1. Adjust your normal smoothing angle. If set to low, itll smooth out your surface. To test if thats it, import both formats and check the wireframe/vert position. If the same, its just an appearance issue.

  2. Thanks for the reply. You were correct. I didn't realize that when exporting, whether to have the vertex normals saved or not is an option.

  3. Close the aperture to increase your depth of field.

  4. Thank you for the reply. I guess I need to add more to my post.

  5. Post a screenshot of the shader graph. Take note if the texture nodes mention srgb, linear or something else

  6. I pulled up the shading workspace to look at the nodes and posted the FBX settings above. The GLB and OBJ models had the same nodes minus the Normal/Map node. For GLB and OBJ, the values for roughness, metallic and specular were the same between the two, but different from FBX. I adjusted those values to match those of the FBX model and the results are also shown above.

  7. Metashape versions 1.7.3 and 1.7.4 and maybe some others have a texel density bug, It usually looks like a perfect square of blurriness in your texture map, not like that what you showed in your example.

  8. Thanks for mentioning the bug. For a while, I was questioning my image quality, thinking the issue was with me. You mentioning this and the results I've gotten with RC and 3DF Zephyr suggest otherwise. And as of Metashape 1.8, it doesn't seem that Agisoft has addressed this issue.

  9. Are you reducing the amount of cameras involved with texture generation first? There's a tool up there somewhere in the menu to remove cameras on a certain degree of overlap, for pre-texture optimisation. Sorry I don't remember exactly what it's called or where it is, I'm not at the computer now.

  10. Thanks for the suggestion. I know the tool you're referring to, although I had never used it. I tried that and generating defocus masks to see if either would make a difference. In both cases, there appeared to be an improvement in the texture, but still not to the quality of what was generated by RC.

  11. I was literally going to type a variation of this question! I’ve done quite a few tests and have had excellent results with straight JPEG capture from camera, but when I shoot RAW, I get worse results. Color accuracy is important so I usually would apply dev settings in bridge then batch process them to jpegs. I’ve tried simply going RAW to JPEG- no actual dev settings- and still get bad results, to the point where I’ve simplified our workflow to only use jpeg straight from camera. I’m interested in what others say on this subject. I do make sure to shoot custom white balance in camera and if need be, I can adjust the png in photoshop later, but all my RAW workflow exp is not helping in this regard; it’s really throwing me for a loop.

  12. I only shoot in RAW. Yes, shooting straight to JPG does simplify the workflow, but you'll always have better control in pre-processing and you can always go back and reprocess the RAW files again and again, if needed, as RAW image files can't be altered.

  13. tru that. if youre not batch processing youre giving yourself a headache.

  14. Thank you for the replies. Definitely batch process. I use Photoshop for adjustments and masking. I was quite hesitant for some time to get the subscription, but Affinity doesn't have the the same capabilities, especially if I'm batch masking; being able to automatically select the subject in a photo is nice (although not perfect).

Leave a Reply

Your email address will not be published. Required fields are marked *

Author: admin