RAW converters: they do 2 things only

It does everything. I have CS but use it only for adding watermarks and printing. CO does everything from start to finish.

I shot over 500 RAW files in the 29th Colorado Renaissance Festival on Saturday. Processed with CO Pro on Sunday in a couple of hours. Printed some and posted to the online gallery. Try doing that in CS for a change.
http://www.pbase.com/nelsonc/2005_co_renaissance_festival&page=all





--
Nelson

You are trying to keep things simple and that's just fine if that's your goal. Your absolutely-no-adjustments method of RAW processing is easy and simple but you're missing the whole point of the advantages of manipulating the image prior to conversion.

I guess with your "pixel is a pixel" point of view Capture 1 and all the other profesional converters are just all hype and no substance. Or worth it only due to their workflow enhancements. That's just not the case and as I suggested, you really should listen to those who know...like Bruce Fraser, who works closely with Adobe and Thomas Knoll.

There is much more happening in a RAW processor than what you think and the tools in ACR only look similar to their PS counterparts. Iif you actualy dig into this and take the time to learn from the true masters of PS and ACR, you'll soon realize that the controls do things differently.

I have enough experience with ACR 3.1 that I have learned to have reluctant respect for the auto adjustments--I almost always tweak by hand but it's pretty good. If simplicity is your goal you'd be better off batching your RAWs WITH the auto-adjustments turned on. That would be similar, but better, than what the camera does with the RAW data when it creates a JPG.

If you use ACR to it's full potential you'll find that all you need to do in Photoshop is sharpen and do specific local edits with masks.

I'd love to hear you debate your perspective on "all that a RAW converter does is..." with the creators of Breezebrowser, C1, ACR, etc. I'd buy tickets to watch that

These are cut&paste responses at the Pixelgenius discusion board (can't link--it's a member-only support board for PK products) from Jeff Schewe and Bruce Fraser addresing specificaly why it's better to do the work in ACR:
Original question:

"Bruce: is there any difference between the contrast adjustment in ACR and the Adjustment Layer, Curves S curve. In your book they seemed identical. Since you can't see the curve in ACR, can you give some guideline between the various ACR settings of contrast and those in Curves S curve. Is there any reason to use an Adjustment layer rather than ACR, other than the capacity to readjust your setting in the adjustment layer? What is your recommendation: use both, or only one of them?"
Reply from Jeff:

"I'm not Bruce. . .(he has a cooler accent) but ever Camera Raw control is essentially better applied at the Camera Raw stage vs. the Photoshop stage because of the way linear captures are processed into gamma adjusted color spaces. You want to do as many Camera Raw adjustments as accurately as possible before the raw conversion-and that includes contrast.

The -ONLY- point at which Photoshop excels over Camera Raw is where adjustments must be made on a local vs. global basis. Camera Raw has no alternative but to process ALL pixels the same way. Photoshop can constrain adjustments through locale selections. In this specific case, making adjustments after the raw process inside of Photoshop can be beneficial. But for global adjustments, do it in Camera Raw if you can."
Reply from Bruce Fraser:

"Since the curve is being applied to linear data, it's pretty hard to visualize—the illustrations in the book show the approximately-equivalent curve in a gamma-encoded space. Note that the midpoint of the S-curve is set by Brightness—it doesn't just stay at the midtone.

Knowing exactly what each setting did numerically (which I don't, though given a week of doing nothing else I could probably figure out) isn't really helpful unless you can readily relate linear capture values to perceived brightness (which I can't, though given a year of doing nothing else....)

A curve adjustment layer in Photoshop is operating on the data that has already been tone-mapped from linear to gamma-corrected space. The ACR sliders control that tone-mapping from linear to gamma-corrected space. So you really want get as close as possible with the sliders, then use Photoshop for fine-tuning if necessary."

AussieDawg said:

Thanks. I'm not aware of Smart Sharpen. How do you access it? I
don't have CS2 available now to check it out.

Click to expand...

Filter -> Sharpen -> Smart Sharpen

Smart Sharpen is the menu item right above "Unsharp Mask". A lot of reviewers and book authors are predicting that "Smart Sharpen" will replace "Unsharp Mask" as the sharpening tool of choice in CS2. I agree with them.

--
Gary Coombs, W9VJ
http://GaryCoombs.com
My Profile contains my Equipment List

A good photograph is knowing where to stand. -Ansel Adams

bullet1 said:

It does everything. I have CS but use it only for adding
watermarks and printing. CO does everything from start to finish.

I shot over 500 RAW files in the 29th Colorado Renaissance Festival
on Saturday. Processed with CO Pro on Sunday in a couple of hours.
Printed some and posted to the online gallery. Try doing that in
CS for a change.

Click to expand...

A pain with CS. But easy with CS2. I've done that and more.

In CS2 Adobe learned from their mistakes and built in an excellent workflow.

--
Gary Coombs, W9VJ
http://GaryCoombs.com
My Profile contains my Equipment List

A good photograph is knowing where to stand. -Ansel Adams

Andy said:

Then they just
create a 16-bit per pixel per color image from the 12-bit raw data.
The highest 4 bits reman as zeros.

Click to expand...

I don't believe this is the case. The conversion process from 12-bit RAW to 16-bit uses all of the available bits in the destination space.

I think you got it a bit wrong in both cases. 12-bit RAW means 12-bits per photosite , not 12-bits per pixel (a "pixel" in this case is comprised of, as you say, green-red-green-blue). Again, in the 16-bit destination color space, that refers to 16 bits per channel , not per pixel.

So that means you have 48 bits per "RAW pixel" (12 bits/photosite * 4 photosites/pixel = 48 bits/pixel) that's getting processed into a 48-bit destination space (16 bits/channel * 3 channels/pixel = 48 bits/pixel).

So, basically, you're going from one 48-bit space to another.

You can see this is true by considering what an 8-bit JPG means. This means for each pixel in the JPG image, there are 256 different values that can be assigned to that pixel's red channel, 256 different values to the green, and 256 different values to the blue. That's 2^8 different values per channel . This means that an 8-bit/channel pixel actually has 24 bits, so it can display 2^24 different colors (16,777,216, about 16 million colors).

So 16 bit color is double the bit-space per channel, meaning each R, G, and B value can have 65,535 different values. This means a single pixel can display about 281 trillion different colors, which should be enough for anybody.

There's another concept in computer graphics when talking about color spaces called the "alpha" channel. This is often used for 3D rendering in games like Doom!. This is 8 bit/channel color using the standard R, G, and B, but it also adds an alpha channel. The 8 bits in that channel specifies 256 different levels of opacity, 0 means totally transparent (invisible) and 256 means 100% opaque (nothing behind it shows through). This is also why in PS you can add a new channel to the channels layer to do masking--in that case, you're adding an 8-bit greyscale mask that can be used to apply effects to different parts of the image differently.

(I don't honestly know if adding a channel layer in PS actually adds an extra channel of bits to each pixel, but that's the concept. It probably uses a different implementation of that concept, because it would be incredibly slow to physically insert 8 bits at the back of every pixel...it most likely just overlays an entirely different image.)

Same discussions as before (years).

There is nothing to discuss, if you work with post processing and have used both methods then you know. There is no reason trying to give the emperical evidence. It is a no-brainer.

I have yet to convert in CS RAW the same level as using a purchased profile for Capture One.

Somebody here mentioned that he has processed 500 shots in a couple of hours in CaptureOne Pro. It's less than 30 seconds per image. Impressive!
Anybody cares to mention what and how they do with their RAW files?

I think we've had enough of bit counting. Now, how do you get from RAW to the print?
--
http://www.pbase.com/andybelov

I've been going back to fred miranda's intellisharpen ii. Seems to work more often for me.

I wonder if intellisharpen iii will use smart sharpen rather than usm.

Magne's profiles are to die for.

I usually use capture one. Magne's profiles are a must for C1. C1 has an awful to learn workflow. Of course once you have it down it is great. Helps that they removed that silly batch limit on the LE version.

RSE is my experimental converter. What that sharpening giveth it taketh away. So I tend to do the "lets see what this will do with it" conversions in it.

And I do a lot of one off images in ACR. If I have a crooked image I usually use ACR just so I can crop and straighten it in one shot.

I wish C1 could run as a CS2 plugin.

Don Glenn said:

If you convert to a linear 16 bit tiff format, then there isn't a
data loss issue. No gamma curve.

Click to expand...

Demosaicing is lossy. So is white balancing. So is RAW RGB -> sRGB or aRGB.

--

John

Andy said:

Guys, think about it. What is an image? It's an array of pixels,
where each pixel has only three things: it's red, green and blue
values. That's it. There is nothing more to change.

Click to expand...

A RAW conversion almost always clips away part of the RAW range of values.

A typical sunlight shot throws away over 75% of the RAW range for the red channel, with normal RAW conversion. You can't get this back by playing with curves in photoshop. Raw values 1200 to 4095 may become one value (255 or 65535) in the 8- or 16-bit output.

--

John

1. Transfer .CR2 files to PC using Downloader Pro. There are no JPEGs since I shoot 100% RAW. The files from my two cameras are named accordingly.

2. Trash the bad ones. dpMagic can display the thumbnails for even trashed RAW files.

3. WB page: Adjust WB (not much to do for outdoor shots). For indoor shots, click on the black, white or neutral gray points. Sometimes I shoot a few reference frames with WhiBal. The auto (magic wand) works well sometimes.
4. Exposure Page:
4a. Adjust EC (exposure) as needed.
4b. Adjust CC (contrast) as needed.
4c. Try with one of 4 film profiles to see which one works the best.
4d. Use the curve tool to fine tune some images.
4e. Use the level tool to fine tune some images.
4f. Adjust CS (saturation) as needed.
5. Focus Page: Fine tune sharpening.
6. Crop the image as needed
7. Use arbitrary rotation tool to fix camera tilt mistakes.
8. Save the adjustment settings by using the archivie settings option.
9. Add one or many images to the processing queue
10. Run IrfanView to resize the output JPEGs for the web.
11. Add watermark in PS or another tool if so desired.

Inside Capture One, at any time, you can apply the settings from WB, Exposure, Focus or all of above to other images.

To see how those adjustments work, look at the online tutorial at http://phaseone.com/Global/Tutorial/Tutorials.aspx

--
Nelson

I use CS2:

1. Create a folder for the job (I call it job even if I'm an amateur). The folder is named "yyyy-mm-dd-name_of_event".

2. Transfer all the raw files from the CFs to a RAW-subfolder under the newly created folder.

3. Open the raw folder in Adobe Bridge. Browse through the thumbnails and tag all usable shots with a star.

4. Open the tagged shots full-screen in Adobe RAW and straighten/crop. Also remove the tag from shots that are OOF (or are junk for other reasons) but that I couldn't spot from the thumbnail.

5. Open the remaining shots once more in Adobe RAW. Now I select groups of shots that are taken in similar light conditions and adjust exposure, WB etc. for the group. Some shots may require individual tweaking.

6. When all shots are adjusted, let Adobe RAW convert and save to the a folder named Original under the job folder. The images are renamed to "yyyy-mm-dd-name_of_event-###" at this stage.

7. Do any editing such as dust removal, if needed, on the images in the Original folder.

8. Run a Photoshop script that creates Medium (for screen viewing) and Small (for posting on-line) folders. The script opens each image in the Original folder, downsamples to the desired sizes and adds a (c) text.

• This is normally done in sRGB and the Originals are stored as JPG with maximum quality. I re-convert some of the shots separately if I need Adobe RGB or 16-bit.

• I create files for print, contact sheets, web galleries, montages etc. etc. separately if needed.

• This is now done 100% in CS2. I used DPP for the RAW part before I got CS2.

• Backup is done automatically by my secound computer. I have one dedicated 160 GB disk for photos in each computer. An automated script copies new stuff to the other computer as soon as both are on at the same time. In addition to that I create two DVD-copies every month of all new stuff. One copy is stored at home and one at work.

Mikke

I have not read all of the responses, so i do not want to duplicate, but to say the least you have oversimplied the importance of the bayer interpolation and what actually goes on in the RAW state.

Each RAW converter uses different algorithms that greatly affect the final image quality, and some like RawShooter even allow you to make adjustments to the bayer interpolation process.

Then of course the ability to make the "curves and levels adjust ments that you speak of on the RAW data rather than the converted data is a huge advantage.

Whatever works for you is fine with me, but the more you can do in RAW the better your image quality will be as well as the faster your workflow.

It is not as simple as you have represented.
--
Regards,

Michael Tapes
Owner
http://www.WhiBal.com
http://www.RawWorkflow.com