Technical: Snow sparkles under a full moon

[GeorgianBay1939]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

The -2EV of EC also helped to prevent blown specular reflections.

Best compromise, I think. What do you think?

I shot about a dozen scenes with the above in mind. I didn't shoot a range of exposures because it was very chilly with the window open (-20ºC).

Here are three of the better ones. Yes, I know that I could do a lot better with a tripod and longer exposures at base ISO! Go full size to see the sparkles better.




orange cast is from my clearance light.











Advice appreciated, especially from those who have tried this sort of thing, day or night.

Many thanks!

PS You'll notice some moon flare in the above. Very difficult to see at night. I also shot the full moon. 400mm EFL, f/4, 1/1250 sec, IS0 125, EC=-2/3. Spot metered. LV ~ 14/15.

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[D Cox]

The exposure seems about right on the first one, but you really do need that tripod.

 

[GeorgianBay1939]

The exposure seems about right on the first one, but you really do need that tripod.

Agree. Next time, if not too cold.

Is my reasoning of choice of ISO and EC to maximize DR and minimize noise correct?

On a tripod: longer exposure will cut down on the noise and drop the ISO to increase DR.

I figure that I can get exposures of between 8 and 16 seconds (500/FL to avoid star trails), giving me 3 to 4 stops over the 1 second exposures in this trial. So I'll take ISO back to base 125 (2.5 stops from ISO 800) to increase DR and use 1-1.5 stops increased exposure to get better S/N.

Make sense? I'd just as soon have this figured out before heading out in the cold with camera on a tripod, hopefully February 22.

Many thanks

 

[gollywop]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

Sensorgen.org doesn't state DR for ISO 125, but it would appear to be closer to 10.5, and at ISO 800 it drops only to 9.5. Read noise, however, is significantly better at 800, so this seems indeed to be a good choice.

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

The -2EV of EC also helped to prevent blown specular reflections.

As you know, you can get rid of this issue by using a tripod, but the increase in SS is the right move if needed for hand holding.

Best compromise, I think. What do you think?

I shot about a dozen scenes with the above in mind. I didn't shoot a range of exposures because it was very chilly with the window open (-20ºC).

Here are three of the better ones. Yes, I know that I could do a lot better with a tripod and longer exposures at base ISO! Go full size to see the sparkles better.


orange cast is from my clearance light.





Advice appreciated, especially from those who have tried this sort of thing, day or night.

These effects are rather spectacular (specular?). You're definitely on the right track. If you keep the first, I'd clone out that contrail.

What do the raw files look like in RawDigger? How well have you done with the sparkle highlights?

The first is compositionally the best IMO. Having the snow covered rocks in the middle helps a great deal, and the fact that the snow is not sullied with footprints is also nice.

Very interesting work.

--
gollywop
http://g4.img-dpreview.com/D8A95C7DB3724EC094214B212FB1F2AF.jpg

 

[Great Bustard]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

No, more light on the sensor *always* results in less noise, even when the higher ISO setting has less electronic noise. But, as you note, greater exposure and/or higher ISO will result in more of the highlights being blown.

The -2EV of EC also helped to prevent blown specular reflections.

Best compromise, I think. What do you think?

I shot about a dozen scenes with the above in mind. I didn't shoot a range of exposures because it was very chilly with the window open (-20ºC).

Here are three of the better ones. Yes, I know that I could do a lot better with a tripod and longer exposures at base ISO! Go full size to see the sparkles better.


orange cast is from my clearance light.





Advice appreciated, especially from those who have tried this sort of thing, day or night.

Many thanks!

PS You'll notice some moon flare in the above. Very difficult to see at night. I also shot the full moon. 400mm EFL, f/4, 1/1250 sec, IS0 125, EC=-2/3. Spot metered. LV ~ 14/15.

 

[GeorgianBay1939]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

Sensorgen.org doesn't state DR for ISO 125, but it would appear to be closer to 10.5, and at ISO 800 it drops only to 9.5. Read noise, however, is significantly better at 800, so this seems indeed to be a good choice.

Thanks gollywop, I had forgotten how I "knew" that the DR of the FZ1000. I will have to review my understanding of noise again. For some reason, I am having difficulty connecting some of the dots. I will try to read Clarkvision to get up to speed.

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

The -2EV of EC also helped to prevent blown specular reflections.

As you know, you can get rid of this issue by using a tripod, but the increase in SS is the right move if needed for hand holding.

I am looking forward to the next trial.... with a tripod! The trick will be the usual: to get sufficient light from the shadows without blowing those speculars. noise vs DR compromise.

Best compromise, I think. What do you think?

I shot about a dozen scenes with the above in mind. I didn't shoot a range of exposures because it was very chilly with the window open (-20ºC).

Here are three of the better ones. Yes, I know that I could do a lot better with a tripod and longer exposures at base ISO! Go full size to see the sparkles better.


orange cast is from my clearance light.





Advice appreciated, especially from those who have tried this sort of thing, day or night.

These effects are rather spectacular (specular?). You're definitely on the right track. If you keep the first, I'd clone out that contrail.

I'm not keeping it except as an example of this trial. Did you notice the flare with a diameter same as the contrail.... mirror image on the left? And the orange colour from my clearance lights? Both would be difficult to fix in post.

What do the raw files look like in RawDigger? How well have you done with the sparkle highlights?

I haven't done that yet. I ran out of the trial version of RD and need to subscribe.

The first is compositionally the best IMO. Having the snow covered rocks in the middle helps a great deal, and the fact that the snow is not sullied with footprints is also nice.

Here are a couple more that are nice compositionally but brutal with noise and jiggle:

Very interesting work.

Yes. I find that trying challenging stuff like this motivates me to learn more from you experts! A nice regenerative feedback loop.

Many thanks for your help!

--
gollywop
http://g4.img-dpreview.com/D8A95C7DB3724EC094214B212FB1F2AF.jpg

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[GeorgianBay1939]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

No, more light on the sensor *always* results in less noise, even when the higher ISO setting has less electronic noise. But, as you note, greater exposure and/or higher ISO will result in more of the highlights being blown.

Yes. Thank you. I had it right in my head but wrote it incorrectly. My bad!

Thanks for your clarification.

And thank you for your patience in dealing with "equivalence-deniers". I, as a relative newbie, really appreciate your explanations.

The -2EV of EC also helped to prevent blown specular reflections.

Best compromise, I think. What do you think?

I shot about a dozen scenes with the above in mind. I didn't shoot a range of exposures because it was very chilly with the window open (-20ºC).

Here are three of the better ones. Yes, I know that I could do a lot better with a tripod and longer exposures at base ISO! Go full size to see the sparkles better.


orange cast is from my clearance light.





Advice appreciated, especially from those who have tried this sort of thing, day or night.

Many thanks!

PS You'll notice some moon flare in the above. Very difficult to see at night. I also shot the full moon. 400mm EFL, f/4, 1/1250 sec, IS0 125, EC=-2/3. Spot metered. LV ~ 14/15.

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[gollywop]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

Sensorgen.org doesn't state DR for ISO 125, but it would appear to be closer to 10.5, and at ISO 800 it drops only to 9.5. Read noise, however, is significantly better at 800, so this seems indeed to be a good choice.

Thanks gollywop, I had forgotten how I "knew" that the DR of the FZ1000. I will have to review my understanding of noise again. For some reason, I am having difficulty connecting some of the dots. I will try to read Clarkvision to get up to speed.

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

The -2EV of EC also helped to prevent blown specular reflections.

As you know, you can get rid of this issue by using a tripod, but the increase in SS is the right move if needed for hand holding.

I am looking forward to the next trial.... with a tripod! The trick will be the usual: to get sufficient light from the shadows without blowing those speculars. noise vs DR compromise.

I gather from your OP that you are actually trying to keep the specular highlights in check, and have chosen moonlight because it likely makes it possible. An interesting thought. But is it really necessary? You've got these very high luminance peaks towering above everything else, like the spikes in Brice Canyon, and you are trying to raise the water level to the top. But, if you lower the level down the peaks, at least for a while, they are still the only thing above the water level. Aren't you just as well off keeping the level lower rather than trying to cover the peaks? Won't the final result be effectively the same?

Best compromise, I think. What do you think?

I shot about a dozen scenes with the above in mind. I didn't shoot a range of exposures because it was very chilly with the window open (-20ºC).

Here are three of the better ones. Yes, I know that I could do a lot better with a tripod and longer exposures at base ISO! Go full size to see the sparkles better.


orange cast is from my clearance light.





Advice appreciated, especially from those who have tried this sort of thing, day or night.

These effects are rather spectacular (specular?). You're definitely on the right track. If you keep the first, I'd clone out that contrail.

I'm not keeping it except as an example of this trial. Did you notice the flare with a diameter same as the contrail.... mirror image on the left? And the orange colour from my clearance lights? Both would be difficult to fix in post.

Not till you mentioned it; it's not a big deal, but, clearly, you're not happy with it. And yes, both would be difficult to fix in post. But the clearance-light glow is definitely easy to fix in the future.

What do the raw files look like in RawDigger? How well have you done with the sparkle highlights?

I haven't done that yet. I ran out of the trial version of RD and need to subscribe.

Oh, for sure.

The first is compositionally the best IMO. Having the snow covered rocks in the middle helps a great deal, and the fact that the snow is not sullied with footprints is also nice.

Here are a couple more that are nice compositionally but brutal with noise and jiggle:

It's too bad these didn't meet your expectations, because, compositionally, they are the picks of the litter.

Very interesting work.

Yes. I find that trying challenging stuff like this motivates me to learn more from you experts! A nice regenerative feedback loop.

I built several regenerative receivers back in my youth. They were fun. But life really picked up with my first superhet (Halicrafters S-40B - I still have it).

Many thanks for your help!

--
gollywop
http://g4.img-dpreview.com/D8A95C7DB3724EC094214B212FB1F2AF.jpg

 

[Detail Man]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

It seems that there is very little (if any) need for retention of spatial resolution within the sparkles.

Everything (of course) looks better at a higher (sensor-level) exposure - but the more tangible concern (might) seem to be whether you could in processing make the tiny specular highlights (likely with a "clipped" Green RAW channel ?) look convincingly white in perceived color by viewers, and not take on a hard to grapple with "magenta" caste (from the non-clipped Red and Blue RAW channels information then, as a result being more prominent relative to Green ?

At any rate, such "gremlins" might likely be easily missed by those who eschew "pixel-peeping".

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

Sensorgen.org doesn't state DR for ISO 125, but it would appear to be closer to 10.5, and at ISO 800 it drops only to 9.5. Read noise, however, is significantly better at 800, so this seems indeed to be a good choice.

Thanks gollywop, I had forgotten how I "knew" that the DR of the FZ1000. I will have to review my understanding of noise again. For some reason, I am having difficulty connecting some of the dots. I will try to read Clarkvision to get up to speed.

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

The -2EV of EC also helped to prevent blown specular reflections.

As you know, you can get rid of this issue by using a tripod, but the increase in SS is the right move if needed for hand holding.

I am looking forward to the next trial.... with a tripod! The trick will be the usual: to get sufficient light from the shadows without blowing those speculars. noise vs DR compromise.

More mechanical stabilization (as well as whatever you electro-mechanical image-stabilization may yield in any single given recorded frame enables the longer shutter-time). However, not only will *everything* (including the peak highlights) be scaled to a higher multiple value, there is additional (true semiconductor type this time) random "shot noise" associated with semiconductor junction leakage-current flows, doubling in the value of random noise-variance every 10 Degrees C (or so).

The image-sensor junction temperature, and resulting random noise increases over the time spent scanning the image-data - causing the long exposure-time approach [of performing a true RAW-level dark-frame subtraction (under UI control in the GH2)] - which is a viable option (absent camera or subject-matter movements taking place during the longer exposure-time).

The "thermal" situation can/does trigger "flaky pixels" to "pop-on" (to go "hot"). Emil Martinec here.

Best compromise, I think. What do you think?

I shot about a dozen scenes with the above in mind. I didn't shoot a range of exposures because it was very chilly with the window open (-20ºC).

Here are three of the better ones. Yes, I know that I could do a lot better with a tripod and longer exposures at base ISO! Go full size to see the sparkles better.


orange cast is from my clearance light.





Advice appreciated, especially from those who have tried this sort of thing, day or night.

These effects are rather spectacular (specular?). You're definitely on the right track. If you keep the first, I'd clone out that contrail.

I'm not keeping it except as an example of this trial. Did you notice the flare with a diameter same as the contrail.... mirror image on the left? And the orange colour from my clearance lights? Both would be difficult to fix in post.

What do the raw files look like in RawDigger? How well have you done with the sparkle highlights?

I haven't done that yet. I ran out of the trial version of RD and need to subscribe.

Indeed - unless you want to download and play with the late Gabor Schorr's ("panopeeper" on the Luminous Landscape forums) "Rawnalyze" (last Version released was 2.10.4.0, for Win OS only) - which, while quite the item of interest and use in its time a while back now, using it is not unlike riding on a bare-bones "tricycle" after de-boarding a "777 jet-liner" (RawDigger):


Rawnalyze 2.10.4.0 - Written, developed, distributed by Gabor Schorr

.

The first is compositionally the best IMO. Having the snow covered rocks in the middle helps a great deal, and the fact that the snow is not sullied with footprints is also nice.

Here are a couple more that are nice compositionally but brutal with noise and jiggle:

Very interesting work.

Yes. I find that trying challenging stuff like this motivates me to learn more from you experts! A nice regenerative feedback loop.

Many thanks for your help!

 

[GeorgianBay1939]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

It seems that there is very little (if any) need for retention of spatial resolution within the sparkles.

Agreed.

Everything (of course) looks better at a higher (sensor-level) exposure - but the more tangible concern (might) seem to be whether you could in processing make the tiny specular highlights (likely with a "clipped" Green RAW channel ?) look convincingly white in perceived color by viewers, and not take on a hard to grapple with "magenta" caste (from the non-clipped Red and Blue RAW channels information then, as a result being more prominent relative to Green ?

I don't know if I even get to that stage. In the daytime, the biggest challenge seems to be the extreme difference in luminance of the speculars compared to the surrounding snow ... never mind the shadows! In the case of the difference between the speculars and the snow it is a compromise: How to make the speculars show at 255 (with a minimal level of oversaturation and "blooming") and for the slope of that end "cliff" to be gradual enough to let the texture of the snow show while not degrading it to a dull grey (beach-sand type of) snow. I am told that "Luminance tone mapping" or "luminosity masks" could be used to tweak the slope of that cliff on the luminosity histogram in post processing. I don't have Photoshop and its use would be a major addition in my meager tool kit .... and my meager brain.

At any rate, such "gremlins" might likely be easily missed by those who eschew "pixel-peeping".

Whereas I use pixel peeping to try to understand what is going on, I am not adverse to accepting some blooming (due to oversaturation of small speculars), some snow greyness, some loss of snow texture and some noise in the shadows. It seems to be a matter of "acceptable" image quality. The difficulty is that my standards keep moving. My sparkles and snow textures are incrementally better now than they were a few years ago . But sometimes I wonder.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

Sensorgen.org doesn't state DR for ISO 125, but it would appear to be closer to 10.5, and at ISO 800 it drops only to 9.5. Read noise, however, is significantly better at 800, so this seems indeed to be a good choice.

Thanks gollywop, I had forgotten how I "knew" that the DR of the FZ1000. I will have to review my understanding of noise again. For some reason, I am having difficulty connecting some of the dots. I will try to read Clarkvision to get up to speed.

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

The -2EV of EC also helped to prevent blown specular reflections.

As you know, you can get rid of this issue by using a tripod, but the increase in SS is the right move if needed for hand holding.

I am looking forward to the next trial.... with a tripod! The trick will be the usual: to get sufficient light from the shadows without blowing those speculars. noise vs DR compromise.

More mechanical stabilization (as well as whatever you electro-mechanical image-stabilization may yield in any single given recorded frame enables the longer shutter-time). However, not only will *everything* (including the peak highlights) be scaled to a higher multiple value, there is additional (true semiconductor type this time) random "shot noise" associated with semiconductor junction leakage-current flows, doubling in the value of random noise-variance every 10 Degrees C (or so).

Ok.

The image-sensor junction temperature, and resulting random noise increases over the time spent scanning the image-data - causing the long exposure-time approach [of performing a true RAW-level dark-frame subtraction (under UI control in the GH2)] - which is a viable option (absent camera or subject-matter movements taking place during the longer exposure-time).

I noticed that the FZ1000 automatically went to dark frame subtraction on the longer trial exposures the other night. Yet to read the manual on that feature.

The "thermal" situation can/does trigger "flaky pixels" to "pop-on" (to go "hot"). Emil Martinec here.

YES! I think that I might have some "hot" pixels in the sky ... in the lowest image in this post.???

Best compromise, I think. What do you think?

I shot about a dozen scenes with the above in mind. I didn't shoot a range of exposures because it was very chilly with the window open (-20ºC).

Here are three of the better ones. Yes, I know that I could do a lot better with a tripod and longer exposures at base ISO! Go full size to see the sparkles better.


orange cast is from my clearance light.





Advice appreciated, especially from those who have tried this sort of thing, day or night.

These effects are rather spectacular (specular?). You're definitely on the right track. If you keep the first, I'd clone out that contrail.

I'm not keeping it except as an example of this trial. Did you notice the flare with a diameter same as the contrail.... mirror image on the left? And the orange colour from my clearance lights? Both would be difficult to fix in post.

What do the raw files look like in RawDigger? How well have you done with the sparkle highlights?

I haven't done that yet. I ran out of the trial version of RD and need to subscribe.

Indeed - unless you want to download and play with the late Gabor Schorr's ("panopeeper" on the Luminous Landscape forums) "Rawnalyze" (last Version released was 2.10.4.0, for Win OS only) - which, while quite the item of interest and use in its time a while back now, using it is not unlike riding on a bare-bones "tricycle" after de-boarding a "777 jet-liner" (RawDigger):


Rawnalyze 2.10.4.0 - Written, developed, distributed by Gabor Schorr

.

No thanks. I am on a MacBookPro (with a large monitor) and really like RawDigger.

The first is compositionally the best IMO. Having the snow covered rocks in the middle helps a great deal, and the fact that the snow is not sullied with footprints is also nice.

Here are a couple more that are nice compositionally but brutal with noise and jiggle:

Very interesting work.

Yes. I find that trying challenging stuff like this motivates me to learn more from you experts! A nice regenerative feedback loop.

Many thanks for your help!

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[GeorgianBay1939]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

Sensorgen.org doesn't state DR for ISO 125, but it would appear to be closer to 10.5, and at ISO 800 it drops only to 9.5. Read noise, however, is significantly better at 800, so this seems indeed to be a good choice.

Thanks gollywop, I had forgotten how I "knew" that the DR of the FZ1000. I will have to review my understanding of noise again. For some reason, I am having difficulty connecting some of the dots. I will try to read Clarkvision to get up to speed.

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

The -2EV of EC also helped to prevent blown specular reflections.

As you know, you can get rid of this issue by using a tripod, but the increase in SS is the right move if needed for hand holding.

I am looking forward to the next trial.... with a tripod! The trick will be the usual: to get sufficient light from the shadows without blowing those speculars. noise vs DR compromise.

I gather from your OP that you are actually trying to keep the specular highlights in check, and have chosen moonlight because it likely makes it possible. An interesting thought.

In daylight the difference between direct sunlit speculars (EV=34) and bright snow (EV=16) is ~ 18 stops. Under moonlight the difference between direct moonlit speculars (EV= 14) and bright snow (EV = -2) is ~ 16 stops. So I have two stops less luminance range to accommodate in the highlights ... never mind the shadows in the trees etc. So I have a less steep slope at the right edge of the luminosity histogram.

But is it really necessary?

In the trial, I seem to get a lot more of a sparkle field. The lower luminance sparkles (because of angles of the planes of the crystals) are no longer hidden in the highlights of the snow. Ie there is more "texture" in the speculars.

You've got these very high luminance peaks towering above everything else, like the spikes in Brice Canyon, and you are trying to raise the water level to the top. But, if you lower the level down the peaks, at least for a while, they are still the only thing above the water level. Aren't you just as well off keeping the level lower rather than trying to cover the peaks? Won't the final result be effectively the same?

I think that the peaks are of varying height (due to orientation of the snowflake reflective surfaces). During the daytime the highest peaks are 18 stops higher than the varying topography in the valleys below. During moonlight the highest peaks are 16 stops higher. I want to include the highest peaks (at minimal blooming), the mid peaks and the "texture" of the lower topography (the snow texture). I am willing to sacrifice the few uppermost peaks to get the medium height peaks.

I think that I am trying to keep the water level low enough to show variation in peaks (and to pick up the peaks) that are closer to the valley floor.

Does this make sense to you?

So in terms of exposure: I want to minimize oversaturation of the direct speculars, but not to eliminate it. So I usually use -EC when spotmetering a bright sparkly snowfield. (ETTR which is controlled by the degree of loss of speculars.)

When bracketing I'll usually select the frame with the lowest exposure to work on.

Then in LR, I usually pull the highlights back by 100% and increase the whites by a bit. I usually increase shadows and clarity.

Here is an example from a couple of days ago. I was experimenting with Low ISO and played with oversaturation and manipulation of WB:



These are the final LR settings ....



... to give this final histogram:



.... for this processed output image.



[ATTACH alt="oversaturation and "blooming" at the top of this crop?? "]1193034[/ATTACH]
oversaturation and "blooming" at the top of this crop??

Best compromise, I think. What do you think?

I shot about a dozen scenes with the above in mind. I didn't shoot a range of exposures because it was very chilly with the window open (-20ºC).

Here are three of the better ones. Yes, I know that I could do a lot better with a tripod and longer exposures at base ISO! Go full size to see the sparkles better.


orange cast is from my clearance light.





Advice appreciated, especially from those who have tried this sort of thing, day or night.

These effects are rather spectacular (specular?). You're definitely on the right track. If you keep the first, I'd clone out that contrail.

I'm not keeping it except as an example of this trial. Did you notice the flare with a diameter same as the contrail.... mirror image on the left? And the orange colour from my clearance lights? Both would be difficult to fix in post.

Not till you mentioned it; it's not a big deal, but, clearly, you're not happy with it. And yes, both would be difficult to fix in post. But the clearance-light glow is definitely easy to fix in the future.

I also had some brake light glow in some of the images. Not used to taking pix at night out of the vehicle!

What do the raw files look like in RawDigger? How well have you done with the sparkle highlights?

I haven't done that yet. I ran out of the trial version of RD and need to subscribe.

Oh, for sure.

Yes!

The first is compositionally the best IMO. Having the snow covered rocks in the middle helps a great deal, and the fact that the snow is not sullied with footprints is also nice.

Here are a couple more that are nice compositionally but brutal with noise and jiggle:

It's too bad these didn't meet your expectations, because, compositionally, they are the picks of the litter.

Very interesting work.

Yes. I find that trying challenging stuff like this motivates me to learn more from you experts! A nice regenerative feedback loop.

I built several regenerative receivers back in my youth. They were fun. But life really picked up with my first superhet (Halicrafters S-40B - I still have it).

In the '50s I learned my electronics at home from an ARRL handbook. Made a TRF from scratch. Then a superhet shortwave from a Heathkit. Then joined the airforce, went to engineering school and had to learn about current controlled devices and junctions etc. Almost as difficult as relearning exposure after being contaminated by the exposure triangle!

Many thanks for your help!

--
gollywop
http://g4.img-dpreview.com/D8A95C7DB3724EC094214B212FB1F2AF.jpg

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[gollywop]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

Sensorgen.org doesn't state DR for ISO 125, but it would appear to be closer to 10.5, and at ISO 800 it drops only to 9.5. Read noise, however, is significantly better at 800, so this seems indeed to be a good choice.

Thanks gollywop, I had forgotten how I "knew" that the DR of the FZ1000. I will have to review my understanding of noise again. For some reason, I am having difficulty connecting some of the dots. I will try to read Clarkvision to get up to speed.

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

The -2EV of EC also helped to prevent blown specular reflections.

As you know, you can get rid of this issue by using a tripod, but the increase in SS is the right move if needed for hand holding.

I am looking forward to the next trial.... with a tripod! The trick will be the usual: to get sufficient light from the shadows without blowing those speculars. noise vs DR compromise.

I gather from your OP that you are actually trying to keep the specular highlights in check, and have chosen moonlight because it likely makes it possible. An interesting thought.

In daylight the difference between direct sunlit speculars (EV=34) and bright snow (EV=16) is ~ 18 stops. Under moonlight the difference between direct moonlit speculars (EV= 14) and bright snow (EV = -2) is ~ 16 stops. So I have two stops less luminance range to accommodate in the highlights ... never mind the shadows in the trees etc. So I have a less steep slope at the right edge of the luminosity histogram.

But is it really necessary?

In the trial, I seem to get a lot more of a sparkle field. The lower luminance sparkles (because of angles of the planes of the crystals) are no longer hidden in the highlights of the snow. Ie there is more "texture" in the speculars.

You've got these very high luminance peaks towering above everything else, like the spikes in Brice Canyon, and you are trying to raise the water level to the top. But, if you lower the level down the peaks, at least for a while, they are still the only thing above the water level. Aren't you just as well off keeping the level lower rather than trying to cover the peaks? Won't the final result be effectively the same?

I think that the peaks are of varying height (due to orientation of the snowflake reflective surfaces). During the daytime the highest peaks are 18 stops higher than the varying topography in the valleys below. During moonlight the highest peaks are 16 stops higher. I want to include the highest peaks (at minimal blooming), the mid peaks and the "texture" of the lower topography (the snow texture). I am willing to sacrifice the few uppermost peaks to get the medium height peaks.

I think that I am trying to keep the water level low enough to show variation in peaks (and to pick up the peaks) that are closer to the valley floor.

Does this make sense to you?

Yes.

So in terms of exposure: I want to minimize oversaturation of the direct speculars, but not to eliminate it. So I usually use -EC when spotmetering a bright sparkly snowfield. (ETTR which is controlled by the degree of loss of speculars.)

When bracketing I'll usually select the frame with the lowest exposure to work on.

Then in LR, I usually pull the highlights back by 100% and increase the whites by a bit. I usually increase shadows and clarity.

Here is an example from a couple of days ago. I was experimenting with Low ISO and played with oversaturation and manipulation of WB:

These are the final LR settings ....



... to give this final histogram:

That's a familiar group of settings.

.... for this processed output image.

[ATTACH alt="oversaturation and "blooming" at the top of this crop??"]1193034[/ATTACH]
oversaturation and "blooming" at the top of this crop??

I'm not completely sure of this, but I think blooming is a CCD phenomenon. There may be an analogous CMOS effect, but I think the phenomenon is held better in check. In any event, I think what you're seeing above is simply a sparkle that is OOF, beyond the DoF.

Not till you mentioned it; it's not a big deal, but, clearly, you're not happy with it. And yes, both would be difficult to fix in post. But the clearance-light glow is definitely easy to fix in the future.

I also had some brake light glow in some of the images. Not used to taking pix at night out of the vehicle!

I'm not surprised. ;-)

What do the raw files look like in RawDigger? How well have you done with the sparkle highlights?

I haven't done that yet. I ran out of the trial version of RD and need to subscribe.

Oh, for sure.

Yes!

The first is compositionally the best IMO. Having the snow covered rocks in the middle helps a great deal, and the fact that the snow is not sullied with footprints is also nice.

Here are a couple more that are nice compositionally but brutal with noise and jiggle:

It's too bad these didn't meet your expectations, because, compositionally, they are the picks of the litter.

Very interesting work.

Yes. I find that trying challenging stuff like this motivates me to learn more from you experts! A nice regenerative feedback loop.

I built several regenerative receivers back in my youth. They were fun. But life really picked up with my first superhet (Halicrafters S-40B - I still have it).

In the '50s I learned my electronics at home from an ARRL handbook. Made a TRF from scratch. Then a superhet shortwave from a Heathkit. Then joined the airforce, went to engineering school and had to learn about current controlled devices and junctions etc. Almost as difficult as relearning exposure after being contaminated by the exposure triangle!

I've still got my 1952 Handbook. I haul it out now and again; it's a real nostalgia trip. I built a Johnson Viking Ranger back in the early '50s when I was a very young teen. I still have it too. I've done a dozen or more Heathkits (radio, audio, TV), but my favorite now-a-days is my Elecraft K2/100. It will see me out. :-O

--
gollywop
http://g4.img-dpreview.com/D8A95C7DB3724EC094214B212FB1F2AF.jpg

 

[Detail Man]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

It seems that there is very little (if any) need for retention of spatial resolution within the sparkles.

Agreed.

Everything (of course) looks better at a higher (sensor-level) exposure - but the more tangible concern (might) seem to be whether you could in processing make the tiny specular highlights (likely with a "clipped" Green RAW channel ?) look convincingly white in perceived color by viewers, and not take on a hard to grapple with "magenta" caste (from the non-clipped Red and Blue RAW channels information then, as a result being more prominent relative to Green ?

I don't know if I even get to that stage. In the daytime, the biggest challenge seems to be the extreme difference in luminance of the speculars compared to the surrounding snow ... never mind the shadows! In the case of the difference between the speculars and the snow it is a compromise: How to make the speculars show at 255 (with a minimal level of oversaturation and "blooming") and for the slope of that end "cliff" to be gradual enough to let the texture of the snow show while not degrading it to a dull grey (beach-sand type of) snow.

I am told that "Luminance tone mapping" or "luminosity masks" could be used to tweak the slope of that cliff on the luminosity histogram in post processing. I don't have Photoshop and its use would be a major addition in my meager tool kit .... and my meager brain.

Dunno. Last I thought much about that stuff was when our friend gollywop started this thread a while ago about that very (relating to one or more clipped RAW-channels) subject:

http://www.dpreview.com/forums/thread/3236013

At any rate, such "gremlins" might likely be easily missed by those who eschew "pixel-peeping".

Whereas I use pixel peeping to try to understand what is going on, I am not adverse to accepting some blooming (due to oversaturation of small speculars), some snow greyness, some loss of snow texture and some noise in the shadows. It seems to be a matter of "acceptable" image quality. The difficulty is that my standards keep moving. My sparkles and snow textures are incrementally better now than they were a few years ago . But sometimes I wonder.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

Sensorgen.org doesn't state DR for ISO 125, but it would appear to be closer to 10.5, and at ISO 800 it drops only to 9.5. Read noise, however, is significantly better at 800, so this seems indeed to be a good choice.

Thanks gollywop, I had forgotten how I "knew" that the DR of the FZ1000. I will have to review my understanding of noise again. For some reason, I am having difficulty connecting some of the dots. I will try to read Clarkvision to get up to speed.

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

The -2EV of EC also helped to prevent blown specular reflections.

As you know, you can get rid of this issue by using a tripod, but the increase in SS is the right move if needed for hand holding.

I am looking forward to the next trial.... with a tripod! The trick will be the usual: to get sufficient light from the shadows without blowing those speculars. noise vs DR compromise.

More mechanical stabilization (as well as whatever you electro-mechanical image-stabilization may yield in any single given recorded frame enables the longer shutter-time). However, not only will *everything* (including the peak highlights) be scaled to a higher multiple value, there is additional (true semiconductor type this time) random "shot noise" associated with semiconductor junction leakage-current flows, doubling in the value of random noise-variance every 10 Degrees C (or so).

Ok.

The image-sensor junction temperature, and resulting random noise increases over the time spent scanning the image-data - causing the long exposure-time approach [of performing a true RAW-level dark-frame subtraction (under UI control in the GH2)] - which is a viable option (absent camera or subject-matter movements taking place during the longer exposure-time).

I noticed that the FZ1000 automatically went to dark frame subtraction on the longer trial exposures the other night. Yet to read the manual on that feature.

(Not "FZ1000 sure", but) ... the GH2 was the only Panasonic that I have owned which allowed the user to, if they prefer, turn off what (in the compact model cases that I am familiar with) is mandatory, machine-controlled dark-frame subtraction for all Shutter Times longer than 1 Second.

The "thermal" situation can/does trigger "flaky pixels" to "pop-on" (to go "hot"). Emil Martinec here.

YES! I think that I might have some "hot" pixels in the sky ... in the lowest image in this post.???

One possibly could pretend they are "stars", Tom, and nobody might (likely) "be the wiser" ...

Best compromise, I think. What do you think?

I shot about a dozen scenes with the above in mind. I didn't shoot a range of exposures because it was very chilly with the window open (-20ºC).

Here are three of the better ones. Yes, I know that I could do a lot better with a tripod and longer exposures at base ISO! Go full size to see the sparkles better.


orange cast is from my clearance light.





Advice appreciated, especially from those who have tried this sort of thing, day or night.

These effects are rather spectacular (specular?). You're definitely on the right track. If you keep the first, I'd clone out that contrail.

I'm not keeping it except as an example of this trial. Did you notice the flare with a diameter same as the contrail.... mirror image on the left? And the orange colour from my clearance lights? Both would be difficult to fix in post.

What do the raw files look like in RawDigger? How well have you done with the sparkle highlights?

I haven't done that yet. I ran out of the trial version of RD and need to subscribe.

Indeed - unless you want to download and play with the late Gabor Schorr's ("panopeeper" on the Luminous Landscape forums) "Rawnalyze" (last Version released was 2.10.4.0, for Win OS only) - which, while quite the item of interest and use in its time a while back now, using it is not unlike riding on a bare-bones "tricycle" after de-boarding a "777 jet-liner" (RawDigger):


Rawnalyze 2.10.4.0 - Written, developed, distributed by Gabor Schorr

.

No thanks. I am on a MacBookPro (with a large monitor) and really like RawDigger.

Right. $20 (or perhaps $40 - if you want multiple selection areas and TIFF export capability) is probably a good investment considering what you want to know/do (relative to camera, lens, other software costs).

The first is compositionally the best IMO. Having the snow covered rocks in the middle helps a great deal, and the fact that the snow is not sullied with footprints is also nice.

Here are a couple more that are nice compositionally but brutal with noise and jiggle:

Very interesting work.

Yes. I find that trying challenging stuff like this motivates me to learn more from you experts! A nice regenerative feedback loop.

Many thanks for your help!

 

[GeorgianBay1939]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

It seems that there is very little (if any) need for retention of spatial resolution within the sparkles.

Agreed.

Everything (of course) looks better at a higher (sensor-level) exposure - but the more tangible concern (might) seem to be whether you could in processing make the tiny specular highlights (likely with a "clipped" Green RAW channel ?) look convincingly white in perceived color by viewers, and not take on a hard to grapple with "magenta" caste (from the non-clipped Red and Blue RAW channels information then, as a result being more prominent relative to Green ?

I don't know if I even get to that stage. In the daytime, the biggest challenge seems to be the extreme difference in luminance of the speculars compared to the surrounding snow ... never mind the shadows! In the case of the difference between the speculars and the snow it is a compromise: How to make the speculars show at 255 (with a minimal level of oversaturation and "blooming") and for the slope of that end "cliff" to be gradual enough to let the texture of the snow show while not degrading it to a dull grey (beach-sand type of) snow.

I am told that "Luminance tone mapping" or "luminosity masks" could be used to tweak the slope of that cliff on the luminosity histogram in post processing. I don't have Photoshop and its use would be a major addition in my meager tool kit .... and my meager brain.

Dunno. Last I thought much about that stuff was when our friend gollywop started this thread a while ago about that very (relating to one or more clipped RAW-channels) subject:

http://www.dpreview.com/forums/thread/3236013

That is very good stuff. The saving grace with the snow sparkles is that they are small (when in focus, but not "nice" when OOF).

When I first started with this challenge I hoped that I would get some "starburst" from the very small reflections by stopping down well into diffraction. No such luck. I also tried filters and software. All not nice.

At any rate, such "gremlins" might likely be easily missed by those who eschew "pixel-peeping".

Whereas I use pixel peeping to try to understand what is going on, I am not adverse to accepting some blooming (due to oversaturation of small speculars), some snow greyness, some loss of snow texture and some noise in the shadows. It seems to be a matter of "acceptable" image quality. The difficulty is that my standards keep moving. My sparkles and snow textures are incrementally better now than they were a few years ago . But sometimes I wonder.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

Sensorgen.org doesn't state DR for ISO 125, but it would appear to be closer to 10.5, and at ISO 800 it drops only to 9.5. Read noise, however, is significantly better at 800, so this seems indeed to be a good choice.

Thanks gollywop, I had forgotten how I "knew" that the DR of the FZ1000. I will have to review my understanding of noise again. For some reason, I am having difficulty connecting some of the dots. I will try to read Clarkvision to get up to speed.

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

The -2EV of EC also helped to prevent blown specular reflections.

As you know, you can get rid of this issue by using a tripod, but the increase in SS is the right move if needed for hand holding.

I am looking forward to the next trial.... with a tripod! The trick will be the usual: to get sufficient light from the shadows without blowing those speculars. noise vs DR compromise.

More mechanical stabilization (as well as whatever you electro-mechanical image-stabilization may yield in any single given recorded frame enables the longer shutter-time). However, not only will *everything* (including the peak highlights) be scaled to a higher multiple value, there is additional (true semiconductor type this time) random "shot noise" associated with semiconductor junction leakage-current flows, doubling in the value of random noise-variance every 10 Degrees C (or so).

Ok.

The image-sensor junction temperature, and resulting random noise increases over the time spent scanning the image-data - causing the long exposure-time approach [of performing a true RAW-level dark-frame subtraction (under UI control in the GH2)] - which is a viable option (absent camera or subject-matter movements taking place during the longer exposure-time).

I noticed that the FZ1000 automatically went to dark frame subtraction on the longer trial exposures the other night. Yet to read the manual on that feature.

(Not "FZ1000 sure", but) ... the GH2 was the only Panasonic that I have owned which allowed the user to, if they prefer, turn off what (in the compact model cases that I am familiar with) is mandatory, machine-controlled dark-frame subtraction for all Shutter Times longer than 1 Second.

The FZ1000 Advanced Manual gives the user the option to toggle the "Long shtr NR" ON/OFF. I suspect that the 1 second applies also. I will have to check that out.

The "thermal" situation can/does trigger "flaky pixels" to "pop-on" (to go "hot"). Emil Martinec here.

YES! I think that I might have some "hot" pixels in the sky ... in the lowest image in this post.???

One possibly could pretend they are "stars", Tom, and nobody might (likely) "be the wiser" ...

LOL!

Best compromise, I think. What do you think?

I shot about a dozen scenes with the above in mind. I didn't shoot a range of exposures because it was very chilly with the window open (-20ºC).

Here are three of the better ones. Yes, I know that I could do a lot better with a tripod and longer exposures at base ISO! Go full size to see the sparkles better.


orange cast is from my clearance light.





Advice appreciated, especially from those who have tried this sort of thing, day or night.

These effects are rather spectacular (specular?). You're definitely on the right track. If you keep the first, I'd clone out that contrail.

I'm not keeping it except as an example of this trial. Did you notice the flare with a diameter same as the contrail.... mirror image on the left? And the orange colour from my clearance lights? Both would be difficult to fix in post.

What do the raw files look like in RawDigger? How well have you done with the sparkle highlights?

I haven't done that yet. I ran out of the trial version of RD and need to subscribe.

Indeed - unless you want to download and play with the late Gabor Schorr's ("panopeeper" on the Luminous Landscape forums) "Rawnalyze" (last Version released was 2.10.4.0, for Win OS only) - which, while quite the item of interest and use in its time a while back now, using it is not unlike riding on a bare-bones "tricycle" after de-boarding a "777 jet-liner" (RawDigger):


Rawnalyze 2.10.4.0 - Written, developed, distributed by Gabor Schorr

.

No thanks. I am on a MacBookPro (with a large monitor) and really like RawDigger.

Right. $20 (or perhaps $40 - if you want multiple selection areas and TIFF export capability) is probably a good investment considering what you want to know/do (relative to camera, lens, other software costs).

Yes.

The first is compositionally the best IMO. Having the snow covered rocks in the middle helps a great deal, and the fact that the snow is not sullied with footprints is also nice.

Here are a couple more that are nice compositionally but brutal with noise and jiggle:

Very interesting work.

Yes. I find that trying challenging stuff like this motivates me to learn more from you experts! A nice regenerative feedback loop.

Many thanks for your help!

I spent a bit of time yesterday trying to compare histograms and zebra striping, with varying metering modes and when using M(anual) and A(perture) Priority Modes. Inconclusive as I had varying scene luminances due to the varying cloud conditions. I'll try again when in a consistent light environment.

As usual, DM, your assistance is very much appreciated.

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[Great Bustard]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

No, more light on the sensor *always* results in less noise, even when the higher ISO setting has less electronic noise. But, as you note, greater exposure and/or higher ISO will result in more of the highlights being blown.

Yes. Thank you. I had it right in my head but wrote it incorrectly. My bad!

Thanks for your clarification.

I should add that if we're talking about very long exposures (where what counts as "very long" is a function of the sensor), thermal noise from the sensor may offset lower noise from more light.

And thank you for your patience in dealing with "equivalence-deniers". I, as a relative newbie, really appreciate your explanations.

Kind of you to say!

 

[GeorgianBay1939]

We had a good opportunity a couple nights ago so I made a trial run at it.

This is my simplistic thinking:

First, definitions:

Light Value (LV) = Exposure Value at ISO 100 or EV (ISO 100) is a measure of scene brightness , not to be confused with EV which is the pair (f/, ss) camera adjustments for Exposure.

The LV of looking at the sun directly is about 34.

See the above handy link for the following:
LV of sunlit snow or white beach sand is about 16
LV of (looking at) full moon is about 15

Using my spot meter:
LV of moonlit horizontal snow is about + 3

So the DR of sparkles on snow during the day is (34-16) or 18 stops (between the snow and the specular reflections of direct sunlight). Impossible to do anything with that range.

And the DR of sparkles on snow during a moonlit night is about (15-3) or 12 stops (between the snow and the specular reflections of direct moonlight). Although difficult to capture in a single exposure I should be able to blend some brackets to get an acceptable HDR shot using Enfuse or Photomatix.

The experiment:
To speed things up I went to a few places close by and shot hand held, well braced. To do it properly I would use longer exposures and use the tripod. : )

At ISO 125 my camera has a DR of about 10 EV or stops. As with all cameras, increasing ISO drops DR. Increasing ISO to 800 on the FZ1000 probably drops DR by ~ 2 stops. Although I lost 2 stops of DR, I decided to use ISO 800 to maximize my analog gain (to minimize my read noise.) So I traded off DR for less noise. right?

Sensorgen.org doesn't state DR for ISO 125, but it would appear to be closer to 10.5, and at ISO 800 it drops only to 9.5. Read noise, however, is significantly better at 800, so this seems indeed to be a good choice.

Thanks gollywop, I had forgotten how I "knew" that the DR of the FZ1000. I will have to review my understanding of noise again. For some reason, I am having difficulty connecting some of the dots. I will try to read Clarkvision to get up to speed.

And I got 2 2/3 stops of shutter speed while shooting wide open (ranging from f/3 to f/4 depending on the Focal Length I used) My shutter speeds ranged from 1 sec to about 4 seconds (2 stops).

BUT then I cranked in -2 EV of Exposure Compensation to keep my shutter interval reasonably short, a second or so. (A 1 second shutter exposure is about my hand holding limit when I am well propped, engine off, dogs quiet in the vehicle.) …. and picked up the read noise again. right?

The -2EV of EC also helped to prevent blown specular reflections.

As you know, you can get rid of this issue by using a tripod, but the increase in SS is the right move if needed for hand holding.

I am looking forward to the next trial.... with a tripod! The trick will be the usual: to get sufficient light from the shadows without blowing those speculars. noise vs DR compromise.

I gather from your OP that you are actually trying to keep the specular highlights in check, and have chosen moonlight because it likely makes it possible. An interesting thought.

In daylight the difference between direct sunlit speculars (EV=34) and bright snow (EV=16) is ~ 18 stops. Under moonlight the difference between direct moonlit speculars (EV= 14) and bright snow (EV = -2) is ~ 16 stops. So I have two stops less luminance range to accommodate in the highlights ... never mind the shadows in the trees etc. So I have a less steep slope at the right edge of the luminosity histogram.

But is it really necessary?

In the trial, I seem to get a lot more of a sparkle field. The lower luminance sparkles (because of angles of the planes of the crystals) are no longer hidden in the highlights of the snow. Ie there is more "texture" in the speculars.

You've got these very high luminance peaks towering above everything else, like the spikes in Brice Canyon, and you are trying to raise the water level to the top. But, if you lower the level down the peaks, at least for a while, they are still the only thing above the water level. Aren't you just as well off keeping the level lower rather than trying to cover the peaks? Won't the final result be effectively the same?

I think that the peaks are of varying height (due to orientation of the snowflake reflective surfaces). During the daytime the highest peaks are 18 stops higher than the varying topography in the valleys below. During moonlight the highest peaks are 16 stops higher. I want to include the highest peaks (at minimal blooming), the mid peaks and the "texture" of the lower topography (the snow texture). I am willing to sacrifice the few uppermost peaks to get the medium height peaks.

I think that I am trying to keep the water level low enough to show variation in peaks (and to pick up the peaks) that are closer to the valley floor.

Does this make sense to you?

Yes.

I just realized that Gibbous Moon either side of Full might change the brightness ratio of speculars to diffuse snowfield from the ratio at Full Moon. Although I would need a larger exposure I might get better "discrimination" at the extreme highlights. Another variable to play with!

So in terms of exposure: I want to minimize oversaturation of the direct speculars, but not to eliminate it. So I usually use -EC when spotmetering a bright sparkly snowfield. (ETTR which is controlled by the degree of loss of speculars.)

When bracketing I'll usually select the frame with the lowest exposure to work on.

Then in LR, I usually pull the highlights back by 100% and increase the whites by a bit. I usually increase shadows and clarity.

Here is an example from a couple of days ago. I was experimenting with Low ISO and played with oversaturation and manipulation of WB:

These are the final LR settings ....



... to give this final histogram:

That's a familiar group of settings.

.... for this processed output image.

[ATTACH alt="oversaturation and "blooming" at the top of this crop??"]1193034[/ATTACH]
oversaturation and "blooming" at the top of this crop??

I'm not completely sure of this, but I think blooming is a CCD phenomenon. There may be an analogous CMOS effect, but I think the phenomenon is held better in check. In any event, I think what you're seeing above is simply a sparkle that is OOF, beyond the DoF.

Not till you mentioned it; it's not a big deal, but, clearly, you're not happy with it. And yes, both would be difficult to fix in post. But the clearance-light glow is definitely easy to fix in the future.

I also had some brake light glow in some of the images. Not used to taking pix at night out of the vehicle!

I'm not surprised. ;-)

What do the raw files look like in RawDigger? How well have you done with the sparkle highlights?

I haven't done that yet. I ran out of the trial version of RD and need to subscribe.

Oh, for sure.

Yes!

The first is compositionally the best IMO. Having the snow covered rocks in the middle helps a great deal, and the fact that the snow is not sullied with footprints is also nice.

Here are a couple more that are nice compositionally but brutal with noise and jiggle:

It's too bad these didn't meet your expectations, because, compositionally, they are the picks of the litter.

Very interesting work.

Yes. I find that trying challenging stuff like this motivates me to learn more from you experts! A nice regenerative feedback loop.

I built several regenerative receivers back in my youth. They were fun. But life really picked up with my first superhet (Halicrafters S-40B - I still have it).

In the '50s I learned my electronics at home from an ARRL handbook. Made a TRF from scratch. Then a superhet shortwave from a Heathkit. Then joined the airforce, went to engineering school and had to learn about current controlled devices and junctions etc. Almost as difficult as relearning exposure after being contaminated by the exposure triangle!

I've still got my 1952 Handbook. I haul it out now and again; it's a real nostalgia trip. I built a Johnson Viking Ranger back in the early '50s when I was a very young teen. I still have it too. I've done a dozen or more Heathkits (radio, audio, TV), but my favorite now-a-days is my Elecraft K2/100. It will see me out. :-O

IIRC, it was the 5 WPM CW that weeded me out!!! I did have to "read" Morse Identifiers on various Navaids in the airforce but reading was a lot easier to learn.

--
gollywop
http://g4.img-dpreview.com/D8A95C7DB3724EC094214B212FB1F2AF.jpg

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[gollywop]

I think that the peaks are of varying height (due to orientation of the snowflake reflective surfaces). During the daytime the highest peaks are 18 stops higher than the varying topography in the valleys below. During moonlight the highest peaks are 16 stops higher. I want to include the highest peaks (at minimal blooming), the mid peaks and the "texture" of the lower topography (the snow texture). I am willing to sacrifice the few uppermost peaks to get the medium height peaks.

I think that I am trying to keep the water level low enough to show variation in peaks (and to pick up the peaks) that are closer to the valley floor.

Does this make sense to you?

Yes.

I just realized that Gibbous Moon either side of Full might change the brightness ratio of speculars to diffuse snowfield from the ratio at Full Moon. Although I would need a larger exposure I might get better "discrimination" at the extreme highlights. Another variable to play with!

Yes, at least the sun stays (more or less) the same size.

I've still got my 1952 Handbook. I haul it out now and again; it's a real nostalgia trip. I built a Johnson Viking Ranger back in the early '50s when I was a very young teen. I still have it too. I've done a dozen or more Heathkits (radio, audio, TV), but my favorite now-a-days is my Elecraft K2/100. It will see me out. :-O

IIRC, it was the 5 WPM CW that weeded me out!!! I did have to "read" Morse Identifiers on various Navaids in the airforce but reading was a lot easier to learn.

Hey, CW doesn't begin to be fun till 30-35 WPM. It's a whole new world.

 

[GeorgianBay1939]

I think that the peaks are of varying height (due to orientation of the snowflake reflective surfaces). During the daytime the highest peaks are 18 stops higher than the varying topography in the valleys below. During moonlight the highest peaks are 16 stops higher. I want to include the highest peaks (at minimal blooming), the mid peaks and the "texture" of the lower topography (the snow texture). I am willing to sacrifice the few uppermost peaks to get the medium height peaks.

I think that I am trying to keep the water level low enough to show variation in peaks (and to pick up the peaks) that are closer to the valley floor.

Does this make sense to you?

Yes.

I just realized that Gibbous Moon either side of Full might change the brightness ratio of speculars to diffuse snowfield from the ratio at Full Moon. Although I would need a larger exposure I might get better "discrimination" at the extreme highlights. Another variable to play with!

Yes, at least the sun stays (more or less) the same size.

I've still got my 1952 Handbook. I haul it out now and again; it's a real nostalgia trip. I built a Johnson Viking Ranger back in the early '50s when I was a very young teen. I still have it too. I've done a dozen or more Heathkits (radio, audio, TV), but my favorite now-a-days is my Elecraft K2/100. It will see me out. :-O

IIRC, it was the 5 WPM CW that weeded me out!!! I did have to "read" Morse Identifiers on various Navaids in the airforce but reading was a lot easier to learn.

Hey, CW doesn't begin to be fun till 30-35 WPM. It's a whole new world.

Whole new world? Brave new world?

Bugs inhabit a world that is alien to me.

That is a word every 2 seconds or so.

I cannot even think that fast!

;-)

 

[Satyaa]

Hi, I have no feedback to offer on your questions (what you described is too technical for me).

What I want to say is that I liked your photos very much, especially the first one and the last two you posted later.

Well done and thanks for sharing.

 

[GeorgianBay1939]

Hi, I have no feedback to offer on your questions (what you described is too technical for me).

Technical for me too!!!

What I want to say is that I liked your photos very much, especially the first one and the last two you posted later.

Well done and thanks for sharing.

Thanks for your generous comments. I like your menagerie at your feeders!