Linda From Maine Loc: Yakima, Washington

CHG_CANON Loc: the Windy City

Maybe some highlights will help you better recognize the misleading or completely wrong aspects??

****

So when you open a raw file in ACR or Affinity Photo or any other program, what you see is that program's rendering of the raw data. If you then make changes and save the file, that saved file may or may not include the original raw data. If you save the changed file as a jpeg file, then the raw data is not saved. If you save a TIFF file, then the raw data is not save, but more data is kept than in a jpeg file. I'm not sure about other file formats.

Some programs may keep a record of changes as a sidecar file or other method. In that case, the program may keep the original raw data. Then, when you reopen the file, the rendering can start with the raw data and reapply the changes for you to see.

****

The 'output' file is the edited and transformed RAW data. The output format is a decision of the human using their digital editor. None of the human's output decision has any impact -- any possible impact -- on the original / source RAW file.

Moreover, RAW files have no sharpening, no Noise Reduction (NR) and no White Balance (WB). RAW files have the original bit-depth of the camera's sensor. The output file format may / may not apply sharpening and NR, certainly a WB, even if just the camera data's WB settings. The bit-depth will be compressed down to an 8-bit format for certain target image formats, or possibly, the sensor bit-depth will be transferred 'as is' into a 16-bit TIFF container. All these transformations have nothing to do with changes to the read-only RAW file.

Blenheim Orange Loc: Michigan

Yes. I saw those passages as more vague and poorly written than wrong. But they could be misleading. Thanks.

Blenheim Orange Loc: Michigan

Yeah that is pretty sloppy. Not sure what he even meant by that.

robertjerl Loc: Corona, California

RAW is easy to work with. No secrets to it.
Just a charcoal grill and some BBQ sauce.


OK, I will show myself out, besides it is time for me to get my feet up and watch the news.

burkphoto Loc: High Point, NC

Think of a raw file as a CONTAINER file. You get the data from the sensor, yes, but you also get a JPEG preview image baked using the menu settings in the camera, plus the EXIF data that accompanies that preview image. The preview is what you see in an electronic viewfinder or LED/OLED display of a digital camera, or in your operating system.

When you open a raw file in post production software, several things can happen. If you're using software from the camera manufacturer (or written for them by a third party), its entirely likely that that software will display the low res preview briefly, while processing the raw data into a 16-bit bitmap, using the settings stored in the EXIF table of the raw file. Once that 16-bit image is processed, a proxy of it is displayed for editing and export. SOME such software will save a NEW preview JPEG into the raw file, made from this bitmap.

Third party post production software such as Adobe Lightroom Classic comes with lots of default "looks" or "profiles." You can choose any as your default, or make your own defaults. In this case, when you "import" a file into Lightroom Classic, it briefly displays the preview JPEG, then switches to a bitmap baked from the default "recipe," based on just some of the parameters in the EXIF. You can change everything about it, of course, in the Develop module. But with LrC, what you get is NOT stored into the original raw container... Instead, the changes are stored as a proxy in Lightroom Classic's catalog database, and they may be written to an external sidecar .xmp file as well. The original, "baked in camera" JPEG preview remains in the raw file. If you open the raw file in another application, that is what is displayed. To see your Lightroom Classic changes, you have to EXPORT from LrC to a new file. The best part about that, is that you can create virtual copies of the original file, and work on each of them separately, and retain the instructions to make each version!

In essence, the RAW DATA in a raw file is never, ever changed. They preview proxy MAY be changed in camera manufacturer's software. But the underlying raw data remains in all cases.

bkwaters

Great explanations about RAW.

Now please explain RGB, sRGB, and CMYK color spaces. Thanks.

burkphoto Loc: High Point, NC

This is a VAST oversimplification aimed at helping you use them:

RGB = The additive color model. It's not a specific profile, but a TYPE of profile. Red + Green + Blue add up to white, when mixed equally.

When mixed unequally, combinations of these three colors create all other colors. (The absence of light is black.)

CMYK = The subtractive color model. It's not a specific profile, but a TYPE of profile. Cyan subtracts red from white light reflected from white paper. Magenta subtracts green from white light reflected from white paper. Yellow subtracts blue from white light reflected from white paper. Because these three add up to a dark brownish gray, black is added as a "Key" color for "Kontrast" and detail.

RGB is a generic, undefined color space. But ADOBE RGB is a very specific set of color boundaries that includes highly saturated colors. It is most often used in high-end printing workflows to send files to ink jet printers using six or more colors. It is sometimes used as an intermediate, or "working" color space when editing images.

sRGB, or "standard RGB" is the lowest common denominator of color spaces. It's the one that all the major manufacturers agreed to use back in the 1990s. As such, you can put your JPEGs in sRGB at the camera, and your calibrated monitor will show you a decent approximation of what the camera saw, warts and all. Your printer, using OEM inks and papers, will reproduce that image approximately the same way. The photo lab you send files to for printing will make a decent reproduction of those files, too. HOW good it looks, depends on the accuracy of your white balance and exposure!

In reality, the "best" profile to use depends on your level of understanding, and the quality of equipment you are using. If you have not taken the time to research and learn about ICC color management, it's probably safest to stick to sRGB. It's not the most accurate color space, but it will give you a pleasing result — again, if you are careful about white balance and exposure. If you set those in reference to a target, your results can be excellent.

If you DO understand ICC color management, then you probably want to use raw data capture at the camera (with the in-camera profile set to sRGB). POST-PROCESSING is then REQUIRED to convert the raw data and to manipulate it for the precise effect you want. Here's my workflow:

> Capture original scene in raw, using an exposure and white balance reference target when accurate color is required/desired. Other camera parameters are set to make the JPEG preview image look as close to what I want as possible.

> I calibrate my monitor monthly, using a SpyderXElite from Datacolor. It's a colorimeter and software package that sets my view of the scene to an international standard. Because I print occasionally, I use these parameters left over from my days in the photo lab world:

Initial color temperature 6500K
Initial black point 0.5 candelas per square meter (cd/m^2)
Initial white point 105 cd/m^2 ± 15 cd/m^2
Initial gamma 2.2

> Import the raw files into Adobe Lightroom Classic. Normally, I use the Camera Natural Profile as my default. It's very accurate, but not at all punchy. Of course, since LrC is a non-destructive editor, I can choose another profile at any time to preview the effect of that.

> I use the sliders in the Develop module until I like what I see on screen.

> Any adjustments I do outside of LrC are done in ProPhoto RGB color space, from 16-bit .PSD or TIFF (.TIF) files. They go back into LrC in that format.

> For FINAL adjustment, I use the Soft Proofing feature of the software. This shows me an approximation of what the output will look like on a specific printer with a specific paper and OEM ink, OR it shows me what my lab of choice is likely to print. (Professional labs will generally share their proofing profiles if you ask them. Paper manufacturers generally make downloadable default profiles for each of their papers used in commonly available PHOTO inkjet printers.)

> When printing to my personal printer, I use the Print module in LrC, setting it precisely to the paper in use and the printer I'm using.

> When using a regular silver halide chromogenic process photo lab, I'll export a JPEG in sRGB.

> For very large prints made by high end boutique/giclee print shops, I'll consult with them to determine the best file type, profile, etc. Some can work in ProPhotoRGB, a super-wide gamut color space capable of containing colors human eyes don't see. I want their profiles, anyway, for soft proofing.

> For CMYK, 25 years ago, I would have opened my RGB file in Photoshop and converted my finally-adjusted file from ProPhoto RGB to CMYK using the closest set of parameters for the type of press and paper in use. However, in 2024, this is a crude and stupid approach. MOST savvy offset printers now prefer to receive files in an RGB color space and make the conversion to the flavor of CMYK that works with their paper, ink, and press conditions. Have a conversation with them to learn what they want! If they want YOU to do the conversions, RUN!!!

The major makers of monitor profiling kits are Calibrite (formerly X-RITE) and Datacolor. If you visit their sites and look around, you'll find a wealth of information on color management best practices.

If, like some here on UHH, you think it's all bull$#!t, just "set everything*" to sRGB and pray! (*Camera, working color space, and output color space for files going to labs) Set your monitor to Gamma 2.2, Color Temperature at Medium or 6500K, and turn the brightness way down*.

*For instance, my monitor can output 400 cd/m^2, but I typically turn it down to 120 cd/m^2. Mine's currently at 20, on the 100-scale brightness slider, with contrast at 70. Those settings work for me, but probably won't work for you! I calibrate and profile my monitor monthly. When it was new in 2021, the brightness was at 13! Monitors do get dimmer as they age. Color shifts... which is why recalibration and re-profiling are needed. Calibration linearizes the output of all three color channels. Profiling allows the color engine in the operating system or Adobe software to match the characteristics of one device as best it can to another device.

Monitor calibration is where it ALL comes together. If your monitor is off, files you share cannot look the same on another person's monitor, or print correctly at your photo lab! The number one reason lab prints look bad is monitor calibration! If your prints don't look like your screen, buy a kit and learn how to use it.

One note to avoid confusion: YES, inkjet printers use a subtractive color model. It's NOT necessary and definitely NOT recommended to convert files to CMYK for an inkjet printer, however! Doing so will severely limit the color gamut sent to the printer. All modern inkjet printers do their OWN conversion in the driver software from your RGB profile to the printer/paper/ink profile needed for that situation.

ICC color management can seem complicated. In reality, it is a fairly simple concept: Various devices reproduce color differently. The "color engine" in your operating system or software attempts to simulate the color space of one device as best it can using another device's more limited capability. It isn't perfect, but it is very, very good. It is, however, extremely easy to leave out a step, or "double profile," or apply the wrong profile. So read carefully and test, test, test.

CHG_CANON Loc: the Windy City

Dikdik Loc: Winnipeg, Canada

"A raw file is just the data collected by the sensor in the camera, nothing more or less."

It's a little more than that. It's the most complete data information available from your camera. This data really cannot be improved on. It can be modified so the end result looks better, but it's still the best data.

gvarner Loc: Central Oregon Coast

As long as the edited file is saved with an extension like jpg or tiff for example, the original RAW file is not changed. It’s that simple. I use Photoshop Elements and save all my edited files as psd files.

DirtFarmer Loc: Escaped from the NYC area, back to MA

The original raw file may be changed when it comes to metadata or the jpg preview, but the original raw data in the raw file ARE NEVER CHANGED. It is not a question of saving the edited file. The original raw data are not changed. The raw data are the important bits from the sensor. Since they are never changed, they provide a consistent starting point for an editing program to apply edits.

If the raw data are changed by computer malfunction (corruption happens) your raw data will start to show problems in the edits. Have a backup.

gvarner Loc: Central Oregon Coast

What part of metadata is changed? How can the embedded preview file be changed by an edit since you save your work with a different file name?

DirtFarmer Loc: Escaped from the NYC area, back to MA

Depending on your editing software, the metadata could be changed to reflect file name changes or changes to the editing parameters. Not all editing programs do that but those things are editable portions of the raw file. If you consider dng a raw file, some editing software can rewrite the jpg preview to reflect the current edits. There's nothing that says the editing software can't change TWO files: the raw file (not the raw data within the raw file) and the output file. The actual raw data are not changed.

Date created and date modified can be changed, but that may be a function of the OS, not the editing software.

bkwaters

Thanks. Great job explaining.