One question that comes up a lot around here is "Should I shot in RAW mode or JPEG Fine?". It seems that every photographer will at some point decide which mode is best for them. For me, I like to shoot in RAW.

A couple weeks ago I was taking pictures with a borrowed Nikon D100 at my amateur radio club's Field Day effort. I took a hand-full of shots in RAW and then switched to JPEG Fine. I was worried that I might run out of storage over the weekend long Field Day effort. Big Mistake!

When I started shooting with the D100 I could tell that the images were overexposed. I dialed in some underexposure and shot away. It wasn't until I got to a computer and downloaded the images that I realized what the problem was. The camera's contrast adjustment was way out of wack. Things that were in shade were ok, but anything in direct light was completely blown out. I should have realized what was going on, but because I was using a new camera, I didn't catch it in time.

So how does all this relate to RAW vs JPEG Fine? The handfull of images I shot in RAW were easily corrected when I brought them into photoshop. The JPEG Fine images on the other hand were almost worthless. The blown out areas contained almost no detail in the JPEGs and I was left with three or four decent images out of 30 or 40 that I took.

So the conclusion is this... Another Field Day has come and gone and once again I only have a half a dozen images I really like. If I had just stuck with my trusty G5 and/or shot in RAW I would have had a lot more usable images.

yep, cletus, a good (hard) story you've related here.

while i strive to get the exposure right ever time, at shoot, i still shoot every image in raw. the cool thing is that i don't have to worry so much when, in my situational street shooting mode, i encounter highly mixed lighting.

i love raw.

I love RAW. Love it. Saved my butt a few times, and others I've just found the flexibilty to be great.

I can relax when shooting my kids and just try to get great shots without having to obsess on exposures.

Originally Posted by DavidTO I love RAW. Love it. Saved my butt a few times, and others I've just found the flexibilty to be great. I can relax when shooting my kids and just try to get great shots without having to obsess on exposures.

I'm using Photoshop CS now and I have mostly been using fine JPEG. I just started shooting some RAW, but I'm not sure about all the choices is have for post processing in RAW before converting to JPEG. I figured out the exposure and white balance pretty well, but tint etc. , I have no clue. Andy would you consider a tutorial on the RAW workflow, including should you send 8 or 16 bit images into PS and what do you do on a 16 bit image before converting to 8 bit.
Thanks,

Originally Posted by tmlphoto I'm using Photoshop CS now and I have mostly been using fine JPEG. I just started shooting some RAW, but I'm not sure about all the choices is have for post processing in RAW before converting to JPEG. I figured out the exposure and white balance pretty well, but tint etc. , I have no clue. Andy would you consider a tutorial on the RAW workflow, including should you send 8 or 16 bit images into PS and what do you do on a 16 bit image before converting to 8 bit. Thanks,

I've also found the RAW conversion thing to be an issue. The IDmkii has a great feature, it will shoot RAW+jpeg super fine. You just get the same image in both formats when you download it. This is great for previewing and sometimes I don't even need to do a RAW conversion, but if I do, I can.

I think the Adobe RAW converter is confusing because, in addition to exposure, it offers a lot of controls that duplicate things that are better done after the conversion using the more powerful tools. Curves and highlight/shadows give a lot more flexibility and offer better results that monkeying with contrast, shadows, color balance, etc at the time of conversion. USM gives a lot more control (including L channel USM) then the converter's sharpening. So, now, I only change the exposure slider, and, well, sometimes make a gross color balance change (like choosing "tungsten" for the nighttime street scenes (but I considered this a quick and dirty first step that needed to be fine tuned wirth curves later.)

RAW is hopeless for high action situations, even with the 1imkii. I tried shooting little league games that way and was always waiting for the camera just when the good shot prestented itself.

I shoot in JPEG, standard mode, standard color. I still use PS7 and I doon't think there's a RAW converter for it. The buffer waiting time is large on the 828...it's frustrating. So, I'm sticking with JPEG for now.

Originally Posted by cletus One question that comes up a lot around here is "Should I shot in RAW mode or JPEG Fine?". It seems that every photographer will at some point decide which mode is best for them. For me, I like to shoot in RAW. A couple weeks ago I was taking pictures with a borrowed Nikon D100 at my amateur radio club's Field Day effort. I took a hand-full of shots in RAW and then switched to JPEG Fine. I was worried that I might run out of storage over the weekend long Field Day effort. Big Mistake! When I started shooting with the D100 I could tell that the images were overexposed. I dialed in some underexposure and shot away. It wasn't until I got to a computer and downloaded the images that I realized what the problem was. The camera's contrast adjustment was way out of wack. Things that were in shade were ok, but anything in direct light was completely blown out. I should have realized what was going on, but because I was using a new camera, I didn't catch it in time. So how does all this relate to RAW vs JPEG Fine? The handfull of images I shot in RAW were easily corrected when I brought them into photoshop. The JPEG Fine images on the other hand were almost worthless. The blown out areas contained almost no detail in the JPEGs and I was left with three or four decent images out of 30 or 40 that I took. So the conclusion is this... Another Field Day has come and gone and once again I only have a half a dozen images I really like. If I had just stuck with my trusty G5 and/or shot in RAW I would have had a lot more usable images.

Originally Posted by rutt I think the Adobe RAW converter is confusing because, in addition to exposure, it offers a lot of controls that duplicate things that are better done after the conversion using the more powerful tools.

I couldn't agree more. When I bring in a raw image, I do three things in the converter:

1. Ballpark the color temp. I just try to get rid of any obvious color casts without trying to get it spot on.
2. Adjust the Shadow level. I bring up the shadow level until just a few pixels are pure black.
3. Adjust the Exposure level. Adjust until just a few pixels are pure white.

After that I bring the image into PS (in 16-bit mode) and use curves and/or levels to do the rest of the color correction and exposure adjustments.

What would be cool is if Adobe (or a Camera maker) could do a RAW converter that didn't alter the image, instead when the image is opened in Photoshop, the converter provides adjustment layers (levels, curves, etc...) that apply the adjustments that the camera and/or converter decided on. For example, if the camera's contrast setting was set to provide more contrast than normal in an image, the image opened in photoshop would automatically have a curves adjustment layer with a curve that upped the contrast.

Originally Posted by tmlphoto I'm using Photoshop CS now and I have mostly been using fine JPEG. I just started shooting some RAW, but I'm not sure about all the choices is have for post processing in RAW before converting to JPEG. I figured out the exposure and white balance pretty well, but tint etc. , I have no clue. Andy would you consider a tutorial on the RAW workflow, including should you send 8 or 16 bit images into PS and what do you do on a 16 bit image before converting to 8 bit. Thanks,

I hate the PS RAW converter. Not that there's anything wrong with it per se, we just don't get along. I much prefer C1. You could download the demo and check it out.

http://www.c1dslr.com/

Originally Posted by DavidTO I hate the PS RAW converter. Not that there's anything wrong with it per se, we just don't get along. I much prefer C1. You could download the demo and check it out.

I picked up C1 for its bulk processing features, and indeed it does a better job than PS and even the Canon File Browser tool (not sure about the tool that comes with their pro-level cameras). But damn, for what it's doing it's awfully expensive and the UI is surprisingly nonintuitive (at least on the Mac).

Originally Posted by jimf I picked up C1 for its bulk processing features, and indeed it does a better job than PS and even the Canon File Browser tool (not sure about the tool that comes with their pro-level cameras). But damn, for what it's doing it's awfully expensive and the UI is surprisingly nonintuitive (at least on the Mac).

I tried C1 and I have to admit that I was almost completely baffeled by it. Maybe a C1 tutorial for dummies would be a good post.

When you say C1 does a better job, what exactly does that mean? Do the images really look better? Is it faster? Easier to convert a lot of similar shots?

Originally Posted by rutt I tried C1 and I have to admit that I was almost completely baffeled by it. Maybe a C1 tutorial for dummies would be a good post.

The thing that's hard to understand is that the processing always happens in the background. You select some images, use the gear pages to set up how you want them converted, then when you hit "process" it will chunk along processing each of them. You can see the status on the multiple gear page.

Oh yea, and you have to drag and drop a folder full of images onto it's "organize" page. There is no concept of "import" except straight from the card.

The interface works fine once you figure it out but it sure is a lot harder to figure out than it ought to be.

When you say C1 does a better job, what exactly does that mean? Do the images really look better? Is it faster? Easier to convert a lot of similar shots?

They look better. It has the best bayer mask interpolation I've seen. Admittedly I've only tried four different ones: Canon File Browser, GraphicConverter (which has horrible quality but is FAST), Photoshop CS, and C1. It is trivial to convert a lot of similar shots.

The annoying thing is even at the $250 price point they're throwing limits on how many shots you can convert and whether or not you convert to multiple image types simultaneously (ie both tiff and jpg). That's getting pretty greedy. Then again I guess they have a captive market.

Originally Posted by cletus After that I bring the image into PS (in 16-bit mode) and use curves and/or levels to do the rest of the color correction and exposure adjustments.

I don't have a strong opinion about the 16-bit mode thing, but my PS guru, Dan Margulis does, see here for more about this than you probably ever wanted to know. Part of his argument is based on the limitations of older versions of PS, so let's forget those. But the part of his argument that is still relevant is:

1. Almost all output devices are 8 bits/channel. So, eventually your image is going to be down sampled to 8 bits/channel. Better to do this yourself than let some printer driver or jpeg converter do it for you.
2. USM is a major weapon in Dan's arsenal. Dan claims that it works much better in 8 bit mode. Essentially, USM works by finding color transitions. Since these can be smoother in 16-bit mode, USM has less to work with.
3. Nobody can really see the difference. Dan has some sort of a challenge out with a complicated set of rules. If you can come up with an example where the 16 bit image really did produce better results, you win. But the rules are complicated and there is a lot of controversy about the relevance of the challenge. Nobody has won it, but given the rules, perhaps nobody could. There is also some sort of double blind testing that is inconclusive.

Suffice it to say, the advantages of 16-bit mode are controversial. I was just wondering if somebody knows something more than me about it and can explain clearly.

Originally Posted by rutt I tried C1 and I have to admit that I was almost completely baffeled by it. Maybe a C1 tutorial for dummies would be a good post.

There is a good basic tutorial here.

Originally Posted by rutt When you say C1 does a better job, what exactly does that mean? Do the images really look better? Is it faster? Easier to convert a lot of similar shots?

For handling a lot of images that require different adjustments it is one of the fastest programs out there that can handle the task. The actually file processing times are not the fastest, but it all happens in the background so you can keep working on other files. It delivers very clean files, lower noise, etc. They have been making a lot of improvements over the last year and it is a much better program than it used to be. Hopefully the next release will be even better.

There is a dedicated forum for C1 here. A 30 day demo can be downloaded from the same site.

Originally Posted by rutt Suffice it to say, the advantages of 16-bit mode are controversial. I was just wondering if somebody knows something more than me about it and can explain clearly.

Frankly the only really significant advantage I see in having 16 bit images is that you get greater exposure lattitude. Dan is right that there are virtually no devices capable of handling 16bpp, even professional quality printers can't (I tried, they complained :-). It'll be downconverted at some point, the only question is when.

So long as I have the opportunity to adjust exposure before downconvert I'm a happy guy.

I find shooting in RAW to be somewhat slow, but very useful. Twice in the last ten days I have deliberatley underexposed my shots by two stops in order to get a faster shutter speed. I was able to bring the shots back because I was shooting in RAW. I also like being able to tweak the color temperature (white balance.)

I do find the extra steps cumbersome. I don't very well understand the software I'm using - it's Canon's free Digital Photo Professional. But that's my fault for not studying it. The other frustration is the in-camera file processing time. The buffer takes a while to empty. I've learned to shoot less than the maximum burst, so I always have a shot or three remaining while the buffer's slowly emptying.

Hey rutt...

Here are my thoughts on each of the points from your post:

Originally Posted by rutt Almost all output devices are 8 bits/channel. So, eventually your image is going to be down sampled to 8 bits/channel. Better to do this yourself than let some printer driver or jpeg converter do it for you.

If the change from 16 to 8 bits becomes critical you could always save off a copy of your 16 bit image as an 8 bit Photoshop file, and make any tweaks to get it just right.

Besides, by bringing RAW images into Photoshop in 8-bit mode, you've already let the RAW converter do the conversion from 16 to 8 for you.

Originally Posted by rutt USM is a major weapon in Dan's arsenal. Dan claims that it works much better in 8 bit mode. Essentially, USM works by finding color transitions. Since these can be smoother in 16-bit mode, USM has less to work with.

From a digital signal processing background, it's hard for me to accept a filter like USM would bennefit from having less input data

If infact the smoother transitions in a 16 bit file cause the USM trouble, you could help the USM with temporary boost in contrast or other adjustments.

Originally Posted by rutt Nobody can really see the difference. Dan has some sort of a challenge out with a complicated set of rules. If you can come up with an example where the 16 bit image really did produce better results, you win. But the rules are complicated and there is a lot of controversy about the relevance of the challenge. Nobody has won it, but given the rules, perhaps nobody could. There is also some sort of double blind testing that is inconclusive.

Not knowing the rules of the challenge, it's hard to comment on this one. It might be the case that anything you can do inside photoshop with a 16 bit image, you can do with careful work in the RAW conversion step and then working in Photoshop in 8 bit mode. For me it's just easier to work in 16 bit inside PS without having to be so precise during the RAW conversion.

Yeah, as I said I'm no expert on the subject. But there does seem to be significat controversy amoung the best qualified experts. In Dan's case, I think he has a rather large set of very carefully tuned tools that work essentially perfectly in 8-bit mode and he doesn't have enough working years left to retune them all for 16-bit mode and doesn't see the point anyhow. If he were starting from scratch right now, probably he'd incorporate more 16-bit into his workflow. It's easy to find Dan's challenge and a lot of rather vituperative pro and con commentary, google for "Dan Margulis challenge".

My own experiences are very inconclusive. I've never really seen a quality difference in my own images for raw vs jpeg let alone 8 vs 12 bit raw conversions. I'm learning to like raw for the exposure flexibility, and as I said, the 1dmkii makes it possible to have both jpeg and raw for each image, which I consider perfection.

Originally Posted by rutt My own experiences are very inconclusive. I've never really seen a quality difference in my own images for raw vs jpeg let alone 8 vs 12 bit raw conversions. I'm learning to like raw for the exposure flexibility, and as I said, the 1dmkii makes it possible to have both jpeg and raw for each image, which I consider perfection.

What do you use for printing? Doing a raw->8-bit tiff conversion versus raw->HQ jpg and then printing using a professional printing service the differences between tiff and jpg are readily noticable even at only 5x7. That was one of the first tests I did, using Canon's File Browser tool to perform the conversions.

I should try it again with my inkjet for output.

Originally Posted by jimf What do you use for printing? Doing a raw->8-bit tiff conversion versus raw->HQ jpg and then printing using a professional printing service the differences between tiff and jpg are readily noticable even at only 5x7. That was one of the first tests I did, using Canon's File Browser tool to perform the conversions. I should try it again with my inkjet for output.

I've had much more experience with out of the camera HQ jpegs that with RAW. I print almost exclusively with an epson 2200. I try to arange not to do a second jpeg compression before printing, so I save post processed images that I like as lossless jpeg 2ks, which is a little slow, but much smaller than any other hi bit lossless alternative.

Originally Posted by rutt I've had much more experience with out of the camera HQ jpegs that with RAW. I print almost exclusively with an epson 2200. I try to arange not to do a second jpeg compression before printing, so I save post processed images that I like as lossless jpeg 2ks, which is a little slow, but much smaller than any other hi bit lossless alternative.

I would not be surprised if the difference between tiff and jpg was not noticable using inkjets. While I'm rather pleased with the output from my Stylus Photo R300 (not sure how that compares to a 2200) its resolution is clearly inferior to the printer my local photo shop uses (Fuji Frontier I believe).

Maybe I'll blow a few bucks tonight printing the same sample images I used with the printing service.