Impressive. Does this only work in raw, or can you do this with jpg as well? I don’t think any of my cameras shoot in raw :\
It won’t work in jpeg. Jpegs throw away most of the data. A jpeg has 256 brightness levels, max, as I understand it. 256 shades of any particular color. A 14 bit raw has about 16,000. There’s simply no way for a jpeg to store all that extra data. If you’re serious about any sort of tricky shots/post processing or getting the most out of images, you have to shoot raw.
Sure, but shooting RAW and using the histograms means you can see where your overexposure areas are and compensate accordingly. Or you can use exposure bracketing to open up the range.
Even my entry-level dSLR will show clipped highlights when you review the picture on the LCD (they flash with a black outline so you can see precisely where the problem is, even without the histogram).
Remember, shoot raw and get the shirt: http://store.froknowsphoto.com/
AFAIK, there is no brand of camera that shows a true raw histogram. You can easily test it for yourself, and overexpose something so that your camera is showing that you are clipping red, green, and blue channels. If you can recover that detail by pulling back the exposure in post, it not showing you true raw clipping.
While I do ETTR, I don’t do it super aggressively. I generally end up exposing about a a third to a full stop over what I would normally expose a scene at. When you ETTR, your post workflow is slightly different. ETTR files tend to need additional contrast and possibly saturation. (I find that whenever you go plus exposure in ACR or Lightroom, i.e. you have an underexposed file, you get a additional contrast and saturation that you sometimes have to back off. When you go minus, it’s the opposite.)
Don’t forget that you have the incredible advantage of being able to try stuff out on RAW format, and not have it be permanent.
If you push/pull film, it’s a one-shot thing, but you can digitally push a RAW photograph, and if you don’t like it, you can dial it up or down and try again.
This is the biggest advantage to digital photography that I see, other than not having to deal with film at all.
Exactly right. If you try that on a JPEG, you’re going to get posterization as you’re trying to spread a small amount of data onto a larger area.
To make it very simple. Imagine all your dark tones have a range of 10 brightness levels, from 0 to 9. (And in SenorBeef’s image, the details in the face are pretty much exactly at 9 or under in each channel. Those will print as pure black on many, if not most, printers.)
So, you want to take those 10 brightness levels and take those 0-9s and pump them up to about 50-110-ish. So what can you do? You have only 10 discrete values that you have to spread across a 60-point range. So you’re going to have holes in your histogram. 0 might become brightnesss leve 50. 1 might become brightness level 56. 2 might be 62. And so on. So you’re missing the smooth gradations, as you don’t have anything with 51,52,53,54, or 55.
With a raw capture, you have 12- or 14-bits of information, meaning that instead of 256 possible brightness values, you have 4096 or 16384 values. So what was 10 brightness levels in that shadowed face in the JPEG might be, say, 128 values. That will fit fine in the 60-brightness level range we need when we convert the raw to an 8-bit file. So you will have much more detail, provided the camera can differentiate such low luminosity values well, and no posterization as you’re not going from a smaller range of values to a larger one, but vice-versa.
Does that make some sense? It’s a little simplified, but not by much. The basic idea is when you take a small amount of values and try to spread them across a larger range, you’re going to have holes in your data.
I never thought of it in those terms. But you still need that original capture to be exposed properly for the data you’re trying to capture. For me, at point of capture, I know how I want the photo exposed. Plus you can do this with JPEGs, too, to a much smaller extent. You can usually “push” or “pull” a properly exposed JPEG at least within a half stop range before the problems mentioned in my previous post become visible. (And, depending on the file, you may be able to go a full stop or so.) Raw just affords you more headroom and more information that you can manipulate.
For me, the biggest advantages of digital over film are: that LCD on the back where you can judge exposure and lighting on the fly. Biggest, biggest advantage. No more Polaroid backs to check your lights. And you can fine tune your exposure like never before. Then, I like the ability to change ISOs on the fly. Thirdly, high ISO performance is insane compared with film. These days, 6400ISO and maybe even higher looks no worse than 400 speed 35mm film. And, of course, no waiting for the film to get back from the processor. On the other hand, we have all this post-processing technology stuff to learn now, which I actually do like, but some see as a pain in the ass.
Oh definitely- I was just saying that assuming your exposure is correct, the ability to tweak your photos over and over is a great advantage. I’m half surprised that there hasn’t been a photographic exhibit of the same exposure that’s been post-processed in a bunch of new and interesting ways.
I also really love the LCD on the back- not necessarily on the fly (still use the viewfinder), but to view the shot I just took and see what it ended up like.
The remarkable performance in low-light afforded by modern technology astounds me- between the high ISO performance of the sensors and the image stabilization of many lenses, it’s possible to take decent indoor photos without flash using something as slow as a f3.5-4.5 zoom.
Oh, same here. I shoot with dSLRs and it’s almost always through the viewfinder. “Live Mode” on dSLRs wasn’t even that common until the last few years. (I would say it was about 5 years ago with the 5D Mark II and the D3 that I began to see it.) The only use I’ve had for it is for composing really awkward angle shots (like right on the ground or overhead), but even there, unless you have a reticulating screen (ETA: or have some kind of external monitor set-up), which the pro-level Canon and Nikons don’t yet, it’s still a bit difficult, and I more than likely will just shoot it blind.
Except Ken Rockwell says raw is linear, while jpg isn’t.
Yes, that is correct. Raw is linear, and jpeg is gamma compressed. I’m not entirely sure what you’re objecting to. I specifically avoided using the term “linear” so as not to get into this somewhat complicated subject, but, hey, we can do it.
ETA: And, oh, I just looked at the Ken Rockwell article you linked to. Let’s just say that (as usual) I very strongly disagree with his opinions. This is among the stupidest things I’ve ever read: “I never shoot raw. Why would I? Raw is a waste of time and space, and doesn’t look any better than JPG even when you can open the files.”
Here’s Adobe’s whitepaper on raw and linear gamma so you could see what’s going on. Page two shows you what a linear distribution looks like. This is why I avoided the term “linear.” It doesn’t look like what I think most people think of as a “linear” distribution of brightness values. The gamma corrected distribution is what I think most people would intuitively think of as a “linear” distribution because, perceptually, that is what it looks like.
Ken Rockwell is not taken seriously amongst the photography community, FWIW. His articles are usually fairly silly. There might be a useful nugget here and there, but…
Pretty much. And, to be fair, that article is from 2009, and I think he’s changed his thoughts on raw vs JPEG. But I can’t figure out what he’s saying here: “Thus in the shadows where this might matter the two are the same, since the full 12 bit resolution in the dark areas is preserved by the non-linear coding. Even if the two formats differed in dark resolution the sensor noise is still greater than one LSB anyway making it a moot point.” It sounds like he might be saying something smart and technical, but beats me what he’s trying to say. They’re most clearly not “the same” as your example shows. You can pull incredible detail from a shadow in a raw file that you simply can’t with a JPEG.