X-T5 and "closer to raw" histogram

[DJF55]

So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx.

What makes you think they’re OOC JPEGs? Given that Erik generally knows his onions and he’s showing the output from an application called RawDigger, not JPEGDigger, I would infer that the first image has blown highlights in the raw file. The second appears “underexposed” of course, but both are simply the result of ETTR using the in-camera histogram, which is based on output from the JPEG pipeline, not the raw data. An implicit consideration of ETTR is generally having to manually adjust tone curves during processing.

Exactly!

Sorry , but Jeff's explanation is still not clear to me. Neither Erik's. Based on advice from 2 Fujifilm reps, I shoot RAW ETTR watching highlights with NLV set to On since over 6 years, with no issues with overexposure. And no, I do not intend to try to convince anyone on this forum about this point.

I do not really care about someone who allegedly "knows his onions" - I make my own experiences and have over time realized that I know a lot when reading help inputs here.

But still, the odd advice pops up, which I try and adopt if useful for my use. Even from "knows his onions" OR " knows her onions" someone ;-)

 

[Erik Baumgartner]

So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx.

What makes you think they’re OOC JPEGs? Given that Erik generally knows his onions and he’s showing the output from an application called RawDigger, not JPEGDigger, I would infer that the first image has blown highlights in the raw file. The second appears “underexposed” of course, but both are simply the result of ETTR using the in-camera histogram, which is based on output from the JPEG pipeline, not the raw data. An implicit consideration of ETTR is generally having to manually adjust tone curves during processing.

Exactly!

Sorry , but Jeff's explanation is still not clear to me. Neither Erik's. Based on advice from 2 Fujifilm reps, I shoot RAW ETTR watching highlights with NLV set to On since over 6 years, with no issues with overexposure.

I think the point here is, you’re watching the highlights with NLV on, and not using the NLV blinkies to manually set exposure. Just trusting the camera to get it right with the NLV view isn’t going to result in a different outcome than with NLV off, but using NLV mode to actively set exposure, will result in overexposure in many daylight situations.

And no, I do not intend to try to convince anyone on this forum about this point.

I do not really care about someone who allegedly "knows his onions" - I make my own experiences and have over time realized that I know a lot when reading help inputs here.

But still, the odd advice pops up, which I try and adopt if useful for my use. Even from "knows his onions" OR " knows her onions" someone ;-)

--
DJF

 

[DJF55]

So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx.

What makes you think they’re OOC JPEGs? Given that Erik generally knows his onions and he’s showing the output from an application called RawDigger, not JPEGDigger, I would infer that the first image has blown highlights in the raw file. The second appears “underexposed” of course, but both are simply the result of ETTR using the in-camera histogram, which is based on output from the JPEG pipeline, not the raw data. An implicit consideration of ETTR is generally having to manually adjust tone curves during processing.

Exactly!

Sorry , but Jeff's explanation is still not clear to me. Neither Erik's. Based on advice from 2 Fujifilm reps, I shoot RAW ETTR watching highlights with NLV set to On since over 6 years, with no issues with overexposure.

I think the point here is, you’re watching the highlights with NLV on, and not using the NLV blinkies to manually set exposure. Just trusting the camera to get it right with the NLV view isn’t going to result in a different outcome than with NLV off, but using NLV mode to actively set exposure, will result in overexposure in many daylight situations.

And no, I do not intend to try to convince anyone on this forum about this point.

I do not really care about someone who allegedly "knows his onions" - I make my own experiences and have over time realized that I know a lot when reading help inputs here.

But still, the odd advice pops up, which I try and adopt if useful for my use. Even from "knows his onions" OR " knows her onions" someone ;-)

 

[Erik Baumgartner]

So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx.

What makes you think they’re OOC JPEGs? Given that Erik generally knows his onions and he’s showing the output from an application called RawDigger, not JPEGDigger, I would infer that the first image has blown highlights in the raw file. The second appears “underexposed” of course, but both are simply the result of ETTR using the in-camera histogram, which is based on output from the JPEG pipeline, not the raw data. An implicit consideration of ETTR is generally having to manually adjust tone curves during processing.

Exactly!

Sorry , but Jeff's explanation is still not clear to me. Neither Erik's. Based on advice from 2 Fujifilm reps, I shoot RAW ETTR watching highlights with NLV set to On since over 6 years, with no issues with overexposure.

I think the point here is, you’re watching the highlights with NLV on, and not using the NLV blinkies to manually set exposure. Just trusting the camera to get it right with the NLV view isn’t going to result in a different outcome than with NLV off, but using NLV mode to actively set exposure, will result in overexposure in many daylight situations.

And no, I do not intend to try to convince anyone on this forum about this point.

I do not really care about someone who allegedly "knows his onions" - I make my own experiences and have over time realized that I know a lot when reading help inputs here.

But still, the odd advice pops up, which I try and adopt if useful for my use. Even from "knows his onions" OR " knows her onions" someone ;-)

ETTR implies in my view primarily using the histogramm, aided by the blinkies when required.

…and you manually adjust exposure so that the brightest important highlight detail is just below blinking using NLV?

And you take every opportunity to voice your NLV Off mantra. Why?

Because NLV doesn’t work as a basis for optimally setting RAW (or jpeg) exposure, and anyone pushing its usage as reliable exposure aid when they’ve been proved wrong on multiple occasions annoys me. I’m pretty tired of people living under a cloud total BS being paraded around as fact these days, with no critical thinking/common sense being applied when someone presents actual facts that are in conflict with their alternate reality.

 

[DJF55]

So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx.

What makes you think they’re OOC JPEGs? Given that Erik generally knows his onions and he’s showing the output from an application called RawDigger, not JPEGDigger, I would infer that the first image has blown highlights in the raw file. The second appears “underexposed” of course, but both are simply the result of ETTR using the in-camera histogram, which is based on output from the JPEG pipeline, not the raw data. An implicit consideration of ETTR is generally having to manually adjust tone curves during processing.

Exactly!

Sorry , but Jeff's explanation is still not clear to me. Neither Erik's. Based on advice from 2 Fujifilm reps, I shoot RAW ETTR watching highlights with NLV set to On since over 6 years, with no issues with overexposure.

I think the point here is, you’re watching the highlights with NLV on, and not using the NLV blinkies to manually set exposure. Just trusting the camera to get it right with the NLV view isn’t going to result in a different outcome than with NLV off, but using NLV mode to actively set exposure, will result in overexposure in many daylight situations.

And no, I do not intend to try to convince anyone on this forum about this point.

I do not really care about someone who allegedly "knows his onions" - I make my own experiences and have over time realized that I know a lot when reading help inputs here.

But still, the odd advice pops up, which I try and adopt if useful for my use. Even from "knows his onions" OR " knows her onions" someone ;-)

ETTR implies in my view primarily using the histogramm, aided by the blinkies when required.

…and you manually adjust exposure so that the brightest important highlight detail is just below blinking using NLV?

And you take every opportunity to voice your NLV Off mantra. Why?

Because NLV doesn’t work as a basis for optimally setting RAW (or jpeg) exposure, and anyone pushing its usage as reliable exposure aid when they’ve been proved wrong on multiple occasions annoys me. I’m pretty tired of people living under a cloud total BS being paraded around as fact these days, with no critical thinking/common sense being applied when someone presents actual facts that are in conflict with their alternate reality.

Big words there.... So what Fujifilm reps and others say on this topic is total BS, so no "onions knowledge" ? Oh well... I am not pushing anything, you are, and I am just telling the audience that I think you are wrong. If you recall, I have adviced all users that are interested in this topic to thourougly test this setting for their style of shooting. Again, and for the last time, the tests that you have published on this forum are in my view not conclusive to prove your point.

---
DJF

 

[Ysarex]

So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx.

What makes you think they’re OOC JPEGs? Given that Erik generally knows his onions and he’s showing the output from an application called RawDigger, not JPEGDigger, I would infer that the first image has blown highlights in the raw file. The second appears “underexposed” of course, but both are simply the result of ETTR using the in-camera histogram, which is based on output from the JPEG pipeline, not the raw data. An implicit consideration of ETTR is generally having to manually adjust tone curves during processing.

Exactly!

Sorry , but Jeff's explanation is still not clear to me. Neither Erik's. Based on advice from 2 Fujifilm reps, I shoot RAW ETTR watching highlights with NLV set to On since over 6 years, with no issues with overexposure.

I think the point here is, you’re watching the highlights with NLV on, and not using the NLV blinkies to manually set exposure. Just trusting the camera to get it right with the NLV view isn’t going to result in a different outcome than with NLV off, but using NLV mode to actively set exposure, will result in overexposure in many daylight situations.

And no, I do not intend to try to convince anyone on this forum about this point.

I do not really care about someone who allegedly "knows his onions" - I make my own experiences and have over time realized that I know a lot when reading help inputs here.

But still, the odd advice pops up, which I try and adopt if useful for my use. Even from "knows his onions" OR " knows her onions" someone ;-)

ETTR implies in my view primarily using the histogramm, aided by the blinkies when required.

…and you manually adjust exposure so that the brightest important highlight detail is just below blinking using NLV?

And you take every opportunity to voice your NLV Off mantra. Why?

Because NLV doesn’t work as a basis for optimally setting RAW (or jpeg) exposure, and anyone pushing its usage as reliable exposure aid when they’ve been proved wrong on multiple occasions annoys me. I’m pretty tired of people living under a cloud total BS being paraded around as fact these days, with no critical thinking/common sense being applied when someone presents actual facts that are in conflict with their alternate reality.

Big words there.... So what Fujifilm reps and others say on this topic is total BS, so no "onions knowledge" ? Oh well... I am not pushing anything, you are, and I am just telling the audience that I think you are wrong. If you recall, I have adviced all users that are interested in this topic to thourougly test this setting for their style of shooting. Again, and for the last time, the tests that you have published on this forum are in my view not conclusive to prove your point.

Eric is entirely correct. If you're going to trust the camera's highlight alert then NLV will lead to unrecoverable highlight clipping and trashed raw files. Trusting the camera's highlight alert is a great idea. With a digital raw file the best possible IQ comes from the maximum amount of sensor exposure and the highlight alert is the ideal tool to determine that. It immediately finds the brightest highlight in the scene and it's much more precise than the camera histogram. In Fuji cameras the highlight alert activates when any two of the RGB channels in the EVF JPEG simulation clip.

I put the camera (X-T4) on a tripod. The weather is heavy overcast with periods of rain. Apart from specular highlights in the chrome and glass of the parked cars the brightest diffuse highlight is in the sky. I activated NLV and started adding EC until the highlight alert came on in the sky. I then backed down 1/3 stop (no alert) and took the photo. Here's that raw histogram:



The raw file is unusable with all three raw file channels highlight-clipped. Yet highlight alert in NLV showed no diffuse highlight clipping.

I then turned NLV off. Highlight alert immediately came on all over the screen. I reduced the exposure until the highlight alert in the sky went out -- 1 full stop!! and I took the photo. Here's that raw histogram:



The raw file is exposed as much as possible and fully useable. Specular highlights in the green channel are clipped and a small section of sky but there is no clipping in either the red or blue channels. Specular highlights should be clipped and the small piece of sky is reconstructable from the red and blue channels. That's a best possible usable exposure.

And here's the photo if you want to see it processed of course from the raw file that wasn't destroyed by NLV.

 

[DJF55]

So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx.

What makes you think they’re OOC JPEGs? Given that Erik generally knows his onions and he’s showing the output from an application called RawDigger, not JPEGDigger, I would infer that the first image has blown highlights in the raw file. The second appears “underexposed” of course, but both are simply the result of ETTR using the in-camera histogram, which is based on output from the JPEG pipeline, not the raw data. An implicit consideration of ETTR is generally having to manually adjust tone curves during processing.

Exactly!

Sorry , but Jeff's explanation is still not clear to me. Neither Erik's. Based on advice from 2 Fujifilm reps, I shoot RAW ETTR watching highlights with NLV set to On since over 6 years, with no issues with overexposure.

I think the point here is, you’re watching the highlights with NLV on, and not using the NLV blinkies to manually set exposure. Just trusting the camera to get it right with the NLV view isn’t going to result in a different outcome than with NLV off, but using NLV mode to actively set exposure, will result in overexposure in many daylight situations.

And no, I do not intend to try to convince anyone on this forum about this point.

I do not really care about someone who allegedly "knows his onions" - I make my own experiences and have over time realized that I know a lot when reading help inputs here.

But still, the odd advice pops up, which I try and adopt if useful for my use. Even from "knows his onions" OR " knows her onions" someone ;-)

ETTR implies in my view primarily using the histogramm, aided by the blinkies when required.

…and you manually adjust exposure so that the brightest important highlight detail is just below blinking using NLV?

And you take every opportunity to voice your NLV Off mantra. Why?

Because NLV doesn’t work as a basis for optimally setting RAW (or jpeg) exposure, and anyone pushing its usage as reliable exposure aid when they’ve been proved wrong on multiple occasions annoys me. I’m pretty tired of people living under a cloud total BS being paraded around as fact these days, with no critical thinking/common sense being applied when someone presents actual facts that are in conflict with their alternate reality.

Big words there.... So what Fujifilm reps and others say on this topic is total BS, so no "onions knowledge" ? Oh well... I am not pushing anything, you are, and I am just telling the audience that I think you are wrong. If you recall, I have adviced all users that are interested in this topic to thourougly test this setting for their style of shooting. Again, and for the last time, the tests that you have published on this forum are in my view not conclusive to prove your point.

Eric is entirely correct. If you're going to trust the camera's highlight alert then NLV will lead to unrecoverable highlight clipping and trashed raw files. Trusting the camera's highlight alert is a great idea. With a digital raw file the best possible IQ comes from the maximum amount of sensor exposure and the highlight alert is the ideal tool to determine that. It immediately finds the brightest highlight in the scene and it's much more precise than the camera histogram. In Fuji cameras the highlight alert activates when any two of the RGB channels in the EVF JPEG simulation clip.

I put the camera (X-T4) on a tripod. The weather is heavy overcast with periods of rain. Apart from specular highlights in the chrome and glass of the parked cars the brightest diffuse highlight is in the sky. I activated NLV and started adding EC until the highlight alert came on in the sky. I then backed down 1/3 stop (no alert) and took the photo. Here's that raw histogram:



The raw file is unusable with all three raw file channels highlight-clipped. Yet highlight alert in NLV showed no diffuse highlight clipping.

I then turned NLV off. Highlight alert immediately came on all over the screen. I reduced the exposure until the highlight alert in the sky went out -- 1 full stop!! and I took the photo. Here's that raw histogram:



The raw file is exposed as much as possible and fully useable. Specular highlights in the green channel are clipped and a small section of sky but there is no clipping in either the red or blue channels. Specular highlights should be clipped and the small piece of sky is reconstructable from the red and blue channels. That's a best possible usable exposure.

And here's the photo if you want to see it processed of course from the raw file that wasn't destroyed by NLV.

You seem to overlook, or ignore? what I wrote in the 1st paragraph here:

https://www.dpreview.com/forums/post/68494490

---
DJF

 

[Ysarex]

So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx.

What makes you think they’re OOC JPEGs? Given that Erik generally knows his onions and he’s showing the output from an application called RawDigger, not JPEGDigger, I would infer that the first image has blown highlights in the raw file. The second appears “underexposed” of course, but both are simply the result of ETTR using the in-camera histogram, which is based on output from the JPEG pipeline, not the raw data. An implicit consideration of ETTR is generally having to manually adjust tone curves during processing.

Exactly!

Sorry , but Jeff's explanation is still not clear to me. Neither Erik's. Based on advice from 2 Fujifilm reps, I shoot RAW ETTR watching highlights with NLV set to On since over 6 years, with no issues with overexposure.

I think the point here is, you’re watching the highlights with NLV on, and not using the NLV blinkies to manually set exposure. Just trusting the camera to get it right with the NLV view isn’t going to result in a different outcome than with NLV off, but using NLV mode to actively set exposure, will result in overexposure in many daylight situations.

And no, I do not intend to try to convince anyone on this forum about this point.

I do not really care about someone who allegedly "knows his onions" - I make my own experiences and have over time realized that I know a lot when reading help inputs here.

But still, the odd advice pops up, which I try and adopt if useful for my use. Even from "knows his onions" OR " knows her onions" someone ;-)

ETTR implies in my view primarily using the histogramm, aided by the blinkies when required.

…and you manually adjust exposure so that the brightest important highlight detail is just below blinking using NLV?

And you take every opportunity to voice your NLV Off mantra. Why?

Because NLV doesn’t work as a basis for optimally setting RAW (or jpeg) exposure, and anyone pushing its usage as reliable exposure aid when they’ve been proved wrong on multiple occasions annoys me. I’m pretty tired of people living under a cloud total BS being paraded around as fact these days, with no critical thinking/common sense being applied when someone presents actual facts that are in conflict with their alternate reality.

Big words there.... So what Fujifilm reps and others say on this topic is total BS, so no "onions knowledge" ? Oh well... I am not pushing anything, you are, and I am just telling the audience that I think you are wrong. If you recall, I have adviced all users that are interested in this topic to thourougly test this setting for their style of shooting. Again, and for the last time, the tests that you have published on this forum are in my view not conclusive to prove your point.

Eric is entirely correct. If you're going to trust the camera's highlight alert then NLV will lead to unrecoverable highlight clipping and trashed raw files. Trusting the camera's highlight alert is a great idea. With a digital raw file the best possible IQ comes from the maximum amount of sensor exposure and the highlight alert is the ideal tool to determine that. It immediately finds the brightest highlight in the scene and it's much more precise than the camera histogram. In Fuji cameras the highlight alert activates when any two of the RGB channels in the EVF JPEG simulation clip.

I put the camera (X-T4) on a tripod. The weather is heavy overcast with periods of rain. Apart from specular highlights in the chrome and glass of the parked cars the brightest diffuse highlight is in the sky. I activated NLV and started adding EC until the highlight alert came on in the sky. I then backed down 1/3 stop (no alert) and took the photo. Here's that raw histogram:



The raw file is unusable with all three raw file channels highlight-clipped. Yet highlight alert in NLV showed no diffuse highlight clipping.

I then turned NLV off. Highlight alert immediately came on all over the screen. I reduced the exposure until the highlight alert in the sky went out -- 1 full stop!! and I took the photo. Here's that raw histogram:



The raw file is exposed as much as possible and fully useable. Specular highlights in the green channel are clipped and a small section of sky but there is no clipping in either the red or blue channels. Specular highlights should be clipped and the small piece of sky is reconstructable from the red and blue channels. That's a best possible usable exposure.

And here's the photo if you want to see it processed of course from the raw file that wasn't destroyed by NLV.

You seem to overlook, or ignore? what I wrote in the 1st paragraph here:

https://www.dpreview.com/forums/post/68494490

This is the first paragraph: "So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx."

I showed readable RawDigger graphs that clearly identify how the camera behaves. They unquestionably backup what Eric explained takes place.

 

[michaeladawson]

Thank you for your service!

I am totally stumped over why some people can’t see Erik’s easy to understand explanation and your easy to see test.

 

[DJF55]

So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx.

What makes you think they’re OOC JPEGs? Given that Erik generally knows his onions and he’s showing the output from an application called RawDigger, not JPEGDigger, I would infer that the first image has blown highlights in the raw file. The second appears “underexposed” of course, but both are simply the result of ETTR using the in-camera histogram, which is based on output from the JPEG pipeline, not the raw data. An implicit consideration of ETTR is generally having to manually adjust tone curves during processing.

Exactly!

Sorry , but Jeff's explanation is still not clear to me. Neither Erik's. Based on advice from 2 Fujifilm reps, I shoot RAW ETTR watching highlights with NLV set to On since over 6 years, with no issues with overexposure.

I think the point here is, you’re watching the highlights with NLV on, and not using the NLV blinkies to manually set exposure. Just trusting the camera to get it right with the NLV view isn’t going to result in a different outcome than with NLV off, but using NLV mode to actively set exposure, will result in overexposure in many daylight situations.

And no, I do not intend to try to convince anyone on this forum about this point.

I do not really care about someone who allegedly "knows his onions" - I make my own experiences and have over time realized that I know a lot when reading help inputs here.

But still, the odd advice pops up, which I try and adopt if useful for my use. Even from "knows his onions" OR " knows her onions" someone ;-)

ETTR implies in my view primarily using the histogramm, aided by the blinkies when required.

…and you manually adjust exposure so that the brightest important highlight detail is just below blinking using NLV?

And you take every opportunity to voice your NLV Off mantra. Why?

Because NLV doesn’t work as a basis for optimally setting RAW (or jpeg) exposure, and anyone pushing its usage as reliable exposure aid when they’ve been proved wrong on multiple occasions annoys me. I’m pretty tired of people living under a cloud total BS being paraded around as fact these days, with no critical thinking/common sense being applied when someone presents actual facts that are in conflict with their alternate reality.

Big words there.... So what Fujifilm reps and others say on this topic is total BS, so no "onions knowledge" ? Oh well... I am not pushing anything, you are, and I am just telling the audience that I think you are wrong. If you recall, I have adviced all users that are interested in this topic to thourougly test this setting for their style of shooting. Again, and for the last time, the tests that you have published on this forum are in my view not conclusive to prove your point.

Eric is entirely correct. If you're going to trust the camera's highlight alert then NLV will lead to unrecoverable highlight clipping and trashed raw files. Trusting the camera's highlight alert is a great idea. With a digital raw file the best possible IQ comes from the maximum amount of sensor exposure and the highlight alert is the ideal tool to determine that. It immediately finds the brightest highlight in the scene and it's much more precise than the camera histogram. In Fuji cameras the highlight alert activates when any two of the RGB channels in the EVF JPEG simulation clip.

I put the camera (X-T4) on a tripod. The weather is heavy overcast with periods of rain. Apart from specular highlights in the chrome and glass of the parked cars the brightest diffuse highlight is in the sky. I activated NLV and started adding EC until the highlight alert came on in the sky. I then backed down 1/3 stop (no alert) and took the photo. Here's that raw histogram:



The raw file is unusable with all three raw file channels highlight-clipped. Yet highlight alert in NLV showed no diffuse highlight clipping.

I then turned NLV off. Highlight alert immediately came on all over the screen. I reduced the exposure until the highlight alert in the sky went out -- 1 full stop!! and I took the photo. Here's that raw histogram:



The raw file is exposed as much as possible and fully useable. Specular highlights in the green channel are clipped and a small section of sky but there is no clipping in either the red or blue channels. Specular highlights should be clipped and the small piece of sky is reconstructable from the red and blue channels. That's a best possible usable exposure.

And here's the photo if you want to see it processed of course from the raw file that wasn't destroyed by NLV.

You seem to overlook, or ignore? what I wrote in the 1st paragraph here:

https://www.dpreview.com/forums/post/68494490

This is the first paragraph: "So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx."

I showed readable RawDigger graphs that clearly identify how the camera behaves. They unquestionably backup what Eric explained takes place.

No. I wrote, as I referenced in the first paragraph: "ETTR implies in my view primarily using the histogramm, aided by the blinkies when required."

More attention to what I wrote would spare you time and effort. I give up with you pros.

---
DJF

 

[Ysarex]

So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx.

What makes you think they’re OOC JPEGs? Given that Erik generally knows his onions and he’s showing the output from an application called RawDigger, not JPEGDigger, I would infer that the first image has blown highlights in the raw file. The second appears “underexposed” of course, but both are simply the result of ETTR using the in-camera histogram, which is based on output from the JPEG pipeline, not the raw data. An implicit consideration of ETTR is generally having to manually adjust tone curves during processing.

Exactly!

Sorry , but Jeff's explanation is still not clear to me. Neither Erik's. Based on advice from 2 Fujifilm reps, I shoot RAW ETTR watching highlights with NLV set to On since over 6 years, with no issues with overexposure.

I think the point here is, you’re watching the highlights with NLV on, and not using the NLV blinkies to manually set exposure. Just trusting the camera to get it right with the NLV view isn’t going to result in a different outcome than with NLV off, but using NLV mode to actively set exposure, will result in overexposure in many daylight situations.

And no, I do not intend to try to convince anyone on this forum about this point.

I do not really care about someone who allegedly "knows his onions" - I make my own experiences and have over time realized that I know a lot when reading help inputs here.

But still, the odd advice pops up, which I try and adopt if useful for my use. Even from "knows his onions" OR " knows her onions" someone ;-)

ETTR implies in my view primarily using the histogramm, aided by the blinkies when required.

…and you manually adjust exposure so that the brightest important highlight detail is just below blinking using NLV?

And you take every opportunity to voice your NLV Off mantra. Why?

Because NLV doesn’t work as a basis for optimally setting RAW (or jpeg) exposure, and anyone pushing its usage as reliable exposure aid when they’ve been proved wrong on multiple occasions annoys me. I’m pretty tired of people living under a cloud total BS being paraded around as fact these days, with no critical thinking/common sense being applied when someone presents actual facts that are in conflict with their alternate reality.

Big words there.... So what Fujifilm reps and others say on this topic is total BS, so no "onions knowledge" ? Oh well... I am not pushing anything, you are, and I am just telling the audience that I think you are wrong. If you recall, I have adviced all users that are interested in this topic to thourougly test this setting for their style of shooting. Again, and for the last time, the tests that you have published on this forum are in my view not conclusive to prove your point.

Eric is entirely correct. If you're going to trust the camera's highlight alert then NLV will lead to unrecoverable highlight clipping and trashed raw files. Trusting the camera's highlight alert is a great idea. With a digital raw file the best possible IQ comes from the maximum amount of sensor exposure and the highlight alert is the ideal tool to determine that. It immediately finds the brightest highlight in the scene and it's much more precise than the camera histogram. In Fuji cameras the highlight alert activates when any two of the RGB channels in the EVF JPEG simulation clip.

I put the camera (X-T4) on a tripod. The weather is heavy overcast with periods of rain. Apart from specular highlights in the chrome and glass of the parked cars the brightest diffuse highlight is in the sky. I activated NLV and started adding EC until the highlight alert came on in the sky. I then backed down 1/3 stop (no alert) and took the photo. Here's that raw histogram:



The raw file is unusable with all three raw file channels highlight-clipped. Yet highlight alert in NLV showed no diffuse highlight clipping.

I then turned NLV off. Highlight alert immediately came on all over the screen. I reduced the exposure until the highlight alert in the sky went out -- 1 full stop!! and I took the photo. Here's that raw histogram:



The raw file is exposed as much as possible and fully useable. Specular highlights in the green channel are clipped and a small section of sky but there is no clipping in either the red or blue channels. Specular highlights should be clipped and the small piece of sky is reconstructable from the red and blue channels. That's a best possible usable exposure.

And here's the photo if you want to see it processed of course from the raw file that wasn't destroyed by NLV.

You seem to overlook, or ignore? what I wrote in the 1st paragraph here:

https://www.dpreview.com/forums/post/68494490

This is the first paragraph: "So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx."

I showed readable RawDigger graphs that clearly identify how the camera behaves. They unquestionably backup what Eric explained takes place.

No. I wrote, as I referenced in the first paragraph: "ETTR implies in my view primarily using the histogramm, aided by the blinkies when required."

The camera histogram is equally negatively affected by NLV and will get you to the exact same error as the highlight alert does but with less precision. I checked when I did the above test. When the highlight alert indicated no diffuse highlight clipping the histogram likewise indicated no highlight clipping -- ALL CHANNELS IN FACT WERE CLIPPED.

Beyond that your view of ETTR is faulty. The histogram your exposing to the right of is the one you see in RawDigger -- the histogram for the raw file. The tools you use on the camera to achieve that result only matter to the extent that you're successful or not.

More attention to what I wrote would spare you time and effort. I give up with you pros.

 

[Ysarex]

So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx.

What makes you think they’re OOC JPEGs? Given that Erik generally knows his onions and he’s showing the output from an application called RawDigger, not JPEGDigger, I would infer that the first image has blown highlights in the raw file. The second appears “underexposed” of course, but both are simply the result of ETTR using the in-camera histogram, which is based on output from the JPEG pipeline, not the raw data. An implicit consideration of ETTR is generally having to manually adjust tone curves during processing.

Exactly!

Sorry , but Jeff's explanation is still not clear to me. Neither Erik's. Based on advice from 2 Fujifilm reps, I shoot RAW ETTR watching highlights with NLV set to On since over 6 years, with no issues with overexposure.

I think the point here is, you’re watching the highlights with NLV on, and not using the NLV blinkies to manually set exposure. Just trusting the camera to get it right with the NLV view isn’t going to result in a different outcome than with NLV off, but using NLV mode to actively set exposure, will result in overexposure in many daylight situations.

And no, I do not intend to try to convince anyone on this forum about this point.

I do not really care about someone who allegedly "knows his onions" - I make my own experiences and have over time realized that I know a lot when reading help inputs here.

But still, the odd advice pops up, which I try and adopt if useful for my use. Even from "knows his onions" OR " knows her onions" someone ;-)

ETTR implies in my view primarily using the histogramm, aided by the blinkies when required.

…and you manually adjust exposure so that the brightest important highlight detail is just below blinking using NLV?

And you take every opportunity to voice your NLV Off mantra. Why?

Because NLV doesn’t work as a basis for optimally setting RAW (or jpeg) exposure, and anyone pushing its usage as reliable exposure aid when they’ve been proved wrong on multiple occasions annoys me. I’m pretty tired of people living under a cloud total BS being paraded around as fact these days, with no critical thinking/common sense being applied when someone presents actual facts that are in conflict with their alternate reality.

Big words there.... So what Fujifilm reps and others say on this topic is total BS, so no "onions knowledge" ? Oh well... I am not pushing anything, you are, and I am just telling the audience that I think you are wrong. If you recall, I have adviced all users that are interested in this topic to thourougly test this setting for their style of shooting. Again, and for the last time, the tests that you have published on this forum are in my view not conclusive to prove your point.

Eric is entirely correct. If you're going to trust the camera's highlight alert then NLV will lead to unrecoverable highlight clipping and trashed raw files. Trusting the camera's highlight alert is a great idea. With a digital raw file the best possible IQ comes from the maximum amount of sensor exposure and the highlight alert is the ideal tool to determine that. It immediately finds the brightest highlight in the scene and it's much more precise than the camera histogram. In Fuji cameras the highlight alert activates when any two of the RGB channels in the EVF JPEG simulation clip.

I put the camera (X-T4) on a tripod. The weather is heavy overcast with periods of rain. Apart from specular highlights in the chrome and glass of the parked cars the brightest diffuse highlight is in the sky. I activated NLV and started adding EC until the highlight alert came on in the sky. I then backed down 1/3 stop (no alert) and took the photo. Here's that raw histogram:



The raw file is unusable with all three raw file channels highlight-clipped. Yet highlight alert in NLV showed no diffuse highlight clipping.

I then turned NLV off. Highlight alert immediately came on all over the screen. I reduced the exposure until the highlight alert in the sky went out -- 1 full stop!! and I took the photo. Here's that raw histogram:



The raw file is exposed as much as possible and fully useable. Specular highlights in the green channel are clipped and a small section of sky but there is no clipping in either the red or blue channels. Specular highlights should be clipped and the small piece of sky is reconstructable from the red and blue channels. That's a best possible usable exposure.

And here's the photo if you want to see it processed of course from the raw file that wasn't destroyed by NLV.

You seem to overlook, or ignore? what I wrote in the 1st paragraph here:

https://www.dpreview.com/forums/post/68494490

This is the first paragraph: "So why does he only show OOC jpeg images, if that is what he is doing, and not jpegs exported from developed RAW images? That would defeat the purpose of shooting RAW, wouldn't it? And, please show exif and readable RawDigger graphs. Thx."

I showed readable RawDigger graphs that clearly identify how the camera behaves. They unquestionably backup what Eric explained takes place.

No. I wrote, as I referenced in the first paragraph: "ETTR implies in my view primarily using the histogramm, aided by the blinkies when required."

I shot it for you. Sorry I don't have a good way to get an image of the camera screen but these have the info you need visible.

First screen shot NLV is on and you can see the histogram display on the screen -- no highlight alert active.



No change to the exposure settings and no change to the image -- all I did is switch NLV off. And we have highlight alert flashing and look at those histograms; green channel nuked, red channel nuked, and luminosity histogram indicating nuked. Camera is telling a very different story now. And the NLV story is the false one.



Took the photo and here's the RawDigger histogram with three clipped channels:

More attention to what I wrote would spare you time and effort. I give up with you pros.

 

[Feugi]

We have gone through this a million times. Erik has the patience of a Saint to deal with the trolling DJF55 keeps putting on.

Tons of evidence has been presented proving DJF55 wrong. It's been done many times before yet we keep going in circles.

https://www.dpreview.com/forums/thread/4774185?page=3#:~:text=3-,The%20results%20are%20in,-.%20Erik%20was%20right

I'm not a religious man, but God please give me the strength to keep my sanity in this thread.

 

[Jeff Biscuits]

I shot it for you. Sorry I don't have a good way to get an image of the camera screen but these have the info you need visible.

First screen shot NLV is on and you can see the histogram display on the screen -- no highlight alert active.

Your results threw me for a moment until it dawned on me you were showing the live view, not the JPEG Clearly slow waking up this morning.

Anyway, FWIW I grabbed a camera and did a similar test.

Image 1: from my low contrast B&W setting (Acros yellow, highlights -2), the highest exposure before blinking cut in. Raw highlights well intact.



Image 2: as per image 1 but 1/3 stop more exposure, blinkies just appearing on the far edge of the keyboard. Raw highlights still intact. I suspect that with vanilla Monochrome or Eterna (and highlights -2 of course) you might get a slightly narrower gap between the raw and JPEG top ends, but even in Acros Y this is close enough to be useful.



Image 3: as per image 1, but with NLV on. This needed an extra 2 1/3 stops to get it to the “not quite blinkies” point. Highlights destroyed. The image, histogram and (lack of) blinkies in the NLV view all indicated (if one were to trust it) that highlights were fine.



Image 4: as per image 2, but with NLV on. Obviously, highlights a little more destroyed than before.

 

[Erik Baumgartner]

I shot it for you. Sorry I don't have a good way to get an image of the camera screen but these have the info you need visible.

First screen shot NLV is on and you can see the histogram display on the screen -- no highlight alert active.

Your results threw me for a moment until it dawned on me you were showing the live view, not the JPEG Clearly slow waking up this morning.

Anyway, FWIW I grabbed a camera and did a similar test.

Image 1: from my low contrast B&W setting (Acros yellow, highlights -2), the highest exposure before blinking cut in. Raw highlights well intact.



Image 2: as per image 1 but 1/3 stop more exposure, blinkies just appearing on the far edge of the keyboard. Raw highlights still intact. I suspect that with vanilla Monochrome or Eterna (and highlights -2 of course) you might get a slightly narrower gap between the raw and JPEG top ends, but even in Acros Y this is close enough to be useful.



Image 3: as per image 1, but with NLV on. This needed an extra 2 1/3 stops to get it to the “not quite blinkies” point. Highlights destroyed. The image, histogram and (lack of) blinkies in the NLV view all indicated (if one were to trust it) that highlights were fine.



Image 4: as per image 2, but with NLV on. Obviously, highlights a little more destroyed than before.

The NLV view can almost seem to work OK if used in a low contrast indoor situation which, I think, has a lot to do with why some folks started pushing it as a great way to determine RAW exposure but, go outside with a bright sky and dark landscape, and you’ll find that you’re overexposing by two full stops.

It’s a shame that Fuji didn’t give us the natural live view, but with the histogram and blinky response of the actual underlying jpeg settings - a much better implementation for both RAW and jpeg shooters.

 

[mahidoes]

I know Eric might disagree, and he could very well be right, too.

But personally, I use Natural Live View set to 'On'. However, I don't check the combined histogram; instead, I use the RGB histogram (which I have a button assigned to). When using the RGB histogram, I see the individual channels.

I don't go right to the edge, but rather keep a third or two-thirds of a stop away because even a strobe or the sun's intensity can change slightly.

So far, I've been fine. It's easy for me because using custom film simulations can affect my JPEGs. Natural Live View simply shows the viewfinder without the simulation but applies the simulation to the JPEG file, which is exactly what I want. I'm not sshooting ETTR or Base ISO all the time. I don't want to fiddle with setting much!

https://www.youtube.com/watch?v=zIxF4pD_4AQ&t=15sThis video explain Natural live view and exposure preview nicely

--
You may subscribe on my substack and see pictures in insta.
https://mahidoes.substack.com/
https://www.instagram.com/mahidoes/
https://www.instagram.com/yuvaanaka/

 

[Truman Prevatt]

I'm aware that the histogram shown by the camera is created based not on the raw image but on a JPG created with the current camera settings, so the histogram clips earlier than what the raw image allows in terms of highlights.

On the X-T3 I had in the past I created a custom setting that lower saturation and contrast and, at the time, I left the film simulation to the standard Provia.

I'm not sure what to do with a few X-T5 settings like Clarity, Sharpness (maybe some of them were in X-T3, I don't remember), should I leave them to 0 or set them at the minimum value. Similarly, should I use the flatter Eterna instead of Provia?

I only shoot raw so what a JPG would look like with the camera settings is not important in itself.

Thank you.

I personally use Eterna with defaults for this purpose and find that while not a raw histo, it gives a good representation. I also keep blinkies turned on, why not? Erik is right about using Rawdigger to “calibrate” whichever setting you choose to help determine when to crank in EC.

 

[mahidoes]

Yes, he is.

One more time....

I shot these yesterday, both using the histogram/blinkies to set RAW exposure.

Have you tried Natural live view -> on and using RGB histogram and monitor individual channels instead of blinkies?

This is the method i'm using. i never tested as you did with Raw digger, but in reality i never had issue so far.

--
You may subscribe on my substack and see pictures in insta.
https://mahidoes.substack.com/
https://www.instagram.com/mahidoes/
https://www.instagram.com/yuvaanaka/

 

[Erik Baumgartner]

Yes, he is.

One more time....

I shot these yesterday, both using the histogram/blinkies to set RAW exposure.

Have you tried Natural live view -> on and using RGB histogram and monitor individual channels instead of blinkies?
This is the method i'm using. i never tested as you did but in reality i never had issue so far.

The only time I can think of that using the RGB histograms is really useful is for SOOC jpeg shooters where the histograms (only with NLV off) accurately track the SOOC jpeg RGB levels (which can be useful when shooting saturated flowers with Velvia).

As the RGB histograms are after the white balance has been set, it is completely useless for RAW exposure unless you recalibrate your camera for UniWB (you’ll be looking at a very green EVF and producing ugly green jpegs). None of that is really necessary for RAW exposure (IMO) as the green channel almost always clips first and the blinkies track the green channel very reliably with normal white balance.

Natural Live View (with its associated histograms and blinky response), unfortunately, doesn’t have anything to do with either RAW exposure or the clipping of SOOC jpeg color. It’s supposed to simulate an optical viewfinder experience (an OVF doesn’t tell you anything about exposure either), it’s a nice “real world” low contrast view but, unlike with NLV off, it unfortunately can’t help you with making informed decisions about how to tweak exposure in situations where the metering can’t be relied upon for getting an optimal exposure on its own.

For RAW shooters, I highly recommend using a low contrast film sim like ProNeg Standard or Eterna (tweaked to taste) for a similar NLV-like view without losing the accurate exposure preview, histogram and blinky response. Not great for jpeg shooters as you’re stuck with low contrast jpegs you probably won’t like.

 

[mahidoes]

For RAW shooters, I highly recommend using a low contrast film sim like ProNeg Standard or Eterna (tweaked to taste) for a similar NLV-like view without losing the accurate exposure preview, histogram and blinky response. Not great for jpeg shooters as you’re stuck with low contrast jpegs you probably won’t like.

let say i'm using Eterna with NLP -> off for ETTR Raw file.
still the histogram / blinkies influenced by white balance. aren't they?