstruction without giving any information about it.

If you want to understand what your camera/sensor is doing, the best way is to use a tool like RawDigger that shows actual raw data.

It is absolutely OK to ignore all this. The folks writing camera software and raw converters are no fools. It is quiet probable that the camera makes a decent choice for you!

If you want to understand the basics, this article is probably the most approachable one regarding that issue: http://www.photonstophotos.net/Emil Martinec/noise-p3.html

Best regards
I can't figure out if you don't read the posts or you do not understand them. Let's try it again: What is DPR talking about in this article:

https://www.dpreview.com/articles/8...on-t-know-what-iso-means-and-that-s-a-problem
Rick Knepper, post: 61474326, member: 1435575"]
https://www.dpreview.com/forums/create?type=discussion&forum=1067

This DPR article suggests that different cameras handle ISO differently. I think I remember it being said the GFX is ISO invariant (I don't think the 645z is ISO invariant) but I don't know how to translate that information into finding the least noisy settings as the article suggests doing.

Any thoughts or help for the befuddled?
If you want some perspective, you can't do any better than ETTR at base ISO. All this other stuff is about what to do when that's impractical.
Essentially, this is what I am doing via bracketing but this passage in the article had me doubting this:

there are times that exposing-to-the-right will result in noisier midtones than you want. In these situations, you have to let the highlights go. However, fixating on JPEG midtones isn't helpful.

This brings us to the biggest problem with using a clumsy metaphor for film sensitivity as the way of setting image brightness in digital: it means we aren’t given the tools to optimally expose our sensors.

ISO ends up conflating the effects of amplification and of tone curve, meaning you have to do your own research to find out what your camera’s doing behind the scenes, and what the best way to expose it is.


This post is my attempt at research and/or obtaining advice. Upon a second reading of the article, it isn't a "how to" article. It's a "beware of this" article.
 
The exposure the camera suggests is based on an average within 1/3rd stop of 18% gray in the image, or better yet an arbitrary (as the standard puts it) good exposure, for an image with color space assigned. Since raw files don't have color space, the exposure value provided by the camera is based on a processed file. I believe it's almost universally a jpeg, and probably srgb by default, but maybe tiff or otherwise? So your EC from a "proper" exposure, or your highlight blinkies, are based on a jpeg with baked in adjustments. It isn't representing the actual headroom you have in the raw file.

The other thing the excerpt seems to be saying really ignores iso itself. For a given scene with relatively bright highlights, if you expose for the mids, the amount of light the aperture & shutter settings allows through to the sensor leads to blown highlights. If instead you expose for the highlights, they will not be blown, but then the amount of light let through isn't sufficient for the signal in the mids to be sufficient, so they're noisier.
The link to the entire article from which the excerpt was taken is below. I inadvertently posted a bogus link in the OP. Basically, this article is warning readers that what you see in the histogram and/or blinkies may not be what you get with the RAW? That's kind of old news isn't it?

https://www.dpreview.com/articles/8...on-t-know-what-iso-means-and-that-s-a-problem

Interestingly, in LL's in-depth review of the Pentax 645z, Michael Reichmann had this to say:

Fortunately the 645z has real-time highlight warning blinkies in Live View mode, and this is taken directly from the sensor. In other words, if something blinks red in Live View it’s blown. If it doesn’t, it isn’t. When displayed along with the histogram in Live View (which also has a red line to the right when anything is blown), you will be able to make technically optimum exposures. (Some experiments have shown that the camera’s real-time indication of “blown” coincides almost exactly with that of the histograms in Lightroom or Rawdigger.
 
https://www.dpreview.com/forums/create?type=discussion&forum=1067

This DPR article suggests that different cameras handle ISO differently. I think I remember it being said the GFX is ISO invariant (I don't think the 645z is ISO invariant) but I don't know how to translate that information into finding the least noisy settings as the article suggests doing.

Any thoughts or help for the befuddled?
If you want some perspective, you can't do any better than ETTR at base ISO. All this other stuff is about what to do when that's impractical.
Essentially, this is what I am doing via bracketing but this passage in the article had me doubting this:

there are times that exposing-to-the-right will result in noisier midtones than you want. In these situations, you have to let the highlights go. However, fixating on JPEG midtones isn't helpful.

This brings us to the biggest problem with using a clumsy metaphor for film sensitivity as the way of setting image brightness in digital: it means we aren’t given the tools to optimally expose our sensors.

ISO ends up conflating the effects of amplification and of tone curve, meaning you have to do your own research to find out what your camera’s doing behind the scenes, and what the best way to expose it is.


This post is my attempt at research and/or obtaining advice. Upon a second reading of the article, it isn't a "how to" article. It's a "beware of this" article.
You have quoted three paragraphs. The first two sentences are on a different topic than the rest of the quote. Do you understand those sentences?

If your difficulty is with the rest of the bolded text and not the first two sentences, what part don't you understand?

Jim
 
https://www.dpreview.com/forums/create?type=discussion&forum=1067

This DPR article suggests that different cameras handle ISO differently. I think I remember it being said the GFX is ISO invariant (I don't think the 645z is ISO invariant) but I don't know how to translate that information into finding the least noisy settings as the article suggests doing.

Any thoughts or help for the befuddled?
If you want some perspective, you can't do any better than ETTR at base ISO. All this other stuff is about what to do when that's impractical.
Essentially, this is what I am doing via bracketing but this passage in the article had me doubting this:

there are times that exposing-to-the-right will result in noisier midtones than you want. In these situations, you have to let the highlights go. However, fixating on JPEG midtones isn't helpful.

This brings us to the biggest problem with using a clumsy metaphor for film sensitivity as the way of setting image brightness in digital: it means we aren’t given the tools to optimally expose our sensors.

ISO ends up conflating the effects of amplification and of tone curve, meaning you have to do your own research to find out what your camera’s doing behind the scenes, and what the best way to expose it is.


This post is my attempt at research and/or obtaining advice. Upon a second reading of the article, it isn't a "how to" article. It's a "beware of this" article.
You have quoted three paragraphs. The first two sentences are on a different topic than the rest of the quote. Do you understand those sentences?

If your difficulty is with the rest of the bolded text and not the first two sentences, what part don't you understand?

Jim
 
I find that highlights are often clipped or lost with today's MF sensors, but shadows have a ton of detail even when they look completely dark.

I try to underexpose a little these days for this reason.
 
https://www.dpreview.com/forums/create?type=discussion&forum=1067

This DPR article suggests that different cameras handle ISO differently. I think I remember it being said the GFX is ISO invariant (I don't think the 645z is ISO invariant) but I don't know how to translate that information into finding the least noisy settings as the article suggests doing.

Any thoughts or help for the befuddled?
If you want some perspective, you can't do any better than ETTR at base ISO. All this other stuff is about what to do when that's impractical.
Essentially, this is what I am doing via bracketing but this passage in the article had me doubting this:

there are times that exposing-to-the-right will result in noisier midtones than you want. In these situations, you have to let the highlights go. However, fixating on JPEG midtones isn't helpful.

This brings us to the biggest problem with using a clumsy metaphor for film sensitivity as the way of setting image brightness in digital: it means we aren’t given the tools to optimally expose our sensors.

ISO ends up conflating the effects of amplification and of tone curve, meaning you have to do your own research to find out what your camera’s doing behind the scenes, and what the best way to expose it is.


This post is my attempt at research and/or obtaining advice. Upon a second reading of the article, it isn't a "how to" article. It's a "beware of this" article.
You have quoted three paragraphs. The first two sentences are on a different topic than the rest of the quote. Do you understand those sentences?

If your difficulty is with the rest of the bolded text and not the first two sentences, what part don't you understand?
I'm good Jim. You answered my question. The rest was idle conversation.
I don't see how anything I've written here since this post helps answer your question about bracketing and the bold text above. If I have answered your question, please quote what I've said that was the answer, so I'll understand.

By the way, I find that bracketing is not a useful strategy for me, as it is practically incompatible with many of the other techniques that I use. The exception to that is bracketed captures for use in making HDR stacks.

Jim
 
I find that highlights are often clipped or lost with today's MF sensors, but shadows have a ton of detail even when they look completely dark.

I try to underexpose a little these days for this reason.
If you are getting clipped meaningful highlights in your raw files, you are not using an effective exposure strategy. What is your strategy?

Jim
 
I find that highlights are often clipped or lost with today's MF sensors, but shadows have a ton of detail even when they look completely dark.

I try to underexpose a little these days for this reason.
If you are getting clipped meaningful highlights in your raw files, you are not using an effective exposure strategy. What is your strategy?

Jim
 
I find that highlights are often clipped or lost with today's MF sensors, but shadows have a ton of detail even when they look completely dark.

I try to underexpose a little these days for this reason.
If you are getting clipped meaningful highlights in your raw files, you are not using an effective exposure strategy. What is your strategy?
I shoot and look at the histograms and try for a U shape. Is that a strategy?

I am trying to expose so that the highlights don't get clipped but it's a work in progress, and the result is that I am shooting a bit dark which seems to work.
What camera are you using? I'll assume a GFX, but if you're using something different, let me know. When you see a histogram with no clipping in the GFX live color histogram, does the shot exhibit clipping when you look at the four raw channels? If so, there are some possible reasons for that, but I'll wait to go into them until I know if that's indeed the case for you.

Jim
 
calirat, post: 61476124, member: 1261637"]
I find that highlights are often clipped or lost with today's MF sensors, but shadows have a ton of detail even when they look completely dark.

I try to underexpose a little these days for this reason.
If you are getting clipped meaningful highlights in your raw files, you are not using an effective exposure strategy. What is your strategy?
I shoot and look at the histograms and try for a U shape. Is that a strategy?

I am trying to expose so that the highlights don't get clipped but it's a work in progress, and the result is that I am shooting a bit dark which seems to work.
What camera are you using? I'll assume a GFX, but if you're using something different, let me know. When you see a histogram with no clipping in the GFX live color histogram, does the shot exhibit clipping when you look at the four raw channels? If so, there are some possible reasons for that, but I'll wait to go into them until I know if that's indeed the case for you.

Jim
 
https://www.dpreview.com/forums/create?type=discussion&forum=1067

This DPR article suggests that different cameras handle ISO differently. I think I remember it being said the GFX is ISO invariant (I don't think the 645z is ISO invariant) but I don't know how to translate that information into finding the least noisy settings as the article suggests doing.

Any thoughts or help for the befuddled?
If you want some perspective, you can't do any better than ETTR at base ISO. All this other stuff is about what to do when that's impractical.
Essentially, this is what I am doing via bracketing but this passage in the article had me doubting this:

there are times that exposing-to-the-right will result in noisier midtones than you want. In these situations, you have to let the highlights go. However, fixating on JPEG midtones isn't helpful.

This brings us to the biggest problem with using a clumsy metaphor for film sensitivity as the way of setting image brightness in digital: it means we aren’t given the tools to optimally expose our sensors.

ISO ends up conflating the effects of amplification and of tone curve, meaning you have to do your own research to find out what your camera’s doing behind the scenes, and what the best way to expose it is.


This post is my attempt at research and/or obtaining advice. Upon a second reading of the article, it isn't a "how to" article. It's a "beware of this" article.
You have quoted three paragraphs. The first two sentences are on a different topic than the rest of the quote. Do you understand those sentences?

If your difficulty is with the rest of the bolded text and not the first two sentences, what part don't you understand?
I'm good Jim. You answered my question. The rest was idle conversation.
I don't see how anything I've written here since this post helps answer your question about bracketing and the bold text above. If I have answered your question, please quote what I've said that was the answer, so I'll understand.
What did you mean by this then?

"If you want some perspective, you can't do any better than ETTR at base ISO."

Then everything I wrote after that was "idle conversation".
By the way, I find that bracketing is not a useful strategy for me, as it is practically incompatible with many of the other techniques that I use. The exception to that is bracketed captures for use in making HDR stacks.
Bracketing works perfectly for me.
--
Once you've done fifty, anything less is iffy.
 
Last edited:
https://www.dpreview.com/forums/create?type=discussion&forum=1067

This DPR article suggests that different cameras handle ISO differently. I think I remember it being said the GFX is ISO invariant (I don't think the 645z is ISO invariant) but I don't know how to translate that information into finding the least noisy settings as the article suggests doing.

Any thoughts or help for the befuddled?
If you want some perspective, you can't do any better than ETTR at base ISO. All this other stuff is about what to do when that's impractical.
Essentially, this is what I am doing via bracketing but this passage in the article had me doubting this:

there are times that exposing-to-the-right will result in noisier midtones than you want. In these situations, you have to let the highlights go. However, fixating on JPEG midtones isn't helpful.

This brings us to the biggest problem with using a clumsy metaphor for film sensitivity as the way of setting image brightness in digital: it means we aren’t given the tools to optimally expose our sensors.

ISO ends up conflating the effects of amplification and of tone curve, meaning you have to do your own research to find out what your camera’s doing behind the scenes, and what the best way to expose it is.


This post is my attempt at research and/or obtaining advice. Upon a second reading of the article, it isn't a "how to" article. It's a "beware of this" article.
You have quoted three paragraphs. The first two sentences are on a different topic than the rest of the quote. Do you understand those sentences?

If your difficulty is with the rest of the bolded text and not the first two sentences, what part don't you understand?
I'm good Jim. You answered my question. The rest was idle conversation.
I don't see how anything I've written here since this post helps answer your question about bracketing and the bold text above. If I have answered your question, please quote what I've said that was the answer, so I'll understand.
What did you mean by this then?

"If you want some perspective, you can't do any better than ETTR at base ISO."
Rick, your question came after that remark by me.
Then everything I wrote after that was "idle conversation".
By the way, I find that bracketing is not a useful strategy for me, as it is practically incompatible with many of the other techniques that I use. The exception to that is bracketed captures for use in making HDR stacks.
Bracketing works perfectly for me.
 
I find that highlights are often clipped or lost with today's MF sensors, but shadows have a ton of detail even when they look completely dark.

I try to underexpose a little these days for this reason.
If you are getting clipped meaningful highlights in your raw files, you are not using an effective exposure strategy. What is your strategy?

Jim

--
http://blog.kasson.com
I shoot and look at the histograms and try for a U shape. Is that a strategy?

I am trying to expose so that the highlights don't get clipped but it's a work in progress, and the result is that I am shooting a bit dark which seems to work.
Hi,

I would guess that camera histograms are a bit conservative. But, you would try to avoid having a spike at the right side of the histogram.

Modern CMOS sensors have very low readout noise. That means that that there is much more clean detail in the deepest shadows than on old generation CMOS or CCD.



The GFX has a modern CMOS sensor designed by Sony. Leica S (typ 007) has a CMOS sensor that is designed for that camera, while the Leica S (typ 006) has an older CCD sensor, going back to old Kodak designs.

The GFX has a modern CMOS sensor designed by Sony. Leica S (typ 007) has a CMOS sensor that is designed for that camera, while the Leica S (typ 006) has an older CCD sensor, going back to old Kodak designs.



 This is a perfect histogram. Note that green channel is just below clipping. There is a small single channel spike on the right, that represents overexposure on the white part of the light in front of the lighthouse.

This is a perfect histogram. Note that green channel is just below clipping. There is a small single channel spike on the right, that represents overexposure on the white part of the light in front of the lighthouse.

The darks reproduce reasonably well.

Also note that the raw histogram here is used with a logarithmic vertical axis. It has been produced by Rawdigger.

Best regards

Erik

--
Erik Kaffehr
Website: http://echophoto.dnsalias.net
Magic uses to disappear in controlled experiments…
Gallery: http://echophoto.smugmug.com
Articles: http://echophoto.dnsalias.net/ekr/index.php/photoarticles
 
https://www.dpreview.com/forums/create?type=discussion&forum=1067

This DPR article suggests that different cameras handle ISO differently. I think I remember it being said the GFX is ISO invariant (I don't think the 645z is ISO invariant) but I don't know how to translate that information into finding the least noisy settings as the article suggests doing.

Any thoughts or help for the befuddled?
If you want some perspective, you can't do any better than ETTR at base ISO. All this other stuff is about what to do when that's impractical.
Essentially, this is what I am doing via bracketing but this passage in the article had me doubting this:

there are times that exposing-to-the-right will result in noisier midtones than you want. In these situations, you have to let the highlights go. However, fixating on JPEG midtones isn't helpful.

This brings us to the biggest problem with using a clumsy metaphor for film sensitivity as the way of setting image brightness in digital: it means we aren’t given the tools to optimally expose our sensors.

ISO ends up conflating the effects of amplification and of tone curve, meaning you have to do your own research to find out what your camera’s doing behind the scenes, and what the best way to expose it is.


This post is my attempt at research and/or obtaining advice. Upon a second reading of the article, it isn't a "how to" article. It's a "beware of this" article.
You have quoted three paragraphs. The first two sentences are on a different topic than the rest of the quote. Do you understand those sentences?

If your difficulty is with the rest of the bolded text and not the first two sentences, what part don't you understand?
I'm good Jim. You answered my question. The rest was idle conversation.
I don't see how anything I've written here since this post helps answer your question about bracketing and the bold text above. If I have answered your question, please quote what I've said that was the answer, so I'll understand.
What did you mean by this then?

"If you want some perspective, you can't do any better than ETTR at base ISO."
Rick, your question came after that remark by me.
I don't see a question nor remember posting any additional questions nor did I mean to imply that I had further questions. I've long accepted ETTR as the best exposure method. Like I said, I'm good. Thanks.
Then everything I wrote after that was "idle conversation".
By the way, I find that bracketing is not a useful strategy for me, as it is practically incompatible with many of the other techniques that I use. The exception to that is bracketed captures for use in making HDR stacks.
Bracketing works perfectly for me.
 
calirat, post: 61476825, member: 1845234"]
I find that highlights are often clipped or lost with today's MF sensors, but shadows have a ton of detail even when they look completely dark.

I try to underexpose a little these days for this reason.
If you are getting clipped meaningful highlights in your raw files, you are not using an effective exposure strategy. What is your strategy?
I shoot and look at the histograms and try for a U shape. Is that a strategy?

I am trying to expose so that the highlights don't get clipped but it's a work in progress, and the result is that I am shooting a bit dark which seems to work.
What camera are you using? I'll assume a GFX, but if you're using something different, let me know. When you see a histogram with no clipping in the GFX live color histogram, does the shot exhibit clipping when you look at the four raw channels? If so, there are some possible reasons for that, but I'll wait to go into them until I know if that's indeed the case for you.

Jim

--
http://blog.kasson.com
645z. Thank you. Much appreciate any suggestions Jim.
[/QUOTE]
I'm not Jim but I own the 645z as well as the GFX. And, you've heard of Michael Reichmann of Luminous Landscape? In his review of the 645z, he makes some interesting statements regarding exposure and the histogram/blinkies which have born out in my own experience with the 645z i.e. keep doing what I think you are doing with the 645z.

https://luminous-landscape.com/pentax-645z-in-depth-review/

Excerpt:

Highlight Warning Disagreement

I will assume that most people interested in an advanced medium format camera understand the principal of ETTR (or Expose to The Right), which is a concept I first wrote about in the early days of digital – following an in depth explanation by Thomas Knoll, one of the inventors of Photoshop and also Camera Raw.

ETTR isn’t anywhere as critical today as it was ten or more years ago. Cameras have much more dynamic range and cleaner shadows. But, the so-called 255 brick wall is still there, where the pixel buckets overflow and all detail is lost if too many photons enter the well.

But using ETTR technique is still a very good shooting discipline and can gain you one or two stops of headroom in a shot, and thus up to an extra two stops of DR and cleaner shadow detail in a wide dynamic range situation.

The Pentax 645z has two primary means of showing you the appropriateness of either its selected exposure, or the one that you manually choose. The first is a histogram and the second is highlight blinking. This is available both while shooting (if in Live View) and also in exposure review.

Here’s the problem. As with almost all cameras the playback histogram and blinkies are based on an in-camera JPG representation of the image, and the colour space that that camera is set to. If you’re set to sRGB you may as well go home and watch TV, because the histogram will bear little useful relationship to the raw image. The sRGB colour space is so much smaller than the colour gamut of the camera that you’ll see highlights shown as blown when you still have two stops of headroom.

If the camera is set to Adobe RBG then you’re somewhat better off. Adobe RGB is a larger colour space than sRGB, but still vastly smaller than the camera’s colour gamut. So, in a word or two, the playback histogram is fairly useless at indicating overexposure, and if you follow it you will be wasting headroom, and thus have shadow areas that are much noisier than they might be otherwise.

Fortunately the 645z has real-time highlight warning blinkies in Live View mode, and this is taken directly from the sensor. In other words, if something blinks red in Live View it’s blown. If it doesn’t, it isn’t. When displayed along with the histogram in Live View (which also has a red line to the right when anything is blown), you will be able to make technically optimum exposures. (Some experiments have shown that the camera’s real-time indication of “blown” coincides almost exactly with that of the histograms in Lightroom or Rawdigger.

I wish though that instead of red blinkies the 645z had adjustable level zebras. They are a more useful and contemporary tool. But more to the point, I would suggest to Pentax that the camera’s raw colour space be added alongside sRGB and Adobe RGB for those of us (most of us?) who shoot raw. That way review histograms and blinkies will be more accurate.

--
Once you've done fifty, anything less is iffy.
 
Last edited:
calirat, post: 61478149, member: 183646"]
I find that highlights are often clipped or lost with today's MF sensors, but shadows have a ton of detail even when they look completely dark.

I try to underexpose a little these days for this reason.
If you are getting clipped meaningful highlights in your raw files, you are not using an effective exposure strategy. What is your strategy?
I shoot and look at the histograms and try for a U shape. Is that a strategy?

I am trying to expose so that the highlights don't get clipped but it's a work in progress, and the result is that I am shooting a bit dark which seems to work.
What camera are you using? I'll assume a GFX, but if you're using something different, let me know. When you see a histogram with no clipping in the GFX live color histogram, does the shot exhibit clipping when you look at the four raw channels? If so, there are some possible reasons for that, but I'll wait to go into them until I know if that's indeed the case for you.

Jim
 
calirat, post: 61476825, member: 1845234"]
I find that highlights are often clipped or lost with today's MF sensors, but shadows have a ton of detail even when they look completely dark.

I try to underexpose a little these days for this reason.
If you are getting clipped meaningful highlights in your raw files, you are not using an effective exposure strategy. What is your strategy?
I shoot and look at the histograms and try for a U shape. Is that a strategy?

I am trying to expose so that the highlights don't get clipped but it's a work in progress, and the result is that I am shooting a bit dark which seems to work.
What camera are you using? I'll assume a GFX, but if you're using something different, let me know. When you see a histogram with no clipping in the GFX live color histogram, does the shot exhibit clipping when you look at the four raw channels? If so, there are some possible reasons for that, but I'll wait to go into them until I know if that's indeed the case for you.
645z. Thank you. Much appreciate any suggestions Jim.
[/QUOTE]
I see Rick's given you some advice specific to that camera. I'll just go with generic adivce.

Since you're using a DSLR, you don't have a live histogram, except in live view. So, set up your shot, make a guess at the exposure, and make a test capture. Now play that capture back, and look at the color histogram. Are the buckets on the right side empty (zero)? If so, the histogram is showing you no clipping. Adjust the camera for the most generous exposure that still has those buckets empty. That's the quasi-ETTR exposure. Take that shot and upload it to a computer with RawDigger running on it. Now look at the raw histogram. Are the right-hand buckets close to empty (they probably won't be completely empty because of outlier pixels)? If so, then you're, to a first approximation, good to go. If not, send the raw file to me.

Lots more details here:


Jim
 

Keyboard shortcuts

Back
Top