Panorama Stitching

Somehow this thread became a duplicate of the one quoted below that deals with a HDR feature request. But this thread then became a request for all multi-shot editing possibilities, i.e. HDR / Focus stacking and panorama stitching. Today’s DPReview’s test of PL4 also points these missing features in PL4.

Please refer to this thread:

2 Likes

I use a similar workflow too, but if the functionality was included in PL I think PL would truly be a complete tool. I would also gladly buy a like add-on module filmpack.

I have been using Hugin for my Panos. It’s not the most user friendly software and I must do an Export to Tiff in PL4 for each image in the Pano. With Multi Row this can be a lot of images and it’s a very cumbersome procedure! Please add the capability to do Panorama’s in PL.

Free code libraries for stitching are, unless I’m mistaken, used by many stitching softwares.
So it would be a reasonable effort to implement this function.

1 Like

It’s mentioned as a negative in more then one review of the Photolab software and a future that a lot of photographers use. Would be great if DXO finally adds this

1 Like

As others have mentioned, Affinity Photo has made a huge effort recently with their Panorama tool. As panoramas are built with already developed images, I’d be keener for DxO to improve the Affinity Photo integration here making it easier to open up a panorama set in Affinity Photo than to build a mediocre panorama tool.

As I’m a huge fan of DxO software, I’d be happy to see a separate Panorama module like ViewPoint (or better yet, Panoramas added to a ViewPoint 4). I’ll happily pay for an integrated panorama tool if it’s good. Making the tool freestanding and not part of Photolab puts more pressure on DxO to assure the quality of a potential panorama tool.

I do not want a half-working panorama tool which is worse than what’s in Affinity or Photoshop or the existing open source panorama tools. This does not mean including every geometric method of panorama but making sure that the methods which are included are the very best.

PS. Same applies for an HDR add-on. Make HDR a full-fledged DxO citizen with a dedicated module. Potentially HDR could be included in a Film Pack 6 but separate module in this case makes more sense to me.

5 Likes

That’s make sense !
I would like pano tools in DxO but with an ergonomic interface ; I love Autopano and as far as it keep on working, I can wait !
I don’t know if it is working on Mac ARM.

Using ICE + Affinity. The first have several projections but don’t import RAW + auto completion have a large problem in skies. The second import RAW but only spherical projection, auto completion is very good.

In 90% of my work, I merge panoramas. So, an add-on in PL would be a the first improvement I would wish. Working in only ONE software, this is a dream !

But dreams are not coming all alone. I have hundreds of RAW for panorama (thousands if I add my old D700 period). If Dxo staff need help to provide RAW to test a panorama programming, I use to make panoramas from 11mm, 14-24, 24-70, 70-200 on D850.

1 Like

Do you mean that the raw converter makes a raw pano which then will be demosaiced?

George

Is that possible at all?
I imaging that any multi image sticking, blending, stacking is done from a tiff in AdobeRGB even if you open a few rawfiles.
Besides the WB floating around in a raw other things as denoising after stacking or lenscorrection can be a thing.

Silkypix v10 does some stacking multi image from rawfiles but i think it’s tiff file based in the end.

Once I also thought “should be possible in RAW” but since I only occasionally do Panos, I’m very happy with Affinity Photo.

Strasse und Rastplatz

That’s from 4 digitized MF Negatives, 30 years old, handheld* withy maybe only 10-15% overlap, with vignetting and negative to positive conversion. For the vignetting there’s no lens profile for the old Mamiya 55/2.8, and even if it was, it would have to be reversed. The middle part could not be auto stitched, AP needed some help, some masking and transforming. So, not something for auto-pano.

And if I have some images of a mountain chain which is in shadow on the left side – why would I like to make the pano unbelievable with too bright shadows? I want DxO to come up with a better DAM - for less than 10 panos per years I don’t need them to put a lot of resources into. But from t he DAM I would benefit each time I start the app.

*[EDIT] no, can’t be, the tripod was out of it’s container and although an underpowered Gitzo Series 1, the head did have a panorama scale.

1 Like

I agree with you in general, but my feeling is that DXO have too many products for today’s market.

Selling an “extra” product seems like a good idea but DXO need to recognise that Photolab is their main product. The NIK collection may be a viable individual product but Viewpoint and Filmpack are unlikely to be viable products on their own. DXO does get sales from FilmPack and Viewpoint from Photolab customers as add-ons. These do add capability to Photolab but it also makes Photolab very expensive compared to the competition. Photolab Elite + Filmpack + Viewpoint costs £363. I think this is unsustainable going forward.

Without Photolab DXO couldn’t exist on Viewpoint, NIK, and Filmpack. If DXO don’t recognise this then they risk making a similar business mistake as when they thought a software company could compete in the hardware business.

DXO need to fold Filmpack and Viewpoint fully into Photolab, to improve the capability and competitiveness of Photolab in a changing market.

DXO have already folded Viewpoint into NIK and should do the same with Filmpack as NIK also needs improving to stay relevant.

Regarding pano and HDR in Photolab, I prefer to use dedicated software but DXO has to recognise where the market is and it needs to offer these capabilities. Even Capture One has now accepted this.

HDR capability could be added using the technology they already own from the NIK collection. Panos as people have already said should be possible using existing library code. For speed it would probably be best to use a “round trip” to external programs (included with Photolab) giving time for an “in program” solution to be developed but providing a “market need” quickly.

3 Likes

I am not sure but maybe you can use viewpoint and pure raw as combination in rawfile modes and some people are using them standalone as adobe plugin.
It’s spreading the market.

I am agreeing that the PL Suite elite : pl, filmpack elite and viewpoint the best solution is to hold there ground against others.
Selling it as full package would be better.

2 Likes

Ian, the idea here is that with core Photolab and five or six modules (FilmPack, ViewPoint, HDR, Panorama, SilverEfex built-into Photolab, DAM) users could choose their feature set and would not be so upset about the price (of course an all modules package would have to enjoy competitive pricing).

Or to simplify:

  • FilmPack/SilverEfex integrated into Photolab (the B&W is SilverEfex is another calibre than what can be done in Photolab)
  • ViewPoint/Panorama
  • HDR
  • DAM

The main reason I need FilmPack is Fine Contrast (that annoys me, DxO shouldn’t be burying core features like that in only vaguely related modules). It turns out I don’t need ViewPoint at all as whatever corrections I need are covered by distortion and rotation. I’d buy ViewPoint for the panorama tools though.

I’d buy HDR as I’d love to have good HDR tools in Photolab. SilverEfex is so good that if that level of B&W processing could be built directly into Photolab, I’d be thrilled and probably buy that too. I would almost certainly not buy the DAM package as I don’t want Photolab DAM features: I’m happy with my external tools.

I’m not crazy about the Basic and Elite distinction. All the Basic versions basically are terrible and a poor introduction to DxO. I could see core Photolab with Elite features kept at a reasonable price. Once photographers come for the core image processing and noise reduction, they will want to do everything in Photolab. DxO could then sell those modules to photographers who do want that additional functionality in Photolab.

I agree that DxO has to be careful with pricing. Their software is not inexpensive. Forcing users to pay for extra Photolab licenses in order to upgrade Nik was also fairly underhanded. DxO policy oscillates between trying to attract new users with customer-friendly policies and trying to squeeze additional revenue out of existing users with customer-hostile policies. As commercial behaviour it’s schizophrenic.

3 Likes

I think it’s possible. I don’t know why not using the used algorithmes but then based on 12 or 14 bit R,G or B pixels.
Color space or WB doesn’t play a role in this project.
Just playing with the idea, a new raw file should be created first so the routines that uses raw info can be used.

George

i looked at the multi image composite function in Silkypix and that’s stacking.
multi exposure several types, moving object removal, stroboscopic motion, DoF stack.
Al this is based on number of raw images which i can process individually and then merge/blended together as “stack” not stitch.
And output 16bit tiff or DNG
So they using the “tiff-like mapping” of the pixels and i think they preview the rawfile’s as a tiff on screen so it can do a lay over of the pixelraster and delete the one’s who gets replaced by pixels of file 2,3,4,5,6,7.
This means there is a “form of recognition software” which decides what replaces what.
This “replacing software” can also be used for stitching.
Still i think the stacking and stitching is done with processing to tiff and then stack one by one wile processing the next one to tiff before blending. So all raw development is done before the stack or stitch
(i can see the processing jumps and 7 images do 7 jumps on the time line)

yes it does as in if the WB isn’t equal in the selected files you get strange effects in the blend/stitch.

i have some freetime in about few weeks i will try(if the time and weather will be good) to create some “stack likable” images.
And start playing with viewpoints and wb and distortion/perspective before stitching just to see what happens. just for the fun of it.

As is feature placement i think the Viewpoint module would be the perfect place for the function stitching and stacking. (like the present one as Viewpoint and the one with stacking and stitching “Viewpoint elite”)

A raw file doesn’t have something as a WB. When shooting JPG it’s advisable to use a fixed WB. The same for the color profile: a raw file is allways in the native color profile of the sensor.

George

A raw file has a wb reference from the camera. Which is used by the rawdeveloper to create a preview. And then you can change it to what ever you desire. So technical rawfiledata doesn’t have wb but as user we speak of WB set by camera.:blush:
That’s why @Joanna uses a fixed camera wb 5500k or so. To equal the preview.

Color profile, same, the attached colorprofile srgb or argb selected in camera or the profiles selected in developer. The raw colorprofile is always the same. Sensors capability’s. Maybe iso variable because of variable DR.

A raw file has an input color gamut. That’s the same for all pictures taken with that camera.
The wb as we are used to talk about is a correction on the output color temperature of the picture after demosaicing the raw file.
Since the input color gamut doesn’t change in a raw file, it makes it suitable for stitching.
Still, just playing with the idea, nothing more.

When using a fixed wb like Joanna it would be better to use 5400K when using PL. I don’t think you will see the difference.

George

Sorry George but there is a difference, when you come from a film photography background.