I am interested in stacking and stitching but there is a but.
1 the wider the options are to choose from the better the navigation and layout or the UI needs to be.(you don’t want to be lost in the the woods searching for a tree…)
2 the more the application needs to load the slower it gets, or you need a beast of a pc.
So segmented loading would be for this kind of work better. (as in a internal exportation of tiff’s to a seperate application, (like nik) so the workload is shifted. (indeed a autoreturn to PL of the result.)
3 this implements a third “sergant” application as like ViewPoint and Filmpack.
I am not implicating that dxo needs to be lauching a new program and charging a leg for this but more it must be not realtime stitching/stacking from raw file IF this implements a slow preformens overall of PL.
Ideal would be a bidirectional connection which allows you to change one of the images wile in stacking stitching mode by going back to PL’s rawfile to change for instance WB or point of view, export again to stack which replaces the old one.
This requires “temp files” which are 16bit tiffs, these needs to be able to clear out once a wile to safe harddisk space.
I am not often in a use of stacking or stitching.
Now i use Silkypix v10pro which has only stacking,(it’s a cheapcprogram if you use the panasonic only version, which don’t swallow DNG’s but tiffs are great.)
Before that i used ICE of windows and then i used the good old stacking app combineZP.( made for “starshooters” in the old day’s) was better then ICE back then.
So to conclude, it has to be flexible, non slowing down main app in iddle mode, above average in quality.
Not an easy task.
I can see one particular problem with DxO trying to include HDR merge/ stacking/panorama stitching…
Taking Affinity Photo as an example, once you import a stack of images into AP for merging, you are presented with a list of layers, that you can choose from, and that can be used to paste parts of one layer onto another in case of incorrect merging.
All this has to happen at the pixel level, not the RAW image level, thus forcing PL to become much more of a pixel and layer editor. It is my feeling that, as nice as these things are to have, implementing pixel layers and editing is a massive investment that could distract from maintaining and improving the RAW side of things for which PL is so well known and loved.
2 Likes
m-photo
( Marc (macOS Sonoma on MBP16" Intel))
15
I like the idea of exporting great DNG files from PL and importing them into a good pixel editor (like Affinity Photo) to finish the job.
But of course if DxO has infinite ressources and also squeeze all the performance juice out of our computers I am okay with having an all integrated solution… one day.
Unfortunately, although I agree with you in principal, and use dedicated software for Pano and HDR if DXO wants to continue to be successful in the raw converter product space adding these functions is important.
DXO was behind the curve with local editing and is playing catch up, and making huge progress. This development cycle was focused largely on improving the dam and providing Fuji support. The improvements to U-Point technology were excellent and have been well received.
The majority of this development cycle was in fringe areas dam and Fuji. Fuji support was definitely necessary and dam improvement worthwhile but did, I am guessing, soak up a lot of development time, that in the short term will not pay back the ROI. DAM has the potential to continue to be a bottomless pit in support time, but I hope I am wrong on this.
The development push for V6 needs to focus entirely on image editing, DXO’s main purpose, building on the layer based UI for local adjustments, mask editing, copying between layers/images, masking from colour and overall editing including pano and hdr. No one is currently doing focus stacking in a raw converter.
Lightroom already has this functionality and Capture One will add this in the next couple of months. C1 is a similar size to DXO so it shows that such functionality is possible.
My point about pano and hdr functionality being needed by the market is reflected in this Jim Nix video. DXO has to stay competitive in relation to the competition. I used to have a PowerPoint slide covering the need for continual improvement with the caption, “survival is not compulsory”. A company has to keep up, or preferably exceed the market increase in performance.
From what I can see, there are a few apps that can stitch/merge/stack from RAW files but you always end up with a bitmap, which might then need “tidying up”, which then means bitmap manipulation, which is something DxO haven’t yet done and, I’m guessing, would have some catching up to do.
There is still quite a list of refinements that need attending to in the RAW editing arena and I, for one, wouldn’t want to see DxO divert any more attention away from what they do best.
Yes and after a year of effort, the DAM is nowhere near as competent as its competitors and several people are finding that it doesn’t yet play nicely with other DAMs. Universal compatibility is a big ask and the majority of discussion on this area sees to be devolving down to “only use one DAM” and “I’ve already got a DAM”.
I’m sorry but I thought Jim’s video did PL5 no favours at all. He skimmed over Control Lines, which, apart from Fuji support, has to be the most powerful, fully working, part of this version. And I didn’t really get anything like what you are saying about pano and HDR.
If there was one thing I am screaming for, more than panos, etc, it’s the simple ability to flip images. I’ve done programming for rotation/flipping and it really isn’t rocket science. If you can rotate, you can flip.
I believe that merge/stack/stitch is done AFTER demosaicing and cannot be done on raw files. A lot of people think Lightroom does this on raw files but this is not the case as the resulting dng is not raw but a tif file in a dng wrapper.
I see this as being no different to doing lens correction and maybe denoising in PL and then sending the files to another program like Affinty Photo to do the merge/stack/stitch.
I don’t want PL to be a jack of all trades and master of none. I use it for RAW conversion and doing that well is what I most care about.
IMO PL could be more helpful when processing multi-shot images for stacks etc. I would like to see a way to define a set of images which need similar processing (because they are a stack or panorama etc) and have PL mostly show and treat the set as if it were a single image. Export to disk to automatically created per set folders would be an important feature. Once I have identified say a set of 5 images as an exposure bracket PL should help me not need to identify that set again.
I think this would be good value for development effort. I don’t see DxO implementing any of the requests in this thread except maybe exposure merge/average because they already have in it Nik, but, why would they want to bake it into PL rather than leave it in Nik?
Please don’t shoot the messenger. Jim Nix revered to the new Control Lines as awesome and I think they are great. However, DXO’s marketing of the new feature could have been better and the simple addition of a “filled” mask option, the equivalent of setting a control line passed the image boundaries (what a kludge), would have made them even better for very little coding time.
Jim’s point was that he reviews all the software titles, expectations of a new version are high and metadata and Fuji, although important are niche, filling in holes in Photolab, but not providing much meat in the new version. Had Photolab implemented pano and hdr the review would have been an upgrade for sure.
I want DXO to flourish but you can’t exist in your own “bubble” without understanding what your competition are doing. Where DXO go in V6 will be critical and I think they must be seen to be meeting market expectations. I might not like the direction, I use dedicated software for pano/hdr, but DXO is not just for me. So yes DXO need to plug some long standing deficiencies as you gave example and this should be done in point updates of V5, but they also need to go where the market expects.
I have said before that DXO should fold ViewPoint and FilmPack into Photolab to make Photolab stronger. I personally don’t think DXO have a viable business model without Photolab and they should recognise this and double down on making it the best product available. Add-on sales were the right strategy 5+ years ago but in business change happens, adapt or…
I don’t know what is a DAM but as a Fuji user, i’m glad that X-Trans was (at last) considered for PL5.
Now, I would love to have a focus / pano / HDR option. The best would be like VP or FM, stand alone products which can be integrated seamlessly in PholoLab. Plug in are maybe cool but having an integrated workflow is much better.
There has been no feedback on that functionality, but I doubt that those features are even being considered in their current plans. Both of those features would require a significant redesign of the browser functionality.
In the streaming 2 or 3 days ago, they said they are working on some of those tools (don’t remember which one(s)) but they are not yet happy with their result because they don’t want to deliver a product not as good as what already exists.
They didn’t say if this will appear soon or not (it seemed to be not soon).
They also told that HDR is now beyond current possibilities of photolab, because of the precision needed (HDR needs to use floating point precision numbers and photolab isn’t coded like this). They didn’t tell if they are looking at this now.
m-photo
( Marc (macOS Sonoma on MBP16" Intel))
27
Where did you hear this ? I mean what do you mean with « in the streaming » ?
I have already asked for panorama stitching for three years (Panorama Stitching).
I am currently using Autopano which a great software but a dead software killed by Gopro.
Some suggestions to answer to other comments :
this year new features are a little limited ; I sustain “His advance on optical corrections or noise management may no longer suffice.”
many stitching software are based upon free software libraries ; why not DxO ?
DxO could buy Autopano code to Gopro, AP user interface was great
stitching from RAW was available in Autopano : is it interesting to work from RAW source, it is possible ?
I feel that pano stitching, HDR and focus stacking have common operating modes ; that would have sense to work on the three subjects ; marketing consideration would could spread the output over several years (stitching first of course !)
)