What format is an un-saved RAW?

As in, when you process your "image" (RAW file) in Lightroom or ACR and then open it in Photoshop, the extension is still the RAW extension - dng or cr2 in my case. I know that Photoshop itself cannot display RAW files, so what are you actually working with until you save?

One of the reasons I'm asking is because I want to know if there is a difference in bit depth between editing in Photoshop 16 bit mode versus working in the RAW processor. I know that a RAW file can only be 12 or 14 bit out of camera, and I read somewhere today that the RAW converter will automatically convert to 16 bit mode for editing. So I'm curious what exactly the difference is if there's 16 bit editing happening in both RAW and Photoshop as well; I'm guessing that opening in Photoshop must generate some sort of subtle quality degradation, or else there wouldn't be so much emphasis on getting things so close to final while still in the RAW processor.

-Randy

When you use the LR Edit in PS command, if you are using a version of PS with a compatible version of ACR (i.e same Process Version) then LR passes all of the current settings that you applied in the Develop module to ACR which then renders the image to an RGB file of whatever bit depth you have chosen in the LR preferences (8 or 16 bit) and also using the chosen colour profile from the same set of preferences. This is exactly what ACR dose when you go from ACR to PS. The only thing is that although ACR processes the image, the ACR window is not opened, as it is taking all of the settings from LR. When you then hit Save in PS it will save the image as either a PSD or TIFF file with all of the layers etc saved again depending on the LR preferences, and automatically import the image into the LR catalogue and stack it with the original RAW file. The PSD/TIFF will be saved in the folder with the RAW file as well.

If you do not have a compatible version of PS/ACR then you can force LR to render the image itself. In this case the PSD/TIFF file will be created by LR and then that file will be opened in PS.The only difference is that in the first case if you close the image in PS without saving you will have no PSD/TIFF file saved on the drive. When LR does the rendering you will have the PSD/TIFF file on the hard drive and in the LR catalogue even if you do not save the image in PS.

HTH

Alan

The advantages of Raw, and the reasons why many of us emphasize "maximizing" your Raw "development" workflow, go beyond the bit depth (comparing say the 16-bit tiff to the 12- or 14-bit Raw). It goes into the nature of the Raw "data" and the latitude you have in processing both colors and tonal data. Yes, for a "rendered" RGB image a 16-bit tiff or psd gives you advantages over an 8-bit file but if you have gotten the most out of your Raw file then you may not actually see meaningful advantages from a 16-bit file. This can make a real difference in how your system resources get used.

In ACR/LR you really are working with RAW data. Once you open into PS it's effectively a PSD/TIFF at the bit depth you select in ACR/LR.

Do everything you can in your RAW converter, it has more data to work with.

I guess I'll add: the Smart Object function can be pretty useful: Open in Photoshop as a Smart Object and if your Camera Raw is compatible with your Lightroom then the image opens with the RGB rendered preview and editing capabilities in its own layer, but that layer can be then re-opened in Camera Raw to do further Raw processing! It's pretty cool!

The two main differences between between editing a Raw and a psd/tiff are:
A.) The Raw data is linear and the RGB image has had a tonal response curve applied to it, gamma 1.8 for ProPhoto RGB, gamma 2.2 foe Adobe RGB and "sort of" gamma 2.2 for sRGB. The TRC skews the data, stretching the shadows and compressing the highlights; so while calculating edits to linear data is relatively straight forward, edits done on gamma corrected data are much more complex and therefore involve more quantization (rounding off) errors.
B.) Raw files are edited once only, at the moment you output the RGB rendering and the workflow is fixed by the designers; pixel editing in PS piles more edits on the original edit and sometimes the order in which this is done can have an important effect.

Note: To avoid giving a false impression I should note that in 12, 14 or even more so 16 bit depth, quantization error, although always present, is normally not a problem. However, it is cumulative and even though PS is working in 16 bits (actually 15 bits, but that's a story for another day) with a lot of editing the effects can pile up and eventually data gaps become wide enough that they remain even in the 8 bit output version.

tzalman wrote in post #16252428
The two main differences between between editing a Raw and a psd/tiff are:
A.) The Raw data is linear and the RGB image has had a tonal response curve applied to it, gamma 1.8 for ProPhoto RGB, gamma 2.2 foe Adobe RGB and "sort of" gamma 2.2 for sRGB. The TRC skews the data, stretching the shadows and compressing the highlights; so while calculating edits to linear data is relatively straight forward, edits done on gamma corrected data are much more complex and therefore involve more quantization (rounding off) errors.
B.) Raw files are edited once only, at the moment you output the RGB rendering and the workflow is fixed by the designers; pixel editing in PS piles more edits on the original edit and sometimes the order in which this is done can have an important effect.

Add to that, in the Canon world the nature of Raw data is not RGB pixels but monochrome pixels, but ones that give you more latitude for adjustments!!

P.S. - if your conversion from Raw is less than optimum, if you have not brought over all the data from the Raw - all the DR for instance or you have clipped the data - you can't fix that in the psd. The only alternative is to go back to LR/ACR, to the Raw file, and fix it there. If you are using DPP, the conversion/transfer will always lose data, so not much you can do there.

That all makes sense. And Tony, I often open as smart objects and remember how floored I was the day I discovered that - especially that it was while I was just learning editing so it was a huge asset to me. These days, if I'm really confident in my raw conversion I will open as a standard image to cut down file size.

So to sum it all up, how's this sound for a game plan: maximize the absolute most dynamic range, contrast and color out of raw converter that you can, BUT leave a little bit of wiggle room on both sides so that you can do some more intricate manipulation in Photoshop and stretch out the remaining areas, i.e. don't push the shadows and highlights too close to clipping/where you desire so you have some room for heavy targeted curves in Photoshop. Sound about right?