[betsypdx]

So people elsewhere are arguing over raw vs. jpeg comparing different shots. One of my favorite features of the K10D is having RAW+ as a feature. As long as you have a lot of SD cards, it's great!

So here are a couple of shots - shot in both jpeg (default settings) and RAW with the K10D. Post-processing: exposure changes and white balance in ACR; levels, color balance, and contrast for the jpegs etc. but nothing major (I'm not a pp pro - perhaps one of the reasons I like taking lots of pics and getting it as "right" as possible in camera). Some sharpening of the RAW before resizing and of both after resizing.

Tell me which is which (and see more from this small batch jpeg vs. raw Photo Gallery by betsypdx at pbase.com- all taken with the Super-Takumar 85mm f1.9).

(And, by the way, not all of these were perfectly exposed - some needed +1 or more EV in ACR, and significant levels "tweaks").

E


F


O


P



E1


F1


If anyone gets all of them (in the gallery) right, they'll get some kind of prize.

[Edvinas]

Can't see the difference.

[anthropas]

Raw photo's are:

E, O, E1

For my prize, can I have your nice Takumar lens? Pretty please!

[roentarre]

Yeah, I am tempted to say E O E1 are raw as well.

For some reasons, the colour showed up better in these 3 images.

I think the true difference is clear after developing these shots. I had a few done and the difference is clear

[Photo Tramp]

I can see some little differences but I wouldn't know which are which Betsy. The all look excellent to me.

[squarerigger]

They are all excellent photographs and I had to really make several comparisons before I too came up with E, O, and E1 as my favorites. Now, I have to qualify this by saying they are may favorites but I have no idea if they were the RAW or the JPEG shot. I think there is also a lot to be said about your ability to take a correctly exposed photo to begin with that can make for few adjustments after the fact either in RAW or JPEG.

Gary

[KennJ]

The correct answer is they are all JPEG's as you can not post a RAW file :0)
seriously, this is a way overblown argument all in all. Photo manipulation after the fact is an art in it's self from what I have seen. What works best for one may not do so well for another person. And even though RAW has more ability to manipulate the more basic aspects of a photograph. It is no panacia for decent shooting in the first place.
I am personally just not that good at the RAW software yet and so tend to only use it for shots I want to be able to fix like bright sky dark earth type situations.And Bracket shots can do just as well butr are also alot of work mixing together.
And I have to agree that E,and O are a bit better but see very little to recomend between the two jukbox shots and find F1 to be a touch better to my eye.
Kenn.

[Peter Zack]

I was talking to my son the other day, a real computer geek who is a digital graphic designer and he called the difference 'dithering'. Because the Jpeg is 8 bit vs. 16 bit in RAW The compression used in Jpeg's creates 'artifacts' added to the Jpeg. It's tough to tell from your images but on my screen with my images, the same RAW + (the only way I shoot BTW), I can see pronounced differences if you look closely. If you take a cloudy sunset picture the RAW image has tons of variations in colour subtleties and the Jpeg will have added random pixels that almost look like small dust spots (more noticable when zoomed in).

The great thing about having RAW + is being able to quickly view all the images taken from the camera in Jpeg and discard the losers. Then take the time to save and work on the good images in the RAW format. Once they are all sorted out i delete the Jpegs as they are just wasting hard drive space. Like you I'm no PP genius and would love to take a course on this subject alone. There is so much that can be done and the leraning curve is massive.

As for your images which are all great and very crisp. The only major thing I can see in the image that helps tell the difference is in the bokeh.

I tried looking at these with out seeing the others answers so I think the Raw images IMO are E,P, F1 but it's hard to tell since they are all Jpeg's when posted here. For example if you look at E1 and F1 theres a yellow graphic in the background on the left. In E1 the red bleeds into the yellow and in F1 there is much less of that bleeding. In the other 2 comparisons the bokeh seems to tell the difference. IMO the bokeh is smoother. But I may be way off base.....

[Kingsofronin]

You cant really tell on the internet on 2 jpegs of similar size 800x600 or lower.

(And when i say you cant really tell, i mean the average everyday person.)

You really need either the original files, or prints to see the difference.

[WMBP]

I would not place a bet, but I'm inclined to guess that the second picture in each case is the one that was converted from raw to jpeg on the computer rather than in the camera. The reason I guess as I did is that the second picture in each case seems slightly less vivid to me, on my computer display. Raw files that haven't gotten much processing typically look a bit washed out.

However, while I hate to be a party pooper, I have to say that this particular test is meaningless. It's like showing someone two prints of the same fairly conventionally exposed photo, in identical frames, and asking an observer from a distance of five feet to tell which was processed at Walgreens and which was processed at the pro photo lab downtown. The possibility that the difference isn't obvious doesn't mean that serious photographers should start getting all their prints made at Walgreens.

You said that you didn't post-process the files much. You may not have done so personally, but remember that the files that come from the camera already converted to jpeg have also already been post-processed in the camera - that's what all those adjustment settings in the menus are about (they don't affect raw files!). So if you further post-processed the jpegs, you were tweaking images that have already been tweaked once by the camera. If you did only basic adjustments to the raw images, those basic adjustments were the first post-processing given to those files.

It's also true that, if the photo was well exposed in the first place and assuming that you are trying to end up with a jpeg that has a conventional "good" exposure, then you won't see a lot of difference. To put it differently, you can certainly post-process a raw file so that it ends up looking pretty much identical to the photo you get from the camera. But what does that prove? A brilliant writer doesn't have to use all the words he knows in everything he writes and he doesn't even have to try hard to limit his vocabulary. A twelve-year old however can't use words he doesn't know.

It's precisely this sort of test that keeps the jpeg vs raw debate sort of artificially alive.

A better test? Take some photos with challenging exposure problems - high dynamic range, or contrasts between dark shadows and brighter brights, or very rich colors. Then post-process the files attentively, making the most of what can be done with the raw file and what can be done with the file that's already been processed in the camera and thus has much less data to work with. Then let's compare the results.

But even that sort of test is misleading, because it tends to suggest that raw is mainly advantageous when the exposure is problematic. Saving the raw data always means means that you get more data to work with on your computer. And sometimes that extra data means that a shot that looks really good starting with a jpeg can be processed to look fantastic if your computer's post-processing software has all of the raw data to work with. The extra data may not matter in every shot. But when do you want to make the decision about whether it was important? Before you shoot the picture, or after?

I'm not a raw evangelist. I don't care if anybody else shoots raw or jpeg. More than that: I am well aware that there are many very good reasons to let the camera do the dirty work for you. It is easier to let the camera do the conversion and initial post-processing for you (although it's not nearly as much easier as jpeg advocates claim). And if you deal only with jpegs, then you can use free, wonderful software like Picasa and may not need to bother with expensive difficult software like Lightroom or PHotoshop. For many types of photos, saving the raw data is overkill - say, for employee ID cards or security photos. And file size difference is indeed a major disadvantage of raw files. The more photos you take, the more of a hassle that size difference becomes. With jpeg you get many more photos per card, and you spend a LOT less storing your photos on your computer.

But that's it. Anybody who thinks that letting the camera throw away a ton of potentially valuable data, and letting the camera decide which data to throw away and which data to keep, is free to do so, but shouldn't claim that this is done in the service of producing the best quality photos at the end of the day. The jpeg-advocate side of this argument logically is about as sound as the argument made by folks who say it's safe not to wear a seatbelt, because they've done it for years and they've never had an accident. God bless 'em.

Will

[WMBP]

Saving the raw data always means means that you get more data to work with on your computer.

Let me anticipate a technical objection to that claim. You might think it's not true if the subject is extremely high-contrast, low dynamic range - say, a photo of a painting that has a perfectly white background with a single perfectly black stripe of paint down the center.

But even then, you get more data. It's like knowing that an important value is 7.000000. Those extra zeros matter. 7.000000 does not equal 7.0. If you've only got 7.0, you don't know that the actual value was 7.014293.

Will

[*isteve]

TO be honest, at this size noone can tell. If you post 100% crops then maybe we could spot the difference?

[tlfaust]

I just think that the use of light in sets 2 and 3 is typical Betsy. Really striking. Your work almost always is eye-catching!
Thanks for posting!

[carpents]

Let me anticipate a technical objection to that claim. You might think it's not true if the subject is extremely high-contrast, low dynamic range - say, a photo of a painting that has a perfectly white background with a single perfectly black stripe of paint down the center.

But even then, you get more data. It's like knowing that an important value is 7.000000. Those extra zeros matter. 7.000000 does not equal 7.0. If you've only got 7.0, you don't know that the actual value was 7.014293.

Will

TO be honest, at this size noone can tell. If you post 100% crops then maybe we could spot the difference?

Actually, both of your responses underscore that for the vast majority of uses, the differences between RAW and JPEG is nil. In other words, if you're only posting the shot online or making small prints, you'll not tell the difference between a properly exposed and shot JPEG and the corresponding RAW.

Of course, the real reason why I shoot RAW is to help me when the shot isn't done perfectly. Odd color casts are much harder to fix in JPEGs than in RAW, as are shots with high dynamic range. It also reduces my workflow, but others with a different flow may find the opposite.

[WMBP]

Actually, both of your responses underscore that for the vast majority of uses, the differences between RAW and JPEG is nil. In other words, if you're only posting the shot online or making small prints, you'll not tell the difference between a properly exposed and shot JPEG and the corresponding RAW.

carpents,

I understand that you are also using raw, so I'm not trying to persuade you of anything.

However, as a technical matter, I disagree with the first sentence, both with the "vast majority" part, and with the word "nil." As I said, the difference between 7.014293 and 7.0 is not nil. As for the words "vast majority," they aren't very meaningful. If you say "the vast majority of consumers don't need to shoot raw," I can agree with that. The vast majority of consumers don't need a DSLR, either. But let's narrow the focus down to people who are serious enough about photography to have spent more than $1000 on their current camera and an extra lens. Moreover, let's talk not about the vast majority of these people - since there are no raw people vs jpeg people - but the photographs that this small set of photographers take. Okay, let me just talk about myself. I just post-processed 400 images from a shoot in a natatorium where the lighting was truly awful. I can say with absolute confidence that the results I have now, while still not great, benefited substantially from the fact that I was shooting raw. Not just some of the images, not even the vast majority of them, but all of them. I know this because the first time I shot in that natatorium, eight months ago, I shot jpeg. The photos I ended up with were unusable.

I should also point out that I disagree in part with i*Steve's comment that you can't tell the differences here because the crops are too small. He and I seem to be allies here, so I don't want to suggest that he's wrong, just that I would personally like to be a bit more precise. He's definitely right that, the smaller the crop, the less the data in either photo, and the less the data, the less range there is for difference, and I take it that was his point. And if the in-camera conversion started out pretty close to the on-computer converted raw file, well, then scaling both images down is indeed likely to produce two files that are nearly identical.

But the technical point I would like to add is, it's just a matter of percentages, and at least some of the time, if you can post-process a raw file well - say, really bringing out those subtle shading differences in clouds at sunset - the differences will remain even when you scale the images down.

It's very simple: the more data you start with, the more options you have. Some people don't WANT options. That's perfectly understandable and I have no problem with it.

Will