Reading mode: Light Dark

What can you do in editing RAW that you can't do in editing JPEGs.

Go to last post
GaryDeM said:
no, i am not wrong. but you are not quite right either. the fact is that a jpeg CAN be just as good as a raw.
Click to expand...
Can be, in theory. Almost never is, in reality.

This isn't hard. The camera sensor and ADC capture a whole bunch of data. That data is what goes into a RAW file.

The data can also be used to generate an in-camera JPEG. However, that's performed by an algorithm with limited adjustments. Furthermore, those adjustments can only be set before you trigger the shutter. It also involves an immediate and irreversible reduction to 8 bits of precision. Data are permanently lost and that is a mathematical fact.

When you shoot RAW, more adjustments are possible, they're possible after the fact, and greater precision can be retained throughout the process until you finally drop back to an 8 bit JPEG.

You may be quite happy with your in-camera JPEG's. They may be outstanding award winning photographs. And, if you (and maybe your "customers") are happy with them, that is just hunky dory. But as a matter of physics and math, the RAW file can virtually always be used to create a "better" image, assuming sound post-processing skills.

Just consider things like noise reduction and sharpening. The algorithms available within, say CS4, are substantially better than those implemented in the camera firmware. And when shooting RAW they can be applied to the original data at full precision versus data than has already been reduced to 8-bits and subjected to lossy compression.

I'll accept that your in-camera JPEG's are simply wonderful; enough to make Ansel Adams cry. But if you shot RAW and processed them carefully, they could be better.
 
malch said:
GaryDeM said:
no, i am not wrong. but you are not quite right either. the fact is that a jpeg CAN be just as good as a raw.
Click to expand...
Can be, in theory. Almost never is, in reality.

This isn't hard. The camera sensor and ADC capture a whole bunch of data. That data is what goes into a RAW file.

The data can also be used to generate an in-camera JPEG. However, that's performed by an algorithm with limited adjustments. Furthermore, those adjustments can only be set before you trigger the shutter. It also involves an immediate and irreversible reduction to 8 bits of precision. Data are permanently lost and that is a mathematical fact.

When you shoot RAW, more adjustments are possible, they're possible after the fact, and greater precision can be retained throughout the process until you finally drop back to an 8 bit JPEG.
Click to expand...
more adjustments are possible. why are those adjustments needed if the image was shot right in the field? that should already be ready to go except for sharpening. any massive adjustment in pping means that the image was shot incorrectly, and now in pping you have to fix it. no thank you, i will shoot my images right in the field. my target when i shoot is 95%(100% on a good day) correct exposure and wb, and of course NO CROPPING. that was done in the camera, it is called composition. i have never understood the endless desire for larger mp sensors, under the view that the pixels are needed for cropping. cropping??? if the thing was composed and shot right the user would not be cropping and would actually use all the pixels he paid for and throwing them out through the deletion process.
Member said:
You may be quite happy with your in-camera JPEG's. They may be outstanding award winning photographs. And, if you (and maybe your "customers") are happy with them, that is just hunky dory. But as a matter of physics and math, the RAW file can virtually always be used to create a "better" image, assuming sound post-processing skills.
Click to expand...
i do not sell my images, ever. they are given to people gratis. your word was create. well that is exactly what i do not want to do. i wish to record the scene. not make/create the image later.
Member said:
Just consider things like noise reduction and sharpening. The algorithms available within, say CS4, are substantially better than those implemented in the camera firmware. And when shooting RAW they can be applied to the original data at full precision versus data than has already been reduced to 8-bits and subjected to lossy compression.

I'll accept that your in-camera JPEG's are simply wonderful; enough to make Ansel Adams cry. But if you shot RAW and processed them carefully, they could be better.
Click to expand...
they would be created better yes. but i have absolutely no desire to create anything. i wish to record the scene. how many times do i have to say this? i want to record what i see, not make what i see later in the pc doing pping. raw's great advantage is in the pping. biut this exactly what i am avoiding by doing what i am suppposed to do with a dslr in the field, shoot the image right. by shooting the image right, the whole idea of pping, except for sharpening, is simply something that the correctly shooting user does not do.
 
GaryDeM said:
more adjustments are possible. why are those adjustments needed if the image was shot right in the field?
Click to expand...
Because almost every component involved in the process is imperfect including the camera and lens. We touched upon sharpening -- that's desirable in part to overcome the effects of the anti-alias filter mounted on the front of the sensor.
GaryDeM said:
i will shoot my images right in the field.
Click to expand...
Good for you. That's equally important when shooting RAW too.
GaryDeM said:
i wish to record the scene. not make/create the image later.
Click to expand...
Fair enough.

But a RAW image will give you a more accurate recording, period.

Whether you shoot RAW or JPEG, you start with RAW data from the sensor. You can't avoid the need to process that. You either do it in-camera or you do it later. In either case, that processing is imperfect and involves some qualitative and somewhat subjective variables.

If the processing is done in camera you (1) have no control over those parameters and (2) no ability to use a better/more accurate algorithm later. Also keep in mind that a camera has very limited processing power and compromises have to be and are made.

If the processing is done after the fact, you have more control over those conversion parameters, you can apply more computational horsepower, and use better algorithms as and when they become available.

Heck, there even exists the possibility of firmware bugs that might contribute losses or artifacts to the in-camera JPEG. This is not entirely theoretical either -- Canon recently coughed up to such a bug:

http://web.canon.jp/imaging/eosd/firm-e/eos7d/firmware.html

[I don't know for a fact that RAW shooters were immune to this specific problem but it's certainly possible].

If your concern is with the "accuracy of the recording", you should definitely shoot RAW. The RAW image is precisely a recording of what the camera "saw". All of the JPEG's you can produce from that, in camera, or on the computer, are merely interpretations of the RAW data. The interpretation that results from the algorithms Nikon decided to embody within the camera firmware are not necessarily the best or most accurate. In fact, they're almost certainly not because they absolutely embody a lot of purely aesthetic choices and the processing power available to them is quite limited.
 
(Other folks have already said this, but I started on this post a while ago.)

One thing a lot of people fail to mention is that every digital sensor exports RAW data -- it doesn't matter if it's in an iPhone or a Mars rover.

What happens to that RAW data depends on a lot of things.

Cell phones and point-and-shoot cameras generally prevent the user from accessing the sensor data. These cameras have a small onboard computer (loaded with firmware) which converts the sensor data into an image using what it considers to be "best guesses" based on how the manufacturer calibrated their camera. These best guesses will generate a 'mostly pleasing' image 'most of the time'. It then automatically converts the data into an easily-accessible format (usually JPEG) and then deletes the RAW data automatically.

The firmware allows the user to adjust a few settings for the JPEG conversion process. More complex cameras allow the user to adjust more settings, but generally these settings include: white balance, brightness, contrast, and sharpness among others (Nikon Active D-Lighting, for instance). These affect the entire image at once, and have a few levels of granularity: about a dozen settings with about a half-dozen variables each.

Higher-end cameras (SLRs, and some medium format cameras and backs) also have firmware which can convert the sensor data into a JPEG file, but they also allow the user to store that RAW data and process it instead of having the camera do it. When it does this, the RAW data completely bypasses onboard computer, except for the simple I/O routines of getting the data onto a memory card.

RAW editing software for a desktop computer (which I will refer to as "RAW software") such as Nikon ViewNX, Adobe Lightroom, Apple Aperture, and Adobe Camera Raw (which is included with Photoshop) can decipher this sensor data the same way as the camera's firmware. Without software for deciphering this data, the file is utterly useless. ('Ease of use' is a big reason some photographers rely on in-camera JPEG conversion.)

RAW software is also configured with "best guesses" in mind. However, if the software was not created by the camera manufacturer, these guesses are different from the guesses put in the camera's firmware. This causes lots of photographers who are new to the concept of editing RAW photos to complain that the RAW image seems "flat" compared to the JPEG. They are seeing the direct result of this data bypassing the camera's firmware.

Just like the camera's firmware, RAW software allows the user to adjust the same settings for the development process: white balance, brightness, contrast, etc. A major similarity is that these changes don't permanently affect the image until it is converted into a legitimate (non-RAW) file format. In other words, these changes are what the digital market calls "non-destructive".

The difference between RAW software and the camera firmware is the sheer number of ways the user can affect the image. Instead of about a dozen settings, RAW software has hundreds. Instead of each setting only having a half-dozen variables, RAW software has hundreds or even thousands. Lately, RAW software is allowing the user to apply RAW changes to portions of the image, instead of altering the entire image at once (a limitation in camera firmware).

These changes are applied at the same step as the in-camera process, but the methods by which the image is altered are massively more complex.

If a photographer can create magnificent images with the relatively-simple computer inside their camera, that is excellent! However, the greater complexity in PC-based RAW software, and the much greater processing power in desktop computers massively increases number of ways the user can develop the image.

In other words, if you're shooting JPEG, you're telling your camera to develop the image. If you're shooting RAW, YOU are developing the image.

Furthermore, JPEG is a lossy format. If you want to record exactly what the camera saw, using the RAW data is the only way to do that.
 
JPEG is 8-bit which translates into 256 colors. If a color does not fit into one of the 256 color "buckets" it get put into the next one over and this causes color shifts and a loss of color fidelity. For many subjects this is not a problem but for people's skin tones or nature photography it is a problem.

With a 12-bit RAW the image file (if it was correctly exposed) can contain up to 4096 gradations, while with 14-bit this goes up to 16,384. RAW data also can be lossless while a JPEG file compression throws away some of the sensor data to save space, which is the point of using JPEG in the first place.
 
GaryDeM said:
by shooting the image right, the whole idea of pping, except for sharpening, is simply something that the correctly shooting user does not do.
Click to expand...
That is unadulterated bull, and you are obviously an idiot. I would be more vociferous in my response, but this forum does not allow it.
--
Anthony Beach
 
Now, now, we don't have to resort to name calling. Gary's comment was an over-generalization for sure, but his point shouldn't be lost on the rest of us which is simply that within tolerances, and with enough skill, the difference between getting it right in camera and doing workstation post processing is often negligible.

Can we always squeeze more out of an image using RAW - most of the time, yes. How much more however varies widely and isn't always worth it if you get everything else right in camera under good conditions and I think that has been Gary's point.

Maybe put another way, too many are substituting using RAW and post processing a crutch for fixing what they should otherwise be nailing in camera and are more often getting lazy.

Me? I try to get it right in camera preferring to shoot fewer frame with higher yeild. I shoot JPEGL+RAW, and post process the RAW when a) I want to squeeze every bit of quality out of an image, even the ones done properly in-camera (OK, call me a-rentive) b) the conditions were beyond what in-camera PP could manage adequately or c) I screwed it up in camera which, after 35 years of shooting, I admit I am still not as good as Gary's batting percentages but not far off either.

There are no absolutes here folks.

Regards,
Mike
 
Michael Firstlight said:
Now, now, we don't have to resort to name calling. Gary's comment was an over-generalization for sure,
Click to expand...
His post was more than that, it left no doubt that anyone that post processes is doing something wrong in the field. That's highly insulting, and I take umbrage to that.
--
Anthony Beach
 
thank you.

i was refering to the old saying with computers, which is gigo. for photography it would sure increase the chances of getting a superior final image with the more shot quality the user can deliver to the pc.

the higher the qulity level the pic is shot in the field the less time the user spends just getting the image back to the normsal or ok range of quality, then on top of that the user has to improve the image to the superior level. my suggestion, and the way i shoot, is to just shoot the image at the very good level first. thus cutting out a lot of the pp time and effort.

i do not care if anyone shoots raw or jpeg. my contention is that the user bought a good dslr to take quality pics. this is far different from buying a p&s to just take snapshots and settle for thsat quality level. the dslr was designed to take good pics, why not do so THEN pp?
 
photonut- you havre always misread what i am saying. i fully realize that there are a lot of users who like to sit at the pc for hrs and play/work with/ adjust the images they take. fine i wish them well. they can create/make all the images they want, more power to them. BUT i do not wish to do this. i have been using computers since 1978 and i have had it. i use the pc as a tool to get a job done, but there certainly is no plesasure of any kind in doing so for me.

if someone wants to shoot raw or jpeg do so. but my view is to shoot the image right first in the field then pp to your heart's content. it makes the pping a lot easier.
 
sure you can control exposure with a jpeg. if i could spend 32yrs shooting film slides, which have ZERO HEADROOM, and only a dr of 4+ stops. then shooting jpeg with a 1 stop or a little less headroom is fairly easy, not to mention the wonderfullness of a wopping 6stops of dr. that is lot better than a film slide with 4+stops of dr.

when i switched from slides to jpegs several yrs ago i simply shot the digital jpeg the same way i shot the slide and it has worked out quite well.

on recent trip out west in august i shot 543 images from 2 dslrs. how many were missed due to the exposure or wb being off? ZERO. and of the 543 how many did i crop? about 10. there was no need to; the composition was right as taken in the dslr on the rest. it can be done but it is harder and requires a lot of time effort work.
 
i got a better questrion- do YOU see the difference. the top pic is the jpeg while the bottom pic is the raw. the raw was converted and that is it. the jpeg is as shot. beyond what i just said no other pp was done including sharpening.



 
GaryDeM said:
you havre always misread what i am saying.
Click to expand...
No, but perhaps you are misstating what you mean. Explain how you are not saying that post processing is not necessary if the shot is taken "correctly" when you wrote, "by shooting the image right, the whole idea of pping, except for sharpening, is simply something that the correctly shooting user does not do." That makes Ansel Adams a hack then, who spent hours in the darkroom fixing his mistakes in the field.
GaryDeM said:
i fully realize that there are a lot of users who like to sit at the pc for hrs and play/work with/ adjust the images they take.
Click to expand...
Some of us can get it done in minutes.
GaryDeM said:
if someone wants to shoot raw or jpeg do so. but my view is to shoot the image right first in the field then pp to your heart's content. it makes the pping a lot easier.
Click to expand...
Right, and the same principle of getting it right in the field applies to both shooting JPEG and RAW. What is most ironic though is that post processing RAW files is easier than post processing JPEGs, and it is more efficacious too.
--
Anthony Beach
 
photonut2008 said:
GaryDeM said:
you havre always misread what i am saying.
Click to expand...
Click to expand...
That makes Ansel Adams a hack then, who spent hours in the darkroom fixing his mistakes in the field.

And I'll bet you stood right there next to Ansel watching him, HUH!
photonut2008 said:
Some of us can get it done in minutes.
Click to expand...
And some of us don't have to be cause its done right the first time.
photonut2008 said:
Right, and the same principle of getting it right in the field applies to both shooting JPEG and RAW. What is most ironic though is that post processing RAW files is easier than post processing JPEGs, and it is more efficacious too.
Click to expand...
So you say
photonut2008 said:
--
Anthony Beach
Click to expand...
 
GaryDeM said:
i got a better questrion- do YOU see the difference. the top pic is the jpeg while the bottom pic is the raw. the raw was converted and that is it. the jpeg is as shot. beyond what i just said no other pp was done including sharpening.



Click to expand...
The RAW shows more shadow detail, but that may be because you have given it slightly more exposure overall in processing. The WB also seems slightly different on my monitor.

However the differences are very slight; your examples vindicate your stance and put the whole argument into perspective.
--
WSSA member #252
 
the raw image in the converter had the shadows adjusted VERY slightly(the windows were colorcoding as needing it), the people's white shirts were shown as blowing our very slightly so the highlight recovery was used very slightly to eliminate that. this was also showing in the raw's histogram.

the jpeg is totally untouched; it is as shot. its histogram in any mode showed no blowouts anywhere. in any event the jpeg was left totally alone. this was shot as a raw+jpeg.

i found it interesting that the raw showed blowouts on both ends of the histogram, while the jpeg did not. and i checked all modes and colors of the histogram for the jpeg. this means that shooting raw for me means that i will be adjusting the image to stop the blowouts, with jpeg this is eliminated and i can just use the image.

inany event, if the obsever of the 2 images was not told which was the raw and which was the jpeg, i doubt if anyone could consistantly accurately tell.
 
If we want to go the comparison route, ok:

Here's the RAW:

This was exposed in such a way that I would not lose any highlights or shadows. At this point, I'm not concerned with capturing what I see; I care about what the camera sees. If I lose any data at this point, it will be gone forever.

I used a 2-stop GND filter to prevent the sky from getting blow out. Since I'm looking into the sun, I'm guessing the scene has between 12-15 stops of light, far too many for the camera to decipher in one go (despite the fact that I could see it with my own eyes just fine).



Since it's RAW, it completely bypasses any adjustments internal to the camera. I could have fiddled with the contrast, brightness, saturation, etc., but they would have been ignored. That's why it looks "flat" -- and therefore not what I remember it looking like.

Here's the resulting JPEG:

If my camera would have had software so I could have done this with its own internal JPEG converter, I would have used it. But I only had a D70s at the time, which is not very powerful compared to Adobe Lightroom.



Sure, there are some specific Photoshop edits which make this not the best example of my RAW-only development process (I removed the blue gas tanks which are no longer there today), but most of the editing was done in Lightroom before it got to Photoshop.

Here's the point I was trying to get across when I wrote that 600-word essay earlier: When you say that I should get the photo 100% right "in the field", that's exactly what I'm doing! But thanks to RAW software, the field now includes my computer at home! Lightroom serves the exact same function as the internals of the camera (which all JPEG photographers are using). I simply intercept the sensor data before it goes to the internal JPEG converter so I can replace it with my own JPEG conversion process.

(aside from some Photoshop edits) I don't use Lightroom to "fix mistakes" in the image. I'm using it to develop the image in the first place.
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top