I went back to the originals for our FR10 inspection examples. A bright stripe called a scratch in the slides is much softer in the source image, and another view of the same faceplate puts it elsewhere. Nobody has linked a physical defect location to these labels. I'd rather stop building the training pack now than teach people this confident nonsense.
The faceplate scratch changes shape in the training slides
HanaBarnes0535 · 2025年10月10日 08:51 UTC
主帖中文翻译暂不可用，以下显示英文原文；回复保留原文。 查看英文原文
17 条回复
Has quality inspected the identified faceplate under its agreed method?
11分Keep the source files beside the altered slide images. Or rather, link the slide images back to them. We lost a useful comparison once because the crop hid which edge was which.
21分Anil, did the crop also change the contrast, or was it purely framing? I have seen both bundled into the innocent phrase made it clearer.
6分Liam, quality has the identified faceplate now. Jamie Brown, our slides had a contrast adjustment as well as cropping. Anil, originals are retained and the pack is paused.
-5分Pausing the pack is sensible, but tell whoever books the training why, otherwise somebody will use the old export because it's the only finished-looking one
18分Jamie Brown, ours was framing in the example I meant. Different problem from Hana's contrast adjustment, although both can get hidden once the picture is in a slide.
13分Use accepted and confirmed rejected samples for the new comparison, including faint defects.
21分Quality accepted the faceplate and identified a separate fine edge scratch on a rejected one. Jamie Allen, training coordinator knows the pack is withdrawn for now. We have a sample list rather than image labels deciding everything.
23分Thanks for checking the physical part. I would try the current light with those samples before ordering another one, though I wouldn't insist on keeping it if the real scratch won't show.
15分Jamie Brown, I'd also want the actual holder in that test; hand-tilting a faceplate until the mark looks nice is a different presentation method
20分Yes, absolutely. I meant a controlled comparison, not somebody finding the flattering angle by eye.
22分Hana, do both original views use the same seat in the holder? The moving stripe could be useful evidence about presentation as well as about the slide edits.
24分They don't. One edge can sit above the locating lip in the temporary holder. Tooling is reviewing that; I can't call the repeat images controlled until the seating is sorted.
11分Retain that poor seat as a check for the revised holder.
15分And let the eventual loader try it, because the designer already knows which edge they're trying not to catch
0分That explains why keeping the originals mattered. Has quality marked the rejected edge in a separate reference view so it doesn't disappear beneath a helpful caption next time?
20分Hana, come back with the revised-holder result when you have it. I'm interested whether the existing light becomes useful once presentation stops wandering.
14分讨论已关闭
此讨论已连续六个月没有活动，现已关闭新回复。 最后活动： .