good histogram = dark images in surf pics...

[msongs]

aloha, have started doing some surf/surfer pics here in hawaii. I use a canon rebel t3i  and shoot in shutter mode and let the camera do the f stop. Tried shooting at 1000. the preview histograms always looked really well balanced and centered. when i got home I converted the canon raws into dng and loaded into CS4. the images were quite dark!. If I opened them in CS4 and did a levels check the histogram was right centered and the only slider adjustment really available made them even darker (moving the left triangle over towards the center to the edge of the histogram).

Of course surf means high contrast, dark waves and white foam so there is a learning curve for sure. I want to use a higher than 1000 shutter speed as the pics were just a bit blurry. the main issue is the dark pics tho. Using the on camera histogram does not seem to be that valuable as a reference. Should I make the settings so the in camera histogram is way over to the right more? the shutter speed is the most important setting. any suggestions?

Marshall aka Msongs

[Vladimirovich]

aloha, have started doing some surf/surfer pics here in hawaii. I use a canon rebel t3i  and shoot in shutter mode and let the camera do the f stop. Tried shooting at 1000. the preview histograms always looked really well balanced and centered. when i got home I converted the canon raws into dng and loaded into CS4. the images were quite dark!. If I opened them in CS4 and did a levels check the histogram was right centered and the only slider adjustment really available made them even darker (moving the left triangle over towards the center to the edge of the histogram).

Of course surf means high contrast, dark waves and white foam so there is a learning curve for sure. I want to use a higher than 1000 shutter speed as the pics were just a bit blurry. the main issue is the dark pics tho. Using the on camera histogram does not seem to be that valuable as a reference. Should I make the settings so the in camera histogram is way over to the right more? the shutter speed is the most important setting. any suggestions?

Marshall aka Msongs

1) in camera histogram (unless you have firmware mod from Magic Lantern which may be available or not for your Canon) = OOC JPG histogram... this is not raw histogram

2) using Adobe raw converters is not the best idea to evaluate raw histogram either - albeit with certain setting and modified profiles you can get close

3) www.rawdigger.com

it is worth it...

[ErikKaffehr]

Hi,

If you would post one of those DNGs it may be easier to help.

Best regards
Erik

aloha, have started doing some surf/surfer pics here in hawaii. I use a canon rebel t3i  and shoot in shutter mode and let the camera do the f stop. Tried shooting at 1000. the preview histograms always looked really well balanced and centered. when i got home I converted the canon raws into dng and loaded into CS4. the images were quite dark!. If I opened them in CS4 and did a levels check the histogram was right centered and the only slider adjustment really available made them even darker (moving the left triangle over towards the center to the edge of the histogram).

Of course surf means high contrast, dark waves and white foam so there is a learning curve for sure. I want to use a higher than 1000 shutter speed as the pics were just a bit blurry. the main issue is the dark pics tho. Using the on camera histogram does not seem to be that valuable as a reference. Should I make the settings so the in camera histogram is way over to the right more? the shutter speed is the most important setting. any suggestions?

Marshall aka Msongs

[Tony Jay]

Hi Marshall.

It seems as if you have discovered a truth about digital imaging: There simply is no predetermined good histogram.
From your description it is clear that you are substantially underexposing.
Your goal is never a pretty looking centred histogram.

The real goal is to maximise the light-gathering capabilities of your camera's sensor.
So, in reality, you want your histogram pushed over to the left but just shy of clipping.
The in-camera histogram is not a good judge of whether clipping has occurred for a variety of reasons, not least the fact that the histogram is derived from a JPEG derivation of the original RAW file.
To help you a bit make sure that the picture style is set to something like "Faithful" that has no changes to contrast factored in.
It makes absolutely no difference to the RAW image but it does affect how the JPEG looks, and hence the histogram that is derived from it.
Ultimately, however, it is only when you load that RAW image into a RAW converter that you will know how much headroom is really available.
My 5D Mark III gives just over a stop of light as headroom between what the in-camera histogram tells me and the histogram in Lightroom.
Your camera will be different - hence the need for experimentation.

I understand that, given the type of shooting you are doing, that you will not be shooting at base ISO even of the light is very bright because of the high shutter speeds required.
So you will have to further determine the trade-off in image quality between ISO-related noise and blurring from shutter speeds that are less than optimal.
Since most cameras that are late models have pretty good noise characteristics I would set the shutter speed to whatever is required and therefore a highish ISO and let the noise take care of itself in post-processing.

Tony Jay

[stamper]

So, in reality, you want your histogram pushed over to the left but just shy of clipping.

Tony you mean the right?

[Tony Jay]

So, in reality, you want your histogram pushed over to the left but just shy of clipping.

Tony you mean the right?

Yes I do! ETTR
R for right!
I cannot believe that I wrote it but I did - unbelievable!
Apart from being distracted a bit while typing that post there are no excuses - just really silly on my part.
Re-read the rest of my post just to make sure that there were no other stupid landmines in there but it seems otherwise OK.

Tony Jay

[LesPalenik]

In a heavy surf, I would dial +1.0 to 1.5 EV, or switch to manual exposure mode.
If you increase the exposure by 1 or 2 stops in postprocessing, you might see some noise, but as Tony mentioned, you can fix it in LR or with some Noise Reduction program.
Topaz Labs is now running a 50% promotion on their excellent DeNoise plugin.  I posted a short review of this plugin with examples on my blogsite at:
www.advantica.wordpress.com

Les


  

[msongs]

hello,

tried to post a 2mb jpg but this site won't allow it. the original dng files are way too big. so let's see if this little tiny file will post. ok that worked ok.

tried adjusting exposure in raw, a few other things like shadow highlight to give detail to the white waters. reading around the internet the basic best exposures seem to be above 1/1000 at f8 and raising iso from 100 if that is needed. and using the highlight warning systems if the t3i has one. I know my Olympus e620 has a good one and it is almost always underexposing pictures lol

[ErikKaffehr]

Hi,

In my favourite raw processor, LR5 I would try to push highlights all way left (providing tone mapping on the surf) and increase exposure to get good mid tones.

Best regards
Erik

hello,

tried to post a 2mb jpg but this site won't allow it. the original dng files are way too big. so let's see if this little tiny file will post. ok that worked ok.

tried adjusting exposure in raw, a few other things like shadow highlight to give detail to the white waters. reading around the internet the basic best exposures seem to be above 1/1000 at f8 and raising iso from 100 if that is needed. and using the highlight warning systems if the t3i has one. I know my Olympus e620 has a good one and it is almost always underexposing pictures lol

[nma]

aloha, have started doing some surf/surfer pics here in hawaii. I use a canon rebel t3i  and shoot in shutter mode and let the camera do the f stop. Tried shooting at 1000. the preview histograms always looked really well balanced and centered. when i got home I converted the canon raws into dng and loaded into CS4. the images were quite dark!. If I opened them in CS4 and did a levels check the histogram was right centered and the only slider adjustment really available made them even darker (moving the left triangle over towards the center to the edge of the histogram).

Of course surf means high contrast, dark waves and white foam so there is a learning curve for sure. I want to use a higher than 1000 shutter speed as the pics were just a bit blurry. the main issue is the dark pics tho. Using the on camera histogram does not seem to be that valuable as a reference. Should I make the settings so the in camera histogram is way over to the right more? the shutter speed is the most important setting. any suggestions?

Marshall aka Msongs

I feel you pain but I would work some more on your exposure with histograms.  As some have noted, the histogram from the in-camera jpeg is sometimes inaccurate. To overcome this type of problem on my Canon 5Dii, I reduce the jpeg parameters with white balance on auto, contrast is low, saturation and sharpening are low. It makes for a crummy image on the preview display, washed out, but now the histogram is much more like what you will see in the raw developer.  You should do a few experiments in the surfing environment to verify what jpeg settings give the most accurate raw histogram, comparing with your raw developer. I find that my jpeg and raw histograms are very close. This means that in the surfer environment I would take a couple of trial shots and view the histograms, making any necessary exposure adjustments so that it is in accord with ETTR. I ignore the preview completely, except for compositional information.This works 100% of the time. It does not follow that the image using default settings in your raw developer will look perfect. All that I guarantee is that all the tonal values will fit in the usable range, with highlights in the appropriate far right of the histogram.  That could result in your raw developer rendering the default image with bright, well exposed, highlights and dark values for most everything else. This is OK. Merely reduce the highlight slider and maybe the whites to taste, while boosting the shadow slider.

Let us know how this works for you.
 

[Tony Jay]

Looking at these images all they require is an increase in exposure and a bit of a tweak in contrast in post-processing.
My goal would be to bring the breaking surf to just shy of clipping.

Tony Jay

[BAB]

Bring your laptop to the beach take a shot crop the area you want the perfect exposure and look at the histogram in lightroom then adjust the exposure on your camera and shoot again until you know what is perfect for your sensor on that camera. If the sunny day is the same sunny day tomorrow and your at the same angle to the light you will know your sensor needs to be set +2.5 EV to make the exposure you want.

I would also suggest a slightly longer focal length

[digitaldog]

when i got home I converted the canon raws into dng and loaded into CS4. the images were quite dark!.

IF you shoot raw, the histogram on the camera is a big fat lie!
This may help understanding why and more about histograms:

Everything you thought you wanted to know about Histograms

Another exhaustive 40 minute video examining:

What are histograms. In Photoshop, ACR, Lightroom.
Histograms: clipping color and tones, color spaces and color gamut.
Histogram and Photoshop’s Level’s command.
Histograms don’t tell us our images are good (examples).
Misconceptions about histograms. How they lie.
Histograms and Expose To The Right (ETTR).
Are histograms useful and if so, how?

Low rez (YouTube): http://www.youtube.com/watch?v=EjPsP4HhHhE
High rez: http://digitaldog.net/files/Histogram_Video.mov

[Vladimirovich]

and look at the histogram in lightroom

www.rawdigger.com , not LR/ACR , is the tool to study raw histograms nowadays...

[digitaldog]

www.rawdigger.com , not LR/ACR , is the tool to study raw histograms nowadays...

Study sure. But if you use LR and have to process images, then that's unnecessary. Learn to use the camera to produce well exposed raw data. I recall spending decades doing this on film when few if any photographers had heard the word Histogram! Then normalize in LR and move on. The histogram is the histogram and histograms really don't provide much info that's useful other than clipping, which LR will easily show you on the actual image. Far more useful to SEE on the image what clips and where then trying to decipher a histogram. At least if your goal is to produce a pleasing image. For some scientific or analytical work, by all means 'study' the histogram.

[bjanes]

Study sure. But if you use LR and have to process images, then that's unnecessary. Learn to use the camera to produce well exposed raw data. I recall spending decades doing this on film when few if any photographers had heard the word Histogram! Then normalize in LR and move on. The histogram is the histogram and histograms really don't provide much info that's useful other than clipping, which LR will easily show you on the actual image. Far more useful to SEE on the image what clips and where then trying to decipher a histogram. At least if your goal is to produce a pleasing image. For some scientific or analytical work, by all means 'study' the histogram.

LR/ACR, at least with PV2012 and the default tone curve, is not good for evaluating overexposure since automatic highlight recovery is employed and an exposure offset is used. With Nikon cameras it +0.5 EV for the D3 and + 0.35 for the D800e. With ACR one can get a reasonable representation of the raw file by adjusting exposure to offset the baseline exposure value and using a linear tone curve by setting the sliders on the main page all to zero and setting the tone curve to linear.

Added in edit (3/18/2014). One must use PV2010 when using this method.

Rawdigger can show overexposed areas as well as the rendered image, and is not limited to presentation of the histogram.

Here is an image taken by the D3 with overexposed yellows as shown by Rawdigger:


And by ACR with the required adjustments. I should have rendered into ProphotoRGB to prevent saturtion clipping rather than into Adobe RGB, but the clipping is reasonably shown.


Here is another exposure that is not clipped in the raw file as shown by Rawdigger:


However clipping is shown in ACR even with the above mentioned adjustments. The clipping in the red channel is likely due to white balance with a red multiplier greater than unity. The green channel shows spurious clipping in ACR.


I agree that Rawdigger is the preferred tool for evaluating raw files.

Bill

[digitaldog]

LR/ACR, at least with PV2012 and the default tone curve, is not good for evaluating overexposure since automatic highlight recovery is employed and an exposure offset is used.

But if that's the process used to render the raw's, how is that a problem?
If one wants to understand the effect of the exposure + processing on their raws, and assuming they want to use LR/ACR, I don't see how this is a problem. The rendering, the clipping (or lack of clipping both exposure and saturation) are the reality of what is currently being seen in the raw converter. Why not simply deal with the processing as we've done in the past with film? That being, your exposure and processing are tied at the hip. Bracket under controlled lighting. Bring raws into LR/ACR. Use PV2012 or not, depending on your preference. What you see is what you get here: exposure plus processing. What am I missing in such an approach?

[Petrus]

If the pictures turn out consistently too dark, exposure metering is doing a wrong evaluation of the subject, really nothing more to it. Quite usual, actually, if the subject is lighter or darker than average, for which the exposure is calibrated. Just simply turn the exposure compensation dial to +1 or +1.5, problem is likely to be fixed, that is why cameras have that function. No need to dwell deeply into conversion theories or purchase better/different RAW converters.

[Vladimirovich]

But if that's the process used to render the raw's, how is that a problem?

the problem is - people shall not think in terms of LR/ACR w/o understanding what LR/ACR does with the data and what is data that LR/ACR deals with... why do you want to cripple their knowledge ?

[digitaldog]

If the pictures turn out consistently too dark, exposure metering is doing a wrong evaluation of the subject, really nothing more to it. Quite usual, actually, if the subject is lighter or darker than average, for which the exposure is calibrated. Just simply turn the exposure compensation dial to +1 or +1.5, problem is likely to be fixed, that is why cameras have that function. No need to dwell deeply into conversion theories or purchase better/different RAW converters.

I agree totally up to the point about better/different raw converters. I don't see how we can separate exposure from processing. But yes, your suggestion about expsoure compensation is exactly how I handle ETTR. The first part is of course gauging via brackting how much to compensate and to do that, at least I have to bring the raw's into LR and see how far I can go before I've really blown out highlights I want to retain. From there, it's exposure 101.