Open William1921 opened 4 years ago
Hi William, thanks for your feedback! It would be interesting to have a look at some of the problematic images (both old and new ones) to better understand what issues you're seeing.
Hi there, sincere apologies for an incredibly delayed reply, I didn't know I had gotten a response, spam folders, we love and loath them.
I will post some images, mainly older ones as they are the medium most worked. Most are neatives digitalised in a single image, incase stiched photos are a problem. The modern images may shed some light on the issues, some parts of the buildings aren't perfectly straight being plaster finished by hand. The house in the back of the older one is the most interest at the moment as plans have already been sourced, at the time these dwellings were only 30 years old and had not suffered subsidence or movement.
Another suggestion would possibly be being able to pin the Box to any of its corners and increase lines on it in an attempt to stighten the match.
Cheers William
I just had a quick look and the images you posted seem a bit tricky to work with when using two vanishing points. However, all three images have a discernable third vanishing point and I got (what appears to be) decent results using all three vanishing points. Have you tried that?
Hello ! Thanks for this great and free tool. I second the request for a scroll zoom if you ever feel like working on fSpy again. It would help to visually check the lines instead of the dots only. Sometimes i'd like to "blindly" place a point and control with the line but it's actually not easy. Cheers
I've used F-spy for a while with mixed results, I'm recreating buildings demolished years ago, most being built before the 1940. Some have few plans available and some don't contain any detailed drawings so using images is a necessity to accurately recreate them. Some images are over a century old and knowing the focal length/camera used to take them is impossible, all images are negatives. Some of my own images from a modern DSLR aren't aligning properly either.
I've got a few suggestions to improve usability and accuracy.
Scrolling into the images instead of the magnifying bubble, large resolution images are almost impossible to see.
Ability to pin the Grid/Box corners to a place on the image.
Alternatively being able to use the box to lineup the perspective completely, Looking back at corner pinning, it's easy to align the X & Z axies. Y is often a problem when working on a single facade not a corner building.
Being able to resize the box would be essential.
I'm happy to donate to get these improvements done. My project is strongly supported project from my local Museums, National & Council Archives.
Cheers William.