Replies: 5 comments 14 replies
|
I don't have the answer to your panorama question, but I imagine the stitching application you're using isn't exporting a raw TIFF file the way something like Maketiff would. Instead, it's probably rendering or otherwise altering the data during export. |
|
I think usually software, unless made specifically for extracting the raw data from a camera raw, will render the exported tiff for display, so changing the rgb data. So nx studio might be worth a shot, but I would not be surprised if it alters the data. Maketiff is made specifically for extracting raw data from camera files, so that is why I recommended that. It's not free, but has a trail mode that never expires (but has some batch limitations). I think that is what negpy does as well when reading raw files, it ignores any profile and what not and just grabs the raw data. |
|
Possibly an even more cumbersome workaround would be to invert both .nef files in NegPy using the exact same settings, then export them and stitch the final result. That's assuming NegPy even lets you do that (i.e., turn off all algorithms that try to evaluate exposure, contrast, etc.). |
|
Ahh okay, thanks for the info on maketiff. I'm not familiar with it but I'll give it a shot. I'm sure someone smarter than me will eventually face the same issue and come up with a better solution. I can always revert to nlp for 120 scans and use negpy for 35mm. I've searched online but haven't really found any guides on negpy as of yet. I understand they're trying to mimic a darkroom flow with the editing process but it's all gibberish to me haha. Even with nlp, it was always a guessing game until I finally understood the sliders. I think I just need to take the time to read the negpy documentation and go step by step to better understand what each slider does. Thanks for taking the time to help with some suggestions on my specific issue |
|
You could always open a feature request. I’m not sure how easy a good panorama stitching feature is to implement. |
|
Thanks. I'm not sure if you follow pushingfilm on YouTube but he just put out a video about pixel shift for scanning film and he mentioned that for something like 6x6, he'll zoom out and scan the whole thing as 1 frame but since it's pixel shift, there's enough resolution to crop and zoom in. So I may take that approach on frames I really care about narrowband scanning and just stick with white light nlp conversions for everything else. I'll look into the feature request though, thanks for the suggestion |
|
can you post files pre and post stitching? i'm wondering what stiching might be doing so that the colors are suddenly off. |
|
Hi. I did my best to include a variety of files here. The individual raw files from my Nikon ZF are img427 and img428. Image Composite Editor is the panorama stitch output from that program. Then I used MakeTiff and NX Studio to export a Tiff of img428 just to compare how the .neg and 2 .tiff files all looked in negpy. Unfortunately the image where the saturation difference is very apparent is an image with people in it and I don't want to share their faces online. I can try to find a different sample if these aren't good enough. |
|
I downloaded the files and noticed a few things during testing. First, the stitched TIFF is only 8-bit. I'm not sure whether that's intentional. Second, the NX View converted tiff it is not just the raw data dumped into a tiff. Third, the TIFF produced by maketiff is tagged as sRGB. I had assumed NegPy would ignore the embedded profile, but it seems to behave differently when the file is tagged as sRGB. The shadows appear to be lifted. Maybe assumes that the file is encoded with srgb trc and tries to encode that back to linear or something? @marcinz606 I thought maketiff would be able to produce a TIFF without tagging a color profile, but that doesn't seem to be the case. So I stripped the profiles using ExifTool. After removing the profiles, the TIFF renders much closer to the original NEF raw file, although it's still not identical. The TIFF is slightly more saturated than the NEF. |
|
I took the two stripped maketiff files and put them them through photomerge in photoshop: https://we.tl/t-fADAden2CFz6YSUa I don't have any experience merging, but the resulting file seems to work quite well with negpy. |
|
Here are nef, maketiff with no profile, maketiff srgb tag, and nx studio run through negpy. 
|
2026-07-21.14-58-38.mp4:) |
|
So there is already a merge feature or did you just add one?:) |
|
I'm adding it |
|
You guys are awesome. I really appreciate the help and discussion. I'm a novice with all this stuff and barely have any free time at the moment. Thank you |
exported result, default conversion, no edits
|
|
stitching merged to main |
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
I am scanning with a Nikon ZF in Raw using the narrowband mode on Jack's Big Scanlight. My workflow has always been to take 2 shots of each 6x6 frame on 120 and stitch together in lightroom then use NLP to convert and process. Lightroom's panorama stitch outputs a dng file which doesn't work with negpy. I haven't been able to find any software that can stitch .nef files and maintain the .nef format. So i've tried a few alternative panorama apps but they all export to tiff. However when I feed that narrowband tiff into negpy it comes out very very saturated. If i import 1 half frame of the 6x6 in .nef format and convert in negpy it looks much more accurate. i'm guess the tiff is locking in some values that are skewing the conversion in negpy. Is anyone else in a similar situation? How are you stitching panoramas and maintaining a raw format that will work with negpy?
All reactions