Why does our FR5 scratch comparison exclude the lip that quality inspects?

NoraAllen0322 · 24 Jan 2026, 00:52 UTC

Closed
NO
NoraAllen0322
I have physically judged accepted and rejected trim plates for this comparison now, but the supplier's proposed image mask removes the bright rolled lip and also removes the faint scuff on our rejected reference; how do I get the reflection problem addressed without quietly shrinking the inspection area?

14 replies

LU
LuisAbbott0086
Replying to NoraAllen0322

Put the required inspected area beside the mask. On one of our comparisons an edge was removed as background even though the drawing still required its finish to be judged. A mask may be a reasonable image-processing choice, but somebody has to account for the actual surface it excludes.

19 points
NO
NoraAllen0322
Replying to LuisAbbott0086

Quality has marked the lip as required inspection area on the drawing, and the scuff is on that lip, so I have returned the masked result with those two things shown together.

9 points
SA
SamAli0246
Replying to NoraAllen0322

Did the supplier know the lip was included?

20 points
TH
ThomasBaker0451
Replying to NoraAllen0322

Also ask what the unmasked image shows. Restoring the region may expose the scuff, or may expose the glare that led them to hide it. Those need different amounts of work.

-2 points
MI
MiaBennett0731
Replying to ThomasBaker0451

Exactly, Thomas. I'd be annoyed about the disappearing area too, but an unmasked white stripe isn't suddenly a useful inspection. Keep that same faint reject in whatever view they propose next.

13 points
NO
NoraAllen0322
Replying to MiaBennett0731

Sam, the original requirement included it, though the demo author had only read the cropped example; Thomas and Mia, the raw view contains the lip but the glare obscures the known scuff, so removing the mask alone does not solve this.

0 points
LU
LuisAbbott0086
Replying to NoraAllen0322

Then ask for an imaging comparison that covers the lip as well as the main face. It could need another view or a different presentation. Keep ordinary accepted finishes in that comparison so the candidate is not chosen solely for making one scuff obvious.

17 points
NO
NoraAllen0322
Replying to LuisAbbott0086

They returned a two-view candidate with the whole required area covered between the views; quality can see the original scuff and the other identified reject, while the accepted examples retain their ordinary finish detail.

14 points
MI
MiaBennett0731
Replying to NoraAllen0322

Were those loaded by the person who will do the job, or by the engineer who knows exactly where to press? The best image isn't necessarily the repeatable one.

22 points
NO
NoraAllen0322
Replying to MiaBennett0731

Trial engineer so far, Mia, and I have asked for repeated loading by both of our operators before we settle on that holder arrangement.

8 points
SA
SamAli0246
Replying to NoraAllen0322

Are the same physical references still available?

24 points
NO
NoraAllen0322
Replying to SamAli0246

Yes, Sam: three accepted plates and two rejects, including the original faint lip scuff, all with quality; the operators repeated the bench comparison and found the intended seat consistently without the engineer correcting it.

19 points
MI
MiaBennett0731
Replying to NoraAllen0322

That sounds more convincing. What does the second view add to the actual job time? I wouldn't want the next quote to recover its attractive number by quietly dropping the lip again.

4 points
NO
NoraAllen0322
Replying to MiaBennett0731

The supplier has revised the time allowance to include both views and the operator loading, and we have retained that candidate for the installed trial; thanks Luis for making the inspected area the first question, because the early discussion was all about how tidy the masked image looked.

15 points

Discussion closed

This discussion is closed to new replies after six months without activity. Last activity: .