Garmin Chrono Software Glitch (Or Am I Missing Something)?

[vh20]

I've been running the Garmin since it first came out, having run in succession Shooting Chrony, Pro Chrono Digital, Magnetospeed V3, and Labradar over the years. The Garmin has been a quantum leap in my opinion. However, I ran into something this week that I just can't make sense of...

I had hit the end of an 8# jug of Retumbo, and opened a new one from a different Lot #. When I went to the range, I ran 10 rounds over the Garmin (300PRC) - the first two rounds were the last from the old lot of Retumbo, the other eight were from the new lot. The first two rounds were right on track with my established AVG MV for the load. However, when I got to the next eight rounds from the new lot, the MV jumped up about 20fps on average. More than I was expecting, but I get it - the new lot is hotter than the old. The ES for ALL 10 was 38, and the SD 12.8. Since the last eight rounds from the new lot were considerably faster than the first two from the old lot, I wanted to find out what the numbers were for just the new lot, so I deleted the first two shots and checked the Session Summary. My ES went way down from 38 to 24. BUT, my SD jumped up from 12.8 to 15.3. The new ES is correct, but I can't see any way the SD could go UP, especially to 15.3 with only a 24 ES. I've never seen such a high SD-to-ES ratio.

I had this happen once before fairly recently, when I deleted a couple rounds from a string that were suspect due to some inconsistencies that had happened during the reloading process. The suspect rounds were marked prior to firing. When I deleted them from the Session, the ES went down and the SD went up. I thought it was a bit strange at the time, but shrugged it off as just the way the velocities distributed themselves. It wasn't anywhere as dramatic of a shift as this last incident was, though. This time I was sure there was a problem in the way the unit re-calculated the SD.

In the past, if I ever deleted any isolated extreme high or low shots (out of curiosity), the SD would come down along with the ES. However, there was a fairly recent software update pushed, and both of the incidents related above have occurred since that update was installed.

Anyone have any insight?

 

[6.5SH]

Does Garmin have the ability to re-flash to an older firmware like LabRadar?

 

[vh20]

Does Garmin have the ability to re-flash to an older firmware like LabRadar?

I don't know. I kind of doubt it, but I didn't know the LR had that ability either. I never had a need to do it, and I ran the LR from the time it first came out to the time the Garmin came out. I don't recall ever having seen the SD go UP when the ES came DOWN when deleting certain shots from the LR, though.

 

[Rob01]

How are these software updates "pushed"? All I do with my Garmin is charge it and use it. lol

 

[cyranis]

Did you math this out by hand to check, before we jump to ghost in the machine firmware land?

 

[vh20]

Did you math this out by hand to check, before we jump to ghost in the machine firmware land?

No, not yet. I do have all the MVs written down, including the ones that were deleted. But just looking at it on the surface, I can't see any way it will come to that SD. As I understand the formula, it would not be possible. When I get a chance I will find my notes and give it a try.

How are these software updates "pushed"? All I do with my Garmin is charge it and use it. lol

Through the phone app. If the app is active on your phone, it will pick up the update and push it to the chrono. The chrono will give you the option to update or cancel. When they first came out, this "push" could happen in the middle of a string while you were trying to record data. The unit would put itself in pause mode and stop recording shots until you answered the question (Update, or Cancel?). If you were behind the rifle concentrating on your shooting and not the chrono, you would keep shooting and not realize it was no longer recording shots and you'd miss getting the data on those. Very frustrating. I reported the issue to Garmin, as I believe several others did, suggesting they not allow it to accept Update queries when the unit was in an active session. I "think" they have fixed this, but I'm not certain. I know that I've never had that happen again since the next update came through. The last couple of update queries have all happened when the unit was not in an active session.
Unrelated to that, one update I have noticed is that the unit now shows the charge level (battery icon with a sliding bar to indicate charge) when you are in an active session. In the beginning, you would see the charge level when you first turned it on, but it went away once you activated a session. I like that much better.

 

[Rob01]

I don’t use a phone app so guess mine will stay as is.

 

[vh20]

I don’t use a phone app so guess mine will stay as is.

I don't either, but I do have the app on my phone, and if I ever look at it and forget to close it out completely instead of just swiping it away, it will stay active and will connect to the unit in the background whenever you turn the unit on. You won't even realize it. The unwanted update queries were avoided by making sure the app was completely closed, but I'm hoping that won't be necessary anymore in case you forget.

 

[Baron23]

No, not yet. I do have all the MVs written down, including the ones that were deleted. But just looking at it on the surface, I can't see any way it will come to that SD. As I understand the formula, it would not be possible. When I get a chance I will find my notes and give it a try.

Through the phone app. If the app is active on your phone, it will pick up the update and push it to the chrono. The chrono will give you the option to update or cancel. When they first came out, this "push" could happen in the middle of a string while you were trying to record data. The unit would put itself in pause mode and stop recording shots until you answered the question (Update, or Cancel?). If you were behind the rifle concentrating on your shooting and not the chrono, you would keep shooting and not realize it was no longer recording shots and you'd miss getting the data on those. Very frustrating. I reported the issue to Garmin, as I believe several others did, suggesting they not allow it to accept Update queries when the unit was in an active session. I "think" they have fixed this, but I'm not certain. I know that I've never had that happen again since the next update came through. The last couple of update queries have all happened when the unit was not in an active session.
Unrelated to that, one update I have noticed is that the unit now shows the charge level (battery icon with a sliding bar to indicate charge) when you are in an active session. In the beginning, you would see the charge level when you first turned it on, but it went away once you activated a session. I like that much better.

Yeah...I have no idea what's up with you SD and but that sure doesn't sound right.

I do manipulation of the chrono data off line in Excel via the csv file you can send out rather than deleting shots in the session

 

[Rob01]

I don't either, but I do have the app on my phone, and if I ever look at it and forget to close it out completely instead of just swiping it away, it will stay active and will connect to the unit in the background whenever you turn the unit on. You won't even realize it. The unwanted update queries were avoided by making sure the app was completely closed, but I'm hoping that won't be necessary anymore in case you forget.

Why not just delete it off your phone?

 

[rustyinbend]

I always pull the raw data off the device, and load it into a spreadsheet. Then if I need to discard shots or manipulate the data, I use the Excel functions to calculate averages and new SD's. I've just found that to always result in numbers that made sense.

 

[vh20]

Did you math this out by hand to check, before we jump to ghost in the machine firmware land?

Ok, real quick calculation using an online calculator (haven't had time to do it by hand):
Using Sample SD, those last 8 rounds come to SD=8.8
Using Population SD, then SD=8.3
This is more like what you would expect just looking at the raw numbers, and the SD went down as expected when shrinking the ES by over 30%, not UP, like it did on the unit.

 

[vh20]

Why not just delete it off your phone?

Well, I guess because that's how the unit gets its updates, and they can be useful, just not in the middle of a session. But as I said, I think they may have fixed that.

 

[cyranis]

The ES for ALL 10 was 38, and the SD 12.8. Since the last eight rounds from the new lot were considerably faster than the first two from the old lot, I wanted to find out what the numbers were for just the new lot, so I deleted the first two shots and checked the Session Summary. My ES went way down from 38 to 24. BUT, my SD jumped up from 12.8 to 15.3.

ES is highest to lowest, this would go down if there was a large variance between your old powder and new, which you stated, so 100% normal.
Your concern seems to mainly be SD.
The standard deviation is a measure of variability in data, and is inversely proportional to the square root of the sample size. This means that as the sample size decreases, the standard deviation increases. Also normal

 

[vh20]

ES is highest to lowest, this would go down if there was a large variance between your old powder and new, which you stated, so 100% normal.
Your concern seems to mainly be SD.
The standard deviation is a measure of variability in data, and is inversely proportional to the square root of the sample size. This means that as the sample size decreases, the standard deviation increases. Also normal

Did you see the actual numbers I posted above when re-calculated?

 

[cyranis]

I did, and I just explained how the changes you described are possible. The only way to know is to do the actual calculations by hand, with all of the original data. Which is why I asked if you did so.

 

[vh20]

So you don’t think the calculator is correct? Maybe so, but I’ve never seen an SD that is more than half the difference between the lowest and highest. I’m not a mathematician, but I do shoot three times a week over this unit and have never seen anything like that before. We shrunk the ES by about 35%, and the population by only 20%. I’m just not sure you can actually get an SD of 15.3 from an ES of 24 no matter how you distribute the velocities, even in a worst-case scenario, assuming we are using realistic velocity values on the order that would come from a typical rifle (around 2900).
For reference, here are the actual numbers:

1) 2896
2) 2900
3) 2929
4) 2910
5) 2934
6) 2919
7) 2916
8) 2933
9) 2921
10) 2931

The first two shots (in red text) were the ones from the old lot and the Avg MV on that load was around 2898. The remaining eight (shots 3-10) were from the new lot, and clearly averaging above 2898, so I deleted the first two to find out what the numbers on just those eight from the new lot looked like.
I can't see any way you can torture those eight numbers to come up with an SD of 15.3, and I don't doubt the validity of the online calculator I used, but maybe it's wrong. I will see what the manual calculation comes up with (after I go back and try to figure it out - I haven't calculated an SD by hand since Statistics in college, and that was decades ago!). But it will be tonight at the earliest before I have a chance to do so.

 

[B y r o n]

For those that don't want to run the phone app, you can download Garmin Express to a laptop or desktop computer. When you connect your Garmin Xero to the computer Express will recognize it and check for updates. After doing this once, Garmin Express will notify you if an update is available when you launch it.

 

[AllenOne1]

 

[spife7980]

Should have been in the OP

Any one else have a data set that they could record, delete some, and then verify if you get the same goofy result of an increased SD with a tightened population? Im a labradar guy myself so I cant.

Id be curious if you deleted some more if the problem would correct itself or worsen.

 

[Baron23]

So you don’t think the calculator is correct? Maybe so, but I’ve never seen an SD that is more than half the difference between the lowest and highest. I’m not a mathematician, but I do shoot three times a week over this unit and have never seen anything like that before. We shrunk the ES by about 35%, and the population by only 20%. I’m just not sure you can actually get an SD of 15.3 from an ES of 24 no matter how you distribute the velocities, even in a worst-case scenario, assuming we are using realistic velocity values on the order that would come from a typical rifle (around 2900).
For reference, here are the actual numbers:

1) 2896
2) 2900
3) 2929
4) 2910
5) 2934
6) 2919
7) 2916
8) 2933
9) 2921
10) 2931

The first two shots (in red text) were the ones from the old lot and the Avg MV on that load was around 2898. The remaining eight (shots 3-10) were from the new lot, and clearly averaging above 2898, so I deleted the first two to find out what the numbers on just those eight from the new lot looked like.
I can't see any way you can torture those eight numbers to come up with an SD of 15.3, and I don't doubt the validity of the online calculator I used, but maybe it's wrong. I will see what the manual calculation comes up with (after I go back and try to figure it out - I haven't calculated an SD by hand since Statistics in college, and that was decades ago!). But it will be tonight at the earliest before I have a chance to do so.

Yep, in Excel using SD sample I get a 13.5 SD for total then 8.8 SD with the two slow shots from prior jug deleted from the data set.

Garmin just fucked it up.

 

[vh20]

Hey, @AllenOne1 and @spife7980 - thanks guys! What an awesome thing to do. Much appreciated, and thanks for confirming my suspicions.

 

[vh20]

@Baron23, Thanks to you too. You posted while I was thanking the others.

 

[vh20]

Should have been in the OP
View attachment 8544039


Any one else have a data set that they could record, delete some, and then verify if you get the same goofy result of an increased SD with a tightened population? Im a labradar guy myself so I cant.

Id be curious if you deleted some more if the problem would correct itself or worsen.

I did it again today. I shot 10 rounds of an exact-same load, then 5 more that were processed slightly differently. I kept them all on the same string and then went back and deleted the last 5, and compared it to the numbers I got when it was just 10.

First 10:

After 15:

Then I deleted the last 5 to get back to the original first 10 again. Watch the SD compared to the same 10 the first time:

Also curious, the AVG didn’t change by even a single tenth of a fps between 10 shots and 15. I haven’t done the math by hand yet. It may work out like that, but that’s very uncommon.

 

[Baron23]

I did it again today. I shot 10 rounds of an exact-same load, then 5 more that were processed slightly differently. I kept them all on the same string and then went back and deleted the last 5, and compared it to the numbers I got when it was just 10.

First 10:


View attachment 8546469

After 15:


View attachment 8546470

Then deleted the last 5 to get back to the first 10 again. (Watch the SD):

You're links don't work and is there supposed to be another link for "then deleted the last 5....etc"?

 

[vh20]

You're links don't work and is there supposed to be another link for "then deleted the last 5....etc"?

There was a problem either with my phone or the site. It was there, then it wasn't...I tried to fix it but it wouldn't respond. I'll try again.

 

[vh20]

Ok, I think it’s fixed now.

 

[spife7980]

No one can prove or disprove any math without the actual data.

 

[AllenOne1]

The variation shown in the pictures above is very small. If you are fretting and losing sleep over this I would suggest you sell the Garmin and move on.

 

[vh20]

The variation shown in the pictures above is very small. If you are fretting and losing sleep over this I would suggest you sell the Garmin and move on.

No sir, that's not the point. The Garmin is the best unit out there, but it has some glitches. If we identify them and report them, others will know what to watch for. If you are fretting over a thread you don't even care about, then I would suggest you don't have to read the thread. If you read the OP, the first time this happened the the unit gave an SD of 15.3 when it should have been 8.3. That is NOT small. It's nearly double, and for me 8.3 is ok for a working load, but 15.3 is not. This one is much smaller (inconsequential to me), but @spife7980 asked for people to post other examples and this one is the next time I saw it happen.

So the POINT of this thread is this:
Currently the Garmin does not allow you to come back to a session after you leave it. If you have some reason to run some rounds over it that, for whatever reason don't really belong with the rest of the data (buddy wants to shoot a few while your gun is cooling, just to get an idea of his MV, etc), the work-around for not being able to start a new session without permanently closing the current one is to record the extra shots in the open session, and then just delete those shots and get the numbers on only the ones that belong in the session. However I've seen on several occasions that this will cause an error in the final calculation (sometimes small, sometimes huge) vs. if you hadn't included, then deleted those extra shots. So if I know this, and report it to other users, they can avoid that by realizing it may create a false answer. Also, Garmin may not know it either, and might be able to fix it on a future update.

 

[vh20]

No one can prove or disprove any math without the actual data.

Yeah, but both can't be "right" (even though they're pretty close to each other). It's the exact same data in both calculations. This one is tiny, but it still gave a different answer with the same 10 shots. I'm just posting it because you asked for more cases.