Hello guys, i have been playing with the software and the results are good but i have a big problem.
I have made professional studio photos of a chair, white background, model centered, rotating in static, good lightning, perfect quality, but the software doesnt recognise all the photos and the details are missing.
I have 73 photos from 3 different angles. See this image, some photos are not recognise by the software so the point cloud is poor.
What can i do? theres some settings i can change to make it better?
Yeah i had a lot of components, i tried to realign various times so that may be the case. But i have a bigger problem now.
I have another set of photos of another object, this time, and i now i really dont know why, the alignment doesnt work at all, i mean, no photo is being align, please take a lot a it if you can, the photos are raw so they have the max quality possible:
It doesnt show anything. I cant go into alignment - analyze - inspect because its greyed out, no image has been allign and they profesional pics. Why this can happen?
could you post some sample images in full resolution?
I guess by “angles” you mean three circles at different hights? That leaves 24 images for each circle which is at the lower limit for ideal objects, which I think an office chair is not neccessarily.
Have you changed some settings? Is by any chance “align components only” set to true?
sorry, you’re right - it’S “merge”. But since it’s “false” it’s not that.
I opened the image you sent in RawTherapee. I have to say I’m not surprised - the chair is basically unicolor. Big no-no for photogrammetry. No features whatsoever. You need to process the images so that the texture on the chair is actually visible in the images - I tried a bit and it might be possible.
So i have to convert that cr2 files into “tif” for example and then it will work?
I have also an archive called .xmp with the cr2 files, i have read about it but i dont know how to use it inside realitycapture. Maybe that can help with the cr2 files?
Hi Gotz, i have converted my images into tiff, then cropped a bit to center the chair and now the results are better, but i have a question. Why there are some photos with a number, and others only with a green tick?
But DO NOT CROP the images!!! It throws RC off balance because it changes the image geometry.
About your other questions: I am afraid they are quite basic and most of them can easily be answered with the in-app help of RC and the forum. Just keep practicing and the experience will come!
Hi Gotz, okey i will not crop the image, just convert to tiff.
I have check the help menu, but i dont recognise that green ticks and numbers on it, can you give me the name of the part that covers that topic in help menu and i will search in it?
btw, JPG with 95% compression is also fine for most purposes and it saves a lot of disk space.
I’m not sure where exactly it’s described but I would suggest that you just read a bit through the basics like GUI and so forth. RC has a somewhat steep learning curve…
hello, it is not necessary to get completely all images aligned at all, it is not unusual,
in general, you can try using control points, add more quality photos or reshoot the scene and realign it again to improve the alignment, however, some objects will still require a special workflow so that they can be captured using photogrammetry and RealityCapture
the number is a count of matches found in the images, the green dot signifies that the image has been aligned - registered in a component, it is depicted in this Help tutorial: Get Familiar with the Application Basics / View Types
