Q: Touch ID being Erratic
-
Nov 20, 2013 1:56 PMby Joe_Fo,kmpa43 wrote:
Thanks joe. I've not had to use the pass code in about 24 hours so far. This new in box replacement is so far so good. Not gonna get excited though. It has missed a few times but hey. It seems to act different than my last phone though. Will report back with more once I've used it for a few days. This new one was made in November so it's fresh.
Good to hear it's mostly working! Any issues with damp hands?
-
Nov 20, 2013 2:02 PMby Joe_Fo,smilleresq wrote:
Joe_Fo, what is unclear to me is why am I now getting almost 100% success with my right thumb (although having to place it with great precision) and 0% with my left thumb and index fingers, all of which were scanned per your instructions. Maybe I did not effectively scan my other fingers as I did the right thumb,
[snip]
It's too awkard for me using my left hand. My guess is you either registered too far towards the tip or during registration, may have moved too far from the last spot scanned.
-
Nov 20, 2013 2:24 PMby Joe_Fo,George Guerrero wrote:
FWIW, I am of the opinion that it is a software issue, at least the one that impacts me specifically.
I get ~90% success rate. When it fails, it does so NOT with a "try again" but as if it doesn't sense the registered finger and eventually moves to "slide to unlock". HOWEVER, if I put a non-registered finger, it correctly provides the "try again" message. So clearly the hardware is sensing something.
It rejects incorrect prints from a non-registered finger, but struggles to fully accept a registered finger 10% of the time. It's as if there is a third option – accept, reject AND "not really sure".
When in the "not really sure" state, it never eventually accepts, but 100% of the time I can get it to work normally when invoking control center and opening and closing calculator.
Incidentally, and I didn't have the phone long enough to say for sure, I started to have issus after the update that added the slight pause before the "slide to unlock" message appeared. Originally, it appeared immediately. Certainly not diagnosing anything by saying so... just curious if anyone noticed the same thing.
Hi George,
Couple of points:
There are several issues, the most easily fixed is "proper" registration and uage. While it does not affect everyone, indicated by reports some have no problems with TID, for those of us that do, it can be overcome provided the H/W isn't defective. This, IMHO, is unrelated to the S/W problems.
The issue you point out is S/W related, and good catch on the "try again" with a non-registered finger. I never tried that until just now, and I see the same thing. The trick I posted earlier in this thread with the Control Panel however does work however. It appears to "fix" a brain dead TID. I'm seeing it on 7.0.4 too.
I never saw it prior to 7.0.3 which introduced the delayed prompt, but then again, I never had any unlock problems from the launch once I changed how I registered a finger. I started having issues when running about 35 apps in the background trying to break the phone to see what others were seeing. I now get the springboard crashes and the brain dead TID issues with Twitter the last active app and a locked screen. That will teach me
There is another thread on ths issue, and there's a registered developer there who reported it to Apple. Hopefully to be fixed soon...
-
Nov 20, 2013 2:27 PMby Joe_Fo,paulfromhere wrote:
George Guerrero wrote:
Goto Settings>Control Center and hit the toggle for Access on Lock Screen
[snip]
I used to be able to do it on 7.0.3, but seem to have lost the control center on the password screen on 7.0.4. I thought it was a bug fix.
It still works for me on 7.0.4
-
Nov 20, 2013 2:47 PMby George Guerrero,Joe_Fo wrote:
I never saw it prior to 7.0.3 which introduced the delayed prompt, but then again, I never had any unlock problems from the launch once I changed how I registered a finger. I started having issues when running about 35 apps in the background trying to break the phone to see what others were seeing. I now get the springboard crashes and the brain dead TID issues with Twitter the last active app and a locked screen. That will teach me
There is another thread on ths issue, and there's a registered developer there who reported it to Apple. Hopefully to be fixed soon...
Hi Joe,
Thanks for your reply. Had not read about the Twitter-specific cause/effect until your note. I just gave it a try and sure enough... brain dead TID. For the benefit of others, just to be clear, I had Twitter open and allowed the iPhone to self-lock after the user-selected 1 minute interval. Simply sleeping the phone using the top button and Twitter as last app does not result in TID purgatory.
Will report back if any other specific use case results in a TID failure. FWIW, I do appreciate your efforts to help debug despite not seeing the issue initially. I'm sure I am not alone in that sentiment.
-
Nov 20, 2013 9:38 PMby BlazeApple,I've had my iphone 5s for about a month now. My sensor has worked flawlessly until a week ago where it would work 1/5 times for a few hours. Things went back to normal after that but I just wanted to mention to you guys who are having problems I registered the same finger twice and find that it increases my recognizability accuracy a good bit I would say my sensor works 9/10 times. Give it a shot and let me know!
-
Nov 21, 2013 5:34 AMby paulfromhere,BlazeApple wrote:
but I just wanted to mention to you guys who are having problems I registered the same finger twice and find that it increases my recognizability accuracy a good bit I would say my sensor works 9/10 times. Give it a shot and let me know!
On my phones they stop working completely within a day, so having extra scans does no good .. still have to just replace them every day.
-
Nov 21, 2013 5:59 AMby smilleresq,paulfromhere wrote:
BlazeApple wrote:
but I just wanted to mention to you guys who are having problems I registered the same finger twice and find that it increases my recognizability accuracy a good bit I would say my sensor works 9/10 times. Give it a shot and let me know!
On my phones they stop working completely within a day, so having extra scans does no good .. still have to just replace them every day.
I tried scanning the same finger twice and that ultimately failed just like a single scan when I was going through my series of TID failures. Unless you scan your finger(s) per Joe_Fo's method, TID will not work properly (unless there is some other software or hardware related issue).
-
Nov 21, 2013 1:31 PMby Dreanmachine1,So a few more days of use on our two newly replaced iPhone 5s 'DN' manufactured units (we have had them almost 10 days). My baby (I know, sick) still has about 98-99 percent success. Most days it is flawless (I registered once initially, with three fingers (no duplicates), and have never registered again). Several days back, I did have a few strings during the day of missing once, but then going right in on the second attempt. Today, I got bold, washed my hands, partially dried them, and yielded pretty damp fingers (they were basically on the wet side). Presto, all three fingers went in right away, in perfect fashion. I was kind of shocked, to be honest.
As far as the wife's unit, she has about a 75 percent success rate. Usually, she misses once (at most twice) and then she is in. She no longer has any complete failures to get her to the point of having to enter her security code. My right thumb registration on her phone has been 100% successful (nada failure). This leads me to believe that she has some constraint still with either her unique fingers, or how she registers. In addition, the registration of her right thumb on my unit has had fairly good success, but not flawless like mine. Again, something in her DNA, possibly. I don't see any hardware issues playing culprit in our units.
I am still looking forward to the eventual software update that will hone this in and allow TID to be more forgiving.
-
Nov 21, 2013 2:08 PMby Joe_Fo,Dreanmachine1 wrote:
[snip]
As far as the wife's unit, she has about a 75 percent success rate. Usually, she misses once (at most twice) and then she is in.
[snip]
Again, something in her DNA, possibly. I don't see any hardware issues playing culprit in our units.
Good to hear it's still going well!
I'm changing my mulyi-step, way too verbose HOWTO, and will add it to my thread, or perhaps start a new one and link to both threads.
Take a close look at her finger(s) and what part she used to register. Looking at fingerprint pics on Google might be a clue as to why some have better results than others. I'm convinced the system need enough unique data to model well.
Looking at mine, there's very little variation at the top of the arches, and I believe that's the reason for my initial failures using the tip, even though I did so flat.
I think I've come up with a good analogy as to why the system gets worse for some people, and hopefully a good guide as to why one has to be methodical with registration, but haven't started writing it yet. Moot point as it really shouldn't be needed, but the question of "why does it get worse over time?" still crops up. Easy answer is GIGO.
