The RAW/JPEG Discussion

[Glenn NK]

Actually this should not be a debate at all, as both file types have their place in digital photography.

One of the best articles I've read on the topic came out today:

http://www.asktimgrey.com/

This daily post is worthwhile, and subscribing to it costs nothing.

Glenn

[Geoff F]

All I get from that link is the basic home page.

However, I would recommend this article for anyone who is considering trying Raw, although this does get a bit detailed rather quickly.

http://www.ronbigelow.com/articles/w...flow_basic.htm

But don't forget this CinC article about Raw shooting.

https://www.cambridgeincolour.com/tu...ile-format.htm

[Glenn NK]

Yes, it's a link to the sign-up area of the site (to receive the daily letter).

The article by Ron Bigelow is excellent in that it deals with the detailed adjustment of a RAW file, but doesn't address the thread topic.

Lately there have been "discussions" about the pros/cons of the differences in RAW/JPEG; the article by Tim Grey addresses these issues without taking sides which so many proponents do.

Glenn

[Manfred M]

I subscribed (and then unsubscribed) the the article; frankly nothing new there that you can't find on dozens of other websites or books. Nicely said, he shoots raw "for insurance"; because he knows that he's got the extra data in RAW, if he needs it. The cost of doing so is the it requires more storage, which when all is said and done, is fairly inexpensive.

Makes sense to me (and I think I have said the same thing myself more than once). In general I will use jpgs with perhaps minor PP work when I post on a website. If it is a problem image, I'll do the deep dive in PP using RAW to get something good out of it.

[ajohnw]

I subscribed (and then unsubscribed) the the article; frankly nothing new there that you can't find on dozens of other websites or books. Nicely said, he shoots raw "for insurance"; because he knows that he's got the extra data in RAW, if he needs it. The cost of doing so is the it requires more storage, which when all is said and done, is fairly inexpensive.

Makes sense to me (and I think I have said the same thing myself more than once). In general I will use jpgs with perhaps minor PP work when I post on a website. If it is a problem image, I'll do the deep dive in PP to get something good out of it.

I honestly can't see the point in working any other way..

-

[jcuknz]

I can see that some may find value and need in shooting RAW but I don't and never do ... while storage is cheap that is the least of my considerations, more important is turn around in camera which was the original reason for not shooting Tiff or RAW. Yet to see any reason to change.

[Manfred M]

I thought I would look at a real-life example of the RAW vs jpg arguement and why I use the capture flow (jpg + RAW) that I do use. I spend a few seconds on the cleaned up images, so had I put in the effort, I could have done a lot better.

This image is a closeup of a jackel that popped up beside us in Etosha National Park in Namibia. It caught me off guard and the image quality shows; while the image is very sharp, it is totally underexposed:





This is the image I created out of the jpg. All I did is clean up the luminosity with the levels control and I turned down the saturation a bit. There is a bit of artifacting near the nose, where the blacks are purple. I did not clean these up, to show what the underexposed jpg is capable of providing.





The final image is where I went first and used the RAW file. Other than a smart-sharpen, there is no other PP work. The colours are perhaps a tiny bit more saturated, especially in the blacks.



Bottom line; yes RAW is a nice insurance policy, but in all honesty, I would be surprised if I use them more than 5% of the time for web bound images. There is theory and then there is reality....

[Colin Southern]

Bottom line; yes RAW is a nice insurance policy, but in all honesty, I would be surprised if I use them more than 5% of the time for web bound images. There is theory and then there is reality....

It really depends on a number of factors; if you're shooting purely reflective scenes with the correspondingly modest dynamic range requirements - and you're nailing your exposures - then RAW offers little benefit (but still essentially no down-side either I might add, IMO). But if one is shooting something like a sunset where the dynamic range requirement is much higher (assuming foreground shadow detail that needs to be preserved) then the extra dynamic range recorded in a RAW capture will yield far far far far superior results over a JPEG (which has already discarded most of the extra information).

Even in the studio - using the far superior dynamic range of the 1Dx over the 1Ds3 - with a RAW capture I can get far more aggressive with the fill light control to reveal shadow detail that just wouldn't be contained within the safety margin of a JPEG shot.

[Letrow]

Interesting read once more and also (once more) it won't change my way of working (which is shoot everything in JPEG and PP those images in preparation for viewing).
I might miss out on an image now and then (because I don't use RAW), but I accept that. My miss rate is actually very, very low if I look at things that can be recovered with RAW.
The photos I do throw away (lots of them) lack sharpness in the subject or just not attractive, so things that can not be recovered with RAW anyway.

I trust on my settings and check them before I start shooting. If the white balance is wrong or exposure too light or dark I can either see that directly on the LCD and correct, or if there are tiny adjustments to be made I can make those in the PP process (black and white point, levels).

[SteveF]

It really depends on a number of factors; if you're shooting purely reflective scenes with the correspondingly modest dynamic range requirements - and you're nailing your exposures - then RAW offers little benefit (but still essentially no down-side either I might add, IMO). But if one is shooting something like a sunset where the dynamic range requirement is much higher (assuming foreground shadow detail that needs to be preserved) then the extra dynamic range recorded in a RAW capture will yield far far far far superior results over a JPEG (which has already discarded most of the extra information).

Remember the days whe we went into the photo store and bought yet another graduated ND filter to cater for just this scenario? Trudging out in the evening or very early morning with a bagful of these frantically trying to get the right one for effect before the light was lost?

Now, at the press of a few keystrokes ... nah, not nearly as exciting!

[Manfred M]

Again Colin, there is theory and there is practice.

My workflow is a lot faster using jpg, if I've set my camera up properly and I do a very quick cleanup in PP. The most common failures I've seen with properly exposed jpgs is banding introduced when a scene has a clear, blue sky, and the other it the one you've pointed out. High ratio lighting in studio portraiture, but only when I'm pushing beyond 3:1 lighting, I've noticed a tiny bit of shadow detail loss, that is easy to recover from a RAW file.

This is of course why I shoot both. 95% of the time jpg works great and I only pull up the raw file when there is a problem.

[Colin Southern]

Again Colin, there is theory and there is practice.

My workflow is a lot faster using jpg, if I've set my camera up properly and I do a very quick cleanup in PP.

This is the bit I'm not getting I'm afraid. If I want to produce a JPEG batch from RAW captures then all I have to do is

- Select all in Bridge and press enter (to open them in ACR)
- Select all in ACR and click on "Save Images" - so probably 1/2 dozen mouse clicks to convert a TOTAL of - say - 500 images.

Can't imagine it getting much quicker than that. Nothing against JPEGs, but I just can't see a single advantage in using them.

[Scott Stephen]

Again Colin, there is theory and there is practice.

My workflow is a lot faster using jpg, if I've set my camera up properly and I do a very quick cleanup in PP. The most common failures I've seen with properly exposed jpgs is banding introduced when a scene has a clear, blue sky, and the other it the one you've pointed out. High ratio lighting in studio portraiture, but only when I'm pushing beyond 3:1 lighting, I've noticed a tiny bit of shadow detail loss, that is easy to recover from a RAW file.

This is of course why I shoot both. 95% of the time jpg works great and I only pull up the raw file when there is a problem.

Shooting both can be the best of both worlds, but for the insane number of files. I did both for a while, but I was coming back in with 400 or 600+ files to wade through, the way I shoot.
But then my little kids move around a lot, so my keeper rate is a lot lower than it would be for less squirmy things, like landscapes or adults or speeding bullets. I really don't mind converting to JPG.

[Colin Southern]

Just did a little test ...

Time taken to start Bridge - Select 284 DNG files - and export them as full resolution JPEGs at the same quality they would have come directly out of the camera ...

... 2 minutes & 39 seconds.

[Manfred M]

This is the bit I'm not getting I'm afraid. If I want to produce a JPEG batch from RAW captures then all I have to do is

- Select all in Bridge and press enter (to open them in ACR)
- Select all in ACR and click on "Save Images" - so probably 1/2 dozen mouse clicks to convert a TOTAL of - say - 500 images.

Can't imagine it getting much quicker than that. Nothing against JPEGs, but I just can't see a single advantage in using them.

Colin – I completely understand your position, you are working with the very best image data your camera can deliver, and there are a lot of merits to your position and a lot of people here on CiC do take the same position. I’m also saying that it’s possible to make a very persuasive argument that jpegs are more than adequate most of the time, especially if an image is properly exposed and people that are more comfortable in shooting jpegs, should continue to do so. The benefits of RAW images are not at great as they might seem.

From my reading, sources vary a bit on the details, but in general there is agreement that we humans can distinguish somewhere between 9 million and 10 million distinct colours, with some decreases as we age. A jpeg file with its 16 million colours is already well beyond what our eyes can make out; in fact we are not able to see almost 40% of the colours that a jpeg image is capable of displaying.

Most of the images we post on our computers are displayed on inexpensive Twisted Nematic (TN) displays. This are actually 6-bit per channel displays (256 colours per channel); or 18-bit which produces 262,144 distinct colours natively. These monitors claim that they do display 16 million colours and this is accomplished by a bit if high-tech trickery called Frame Rate Control (FRC) to simulate the full range of colours. I doubt that anyone other than the screen manufacturers marketing departments really believe the numbers. The colour accuracy of these displays is not good. What the real colour output is somewhere between the two extremes, and the fact that I do notice issues makes me think that the actual quality falls within or perhaps below our visual threshold, i.e. less than 10 million distinct colours.

Higher end In-Plane Switching (IPS), on the other hand does display true 24 bit colour and is natively capable of displaying the sRGB colour space. These too use FRC to “extend” the colour space. Really my bottom line is that most screens that people use can at best handle the colour capabilities of a jpeg file, even though our eyes cannot resolve the colour detail to that level.

Printers have an even lower range of the number of distinct colours that can be reproduced, and this information is not readily available. The numbers I’ve seen tend to suggest that photo quality inkjet printers are capable of producing between 200,000 and 300,000 distinct colours.

So, our eyes can’t see all of the colours that a jpeg is capable of displaying. Some of the better monitors can reproduce it, but the most commonly used ones cannot, and printers can certainly not reproduce anywhere near the colour range of a jpg. The real question is why do we bother at all with RAW?

The only reason that seems to hold up to technical scrutiny is that in cases where we do significant post-processing and the algorithms in our PP programs can introduce artifacts because we do not have enough data in the jpeg file. This can come from improper exposure in the initial image, extreme changes in some areas of our images, etc. Having the additional data allows the PP software to create more subtly and in certain instances we can see that in our final output. In my experience, RAW is a bit like an insurance policy, if everything goes well, you don’t need it, but when things go wrong you are glad you have it.

So, why use jpegs rather than RAW? Simply for the technical reasons I have outlined; they are good enough most of the time. I have no issue that you or anyone else uses RAW all the time, there certainly is no harm in doing so, but there is usually no technical advantage to doing so either. There are limitations to using jpegs and the people that only shoot jpeg need to be aware that there are times that their strategy will result in compromised image quality.

So bottom line, why shoot jpeg? Simply because they usually good enough.

[Colin Southern]

The real question is why do we bother at all with RAW?

See YOUR answer below.

In my experience, RAW is a bit like an insurance policy, if everything goes well, you don’t need it, but when things go wrong you are glad you have it.

Exactly. Or put another way, when we're offered "free insurance" with no down-side, why not just take it? Sure, shooting JPEG may be (or may NOT be) just fine 99% of the time (for some shooters) compared to RAW -- and if that's what folks want to do then more power to them -- I just can't for the life of me understand when offered a choice between "99%" (JPEG) and 100% (RAW) why anyone would choose the "99%" option. Just doesn't make any sense to me.

[pnodrog]

Exposure adjustment and Insurance is my main reason. Below is a thread that illustrates why I shoot using RAW files.

Thread: N.Z. Harrier vs Duck – RAW and PP vs Over exposure

by pnodrog

[Glenn NK]

I’m also saying that it’s possible to make a very persuasive argument that jpegs are more than adequate most of the time,

Manfred:

The problem I have is that I'm never sure when "most of the time" starts.

If I knew when it was going to be "most of the time", I'd switch over to JPEG. But then I'd have to know when to switch back again. So like Colin, I take the insurance route.

HOWEVER. The whole point of starting this thread wasn't so people could trumpet their method (which unfortunately happened - I should have known better), but so that people would find the link, and read the article entitled "Bit Depth Limitations". I get the impression that some didn't read it, but still jumped in to defend their mode of shooting. Pity.

Glenn

[pnodrog]

Sorry but I did not find the link to the article - but I did sign up out of interest.

[Letrow]

Manfred:
HOWEVER. The whole point of starting this thread wasn't so people could trumpet their method (which unfortunately happened - I should have known better), but so that people would find the link, and read the article entitled "Bit Depth Limitations". I get the impression that some didn't read it, but still jumped in to defend their mode of shooting. Pity.

Glenn, the article was interesting, but unfortunately I couldn't access the original article he referred us to (that of the previous week).
The method trumpeting is actually quite a good read I think, as it gives me more insight in the pro's and con's of both methods. It is a hot item still apparently (and undecided for the time being) and this thread proves it.

And it might get some people to rethink what method they use. Yesterday for instance I installed the newest Gimp (2.8 by now and it does look good) and (because I had just seen this thread and a few others in the last week) the newest UfRaw, so that I could try out some possibilities for myself. I have used UfRaw in the past, so I wasn't new to it by the way.

I shot a few photos in RAW and used Ron Bigelow's method on them (see the 2nd post by Geoff). So: white balance and colours in the RAW editor and then the usual stuff (levels, contrast, sharpening) in Gimp. It produces nice results and the work is easy enough, but the end result isn't really different from what I get with JPEGs (assuming that camera settings such as white balance and exposure are correctly set to begin with of course).
I did configure one of the custom settings on my D7000 to shoot in RAW though, so that I can experiment a bit going forward.

I don't want to see one method as an insurance policy and in that sense as better than the other method. That sounds like a fallacy to me. You need insurance when something goes really wrong, but (again) assuming you have set your camera correctly that should not happen.
On a side note: we Dutch have a reputation of being over insured (we like to feel save), compared to people in other countries. The way to look at insurance apparently is that you try to figure out whether your bank account can easily take the burden of replacing something. If that is so, you don't need insurance (e.g. a bike can be easily replaced), but if not, then you do need insurance (e.g. your house might stretch the bank account a bit too much). This way of thinking can be applied to photography as well of course and will be dependent on the particular occasion.
And even then, both methods (RAW and JPEG) allow for recovery, each up to a certain point of course.
It depends a lot on what you shoot and what you need the photos for I think and both methods will have their uses and supporters.

And to reiterate one of my own weak points (which is for me personally something that plays a role in this debate): most of the shots that I have to throw away are shots that I have myself or the wind to blame for (poor composition, camera shake, wind blowing my macro subject out of focus, etc.). I have tried to recover these errors in both RAW and JPEG, but I can't seem to get the results I want.

I don't feel the need to convince others to do it my way (about this 'debate' or about Nikon versus Canon or Sony) and I hope others can see it like that as well, but I do like the relevant bits of information I get out of this.