Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1       2       3              6      
7
       8       end
  

How much post processing?

  
 
AmbientMike
Offline
• • • • • •
Upload & Sell: Off
p.7 #1 · p.7 #1 · How much post processing?



gdanmitchell wrote:
When you "blow out" something that is too bright, you cannot "get it back in post." A blown our area is one in which the signal was so bright that it simply maxed out the luminosity values.

When an area is truly blown out, there is nothing left to recover. Let's say that at an exposure that doesn't blow the area out there would be luminosity values between (to make up numbers) 235 and 255 for the pixels in the area. When the area is blown out the camera might record the same 255 level for all pixels in that
...Show more

Of course you can get blown out areas back in pp if you shoot raw. That's a major reason I use raw, if I use it. For one thing, the blinkies on the lcd are indicative of the jpeg. I guess you could argue that if you overexpose it enough even the raw is blown, but you don't really seem to be doing that?

You have to be a bit careful, some stuff you don't want to blow out, but raw processors don't seem to have much trouble with sky. I've read to blow out nothing in reputable publications, I think Fred had a page about really blowing a lot of stuff out at least on the lcd using a different brand of camera presumably since more exposure gives less noise and higher shadows. I'm in between those two.

Most pro images on film color slides. Yes you could do a lot on b&w but pretty much no pp on color. On prints film you mostly tried to find a lab paying attention to get the exposure and color right. Slides you basically didn't have any pp. Sure you could sandwich using a pin registered back but probably less than 1% actually did that. Exceedingly rare. Obviously in a studio lighting controlled byt not pp on slides. So if you know of people doing a lot of manipulation I'd be curious about how they did that on color. B&W, yes but color few pros had a darkroom to print slides or even print film

Even the contrast adjustment on the cheapest software is more than you had on film. You couldn't control contrast in pp on color film, basically.

I think I gave up on trying to capture the scene I actually saw long ago. I find it odd that people do that. I guess if you shoot at midday trying to do that you need raw and massive amounts of DR.

Mostly I just don't like pp. It's miserable thinking about the pp in the field. Can't avoid it all even on jpeg though, you have to consider if you blow it out its gone and underexpose a bit more.

For landscape if I blow out a bit of clouds or something it's probably not a big deal to get it back in pp. Pretty easy to blow something out in the sky given the range if tones on a landscape. So usually use raw there. Honestly I don't look at luminance histogram so much anymore anyway if using a camera that has the other stuff, other more important measures imo.


Sometimes conditions aren't good and you pass. Just like film.



May 28, 2023 at 11:36 AM
DailyShooter
Offline
• •
Upload & Sell: Off
p.7 #2 · p.7 #2 · How much post processing?


AmbientMike wrote:
Of course you can get blown out areas back in pp if you shoot raw. That's a major reason I use raw, if I use it. For one thing, the blinkies on the lcd are indicative of the jpeg. I guess you could argue that if you overexpose it enough even the raw is blown, but you don't really seem to be doing that?

You have to be a bit careful, some stuff you don't want to blow out, but raw processors don't seem to have much trouble with sky. I've read to blow out nothing in reputable publications, I think
...Show more

Sure a jpg can show blown highlights that might not be represented in the corresponding raw image.

What is a blown highlight? Its a part of the image where the blue channel value is 255, red is 255, and green is 255 and the combo = white.

But for blown highlights in the raw file, you will not find a PP guru or professional photographer anywhere in the world that would agree with you, regarding the recovery of blown highlights. You lower the tonality of a blown highlight (hexadecimal 255) it ALWAYS lowers to a gray color, NEVER the actual color found in that scene. Was it a Red, Green, or Blue channel or a combination of those that got blown? Your file has no info on this, and the PP program has even less.

It seems you lack understanding of the math behind the histogram, and what happens when one tries to normalize a blown highlight...when you attempt to recover it, what color results? What shade of Blue, Red, or Green or a combination thereof? The PP program has no clue, your raw image file has no such information...so is there magic involved?

Your attempts to recover blown highlights will in fact result in a sordid color if not sordid color casts. And that "recovered" part of your image will look NOTHING like the actual scene.



May 28, 2023 at 11:49 AM
Mike_5D
Offline
• • • •
Upload & Sell: Off
p.7 #3 · p.7 #3 · How much post processing?


DailyShooter wrote:
Sure a jpg can show blown highlights that might not be represented in the corresponding raw image.

What is a blown highlight? Its a part of the image where the blue channel value is 255, red is 255, and green is 255 and the combo = white.

But for blown highlights in the raw file, you will not find a PP guru or professional photographer anywhere in the world that would agree with you, regarding the recovery of blown highlights. You lower the tonality of a blown highlight (hexadecimal 255) it ALWAYS lowers to a gray color, NEVER the actual color found
...Show more

There's blown out, and there's BLOW OUT. Sure, if it's really gone, ie 255/255/255, then you won't get any color info back. But a raw file has more DR than the JPG preview on which the histogram and blinkies are based. So if you're just starting to get blinkies, which usually happens first in the whitest part of the scene, then yes, you will be able to recover. I just shot into the brighest part of an overcast sky. The camera, with the picture style set to faithful) says nearly the entire sky is completely blown out. Lightroom says the brightest part is 97-98%. I have to push the whites slider to +30 before Lightroom's overexposure slider matches the camera's.



May 28, 2023 at 12:09 PM
AmbientMike
Offline
• • • • • •
Upload & Sell: Off
p.7 #4 · p.7 #4 · How much post processing?




DailyShooter wrote:
Sure a jpg can show blown highlights that might not be represented in the corresponding raw image.

What is a blown highlight? Its a part of the image where the blue channel value is 255, red is 255, and green is 255 and the combo = white.

But for blown highlights in the raw file, you will not find a PP guru or professional photographer anywhere in the world that would agree with you, regarding the recovery of blown highlights. You lower the tonality of a blown highlight (hexadecimal 255) it ALWAYS lowers to a gray color, NEVER the actual color found
...Show more

This is mostly just incorrect.



May 28, 2023 at 12:35 PM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.7 #5 · p.7 #5 · How much post processing?


Mike_5D wrote:
There's blown out, and there's BLOW OUT. Sure, if it's really gone, ie 255/255/255, then you won't get any color info back. But a raw file has more DR than the JPG preview on which the histogram and blinkies are based. So if you're just starting to get blinkies, which usually happens first in the whitest part of the scene, then yes, you will be able to recover. I just shot into the brighest part of an overcast sky. The camera, with the picture style set to faithful) says nearly the entire sky is completely blown out. Lightroom says the
...Show more

[Fixed a bunch of typos in this post.]

It is true that the blinkies (or the right edge of the histogram, to be honest) do not always indicate that all three channels have recorded their maximum luminosity values in the captured raw file. These indicators are generalizations to some extent, and there's some nuance to the whole issue. (For example, I earlier mentioned that blowing tiny specular highlights may not be an issue in some cases and that it is possible to think of cases where a creative decision to allow highlights to blow out can be fine.)

And, indeed, you can sometimes recover (or amplify) some remaining detail from areas in some raw files that appear to be blown out. Note that the words "appear to be" are doing a lot of work here. If the luminosity data of the three channels have actually hit their max for pixels that should ideally be recorded as below-maximum levels, the detail is gone and you cannot get it back. But raw files seem to have a bit of "space" between what the camera indicates to be blown out and that unrecoverable state. You'll have to learn about this through experience with your own camera.

But the "protect the highlights" rule still applies!

Since we're getting into nuance now, I'd disagree with the earlier suggestion that all three color channels must hit the maximum luminosity level before we have a "blowing out" problem. That is, indeed, a bad thing and it results in the camera essentially recording pure white where there should be something else. But it can actually happen when we hit the maximum and can't accurately record the value for a single channel.

Accurate color (with the exception of pure red, blue, or green) comes from recording an accurate balance among the three channels. (It is a bit more complicated than that, but I digress...) Imagine that some very slightly pinkish almost-white color comes from R:255, G: 250, B: 250. The ratio between those three values is what produces that subtly pink almost-white color.

Now let's overexpose that a bit so that an accurate version of that color would require (and, again, I'm simplifying) R:260, G: 255, B: 255. See the problem yet? In this case we'll blow out the R channel since it now requires a value higher than 255. Now all three channels would be 255... and the pink tone is gone and we just get pure white. The color shifted.

There are plenty of obvious real-world examples of this in photography that we see all the time on the internet and, unfortunately, even in prints from people who should know better.

Let's consider the ubiquitous sunset or sunrise photographs, in which the red channel can be extremely hot. In this case your camera may not even show that the image is blown out — no blinkies, and no "past the right edge" tip of the luminosity curve. You might be tempted to push to almost the right, just below where the camera indicates that you are going too far.

But don't.

The camera is (unless you use an accurate three-color histogram and look very carefully) basing its indication on a sort of average of the three color channels... and if you push all the way to the right you can easily blow out just the red channel even thought the blue and green channels are well below that level.

Here's the problem. In your photograph there are things like sunset-lit clouds that are both very bright and very hot in the red channel. You'll lose important details in those clouds if you let the red channel blow out. I regularly see photographs with brilliant and overly-intense clouds... that have lost their details. (It could happen in the other two channels, too, but we see it crop up most commonly with overly hot red channel images.)

In addition, let's not forget that color balance issue that I described above. Not only do we lose detail in that single hot channel, but color balance among the RGB channels is now thrown off and the color of that brilliant sunset doesn't look right either. (Another place where this problem shows up a lot is with fall color photographs, where letting the red channel get too hot results in leaves that lose detail and may have odd color shifts.)

Again: protect the highlights, whether considering the composite white or the individual RGB channels.

Dealing with these issues can require attention at the time of exposure, and sometimes that means considering how the image will work best in the post-processing phase and then adjusting exposure with that in mind. If that's too much work, well that is a preference that anyone is entitled to have. But some folks feel that the better results are worth the extra work.

Oh, one more "trick" for those who do find themselves with something blown out in these ways. It is sometimes possible to transfer some texture to the blown to area of the image from another part of it. On occasion I've had very bright rocks or small sections of a sunlit cliff blow out in an image that otherwise looked fine... and that pure white area sticks out like a sore thumb. One solution can be to clone a bit of texture from a non-blown area and apply it to the blown out spot, perhaps adjusting the opacity of the clone so that it is very subtle. There are other tricks that can apply, too.

YMMV.

Edited on May 29, 2023 at 05:50 PM · View previous versions



May 28, 2023 at 12:45 PM
DailyShooter
Offline
• •
Upload & Sell: Off
p.7 #6 · p.7 #6 · How much post processing?


AmbientMike wrote:
This is mostly just incorrect.


What is "mostly incorrect"? I bet you can't explain yourself here...so you make a lame response



May 28, 2023 at 02:54 PM
DailyShooter
Offline
• •
Upload & Sell: Off
p.7 #7 · p.7 #7 · How much post processing?


gdanmitchell wrote:
It is true that the blinkies (or the right edge of the histogram, to be honest) do not always indicate that all three channels have recorded their maximum luminosity values in the captured raw file. These indicators are generalizations to some extent, and there's some nuance to the whole issue. (For example, I earlier mentioned that blowing tiny specular highlights may not be an issue in some cases and that it is possible to think of cases where a creative decision to allow highlights to blow out come be fine.)

And, indeed, you can sometimes recover (or amplify) some remaining detail
...Show more

I agree with what you wrote, but I just want to add that if 1 or 2 of the 3 channels are blown (x'255'), recover is not possible...one tries and then gets a strange color cast.



May 28, 2023 at 02:59 PM
Zenon Char
Offline
• • • • • •
Upload & Sell: On
p.7 #8 · p.7 #8 · How much post processing?


We did touch on pleasing vs accurate colour. I did a search and found this. Jim Kasson is a colour scientist. Like I said. I stopped worrying about accurate colour long ago.

https://www.dpreview.com/forums/post/66770635

https://www.dpreview.com/forums/post/66748872

https://fstoppers.com/education/what-color-science-and-should-it-actually-matter-photographers-298843



May 28, 2023 at 05:04 PM
Ferrophot
Offline
• • •
Upload & Sell: Off
p.7 #9 · p.7 #9 · How much post processing?


I have to support DailyShooter in the issue of blown highlights. My experience is once they are blown they are gone and any attempt to just wind back the exposure ends up with something less than satisfactory. Better to just clone in something from a part that is not blown. This is especially true when shooting in cloudy and overcast weather where the sky or clouds are much brighter than the subject. I have found it much better to under expose and then, holding the sky constant, to lift the other parts of the image as needed. As for the blinkies, I find they are a good warning that that part of the image is in danger. Sure, sometimes they give a premature warning, but if gradual changes in tones are desired then it is best to underexpose to have none of them. One reason why I always take test shots before taking my subject. My R6 has a good exposure meter but it is only truely reliable in good lighting conditions.


May 28, 2023 at 07:36 PM
Zenon Char
Offline
• • • • • •
Upload & Sell: On
p.7 #10 · p.7 #10 · How much post processing?


Zenon Char wrote:
We did touch on pleasing vs accurate colour. I did a search and found this. Jim Kasson is a colour scientist. Like I said. I stopped worrying about accurate colour long ago.

https://www.dpreview.com/forums/post/66770635

https://www.dpreview.com/forums/post/66748872

https://fstoppers.com/education/what-color-science-and-should-it-actually-matter-photographers-298843


I did get the pleasing colour part right I think. I always thought accurate meant passing files to several other parties during processing so they all see the same thing. While important it seems it's a little more complicated than that.



May 28, 2023 at 09:52 PM
 


Search in Used Dept. 

AmbientMike
Offline
• • • • • •
Upload & Sell: Off
p.7 #11 · p.7 #11 · How much post processing?




DailyShooter wrote:
What is "mostly incorrect"? I bet you can't explain yourself here...so you make a lame response


Just didn't have time for this garbage. Better things to do on a holiday weekend.....



May 28, 2023 at 10:00 PM
AmbientMike
Offline
• • • • • •
Upload & Sell: Off
p.7 #12 · p.7 #12 · How much post processing?


One thing, I've been out shooting. Oh dear, got a little bit of blinkies. In unimportant areas mostly blue/white sky. Not any kind of oddball color the software is going to have issues on.

So I left them rather than plunge important parts of the scene into darkness. Yes, you can get areas back.



May 28, 2023 at 10:18 PM
DanielScott
Offline
• • •
Upload & Sell: Off
p.7 #13 · p.7 #13 · How much post processing?


AmbientMike wrote:
Just didn't have time for this garbage. Better things to do on a holiday weekend.....


Yikes.



May 29, 2023 at 12:37 AM
DailyShooter
Offline
• •
Upload & Sell: Off
p.7 #14 · p.7 #14 · How much post processing?


AmbientMike wrote:
Just didn't have time for this garbage. Better things to do on a holiday weekend.....


But you have the time to make more lame responses...its seems you cannot back up your claims.



May 29, 2023 at 09:10 AM
DailyShooter
Offline
• •
Upload & Sell: Off
p.7 #15 · p.7 #15 · How much post processing?


AmbientMike wrote:
One thing, I've been out shooting. Oh dear, got a little bit of blinkies. In unimportant areas mostly blue/white sky. Not any kind of oddball color the software is going to have issues on.

So I left them rather than plunge important parts of the scene into darkness. Yes, you can get areas back.


More of your claims with no back up, no cogent argument.

So if your blue channel is 200, your red is 254 and your green is "350 aka solid white", how are you going to recover that area to the same color found in the scene?

You. Can. Not.

The best you can do is clone that part out or accept the color cast you get when you attempt to recover.

This is basic Digital Photography 101 stuff here guys...but please do double-down in your mythology...



May 29, 2023 at 09:15 AM
AmbientMike
Offline
• • • • • •
Upload & Sell: Off
p.7 #16 · p.7 #16 · How much post processing?


Turns out you probably don't know if your raw is blown. Unless you use RawDigger or a couple other methods. I'm sure I don't know anything so here's Jim

https://www.lensrentals.com/blog/2023/05/how-to-expose-raw-files-part-2/comment-page-1/



May 29, 2023 at 11:02 AM
AmbientMike
Offline
• • • • • •
Upload & Sell: Off
p.7 #17 · p.7 #17 · How much post processing?


DailyShooter wrote:
More of your claims with no back up, no cogent argument.

So if your blue channel is 200, your red is 254 and your green is "350 aka solid white", how are you going to recover that area to the same color found in the scene?

You. Can. Not.

The best you can do is clone that part out or accept the color cast you get when you attempt to recover.

This is basic Digital Photography 101 stuff here guys...but please do double-down in your mythology...


You'd actually have to read my posts to see the arguments. But even if green is blown red and blue are multiplied. If green is even blown which LR can't tell you, if you read the article above, in camera histogram can't either, unless you use uni WB, probably not fun



May 29, 2023 at 11:14 AM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.7 #18 · p.7 #18 · How much post processing?


DailyShooter wrote:
I agree with what you wrote, but I just want to add that if 1 or 2 of the 3 channels are blown (x'255'), recover is not possible...one tries and then gets a strange color cast.


Perhaps you missed "this part...?" ;-)

gdanmitchell wrote:
Since we're getting into nuance now, I'd differ with the point made earlier that all three color channels must hit the maximum luminosity level before we have a "blowing out" problem. That is, indeed, a bad thing and it results in the camera essentially recording pure white where there should be something else. But it can actually happen when we hit the maximum and can't accurately record the value for a single channel.


Accurate color (with the exception of pure red, blue, or green) comes from recording an accurate balance among the three channels. (It is even a bit more complicated than that, but I digress...) Imagine that some very slightly poking gray color comes from R:255, G: 250, B: 250. The ration between those three values is what produces that subtly pink almost-white gray color.

Now let's overexpose that a bit so that an accurate version of that color would require (and, again, I'm simplifying) R:260, G: 255, B: 255. See the problem yet? In this case we'll blow out the R channel since it now requires a value higher than 255. Now all three channels would be 255... and the pink tone is gone and we just get pure white.

There are plenty of obvious real-world examples of this in photography that we see all the time on the internet and, unfortunately, even in prints from people who should know better. Let's consider the ubiquitous sunset or sunrise photographs, in which the red channel can be extremely hot. In this case you camera may not even show that the image is blown out — no blinkies, and no "past the right" tip of the luminosity curve. You might be tempted to push to almost the right, just below where the camera indicates that you are going too far.

But don't. The camera is (unless you use the three-color histogram and look very carefully) basing its indication on as sort of average of the three color channels... and if you push all the way to the right you can easily blow out just the red channel even thought the R and G channels are well below that level.

Here's the problem. In your photograph there are things like sunset-lit clouds that are both very bright and very hot in the red channel. And you'll lose important details in those clouds. I regularly see such photograph with brilliant and overly-intense clouds... that have lost their details.

And, in addition, let's not forget that color balance issue that I described above. Not only do we lose detail in that single hot channel, but any balance among the three color channels is now thrown off and the color of that brilliant sunset doesn't look right either. (Another place where this problem shows up a lot is with fall color photographs, where letting the red channel get too hot results in leaves that lose all detail.).

Yes, a lot of words. But perhaps the main point of that post, too!

Getting past some of the (often repeated) naive misconceptions about digital imagery sometimes requires more than a few words. As you may have figured out, we can reply to nonsense about things like what it means to "blow out" an image or how to address this by typing something quick and pithy like, "You are wrong!," or fire off another ad hominem distraction (see elsewhere in this thread), or point out that "blowing out only one or two color channels can create problems, too."

But I like to imagine that at least some FM posters are interested enough in understanding this stuff that they might take the time to read a more thorough explanation.

I also do understand that others won't. I can't make them.

They are free to ignore the posts. We are free to ignore them, too. Which in some cases (examples in this thread) is the best approach. In that regard, another favorite quote comes to mind:

“To argue with a person who has renounced the use of reason is like administering medicine to the dead.” -Thomas Paine

YMMV.




May 29, 2023 at 12:02 PM
DailyShooter
Offline
• •
Upload & Sell: Off
p.7 #19 · p.7 #19 · How much post processing?


gdanmitchell wrote:
Perhaps you missed "this part...?" ;-)







Yes, a lot of words. But perhaps the main point of that post, too!

Getting past some of the (often repeated) naive misconceptions about digital imagery sometimes requires more than a few words. As you may have figured out, we can reply to nonsense about things like what it means to "blow out" an image or how to address this by typing something quick and pithy like, "You are wrong!," or fire off another ad hominem distraction (see elsewhere in this thread), or point out that "blowing out only one or two color channels can create
...Show more

Often, my curt "you are wrong" responses are less for the one I am responding to (their minds will rarely change...we see this in the MAGA crowd), and more for the potentially thousands of readers that read bogus and ignorant claims about highlights, and such.

I think your explanations are correct and most who are in the know would agree with you. But it's sad that there are some with decades of experience that still harbor bogus opinions, and on such basic things.

We often cannot change such minds...but the readers of these exchanges...I believe that their minds often can be...and it is those I respond to...



May 29, 2023 at 12:32 PM
AmbientMike
Offline
• • • • • •
Upload & Sell: Off
p.7 #20 · p.7 #20 · How much post processing?


As far as Thomas Paine, I'm guessing you didn't read the Jim Kasson article I posted a link to. If you didn't you'd be the one renouncing reason and all that. But, just keep on blathering, hard to believe your (inaccurate) comments aren't aimed at me, am I incorrect in thinking they are?

Pardon me for not wasting my time on another (stupid) argument on here, yesterday.

If you had any interest in learning about raw files you'd be reading the link I posted. Have either of you founts of knowledge, thirsting for more, read it.



May 29, 2023 at 12:36 PM
1       2       3              6      
7
       8       end






FM Forums | Canon Forum | Join Upload & Sell

1       2       3              6      
7
       8       end
    
 

You are not logged in. Login or Register

Username       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.