Help!! what did I do wrong? :( - DPChallenge Forums

Tamron SP AF 28-75mm f/2.8 XR Di for Canon

12/25/2007 10:31:11 PM · #1

So lately I have been noticing that my studio shots come out looking pretty awful after converting from raw. My pictures of girls normally don't come out looking nearly as gray...may just be a coincidence but can anyone shed some light on why these look so gross? It is not my converter because I have had plenty of pictures come out looking stunning in every way that are naturally lit right out of the converter.

Here are before afters of the really dull images:

[thumb]625780[/thumb]

[thumb]625781[/thumb]

Here is an example of a picture converted from raw with NO PROCESSING at all that I took that does not have this problem at all:

SamDoe1

Canon EOS-350D Rebel XT

12/25/2007 10:39:13 PM · #2

Use your histogram to see if the exposure is dead on or not, LCD screens can lie. An ideally exposed image should have a histogram looking like a bell curve. From the way it looks from the first shot, it just looks underexposed to me. The second one looks pretty normal to me, but I don't do portraits... The color settings on your camera may have been changed though, might want to check up on that.

Tamron SP AF 28-75mm f/2.8 XR Di for Canon

12/25/2007 10:41:33 PM · #3

Oh interesting, I have NEVER used my histogram before....i should start then. Just to double check, is this what you mean:: bell curve

SamDoe1

Canon EOS-350D Rebel XT

12/25/2007 10:44:41 PM · #4

Originally posted by lovethelight:

Oh interesting, I have NEVER used my histogram before....i should start then. Just to double check, is this what you mean:: bell curve

Yup.

The histogram shows the level of light values falling from darkest to lightest going from left to right. Having a bell curve shows that no part of the image is blown out and no part is completely shadows. Shifting the histogram to the right over exposes, and to the left underexposes. Not really easy to get the perfect one (at least not for me, haha), but that's the best way to make sure you have the proper exposure.

Edit to add:

Here's some more information for you.

Message edited by author 2007-12-25 22:45:57.

scalvert

Canon EF 24-105mm f/4.0L IS

12/25/2007 10:45:12 PM · #5

The dull shots were very light / white shirts on a light background. I suspect it's throwing off your metering.

12/25/2007 10:47:03 PM · #6

I don't meter though, I just adjust the lights until it looks right. How could I do better with it in the future?

Tamron SP AF 28-75mm f/2.8 XR Di for Canon

12/25/2007 10:48:41 PM · #7

Originally posted by SamDoe1:

Originally posted by lovethelight:
Oh interesting, I have NEVER used my histogram before....i should start then. Just to double check, is this what you mean:: bell curve

Yup.

The histogram shows the level of light values falling from darkest to lightest going from left to right. Having a bell curve shows that no part of the image is blown out and no part is completely shadows. Shifting the histogram to the right over exposes, and to the left underexposes. Not really easy to get the perfect one (at least not for me, haha), but that's the best way to make sure you have the proper exposure.

Edit to add:

Here's some more information for you.

But if you are shooting with a white background, wouldn't it shift way to the right? How could you tell if it is right in that situation?

SamDoe1

Canon EOS-350D Rebel XT

12/25/2007 10:55:32 PM · #8

That I'm not sure about. I would think that the values would still be evenly distributed, just that the histogram would fall off quickly towards the right side.

Kind of like this.

Message edited by author 2007-12-25 22:56:07.

scalvert

Canon EF 24-105mm f/4.0L IS

12/25/2007 10:56:01 PM · #9

Originally posted by lovethelight:

I don't meter though, I just adjust the lights until it looks right. How could I do better with it in the future?

Yup, the histogram is your friend. The left side is black and the right side is white. In this case your subject is very light, so most of the data in your histogram curve should be on the far right side (almost white), with a sharp dropoff just before the right edge. If the data is all the way to the right and gets cutoff, then you're blowing your highlights (a small cutoff is OK with RAW). If there's a gap on the right edge with no data in this situation, then you're underexposing.

MelonMusketeer

Nikon D7000

Nikon MF 50mm f/1.4

12/25/2007 11:03:13 PM · #10

I agree with

scalvert about the white shirts fooling the camera's metering system.
Since you don't meter, find out about the "blinkies" . It's called different things by different manufacturers. The LCD display blinks the highlights that are blown out when it shows the "preview" screen right after the shot is captured.
I use it a lot with the Fuji S3 when using older lenses that don't meter with the S3 body in order to keep from overexposing. I just click the exposure up until the blinkies start, then back off a click and it comes out right on target.

scalvert

Canon EF 24-105mm f/4.0L IS

12/25/2007 11:04:46 PM · #11

Look at the histograms of your before and after shots to see the difference. Using your second shot as an example, the "before" histogram looks like this:

It's underexposed. The point I circled in red (just before any significant data) should be all the way over on the right edge.

doctornick

Apple iPhone 13 Pro

Canon 12x IS USM (S3 IS built-in Lens) -- 36-432mm (35mm EQ)

12/25/2007 11:11:25 PM · #12

Next on your shopping list or gift wishlist is a Light meter, it's especially important when you are using strobes.

GeneralE

Canon PowerShot S3 IS

12/26/2007 12:26:18 AM · #13

Originally posted by lovethelight:

But if you are shooting with a white background, wouldn't it shift way to the right?

No, your meter tries to make the average of all the light coming in come out to 18% gray, so you have to over-expose to compensate and make the whites actually white -- you get the same situation shooting scenes with snow.

dwterry

Canon EOS-1D Mark III

12/26/2007 12:58:19 AM · #14

Wow... you've been doing everything without a light meter?!? I'm impressed!!! So I think you've learned a lesson here, but wow, I gotta say it again. Based on what I've seen you produce, I'm really, really impressed!

Now go out and get yourself a light meter! :-)

Or at least ... shoot an 18% gray card and use it to check your in camera metering.

Pentax FA 28-80mm 3.5-5.6 AF - Silver

12/27/2007 12:36:58 AM · #15

i have no idea how to use a light meter.... :(

Puckzzz

Pentax K10D

12/27/2007 01:12:20 AM · #16

Originally posted by lovethelight:

i have no idea how to use a light meter.... :(

me neither. I guess I work the same way you do. I just adjust the lights and settings until it's to my liking.

Interesting thread! I'll try to do my next shoot using the histogram.

NikonJeb

Nikon D850

Nikon AF-S Nikkor 28-300mm f3.5-5.6G ED VR

12/27/2007 01:53:58 AM · #17

Originally posted by lovethelight:

i have no idea how to use a light meter.... :(

Originally posted by Puckzzz:

me neither. I guess I work the same way you do. I just adjust the lights and settings until it's to my liking.

Interesting thread! I'll try to do my next shoot using the histogram.

Yeah, what is it with these people that know how to use the features in the camera anyhow???????

I keep having them tell me about these things and it helps!

Sheesh!!! LOL!!!

Message edited by author 2007-12-27 01:54:37.

eschelar

Canon EOS-7D

Canon EF-S 17-55mm f/2.8 IS USM

12/27/2007 03:28:22 AM · #18

Originally posted by dwterry:

Now go out and get yourself a light meter! :-)

Or at least ... shoot an 18% gray card and use it to check your in camera metering.

I second this.

I was hanging out at the local photo club eating lunch and one of the more experienced guys sharted shooting some odd looking sequences of a girl with a peice of black paper.

Turns out that he's teaching a class on 'metering with the built in camera meter'. His demonstration shots contained a girl with darkish skin, but the shots were metered alternately off a peice of paper, and her skin.

Knowing a few things as I do, I made a few extra suggestions.

#1, for his demonstration tests, I suggested that he meter off a flat black subject, a flat white subject and a grey master. Because this was a class on practical usage, I suggested using a foam divider from his lowepro camera bag. I found it pretty reflective, so I also made a suggestion that he make a 'cheater' grey card by printing a page of grey made in PS or Paint with a fill of RGB R:127 G:127 B:127. It won't be perfect, BUT it does stand a chance of being fairly consistent at least for a few months and this is the purpose of such a thing.

I also posed in the pictures with the girl, and by alternately metering off of her skin and off of my skin, a further comparison was made. I also made sure he metered off her dark asian eye as well. That one used to throw my shots off now and again.

Simply by going through the motions of this exercise will give you a really good idea how your camera's light meter works.

What kinda sucks for you though is that this really only works for constant lighting (studio), natural lighting (outdoors) and non-studio ambient lighting (in the house, at the mall, wherever).

Even worse, the in-camera meter is not actually looking at the light, it's looking at the light reflected from the subject. All kinds of things can throw it off. If the subject is angled, if the subject is more reflective or less reflective, all of a sudden, the subject of your metering becomes the standard by which everything else in frame gets measured by in a relative way. This amount of variance doesn't always make a big difference, but it surely accounts for at least SOME user error that ends up being the cause of odd results.

While the above is very useful for learning the in-camera meter, and is invaluable for many situations, it's actually a fair bit more complicated than using a light meter. That should be good news.

The light meter is like a tiny little 'camera' that only takes a picture of the light. Instead of a photo though, it gives you information. What makes it truly invaluable in a studio is that it can work WITH your strobes which is very difficult to do accurately even if you have modeling lights on your strobes.

First, you set up your light meter like a camera. Give it ISO and aperture (some light meters have a little wheel that you can use to tell it the aperture you are using after it has taken a reading). It will give you information about shutter speed and more. The manual of whichever model you end up getting will tell you more, but it's not that difficult.

All you actually have to do is put the light meter near your subject and fire the light. Your light meter will tell you information about what it saw. This is best done on one light after the other.

After that it's really quite simple. You just compare the numbers of what you got from each source and that tells you what kind of balance you have.

Some people never use them, and you might be able to do a lot without one, but those who don't use them usually fall into two categories: Those who have used them and learned and moved on, and those who have never used them and have a few uncertainties - which can easily lead to confusion and problems.

If you have modeling lights, you might be able to use them to balance your lighting, and using the in-camera meter - especially with the histogram, you might also be able to manage your exposure overall, but it's still well worth it to get a light meter.

fotomann_forever

12/27/2007 03:33:30 AM · #19

I have to say, if you are using a single strobe and perhaps a reflector you can pretty much always get away with just a histogram reading. If you're going any more complex than that, you NEED a flash meter.

Just like a relationship with someone you love, if you lovethelight, you WILL learn how to read it. :-)

eschelar

Canon EOS-7D

Canon EF-S 17-55mm f/2.8 IS USM

12/27/2007 03:36:46 AM · #20

If I recall correctly, Ilove has two strobes and a background light. Can't exactly remember. Got a headache building too, so don't feel like checking back.

fotomann_forever

12/27/2007 03:43:47 AM · #21

Yeah, with main, fill, b/g... you really do need to meter the strobes to get any type of predictable results.

It's simple enough to manage a single strobe with a histo reading, but one may never see a really hot hair light or b/g light on a histogram. And fill ratios: forget about trying to make a guess from a histogram.

dwterry

Canon EOS-1D Mark III

Leica 18-56mm f/3.5-5.6 Vario Elmar-T ASPH

12/27/2007 09:36:24 AM · #22

Originally posted by fotomann_forever:

Just like a relationship with someone you love, if you lovethelight, you WILL learn how to read it. :-)

lol ... good one! :-)

MAK

Leica CL

12/27/2007 10:03:09 AM · #23

I wondered what that wiggley line was for on my LCD.
Guilty as charged though, I never use the histogram it is the one thing I really need to start doing even though

Mr_Pants has told me numerous times before..

what if the point in the middle goes off the top, what is happening there?
I have seen this before on other pics I shot from ages ago.

dwterry

Canon EOS-1D Mark III