RAW mode to prevent blown highlights ?

..how exactly is this shot ruined by blown highlights ?!? To me it looks perfectly good. You NEED some highlights in a picture. Otherwise it's gonna look dull. Just like you need sweet moments in life to help you deal with the sour ones... ;o)

seriously - you seem to be complaining about a picture that could hardly be exposed any better. Take a good look at it again and then come back here.

David
 
I am a little puzzled by the example. When I load it into
Photoshop neither the Histogram nor Levels show any pixels at 255
or more.

Frank B
It doesn't in mine either. It opens it as generic RGB. That is the color space I use too. It appears a bit darker than it does in the web browser. There are no blown highlights in the color space on my machine.

Even if it appeared the same as the browser, I would think it would correct well with Photoshop, RAW or no.
--
Dave Lewis
 
Whether your in RAW mode or JPG mode, if the exposure is blown out, it's blown out. The file mode shouldn't make any difference. True, RAW mode retains more information than JPG, but if it's blown out information, it's still just more blown out information.

Or are you looking at it from the angle that some of it may be converted down from 12 bit color to blown out information because of the 8 bit color (JPG) vs. the saved 16 bit color (really still 12 bit RAW) TIF format? Am I off here? I can't see that it would make any appreciable difference.

VES
Ive mostly always shot in fine jpeg, but lately ive been more and
more annoyed with minor blown highlights in my shots. Usually its
something like just where the sunlight happened to reflect off an
animals back or something minor like that. It just seems to ruin
what would be an otherwise perfect shot.

My question is IF i shoot in raw mode, would i have that much
success in dealing with these blown highlights ?

heres an example photo

http://www.pbase.com/image/21192229

Would i of been able to recover where its blown out on the lions
face ?

I usually try to use the clone stamp or selective color to tone
down the glaring whiteness of those spots but I am hoping RAW is a
better way.

Typically shooting action shots like that, i dont have time to look
at the histogram each shot and while i can get the shot close, its
always just some little thing that still gets past me. The limited
dynamic range of digital can be disapointing like that, as can the
10d's metering. Still a great camera though i think.
--

My pictures may only be worth 500 words, but I'm taking a Photographic English Composition course.

Grateful for any constructive criticism regarding my photos, composition, lighting, technique, etc.

http://www.pbase.com/vsteven
 
It's not that simple. All the user sees (normally) are the JPG file, which is the final cumulative result from the raw sensor information, with the conversion performed in the camera.

It appears that, at least for the 10D, the in-camera conversion is partially responsible for the blown-out highlights. Thus the problem can be reduced by using a better program to do the conversion off-line.
Whether your in RAW mode or JPG mode, if the exposure is blown out,
it's blown out. The file mode shouldn't make any difference.
True, RAW mode retains more information than JPG, but if it's blown
out information, it's still just more blown out information.

Or are you looking at it from the angle that some of it may be
converted down from 12 bit color to blown out information because
of the 8 bit color (JPG) vs. the saved 16 bit color (really still
12 bit RAW) TIF format? Am I off here? I can't see that it would
make any appreciable difference.

VES
 
Okay, I can see what your saying, that perhaps in marginal cases the processing can make a difference. But I think it still stands that if the information captured by the lens and exposure is optically blown out, no conversion can create (or lose) the lost information.

So the real answer to his question would be, It depends. Was the information actually lost in the exposure, or was it lost in the conversion/compression process? The conversion process of which you speak, isn't it the same as what I was suggesting about the conversion of a 12 bit color image to an 8 bit color image?

VES
It appears that, at least for the 10D, the in-camera conversion is
partially responsible for the blown-out highlights. Thus the
problem can be reduced by using a better program to do the
conversion off-line.
Whether your in RAW mode or JPG mode, if the exposure is blown out,
it's blown out. The file mode shouldn't make any difference.
True, RAW mode retains more information than JPG, but if it's blown
out information, it's still just more blown out information.

Or are you looking at it from the angle that some of it may be
converted down from 12 bit color to blown out information because
of the 8 bit color (JPG) vs. the saved 16 bit color (really still
12 bit RAW) TIF format? Am I off here? I can't see that it would
make any appreciable difference.

VES
--

My pictures may only be worth 500 words, but I'm taking a Photographic English Composition course.

Grateful for any constructive criticism regarding my photos, composition, lighting, technique, etc.

http://www.pbase.com/vsteven
 
I think the problem is that he posted an image he worked on heavily to remove the seriously blown highlights.

If he had posted the original image, it probably would have been clearer. You can see a loss of detail in the lions face but it's been toned down and is no longer bright white as I assume it was in the original.
..how exactly is this shot ruined by blown highlights ?!? To me it
looks perfectly good. You NEED some highlights in a picture.
Otherwise it's gonna look dull. Just like you need sweet moments in
life to help you deal with the sour ones... ;o)

seriously - you seem to be complaining about a picture that could
hardly be exposed any better. Take a good look at it again and then
come back here.

David
--
Bob Lindabury
 
The "marginal cases" is about 1/2 to 1 stop in exposure, i.e., if the JGP appears to be slightly blown in highlights, reducing the exposure 1/2 stop can eliminate the problem. Using 12 bit or 8 bit by itself does not give you this capability, i.e., if you convert an overexposure picture fron RAW to 48-bit TIFF and then trying to reduce the expoure, you still cannot rescue the highlight.

This is very visible when you examine the histogram. Blown highlights mean that there is a spike at the upper end (255); reducing the exposure in JPG or TIFF (regardless the number of bits) would only shift the spike from 255 to 245 (say), with no data above 245. Reducing the exposure in RAW can eliminate the spike and there will be continuous data between 245 and 255.
Okay, I can see what your saying, that perhaps in marginal cases
the processing can make a difference. But I think it still stands
that if the information captured by the lens and exposure is
optically blown out, no conversion can create (or lose) the lost
information.

So the real answer to his question would be, It depends. Was the
information actually lost in the exposure, or was it lost in the
conversion/compression process? The conversion process of which
you speak, isn't it the same as what I was suggesting about the
conversion of a 12 bit color image to an 8 bit color image?
 
I would assume that a good way to work would be to perhaps convert
the image twice, once to bring up the shadows and then again for
the highlights and do a blended composit in photoshop ?
This would certainly work ... but it's not the way I approach the problem. It takes a lot of time to blend the two files back together properely, and I think you can get similar results with less work. ( There are actions to combine the two files, but from everything I've seen, none of them do a good job. The images come out looking unnatural, and with most of the contrast lost in the blend. )

Personally, I convert a 16-bit tiff file, then do some tricks in Photoshop, either a complex curves adjustment, a careful selection, or both. You can use the select color feature to select the brightest highlights in your picture, then invert ( and feather!! ) the selection, to get the shadows and mid-tones ... and then apply the curves to everything but the highlights you don't want to "blow out" in post. ( This takes some doing in 16-bit mode, but if you look back over my Workflow page, I explain how to get around Photoshop's 16-bit limitations... )

Now, I'm a landscape photographer, and keeping a deep blue sky typically means underexposing the foreground by about two stops ... between these two tricks, I'm able to do a good job of pulling the shadows back up, and keeping the whole image looking natural.
Disapointingly now i guess my 1gig CF card isnt going to seem as
big, and theres yet another software title to purchase, I belive
the Capture One is the best raw convertor out right now ?
Don't know; I'm still using Canon's RAW converter. I don't know whether C1 is supposed to be the best converter for image quality , or just for workflow ... I've got my own workflow, and I only use RAW when I'm expecting a really good photo, or when I need the extra tonal range to pull up shadows without overexposing highlights ... so for me the issue is dealing with all these files. I extract the embedded jpeg for browsing and previewing ( any thumbnail software works for this ), then when I'm ready to use that file, I pull up the CRW and build a 16-bit tiff. I don't think C1 would revolutionize my workflow, and so far, it hasn't interested me more than $500 does...
Thanks for the complements on the image as well. Ive got about
4000 shots of my trip to pittsburgh so look for alot more once i
get around to resizing and everything
Excellent! I'm sure I'm not the only one waiting to see them...!
 
In terms of compensation, i almost always shoot at least -1 or
sometimes even more. I find without any comp there are always to
many blown highlights
This is what you'd do in RAW, to prevent blown highlights. It's
just you can get away with more of it; when you underexpose in
jpeg mode, there's only so much you can bring the shadows back up.
You'll have a lot more shadow detail, and the colors in the shadows
will be a lot more accurate, shooting in RAW. But you'll still
have to underexpose, and then pull the image up, without blowing
the highlights in post.
I agree with Forrest, and have until recently always erred on the Dark Side, in order to avoid overexposure, which is kinda overcompensating in a rational method. The problem is noise. When bringing up the shadow areas I do indeed find noise present to 'ruin' the image

this Luminous Landscape article points out WHY noise is heavier in shadow areas and why underexposing CAN be a problem more than overblowing highlights

http://www.luminous-landscape.com/tutorials/expose-right.shtml

While I ALSO agree with the article I know that there is a happy medium and there's no shame in getting whats RIGHT in any method possible (especially by cloning in to cover the odd small blown out highlight) also note that most cameras are EXTRA touchy when they do the 'flashy overexpose marker' a bit too easily so you can get away with a bit more than you think sometimes)

All in, I hate overblowing highlights, but I know I can repair many of them, and so long as they PRINT ok, thats all that matters :D
 
Vern

Blown 'highlights' are subjective to several factors. Your camera type and conversion method are the 2 main factors though so with that in mind, take a look at what is commonly referred to as a blown highlight with a D30 RAW image being processed in C1:
(WARNING!! Large images of 3-500k)
http://www.vizualgroove.com/c1/blowout1.htm

The preview is configured to show blown highlights within 3 points of max using a red overlay and blown shadows within 3 points of max using a cyan overlay. By adjusting EC down a bit, the blown highlights are under the range I specified.

The values of what the software considers over or underexposed can vary greatly based on individual channels. Also, typically blown highlights as are most often discussed here usually mean RGB values of 255 respectively. In the next example, all RGB values are 255, yet a minor negative EC brings the blown highlights within my set specs.
http://www.vizualgroove.com/c1/blowout2.htm

So you can see there is more to the process than meets the eye and that is why there is so much confusion over the subject. Unless there are extreme cases of overexposure, a RAW file, at least a Canon RAW file coupled with C1 can usually bring out some sort of detail.
HTH,
VG
So the real answer to his question would be, It depends. Was the
information actually lost in the exposure, or was it lost in the
conversion/compression process? The conversion process of which
you speak, isn't it the same as what I was suggesting about the
conversion of a 12 bit color image to an 8 bit color image?

VES
It appears that, at least for the 10D, the in-camera conversion is
partially responsible for the blown-out highlights. Thus the
problem can be reduced by using a better program to do the
conversion off-line.
Whether your in RAW mode or JPG mode, if the exposure is blown out,
it's blown out. The file mode shouldn't make any difference.
True, RAW mode retains more information than JPG, but if it's blown
out information, it's still just more blown out information.

Or are you looking at it from the angle that some of it may be
converted down from 12 bit color to blown out information because
of the 8 bit color (JPG) vs. the saved 16 bit color (really still
12 bit RAW) TIF format? Am I off here? I can't see that it would
make any appreciable difference.

VES
--
My pictures may only be worth 500 words, but I'm taking a
Photographic English Composition course.


Grateful for any constructive criticism regarding my photos,
composition, lighting, technique, etc.

http://www.pbase.com/vsteven
 
I apprecaite the explaination and examples. Yet another reason to use C1LE now. :)

VES
So the real answer to his question would be, It depends. Was the
information actually lost in the exposure, or was it lost in the
conversion/compression process? The conversion process of which
you speak, isn't it the same as what I was suggesting about the
conversion of a 12 bit color image to an 8 bit color image?

VES
It appears that, at least for the 10D, the in-camera conversion is
partially responsible for the blown-out highlights. Thus the
problem can be reduced by using a better program to do the
conversion off-line.
Whether your in RAW mode or JPG mode, if the exposure is blown out,
it's blown out. The file mode shouldn't make any difference.
True, RAW mode retains more information than JPG, but if it's blown
out information, it's still just more blown out information.

Or are you looking at it from the angle that some of it may be
converted down from 12 bit color to blown out information because
of the 8 bit color (JPG) vs. the saved 16 bit color (really still
12 bit RAW) TIF format? Am I off here? I can't see that it would
make any appreciable difference.

VES
--
My pictures may only be worth 500 words, but I'm taking a
Photographic English Composition course.


Grateful for any constructive criticism regarding my photos,
composition, lighting, technique, etc.

http://www.pbase.com/vsteven
--

My pictures may only be worth 500 words, but I'm taking a Photographic English Composition course.

Grateful for any constructive criticism regarding my photos, composition, lighting, technique, etc.

http://www.pbase.com/vsteven
 
I agree with Forrest, and have until recently always erred on the
Dark Side, in order to avoid overexposure, which is kinda
overcompensating in a rational method. The problem is noise. When
bringing up the shadow areas I do indeed find noise present to
'ruin' the image
I think if the contrast range in the scene is limited enough, it's probably better to overexpose a bit, then pull the image down , to control noise, and get better definition throughout the image. It's very true that there's going to be far more noise in the shadows, then in the highlights ... and this is why you're better off using ISO 400, then trying to push 200.

But, like you said, it's about finding that happy medium ... if your scene has a lot of contrast, this isn't much of an option, and you have to choose between some noise, and big patches of pure white, or at least too bright.
this Luminous Landscape article points out WHY noise is heavier in
shadow areas and why underexposing CAN be a problem more than
overblowing highlights
This is a very interesting article ... but I think the underlying point is about how RAW converters should be , and not how they really are.
possible (especially by cloning in to cover the odd small blown out
highlight) also note that most cameras are EXTRA touchy when they
You're the second person to suggest this -- cloning detail back into blown highlights. I'm curious about this, can't fathom how it would work in a landscape. Could you tell me how I would go about cleaning out these blown highlights???



( By the way, the original was a couple stops darker than this looks, to avoid blowing the highlights any worse, and I can't see any noise in this image ... although others show a bit of it. )
 
the processing can make a difference. But I think it still stands
that if the information captured by the lens and exposure is
optically blown out, no conversion can create (or lose) the lost
information.
Well, you're half right. Highlights aren't blown optically by the lens; this is the sensor, and the exposure. Remember, the lens controls the aperture, and the camera uses a shutter to expose the sensor to just the right amount of light, at that aperture. But this is totally off topic...

The way you're going to see a difference between RAW and jpeg modes, with regard to lost highlights, is that RAW mode gives you quite a bit more lattitude in the shadows ... you can get away with significant ( 2 to 3 stops ) underexposure, then pull the shadows back up, while preserving very high image quality. The initial underexposure preserves the highlights, then your Photoshop magic normalizes the image.
 
VES
So the real answer to his question would be, It depends. Was the
information actually lost in the exposure, or was it lost in the
conversion/compression process? The conversion process of which
you speak, isn't it the same as what I was suggesting about the
conversion of a 12 bit color image to an 8 bit color image?

VES
It appears that, at least for the 10D, the in-camera conversion is
partially responsible for the blown-out highlights. Thus the
problem can be reduced by using a better program to do the
conversion off-line.
Whether your in RAW mode or JPG mode, if the exposure is blown out,
it's blown out. The file mode shouldn't make any difference.
True, RAW mode retains more information than JPG, but if it's blown
out information, it's still just more blown out information.

Or are you looking at it from the angle that some of it may be
converted down from 12 bit color to blown out information because
of the 8 bit color (JPG) vs. the saved 16 bit color (really still
12 bit RAW) TIF format? Am I off here? I can't see that it would
make any appreciable difference.

VES
--
My pictures may only be worth 500 words, but I'm taking a
Photographic English Composition course.


Grateful for any constructive criticism regarding my photos,
composition, lighting, technique, etc.

http://www.pbase.com/vsteven
--
My pictures may only be worth 500 words, but I'm taking a
Photographic English Composition course.


Grateful for any constructive criticism regarding my photos,
composition, lighting, technique, etc.

http://www.pbase.com/vsteven
 
While your statement regarding lattitude in the shadows is correct, that is not what reduces or eliminates blown highlights.

We are talking about a case of identical exposure being used between JPG and RAW, not underexposing in RAW and then pull it back up.
The way you're going to see a difference between RAW and jpeg
modes, with regard to lost highlights, is that RAW mode gives you
quite a bit more lattitude in the shadows ... you can get away with
significant ( 2 to 3 stops ) underexposure, then pull the shadows
back up, while preserving very high image quality. The initial
underexposure preserves the highlights, then your Photoshop magic
normalizes the image.
 
We are talking about a case of identical exposure being used
between JPG and RAW, not underexposing in RAW and then pull it back
up.
I understand this, but I think it's mostly accedemic. Unless I'm doing this completely wrong, my experience has shown that using RAW instead of Jpeg format gives you about another 1/2 stop of lattitude for the highlights, versus another 2 or so stops going after the shadows. ( That's a total of maybe 1 stop over, and 3 to 3 1/2 under. )

So, in practice, I don't think just wrapping your pictures up in a better file format is going to make a great deal of difference ... but using RAW mode, a darker exposure, and some Photoshop processing will. It's the end result that matters, and getting that end result ( a print or a web image ) as good as you can make it is a process.

Anyway, thems my two cents...
 
There is a limited edition of capture one thats $99. I guess it offers most all the same features. Im downloading it today but from the site it shows you can do levels and curves to the actual conversion.

I try shooting some RAW shots and see what i can get out of it. I really dont have the experience and knowledge to make a good report on how useful it may be, but can share my impressions of the images from an asethic standpoint
 
That shot was after it was toned down and had some going over with the clone tool. I guess from alot of the opinions that I must of done a pretty good job.

I was just really curious if some sorta raw convertor would of allowed me to bring back some details in that region as it may not of been 100% blown, but at the point where it looks pure white to the eye.
 
I've followed this thread, and had started a similar one like it in this forum about three weeks ago (the focus of that thread was how to deal with the situation where your dynamic range was greater than your camera could handle).

I just came across a piece of software that combines multiple exposures (one for highlights, one for shadows, etc.) using different algorithms and you pick the best one for your image. It automates what could be complex masking. I've downloaded the eval copy, and will try it out this weekend, so no recommendation yet:

http://www.multimediaphoto.com/photomatix/index.html
Ive mostly always shot in fine jpeg, but lately ive been more and
more annoyed with minor blown highlights in my shots. Usually its
something like just where the sunlight happened to reflect off an
animals back or something minor like that. It just seems to ruin
what would be an otherwise perfect shot.

My question is IF i shoot in raw mode, would i have that much
success in dealing with these blown highlights ?

heres an example photo

http://www.pbase.com/image/21192229

Would i of been able to recover where its blown out on the lions
face ?

I usually try to use the clone stamp or selective color to tone
down the glaring whiteness of those spots but I am hoping RAW is a
better way.

Typically shooting action shots like that, i dont have time to look
at the histogram each shot and while i can get the shot close, its
always just some little thing that still gets past me. The limited
dynamic range of digital can be disapointing like that, as can the
10d's metering. Still a great camera though i think.
 
Looks intresting, I will try out the demo as well.

Now if only we could get a sensor that could record a 10 stop range....other than the one called our eyes.
 

Keyboard shortcuts

Back
Top