In/Out Device

eelhc

Hall of Fame
Timbuk2 Gearbox makes an awesome case for the In/Out. Hardshell case with plenty of room. I like Timbuk2 products.

20171211_171520-S.jpg
20171211_171606-S.jpg
20171211_171624-S.jpg
 

ElMagoElGato

Semi-Pro
I tested the service practice briefly today with the latest firmware.

The court calibration was fairly good in an indoor court. The court was found within a few attempts. I guess it helps if you move the camera just a little bit when the court is never found.

It was only 10 to 20 shots but calls were missed or ignored more than 50%. It'll be next major improvement if missing calls happen this much. There wasn't any false positive calls. Calls made were pretty much spot on.
 

Tennis_Monk

Hall of Fame
I tested the service practice briefly today with the latest firmware.

The court calibration was fairly good in an indoor court. The court was found within a few attempts. I guess it helps if you move the camera just a little bit when the court is never found.

It was only 10 to 20 shots but calls were missed or ignored more than 50%. It'll be next major improvement if missing calls happen this much. There wasn't any false positive calls. Calls made were pretty much spot on.
Did you use a tripod or netpost?
 

newpball

Legend
Wow, more censorship, it seems any hint of criticism is removed here.

All it will do is to hurt the poor suckers who buy this product based on the obviously biased reporting here!

Tennis-warehouse loses credibility here.

:(
 

Tennis_Monk

Hall of Fame
Wow, more censorship, it seems any hint of criticism is removed here.

All it will do is to hurt the poor suckers who buy this product based on the obviously biased reporting here!

Tennis-warehouse loses credibility here.

:(
All I can say, having been a member for so many years, is that the moderators are very fair.

Stick to their policies and all is well.

Criticism is also allowed. As long as it within bounds of policies.
 

newpball

Legend
All I can say, having been a member for so many years, is that the moderators are very fair.

Stick to their policies and all is well.
Criticism is also allowed. As long as it within bounds of policies.
Indeed it is generally fair, but certainly not in this topic!

You would not know because you are one of the cheerleaders and I bet none of your postings here are deleted!

:(
 

Tennis_Monk

Hall of Fame
Indeed it is generally fair, but certainly not in this topic!

You would not know because you are one of the cheerleaders and I bet none of your postings here are deleted!

:(

Personally I never had any posts deleted, ever.

Obviously I wouldn't know, but I fail to see any specific reason beyond policies for posts deletion.

Tw doesn't sell or market this product.
 

ElMagoElGato

Semi-Pro
I used in doubles matches today.

The court calibration took more than 20 attempts. It was all cloudy at an outdoor court. I made a lot of calibration reports. But I wasn't fair because I always did it when 4 players were playing on court. A strange thing is that once the court gets found, the calibration becomes a lot faster even the device is turned off.

The main purpose today was to check false positive calls. It was significantly improved even though it was double matches with the singles practice mode. I didn't check all the calls but there were only a few junk calls in 2 hours. Points were tracked pretty well overall. By the way, everybody should experience the point replay with video on the cloud application. It makes video review a lot easier.

So for me, the singles practice mode is good enough to have spent $200. I'd charge at least $50 dollars if someone ask me to turn 1 hour practice video into a shot placement database.

I wonder how Gregoire makes a match mode. It won't be easy if he intends to track scores. I wish he make it simpler like only detecting the service side and making line calls.
 

amalguy

New User
I used in doubles matches today.

The court calibration took more than 20 attempts. It was all cloudy at an outdoor court. I made a lot of calibration reports. But I wasn't fair because I always did it when 4 players were playing on court. A strange thing is that once the court gets found, the calibration becomes a lot faster even the device is turned off.

The main purpose today was to check false positive calls. It was significantly improved even though it was double matches with the singles practice mode. I didn't check all the calls but there were only a few junk calls in 2 hours. Points were tracked pretty well overall. By the way, everybody should experience the point replay with video on the cloud application. It makes video review a lot easier.

So for me, the singles practice mode is good enough to have spent $200. I'd charge at least $50 dollars if someone ask me to turn 1 hour practice video into a shot placement database.

I wonder how Gregoire makes a match mode. It won't be easy if he intends to track scores. I wish he make it simpler like only detecting the service side and making line calls.
Did you succeed to record a 2 hours "Line Call + Record" session ? No overheating with bad behavior like not stable video or increased missed calls. No freeze of device screen freeze, no empty battery.
Which size and format kind of sd card did you use ? External battery plugged ?
FYI, I succeded to record to 5Gb video (1h13m) on my sdcard formated with exFat. With Fat32, the device is programmed to stop a recording at 3.5GB and restart it (because of FAT32 limitation).
Thanks.
 

ElMagoElGato

Semi-Pro
Did you succeed to record a 2 hours "Line Call + Record" session ? No overheating with bad behavior like not stable video or increased missed calls. No freeze of device screen freeze, no empty battery.
Which size and format kind of sd card did you use ? External battery plugged ?
FYI, I succeded to record to 5Gb video (1h13m) on my sdcard formated with exFat. With Fat32, the device is programmed to stop a recording at 3.5GB and restart it (because of FAT32 limitation).
Thanks.

I didn't record the 2 hour session straight. Someone accidentally touched the device a couple of times. I felt calls were missed too often sometimes. I restarted recording each time, even rebooted the device once, and ended up with 7 sessions.

None of the videos were broken. There was no fatal mal-function like a freeze. It was chilly outside. Obviously there wasn't any overheating.

One battery fully charged made it all the way. I didn't use an external battery. I've got a second battery instead. If a battery outage broke a video, I could test the fixing feature. That was my plan.

My SD card is 64GB and formatted by exFat. I can use the rec only for a few hours if I'm merely testing the file size. But I don't think I'll do it until I see a problem.

Overall, my concerns are narrowed down to frequency of missing calls at the moment.
 
Last edited:

amalguy

New User
I didn't record the 2 hour session straight. Someone accidentally touched the device a couple of times. I felt calls were missed too often sometimes. I restarted recording each time, even rebooted the device once, and ended up with 7 sessions.

None of the videos were broken. There was no fatal mal-function like a freeze. It was chilly outside. Obviously there wasn't any overheating.

One battery fully charged made it all the way. I didn't use an external battery. I've got a second battery instead. If a battery outage broke a video, I could test the fixing feature. That was my plan.

My SD card is 64GB and formatted by exFat. I can use the rec only for a few hours if I'm merely testing the file size. But I don't think I'll do it until I see a problem.

Overall, my concerns are narrowed down to frequency of missing calls at the moment.
Ok thanks. I succeded to use the broken video repair feature. But is seems that you need, on your sd-card, a free space equal to the final video size.
 

jwocky

Rookie
resolution(360/720/1080), frames per sec, write speed of the In/Out, SD card class the device can handle, etc. will be some of the determining variables

Have any of the details on this been published by the inventor?
 

mrmoonball

New User
Did you succeed to record a 2 hours "Line Call + Record" session ? No overheating with bad behavior like not stable video or increased missed calls. No freeze of device screen freeze, no empty battery.
Which size and format kind of sd card did you use ? External battery plugged ?
FYI, I succeded to record to 5Gb video (1h13m) on my sdcard formated with exFat. With Fat32, the device is programmed to stop a recording at 3.5GB and restart it (because of FAT32 limitation).
Thanks.

Mine still crashes and starts freezing, shutting itself off randomly after like 30-40 min. Never got anything longer than 30 min.
 

Greg G

Professional
On the latest firmware 12-05-2017.

The good:
Court detection seems better- the court that it had previously failed to detect was finally detected, still with some difficulty. Took about 10 tries.

The bad:
Line calls were more inconsistent, but I suspect it may be the poor lighting conditions of the court. The mapping of ball placement was pretty good though.
 

beanie

New User
Firmware 2017-12-05.

Court detection seems better - It recognized the lines at first try. However, I went to a court that was used for kiddy tennis and there were extra lines on the court (from the peeled off tape). In/Out consistently detected the kiddy lines on one side of the court where the baseline and sidelines were much shorter. LOL. I did "Find Court" multiple times and for a few times, it detected the correct lines but green lines were poorly drawn. I would try again and it would return to detecting the kiddy court lines.

Non-ball items - I did not experience these calls at all - yay - it did not mistake my white shoes, navy blue sweat pants, shadow on lines for a tennis ball.

Has anyone thought about doing some precision testing? It is a challenge. I notice In/Out will not call a vertically dropped ball, e.g. when I start a practice rally. It "knows" to call balls that travel over the net. Thus, it's a challenge to test for precision. I am not good enough to consistently toss or hit balls to a specific spot on the other side of the net.
 

ElMagoElGato

Semi-Pro
Firmware 2017-12-05.

Court detection seems better - It recognized the lines at first try. However, I went to a court that was used for kiddy tennis and there were extra lines on the court (from the peeled off tape). In/Out consistently detected the kiddy lines on one side of the court where the baseline and sidelines were much shorter. LOL. I did "Find Court" multiple times and for a few times, it detected the correct lines but green lines were poorly drawn. I would try again and it would return to detecting the kiddy court lines.

Non-ball items - I did not experience these calls at all - yay - it did not mistake my white shoes, navy blue sweat pants, shadow on lines for a tennis ball.

Has anyone thought about doing some precision testing? It is a challenge. I notice In/Out will not call a vertically dropped ball, e.g. when I start a practice rally. It "knows" to call balls that travel over the net. Thus, it's a challenge to test for precision. I am not good enough to consistently toss or hit balls to a specific spot on the other side of the net.

Maybe you can just play with it and check close calls on the device when it happens. You can report them, too, this way.
 

amalguy

New User
Firmware 2017-12-05.

Court detection seems better - It recognized the lines at first try. However, I went to a court that was used for kiddy tennis and there were extra lines on the court (from the peeled off tape). In/Out consistently detected the kiddy lines on one side of the court where the baseline and sidelines were much shorter. LOL. I did "Find Court" multiple times and for a few times, it detected the correct lines but green lines were poorly drawn. I would try again and it would return to detecting the kiddy court lines.
For the kiddy tennis lines recognition, it's anormal because the soft should have a pattern of a standard tennis court. Did you submit a ticket for it ?
 

beanie

New User
Yes, I submitted a support ticket. Thank you.

From "Troubleshooting on Court" web page:

Some courts more problematic than others
No tennis court is perfect but some conditions can make the detection much harder.
Additional lines (for kids courts or for basket ball playground) are not really supported at the moment. It will be supported in a future upgrade but it's not a priority at the moment.
 

ElMagoElGato

Semi-Pro
I updated the device to the beta 2018-01-01 but I don't have a chance to use it until one week later. Has anyone tested the ball speed and the net clearance height? It'll be a great aspect to add to the statistics. I'm also interested in the export data of the stats. I wish the data have time stamp on each call.

Personally, the improvement on the video audio was a good thing. I use a Mac that is basically based on QuickTime Player. Now I can hear the sound without switching to Windows.
 

amalguy

New User
I updated the device to the beta 2018-01-01 but I don't have a chance to use it until one week later. Has anyone tested the ball speed and the net clearance height? It'll be a great aspect to add to the statistics. I'm also interested in the export data of the stats. I wish the data have time stamp on each call.
I tried today the new firmware.
The ball speed is correclty shown in the line call screen but not the height over the net if I didn't miss it.
And displayed in stats like this :
zilpc9.jpg

The corresponding csv file is the following :
Code:
Firmware version: 2018-01-01
Date session: 20180104-152820
Type session: 0
Unit: Imperial
Format:
inout    x    y    speed    netheight
1    -33.39    5.80    41    3.7
1    36.75    -10.11    55    9.2
0    -25.29    16.67    85    37.6
The first column tell if the ball is in (1) or out (0)
The x,y is the coordinate of the ball's bounce in feet with the following coordinates system :
2qtcmms.jpg

The speed is in Mph. For the net height,I don't guess the unit.
And the timestamp is missing.

And there also a zoom feature in the bounce review on the device.
 

ElMagoElGato

Semi-Pro
I tried today the new firmware.
The ball speed is correclty shown in the line call screen but not the height over the net if I didn't miss it.
And displayed in stats like this :
zilpc9.jpg

The corresponding csv file is the following :
Code:
Firmware version: 2018-01-01
Date session: 20180104-152820
Type session: 0
Unit: Imperial
Format:
inout    x    y    speed    netheight
1    -33.39    5.80    41    3.7
1    36.75    -10.11    55    9.2
0    -25.29    16.67    85    37.6
The first column tell if the ball is in (1) or out (0)
The x,y is the coordinate of the ball's bounce in feet with the following coordinates system :
2qtcmms.jpg

The speed is in Mph. For the net height,I don't guess the unit.
And the timestamp is missing.

And there also a zoom feature in the bounce review on the device.

You clocked 136km/h? That's really impressive. The net height is a sort of mystery. Hmm.

The review mode of the companion app obviously uses time stamps, so they're stored somewhere. I wish Gregoire add them in the future.
 

eelhc

Hall of Fame
I tried today the new firmware.
The ball speed is correclty shown in the line call screen but not the height over the net if I didn't miss it.
And displayed in stats like this :
zilpc9.jpg

The corresponding csv file is the following :
Code:
Firmware version: 2018-01-01
Date session: 20180104-152820
Type session: 0
Unit: Imperial
Format:
inout    x    y    speed    netheight
1    -33.39    5.80    41    3.7
1    36.75    -10.11    55    9.2
0    -25.29    16.67    85    37.6
The first column tell if the ball is in (1) or out (0)
The x,y is the coordinate of the ball's bounce in feet with the following coordinates system :
2qtcmms.jpg

The speed is in Mph. For the net height,I don't guess the unit.
And the timestamp is missing.

And there also a zoom feature in the bounce review on the device.

Haven't had a chance to play lately week so I've not had the chance to test the new firmware but this stuff is getting me excited. I've maintained since the very beginning the line calling feature is less interesting than the stats tracking. Already, this is much more useful information than the tennis sensors.
 

amalguy

New User
Is it like showing data with the video replay on the device? That’ll be really nice.
Today, these new stats (ball speed and height over the net) are not displayed in the replay in the wep application.
And the unit for the height over the net is feet. I tested it yesterday and it had been quite accurate under 1 meter. For this, I put a panel against the net with 1 meter above the net.
I made shot near the panel and remember where the ball cross it and measure the height.
But not sure it's accurate all the time and especially above my panel (not very easy to check)
 

amalguy

New User
Yes just like in this screenshot you can see serve speed 74 MPH https://i.imgur.com/pPRWqKB.png

My device is still crashing. My gut feeling says its overheating.
Yes the ball speed is displayed on Line call screen in practice simple and serve training modes. And you can swtich between ball speed and height over the net by clicking on the actual displayed stat.
FYI I got these checks about the height over the net stat :
In/Out device -> My measure
1,1m -> 0,9m
0,8m -> 0,8m
3,3m -> ??
0,8m -> 0,6m
0,2m -> 0,2m
 

mrmoonball

New User
Yes the ball speed is displayed on Line call screen in practice simple and serve training modes. And you can swtich between ball speed and height over the net by clicking on the actual displayed stat.
FYI I got these checks about the height over the net stat :
In/Out device -> My measure
1,1m -> 0,9m
0,8m -> 0,8m
3,3m -> ??
0,8m -> 0,6m
0,2m -> 0,2m

I had to enable it in the hidden menu. Will try again tomorrow. Are your devices also crashing randomly, freezing up and shutting itself off after a while?
 
I hopefully get mine back soon. He said I inserted the battery upside down. The device should prevent this, if true.

The root cause is the battery pulled out of the device when I unplugged the USB charger. That is another design flaw.
 
All I can say, having been a member for so many years, is that the moderators are very fair.

Stick to their policies and all is well.

Criticism is also allowed. As long as it within bounds of policies.
That is a troll. A troll that posted personal attacks and insults and rarely anything of true value to potential buyers. Best to ignore.
 

mrmoonball

New User
Mine crashes sometimes, too. My impression is a memory leak because it happens even on a cold outside court.

Theres a hidden menu that displays the temp. When you just start the device and its IDLE its around 45 C for me. It ALWAYS works good for 10 first minutes after that it starts to crash, touch screen doesnt work or becomes sluggish, it jumps front and back from menus by itself e.g. If I let it cool down for 10 min and then start again. It works for a while. It's overheating.
 

ElMagoElGato

Semi-Pro
Theres a hidden menu that displays the temp. When you just start the device and its IDLE its around 45 C for me. It ALWAYS works good for 10 first minutes after that it starts to crash, touch screen doesnt work or becomes sluggish, it jumps front and back from menus by itself e.g. If I let it cool down for 10 min and then start again. It works for a while. It's overheating.

Yes, sounds like overheating. Unfortunately I won’t be able to test it because my courts are basically cold now.
 

amalguy

New User
Yes, sounds like overheating. Unfortunately I won’t be able to test it because my courts are basically cold now.
As said in Gregoire support, this temperature is internal to the processor :
You can read the temperature of the processor in ° Celsius. It's normal if this temperature becomes very high (80°C to 100°C) as it measures the internal temperature of the processor - this is not the temperature of the device casing.
In my test, with the device only on with no record/line call, the temperature grows like in the following diagram :
ve9ilh.jpg


Then for me no crash during 120 minutes with no record/line call.
When recording only (no line call), I've got these results :
  • Initial t° : 35°c
  • Record only lanched
  • Screen freezed after 16 min 27 sec.
  • I've stoped the video after 25 min 48 sec.
  • Video seemed to be recorded
  • T° : 86 °c
  • When I tried to replay the video, I've got a black screen and the device was freezed.
  • Hard reboot
  • T° 79°c
  • Record only lanched again
  • Screen freezed again after 16 min 41 sec.
These tests are done, at home, with the previous firmware 2017-12-05.
I will do these tests agin with the new firmawre.

mrmoonball, your crash are only with the new firmware ? during recording ?
 

ElMagoElGato

Semi-Pro
It's a binary file, then we can't do anything with it

Only Grégoire can use it, I guess. Firmware writers sometimes are very good at finding problems through debug files. Though I don’t put priority on it myself at the moment, it has to be worked out somewhere on the road.
 

CarterStC

New User
I just flashed the January firmware, and my touch screen locks up when I try to change the time. Then, I force rebooted it and it won't respond to pressing Record Only. Also, if I try to detect the courts, everything turns black. What a mess!
 
Top