RAW Development

[JeffreyWalther]

I wished it was  possible to:

- also save during RAW development
- edit a picture and still have access to the RAW data (no baking in)

[JeffreyWalther]

- also save the RAW settings back to the RAW file (instead of creating a new Affinity Photo file)
- cancle RAW development by [X] without first cancle development [YES/NO] and then again close the window[YES/NO].

[John Rostron]

No image editor will change the raw file (not even Adobe Raw). Many image editors save a record of edits that have been made as sidecar files. Affinity does not. If a program saved your changes back to a raw file, it would no longer be a raw file.

Many users have requested that Affinity implement sidecar-like files. So far Serif has not committed itself.

John

[JeffreyWalther]

3 hours ago, John Rostron said:

If a program saved your changes back to a raw file, it would no longer be a raw file.

Well, that is actually not true.
The raw file would still have no baked in data. The changable settings would just be updated.

But I do not care.
I really wish it was possible to develop a picture, save the raw state and open it again at any time to continue.
But it would be great, if the raw settings would never baked in so that you can change them at any time.
This is already possible with raw footage of cinema cameras. And this is much more data to process.

 

 

[fde101]

On 10/3/2019 at 11:32 AM, John Rostron said:

No image editor will change the raw file (not even Adobe Raw).

There are a few camera-specific RAW developers which will do this for some cameras.  They do not change the image data, but as @JeffreyWalther suggested, simply update the metadata stored in the file indicating what the "camera" settings were.

I don't realistically see this happening for a generic application that deals with a range of cameras across multiple companies.

I don't think most of the "big" camera companies tend to offer this either, but SIGMA Photo Pro does this for example.

 

On 10/3/2019 at 2:44 PM, JeffreyWalther said:

But it would be great, if the raw settings would never baked in so that you can change them at any time.

This is how non-destructive RAW development software works - DXO Photolab, Capture One, On1 Photo RAW (as long as you are not using the "layers" feature), etc.

Affinity Photo is a different kind of program.  Once you start editing the image data the layer becomes a "pixel" layer and attempting to re-develop the RAW data would wipe out the edits that you made.

[Ron P.]

I think Serif needs to strongly consider adding XMP (side-car files) to Affinity Photo. Since they do have a raw developer. Most other raw develop apps do include this. One such app is Corel's Paintshop Pro. It has a RAW developer built in, and comes bundled with Aftershot. Paintshop Pro is in the same price range as Affinity Photo. It's constructed about the same as Affinity Photo, where the RAW developer is somewhat simple, and main emphasis is on Raster (pixel) image editing.

[JeffreyWalther]

Coming from the film business, I also edit RAW footage. Most of the time I use DaVinci Resolve for its nice node workflow.

There you have your raw footage, where you also can change the raw settings (ISO, whitebalance, shadows, highlights,...) afterwards. They are not baked in and never will.
But then, any further edit on this is a non-destructive process. You places nodes after nodes doing something with the picture. This can be masking, grading, color shifting. The process is stored in the nodes and applied live while watching the footage. And those can be 4K at 60p or more and it works on beefy machines.

This is something I would really like to see on simple (single) pictures. Having the raw image with changable raw settings and then live adjustments on top without crushing the original image. Even pixel edits should work that way (process instructions stored in an adjustment layer or a node as in DaVinci).

Only at the time, when the final picture is exported to PNG, TIF or whatever, then it all should be baked in.

This would be the best feature ever that would have been introduced in Photo.

[fde101]

Currently, the develop persona only works on a pixel layer.

For the functionality you are requesting, it would need to work on an image layer, which would need to track the settings and behave almost as if the develop settings were in effect a "hidden" adjustment layer integrated into the image layer.

As long as you then kept it as an image layer and did not do anything (such as using the inpainting brush...) which would rasterize the layer, you could go back into the develop layer with the settings intact.

If the layer were rasterized into a pixel layer, Serif would no longer be able to offer that feature.

[JeffreyWalther]

5 hours ago, fde101 said:

Currently, the develop persona only works on a pixel layer.

For the functionality you are requesting, it would need to work on an image layer, which would need to track the settings and behave almost as if the develop settings were in effect a "hidden" adjustment layer integrated into the image layer.

As long as you then kept it as an image layer and did not do anything (such as using the inpainting brush...) which would rasterize the layer, you could go back into the develop layer with the settings intact.

If the layer were rasterized into a pixel layer, Serif would no longer be able to offer that feature.

As mentioned before, the technology is already there and existing, e.g. in DaVinci Resolve.
I see no difference in image processing. Photo is single frame editing, Resolve is a sequence of frames which requires much more technology and know how.

[fde101]

4 hours ago, JeffreyWalther said:

e.g. in DaVinci Resolve.

This isn't really the best comparison.

Node-based workflows are an entirely different beast from layer-based ones, destructive or otherwise.

 

A better comparison would be just about every DAM and RAW-development package on the market - Capture One, etc.

That is an entirely different type of program from what Photo is, and as Serif is developing one for the Affinity line, I would hope to see this kind of non-destructive capability in that DAM - whether or not it gets added to Photo is another story.

 

The existing differentiation between an image and a pixel layer does in fact provide a mechanism on which this could be built within Photo and quite reasonably.  This would be a change and it would be different from anything else out there which fits the design pattern of this software - but that isn't necessarily a bad thing.

 

This type of program is destructive by nature, so whether or not this will ever get added here or what kind of priority it will be given I would consider to be up in the air.  From one perspective I do think it would be a clever capability to incorporate and would offer (with a minimal but important change to the current behavior on the surface of the app) a level of flexibility that simply doesn't exist in other similar programs, but people already get confused between the two layer types (as evidenced on other recent threads, and older threads) and complain about the way they interact, so I don't know if Serif will be too quick to add yet another layer of complexity to the interaction model in support of what you are requesting.  Image and pixel layers are quite different even if they superficially look similar for many uses, and many users don't quite seem to get it.