Focus Stacking

Persoanlly i dont see Focus Stacking as a feature of C1, its best left to something like Helicon Focus or PS.

Hi Christian G,

https://support.captureone.com/hc/en-us/community/posts/7307154882205/comments/12732456793117

Please look at this process:

Hello, I'm a professional photographer and I work a lot with focus stacking for jewelry. I shoot with a Phase One IQ4 150mp. What I'm suggesting here is not a problem with the capture one but is a missing tool for using with the PhaseOne XF body.

What problem do you see this solving?

I will describe my workflow, I set my XF camera for focus stacking, open live view on capture one, focus in the near point, go to the camera and click the first botton, back to the capture one and focus the far focus point, go to the camera again and click the second Botton and then I can take the shot. When you have the camera high on a foba stand go and back is very annoying, especially when you take hundreds of photos like that per week. I imagine a dedicated tool on capture one where I can focus and choose the near and far focus point and take the shot without the need to touch the camera. 

 

When was the last time you were affected by this lack of functionality, or specific tool?
I never found that functionality on capture one. 

Are you using any workarounds or other solutions to achieve your goals in Capture One?
I need to do it manually, but since this is the native software of a 50K camera I think this tool is very needed. 

Are you happy to be contacted further about this suggestion or request?
Yes, I will very happy to help solve that problem. 

Best regards,

César Oliveira

https://support.captureone.com/hc/en-us/community/posts/14117230065437-Focus-stacking-interface-for-Phase-One-XF-IQ4?page=1#community_comment_15312676446877

368752449678

My comment was to steer the conversion back to the topic, not a comment on feasibility, timing or otherwise. 

For the record in this thread, I'm interested in a traditional focus stacking solution that has much more friendly workflow than what we have to do now to use a third party tool.  I wouldn't mind using a third party tool that was well integrated with C1 if the workflow was a lot simpler than it is today (auto generation of temporary files, auto import of result back into same folder with an appropriate filename that sorts with source files, auto cleanup of temporary files, etc...).

My personal preference for a third party tool to integrate with would be Affinity Photo since I already own that and it seems a popular companion with Capture One since many Capture One customers are not partaking of Adobe tools, some edits just require pixel editing beyond what Capture One can do and Affinity Photo is one of the best bargains out there in photo editing tools (you get a highly capable pixel editor for not a lot of money).

I'm not in favor of some AI Focus tool that attempts to recreate in-focus looking content where the out-of-focus areas were in a single shot.  To me, that is a different tool that might have a bit of market overlap with focus stacking, but is not the same solution at all and is not what I'm looking for.  Since you cannot just "sharpen" or even "AI sharpen" seriously out of focus details in an image, the only way to make them look sharp would be to generate new content (generative AI).  That is making up detail that may look like it belongs (if done well), but it's not necessarily the same detail that was actually there in the original shot.  I'm interested in the actual detail which requires traditional focus stacking.

375613028478

"While I truly enjoy the passion and engagement I'm seeing, let's keep this thread on the original topic of Focus stacking."

Should we understand that proposal to imply that, if Stacking might become an active development project, you would actively tend to favour the traditional technical approach over some sort of fully AI driven development (at this point in time)?

The traditionalist in me would also favour such an approach.

However, I do wonder whether some camera manufacturers and software developers might be looking in a different direction.

Hi guys, 

While I truly enjoy the passion and engagement I'm seeing, let's keep this thread on the original topic of Focus stacking. 

383401430157

I understand you prefer C1 to focus their development effort on implementing AI sharpening tool instead of focus stacking. Fair enough. 

Feel free to create a new thread in Feature Requests for AI sharpening tool and promote your idea there. The topic of this thread is focus stacking, not AI sharpening.

368752449678

You're stretching some things in your hypothesis.

If images of a static subject are taken from a single position in theory it must be possible to replicate, in software, what the lens can do by shifting focus.

Why do you think that is true?  If something is significantly out of focus, the detail for that part of the image was simply NOT recorded at all.  This is not like one of those gimmicky cameras from a few years ago that actually had a ton of sensors that was recording images from a whole bunch of different depths that could then refocus the image afterwards.  An image capture with a single sensor only captures full detail for the parts of the subject in focus.  Heavily blurred areas simply don't capture much detail and the only way to try to reconstruct that detail is to guess.  A good AI implementation may be good at guessing, but its still guessing.  It's not the real detail.  It can be made to look pretty good to someone who doesn't know what the real detail should look like, but those guesses don't necessarily represent what was really there.

(Moreover, a single shot made with a relatively fast shutter speed, stopped-down lens, and some sort of high ISO capable sensor, might normally be preferable for the user since the movement of content matter in the image between frames could be largely eliminated.)

Moving content does create its own challenges for focus stacking.  Still one has to decide if AI generated content that may not be what was really there is OK for your use.  You will note in my earlier post, I didn't dismiss AI generated content creation for out of focus areas as entirely useless for all cases, just not what several of us wanted for our particular purposes.

This would not imply any AI-created content. Just the use of AI learning to better adjust content pixels for perceived sharpness and details that already exist in the image but in imperfect form compared to how the photographer wished them to be displayed.

This sounds like you're talking about sharpening fixup rather than what we're really talking about.  When I take a landscape shot that I want sharp from 3ft to infinity at f/5.6 at 80mm, there is no photo to be had where things are just slightly out of focus near or far.  Without focus stacking, one of the other of foreground or background is going to be seriously blurred, such that content detail is not just slightly out of focus, but is lost.   This is where I find focus stacking most useful - particularly when I want something close to the camera in sharp focus along with a far away background.

I almost hesitate to write this but most human eyes and supporting processing engines (the brain) do not offer the level of DOF, at any instant of light capture, that one normally tries to present in a photograph. 

Yeah, but human eyes have the ability to scan the scene and focus on whatever they want to.  Those same eyes cannot do that in a photo so the entire photo needs to be sharp if you want the viewer to be able to see any part of it sharply.

So, philosophically speaking, most photographs are, and always have been, interpretations of visual "fact". So any processing at all, no matter the source of the process, would always be adding to (or eliminating) data as far as the observer is concerned.

What are you trying to do here?  Are you trying to convince us that there's no value in focus stacking, that we shouldn't be striving to offer certain types of photos that have a large depth of field so the viewer can look at any part of the photo and see a sharp view?

However, this may be speculation about future development possibilities that are too far ahead of the present to influence whether focus stacking, as it is currently performed, is something that might become considered obsolete before a functional version of it could be developed and released.

Again, not sure what this is about?  It sounds like you theorize that some future technology will supplant focus stacking so perhaps we shouldn't even ask for a better focus stacking solution now?

Please keep in mind that all I asked for in this thread was for better workflow integration with existing focus stacking tools so it's not so much a pain to use today's focus stacking technology using other existing tools like Affinity Photo.  I'm not asking C1 to go invest a whole lot of their development team recreating today's focus stacking features in their own software.  Just give me decent workflow integration with existing tools and I'll be fine with that.  Why are you arguing so hard against this?

381975022417

My previous thoughts on AI sharpness calculations for a single image (or maybe a very limited number of images ) vs tens of images were with studio-based product photography in mind. Or maybe macro shots. Or perhaps even Jan's models. 

If images of a static subject are taken from a single position in theory it must be possible to replicate, in software, what the lens can do by shifting focus. Indeed since moving the focus might also imply other concerns about detailed image changes the use of a single image and s single shot might offer benefits.

Knowledge of the distance to various parts of the image could help (probably already known in the file)  and a well-trained AI might assist with making the size and reflectiveness of components of the image in the correct proportions to the primary point of focus at the same time eliminating out-of-focus artefacts.

Use for landscapes might require a different approach but, in theory, landscapes typically would not have extreme, mm by mm, critical sharpness requirements. So the idea of AI-trained sharpening of relatively distant (compared to studio product images) points being made less out-of-focus looking may have even greater validity, from a purely technical POV.

(Moreover, a single shot made with a relatively fast shutter speed, stopped-down lens, and some sort of high ISO capable sensor, might normally be preferable for the user since the movement of content matter in the image between frames could be largely eliminated.)

This would not imply any AI-created content. Just the use of AI learning to better adjust content pixels for perceived sharpness and details that already exist in the image but in imperfect form compared to how the photographer wished them to be displayed.

I almost hesitate to write this but most human eyes and supporting processing engines (the brain) do not offer the level of DOF, at any instant of light capture, that one normally tries to present in a photograph. 

So, philosophically speaking, most photographs are, and always have been, interpretations of visual "fact". So any processing at all, no matter the source of the process, would always be adding to (or eliminating) data as far as the observer is concerned.

However, this may be speculation about future development possibilities that are too far ahead of the present to influence whether focus stacking, as it is currently performed, is something that might become considered obsolete before a functional version of it could be developed and released.

Plus that part where Affinity is a meh RAW editor (imo) but does quite well as a pixel editor - much like Photoshop and Lightroom.

369182181677 – I could imagine for Affinity Photo it might be more opportunity to steal users away from Adobe. The integration between Adobe and Capture One is non-existent. If Affinity Photo took this on it might be at least a small win for them.

Keep in mind that the user base for Capture One is somewhere in the range of 260K licensed users supported by a staff of about 150 across development, management, support, etc. Adobe is a multi-billion dolllar company with ten times the user base supported by 30K staff? The two companies are in completely different leagues in terms of user base and employees. Adobe makes the vast majority of their revenue from publishing. Photography is a small fraction of their revenue.

I provide these stats only to build the picture (pun intended) that Capture One has to make much more discerning decisions about where to expend their resources based on what they believe will deliver the most value for their users.

Personally I believe Capture One needs to pay more attention to deficiencies that cast negative light on the company. These deficiencies may be costing them as much in reputation and users as new features might bring them. I 100% agree they need a major overhaul in DAM capabilities and core code. I sense their code base is woefully dated and needs major transformation and that this is why it is such a major challenge for them to tackle these long standing issues.

When it comes to Focus Stack Merge there are exceptional products out there. Encouraging those companies to provide plugins to Capture One might be a faster path to success. Given that Capture One has marked this feature as Not Currently Planned they must have determined that the effort won't bring the needed value to enough customers to warrant the investment.

369182181677

I just don't see what's in for Affinity? You expect them to do the work C1 is not capable to deliver?

Lightroom and Photoshop have a synergistic existence.  Many who have Lightroom use Photoshop with it to do things that you cannot do in Lightroom.  The same kind of relationship can exist for Capture One and Affinity Photo.  For things that you may want to do to your photos that Capture One cannot do, you can take them into Affinity Photo to do so.

To name just a few things:

• Focus stacking
• Skewing or stretching panos to correct edges/corners or fit a crop better
• Much more advanced masking
• Much more advanced cloning, healing or eventually generative AI fill
• Upsizing and then applying advanced sharpening
• More advanced pano merging
• More advanced HDR merging

So, what's in it for Affinity?  The more Capture One users are aware of the advanced things in Affinity Photo and the cleaner the workflow is between Capture One and Affinity Photo, the more Capture One users are likely to purchase Affinity Photo which is perhaps the greatest single bargain (bang for the buck) in photo-related software.  There could be co-promotions, co-marketing, etc... that could benefit both parties.

As I said elsewhere, Lightroom can fallback to Photoshop for the things it isn't capable of doing, it would benefit both Capture One and Affinity Photo if they had the same kind of relationship.  But JUST marketing it without streamlining the implementation and workflow would be hogwash.  You have to deliver the goods on making them work well together, which is not a big development project (much of the work can be done in one or more Capture One plug-ins much like Helicon Focus has already done in their plug-in).

"Affinity has promised digital asset management for years and has yet to deliver." Correct. And no difference to what C1 dares to call DAM. It's management, Walter and C1 sucks here big time as well. For a large multiple of the price. Anyway, that is nothing what any company should prevent from developing a decent focus-stacking app or plug-in. Except of course, the energy and time going into an interface for old C1.

Which still would not result in any progress of a DAM on both sides, just saying. But if ever Affinity comes up with some kind  of DAM, you could bet on Affinity's developers would not dare to forget implementing a decent search function, which apparently is "too much asking for" on C1's side of things. Btw. they are lacking of that feature for quite a longer while than Affinity is promising a browser or DAM.

I just like to know, what's in for Affinity? The benefits are all going to C1. But why not asking for it... 

369182181677 – Affinity has promised digital asset management for years and has yet to deliver. Helicon Focus has a C1 plugin for Focus Stacking. If Affinity Photo provided that kind of interface for Focus Merging, HDR Merging, and Pano Stitching, then C1 could get rid of their own features that they clearly ceased to develop after promising to improve them. I would be OK with that.

Admittedly I'm a bit bewildered by the idea to let Affinity Photo do the job C1 describes as "not currently planned" (which reads to me like "we can't do it better than special apps" and I understand that well).

You realize what Affinity Photo is already able to do? For less than ¼ of the license costs the app already can do much more on multiple fields than C1 ever will be able to do. One has to ask how long the quality gap in RAW editing between C1 and AP will remain...

I just don't see what's in for Affinity? You expect them to do the work C1 is not capable to deliver? 368108974477 how about financing this work and still justify C1's steep prices?

368108974477

I posted to your Affinity thread, but because I'm new to those forums my post is awaiting moderator approval.  I attempted to document the current workflow and what an improved workflow could be.

381975022417 – See this post.

https://forum.affinity.serif.com/index.php?/topic/194900-capture-one-affinity-photo-plugin/

368108974477

 I too would like to see some great integration between Affinity Photo and Capture One. Perhaps we should hit the Affinity Photo boards and suggest that they work with CO to create some plugin features with Capture One.

Excellent idea.  Let's hit their boards.  See you over there.  I'm still hoping that someone at Capture One in product management or business development is listening here too.

381975022417 – I too would like to see some great integration between Affinity Photo and Capture One. Perhaps we should hit the Affinity Photo boards and suggest that they work with CO to create some plugin features with Capture One.

368752449678

I would not be surprised if someone out there running an AI project is looking at ways to simple take one image and create a "sharp" version, total depth of field if required, from something with significant areas of out-of-focus image.

I've seen a couple of sales pitch samples that seem to suggest that a well-trained AI is now able to take a very blurry image and make it remarkably sharp. (How it might do that is probably a discussion for elsewhere.)

If that really is possible to a point where a typical viewer would not be able to tell whether the image was a single shot with AI sharpening applied or a stack of several images, then perhaps the traditional concept of multi-frame focus stacking and processing is coming close to being superseded quite soon. At least as far as the mass perception of a successful outcome is concerned. 

I don't consider AI generated content (that attempts to recreate detail that doesn't really exist in one captured image due to lack of DOF) for my landscapes as a true substitute for focus stacking.  It may have its own uses, but it's not what I'm looking for for landscape image focus stacking.  I want actual focus stacking. Jan seems to have the same opinion for their scale model photography and I could image that might be true also for other product and food photographers who want real detail, not AI manufactured detail.

And your own suggestions with regard to an Affinity plug-in form of solution makes sense. Something to add the the existing Helicon plug-in solution. 

So the simple solution providing a satisfactory result and "uncomplex" workflow seems to be favoured so far.

I can't speak for others.  But, I already own Affinity Photo (it's my pixel editor when C1 can't quite do what I need) and its focus stacking is quite good.  So, my main complaint with using focus stacking via Affinity Photo is that the workflow is cumbersome and involves a bunch of manual steps which just makes the whole process take a lot of time and its easy to make mistakes with where things are put and what you can and can't delete when you're done.  So, I'd definitely be OK with just making the integration smarter so the workflow was easy.  If that workflow was fixed, doing the actual stacking in Affinity Photo is fine with me.

I'll repeat something I've said earlier.  I really think Affinity and Capture One need to put their heads together and make more great integration points between these two products.  If Capture One continues to avoid doing their own focus stacking implementation, then what does anyone have to lose by doing a really good workflow integration with Affinity Photo for focus stacking.  It just makes both products more attractive when used together.  And, Capture One customers must already be in the sweet spot of Affinity Photo's target customer base since, somewhat by definition, they are far less likely to be Photoshop users (already selecting a non-Adobe product and perhaps a non-subscription product).

And, both Affinity Photo and Capture One are completing head-on with Adobe's Lightroom + Photoshop built-in integration and photography plan.  Why not make a partnership that has a similarly compelling integration?