Why are TIFs from RT opened in Photoshop with layers merged?

What’s happening is this.

  1. Export an edited RT raw file to Photoshop as a tif
  2. Edit the tif in PS with layers
  3. Re-open the tif in RT for further edits
    a. RT opens the tif file as a copy showing edits but, without layers.
    b. RT exports the tif copy to PS as a new copy, without layers.

So, RT is not opening a layered tif with the original file name and as such will not export a saved layered tif to PS with the original file name.

You can only edit the layered tif in RT as if the layers were merged in PS. So, it appears that moving between the two applications is not going to be a fluid operation whereby the original tif file is being developed further.

Yes, and that’s the problem. If I try to “open…” the TIF “…with Photoshop” from the File Browser I get this message “This command requires all selected images to be queue-processed first.”
This is not usable, when the only purpose with opening the TIF in Photoshop is to examine the status of prior editing, i.e. having the layers listed. I am talking about TIF files initially listed in the File Browser and not yet applied with a pp3 file.

Unfortunately an opening of such a layered TIF file in the Editor immediately applies a pp3 file, and with that the TIF is now deemed a new copy and will open in Photoshop with all prior layers collapsed.

This means that the Editor CAN’T be used as a tool to compare a RAW file with its prior TIF variants the way the arrow icons (F4 and Shift-F4) suggest, as this will turn all prior TIF variants into new copies.

In my opinion this is an unexpected and very unwanted consequence of RT applying a pp3 file to all files as soon as they are loaded into the Editor.

The arrow icons below the Editor suggest that the editor can be used for a one step at the time slide show including all images in the current folder. This is how it works in Bridge (though automated). Nobody expects a slideshow to change anything about the current editing status for any of the files in the current folder - editor or not.

In my opinion a slideshow function - as attempted in the Editor of RT - logically belongs in the File Browser of RT (but is currently missing) - OR the editor must stop creating pp3 files up front and assume that loaded files which are not RAW files remain untouched until “Save as new copy Y/N?” (or likewise) explicitly is confirmed.

I wish to add a few comments here:

  • you’re right that the Editor arrows tooltips are misleading: where they say Navigate to the next/previous image, they must really say Open in the Editor the next/previous image (which is what they do)
  • when you open an image in the Editor, it is always processed by the engine («the program») and all settings applied to the original image are saved into a pp3 file. It’s not that RT applies a pp3 to the image
  • you can imagine a raw file as an input image with 3 layers, one for each RGB channel: when RT opens it in the Editor, it processes all 3 layers to generate a single layer output image that you can see on your display. Same goes for any image with multiple layers: RT process it to generate a single layer image. That’s what it does, as that is what it has to be for a raw file
  • RT is not meant to replace a full fledged DAM. It has a simple file browser to help you open files in the Editor, among other minor things. That’s it
  • you can open an issue on github and request that feature there. Then maybe it will be implemented in a future version (just maybe)

You can easily switch of that TIFFs are shown

Karlheinz,
Yes, but that’s not solving my problem, I don’t want to be without the TIFs, until my new RAW version shows to be competitive (please read my explanations above and below).

Xavier,
I am sorry to hear, that this suggestion to create a slide function in the File Browser in order to be able to compare a RAW file with its prior TIF versions directly in a large format and without changing anything - is not considered interesting. This could prevent me and others to load multilayered TIFs into the editor by mistake (once everybody know the catch).

After all - as more and more photographers find it interesting to test RT as an alternative to Lightroom or ACR, such a functionality is an important competitive parameter, as everybody will be familiar with the similar function in Lightroom and Bridge and most likely miss it fullhearted at some point.
I didn’t know the existence of this barrier until now when finally becoming more confident with RT I recently began to re-edit my older RAW files in RT and in that process needed to compare the result with existing TIFs based on the same RAW files (only created via ACR and Photoshop).

I will not use time on github with the prospect you imply. I must use Bridge parallel with RT - regrettably.

Well, you might volunteer as a developer for multi frame TIFFs … on the other hand, I never missed this feature. And Adobe (or some investors) pay for the development of Lightroom. Everybody here is an unpaid freelancer implementing things in their spare time.

This is not true at all. The multi layered tif file remains unchanged, that’s the promise of a non-destructive editor such as RT. The problem arises when you send the file to an external editor. You’re not using that functionality correctly. It is not a substitute to open that exact same file in an external editor, it is a way to send a processed file to an external editor.

1 Like

The fact remains that when a pp3 file is created for a multilayered TIF, the TIF will not open in Photoshop as such but as a flattened version of the original file.

I need it to open as is, i.e. as a multilayered TIF.
So, this functionality would logically be a part of the File Browser. If this was possible the current processing in the Editor would make sense too.

Also a slide function belongs in the File Browser and would be very useful.

I am arguing for a RT that supports expected functionality when the more “heavy” photographers begin to consider RT an alternative to ACR, Bridge and Lightroom.

I am not “heavy” - just trying to imagine what will happen.

No. Try opening the tif from your operating system, it will still be multi layered.

@Eigil_Skovgaard I would like to help identifying some mis conceptions.

There is no such "presentation " fonction in RT. nothing is presented.
What is done by the “painters pallet” is to process the input file according to the processing parameters, put the output in a temporary file, launch the external editor with it (GIMP,…).
A PP3 will be saved alongside the input file.
The input file is not processed. the output file is a photo in whatever format you want (jpeg, tiff, PNG)

What you want I think : you view your multilayered tiff in viewer ( but what viewer can view a multilayered TIFF?) and if you choose sent the file to photoshop.

I suppose you mean the input multilayer TIFF. As there is NO notion of layer at all in RT and therefore, there is NO merge function
RT opens the upper layer, I suppose. It processes it, displays it and export it in your preferred format (TIFF with one layer).

Remember you are in the editor. What you view in the editor is not a “presentation” but a preview of the output photo.

remark: the tooltip says “edit current image in external editor”. Current image is the output image.

  • So it seems you need a SW that can take into account multilayered TIFF, display a chosen layer, edit the layer, export the processed Multilayered as a multilayered TIFF and launch phtoshop.
    Isn’t photoshop itself this SW? or perhaps GIMP

RT doesn’t manage anything. It is definitely NOT a DAM.
RT strips nothing from the multilayered TIFF, it is unchanged and present in the same place.

Why you need RT passes a file to photoshop. It is the role of a DAM.

RT doesn’t send anything from the file system. It sends the processed output file. Its’ not a DAM

I suppose RT was never envisioned as a post-processing for photoshop although the contrary is true. see @stuntflyer post.
It has to be considered as a raw processor with also interesting possibilities to process demosaiced photos . It not the swiss knife of graphical design
It has no bitmap
capability.

So to get a photo (displayable file) you still need to merge the different layers with specific blending methods. Is it true?

What worries me in this sentence is that the file browser never displays the “original RAW file”.
If there is no PP3, it displays the embedded jpg, otherwise a thumbnail of the output file.

That is sure and the expected behaviour.

That is a problem of photoshop. What is the meaning of the error message?
It could mean that photoshop cannot open multilayered TIFF?
RT is not the way to solve the issue.

The PP3 sidecar file has no impact, only when you open the RT editor.

There is NO slideshow concept in the editor. I should say it is a film strip.
Each time you select a photo, it is processed in order to be displayed.
You can find slideshow in DAM, for instance BRIDGE.

There is no attempt as there is no slideshow.

There is no harm. the multilayered TIFF is unchanged

Not sufficiently accurate. You don’t speak about the same thing.
I rewrite correctly the sentence.
The fact remains that when a pp3 file is created for a multilayered TIF, the output TIFF created by RT will not open in Photoshop as such but as a flattened version of the original file.
That is the intended behaviour. The multilayered TIF is UNCHANGED and thus can be viewed as a multilayer file in photoshop.

Conclusion : use a DAM to manage your different assets (photos, internall formats as PSD or multilayered TIIFF, sidecar files, photomontages, slideshows…).
use RT for what it is developped: process raw and photo formats (jpeg, TIFF,PNG)

There are FOSS DAMs also.

7 Likes

Dear God!
If you really mean to drown initiatives from your motivated users in this way, RT will never move towards an alternative to the widely used raw editors - and I’ll not waste my time trying to explain what is obvious.

There seem to be a couple of fragile egos in the way for progress here.

@Eigil_Skovgaard you recieved the replies that you received because you haven’t made it clear that you understand why the tiff file wasn’t layered in your raster editor. But then you tried to twist your misunderstanding into a feature request.

I’m sorry didn’t not recieve a reply you found satisfactory.

If it’s an answer to me, you missed the point. I am neither a dev or a present user of Rt.
I just tried to explain and to help, but I see to no avail.
Sorry for the annoyance I caused.

Edit: I took some time, in a friendly manner, to help you. I don’t think I deserved such reply.

Now, As I hope, you understood RT in present state cannot do what you expect, you could clearly and thoroughly explain what is your use case in order that some geek can advise on a preferred Foss SW or perhaps develop a new function…

Best regards

1 Like

You guys and girls seem to be in the need of opponents you can argue with without honestly trying to understand the problem presented to you (I exclude Xavier from that observation).

I am not bored and seeking fights. I don’t need to fight anybody, and I find no pleasure in arrogant misinterpretations to avoid offering a serious explanation. I am asking and looking for solutions!

I will still be using and promoting RT, but after this it has become harder to promise anybody that a real breakthrough that could position RT among known commercial products is within reach in any near future.

Not sure we have read the same thread.

Why would you promise that in the first place?

Hey dude, be respectful for the old ones :wink:

RawTherapee does not work in or with layers. There are no adjustment layers or duplicate layers with blend modes or anything like that. If you have a multilayered tiff the first thing that RT has to do is internally merge and flatten the layers. Then you make whatever adjustments you please (say some wavelet contrast adjustments. It makes the adjustment on the internal flattened image. You then save the image as a new single layer tiff.

You seem to be asking for there to be some way for RT to be able to deconstruct how the (in my example wavelet contrast) adjustment would look on all the individual layers of the original tiff taking into account the blend mode interactions and then generate a .pp3 (sidecar file) that show the adjustments on a layer by layer basis so that it can then generate a new multi layer tiff.

The alternative would be for RT to be completely rewritten to use a layer model so that it could do the adjustments on a layer by layer basis. That is just not going to happen.

@Eigil_Skovgaard I already provided you with the exact solution to solve your feature request early on in the thread. I also explained why this would not be a very high priority change to make (yours is not a common workflow imho).
I don’t know what you expect more from this discussion…

3 Likes

No raw processor can develop a multilayered tiff because there is no ‘single layer’ it can work on.
Not even lightroom can do it. Lightroom treats mulilayered tiffs as a virtual stamp visible/flatten image product - yes you can add lightroom adjustments to it - but if you try to reopen in PS you can only open a flattened copy or the original layered version with no Lr adjustments.

Raw files are like a cake mix - something you can be creative with. But a TIFF is a baked cake fresh from the oven.
Yes we can process it by adding adjustment layers etc - icing, sprinkles, jam, cream etc. BUT if we forgot to put the dried fruit in we’re screwed - that’s something we needed to do when the cake mix was RAW.

In the Andy Astbury school of photographic workflow there is only one reason for taking a tiff file back into RT - that would be a flattened version of a fully processed multi-layers tiff for upsmpling via Lanczos.

1 Like

@Eigil_Skovgaard i tried, but cannot reproduce this. Opening a multilayered tiff in RT will create a sidecar pp3. Yes. Exporting this opened file will create new tiff including a counter in the filename like pic-1.tiff. Its a new single layered file and not the original tiff as long as you dont overwrite the original one.
Of course, Photoshop will open both, the original one (still with multilayers) and the exported one with single layer.
Result will be: you got the original wihthout any modifications and an new file including all modifications as a single layer. Why not put this file as an additional layer to the original?

marter