New Flir products

[Fraser]

I have received a response from Bill on the frame rate issue on some Android platforms.

When developing the Android variant of the FLIR ONE, FLIR used the Samsung S5 as the development platform. They worked to ensure that the frame rate was similar to that of the FLIR ONE GII on the iOS platform.

The fact that there are many very different Android platforms has proved challenging for the developers as their code does not perform the same on all of them. Bill has stated that they are very aware of the frame rate issue on some platforms and continue to work on it.

Bill helped me out with the supply of some professional software when FLIR Customer Services could not. He is a very friendly and helpful gentleman.

Aurora

[Bud]

Can anyone identify the IR protective window material ?

[Fraser]

Silicon, the same as for the lens used in the LEPTON.

I have not worked with Silicon lenses and windows before but they appear to work OK. How good they are compared to Germanium or GASIR, I do not know.

Aurora

[Spike101]

I got mine yesterday, but the micro-usb connector on my Nexus 5 is backwards, so i can only take thelfies... I actually got a micro-usb extension cable, but for some reason it does not work (the Flir One app does not recognize the device if it's plugged in via the cable). However if I use the micro-usb extension cable together with my charger cable and use it to connect the phone to the pc, i can even transfer data, so the data lines seem to be ok. What else could be wrong. Is there something i need to consider when buying a cable?

[marshallh]

I got mine yesterday, but the micro-usb connector on my Nexus 5 is backwards, so i can only take thelfies... I actually got a micro-usb extension cable, but for some reason it does not work (the Flir One app does not recognize the device if it's plugged in via the cable). However if I use the micro-usb extension cable together with my charger cable and use it to connect the phone to the pc, i can even transfer data, so the data lines seem to be ok. What else could be wrong. Is there something i need to consider when buying a cable?

Yes, it must be an OTG-compliant extension cable.
In such cables, the ID pin is shorted to GND to let the phone know that it should be a host instead of a device. But most cables out there leave it unconnected internally.
You can carefully insert a small strand of wire across the contacts and get it working.

[Redshift1340]

Here's a short video taken with my F1 G2 for iOS if anyone is interested in seeing the frame rate that the iOS version is getting and wish to compare with their android's frame rate.

https://www.dropbox.com/s/gpoa6b766efneeg/video%20jul%2029%2C%206%2019%2057%20pm.mov?dl=0

[Spike101]

Yes, it must be an OTG-compliant extension cable.
In such cables, the ID pin is shorted to GND to let the phone know that it should be a host instead of a device. But most cables out there leave it unconnected internally.
You can carefully insert a small strand of wire across the contacts and get it working.

Thanks, I'll try that. But I don't understand how this is necessary, because when I plug it in directly, without any cable, then the pins are not shorted either (or are they shorted internally in the flir one? but then a straight cable should work as well?)

[tomas123]

Here's a short video taken with my F1 G2 for iOS if anyone is interested in seeing the frame rate that the iOS version is getting and wish to compare with their android's frame rate.

https://www.dropbox.com/s/gpoa6b766efneeg/video%20jul%2029%2C%206%2019%2057%20pm.mov?dl=0

thanks
I see more dropped frames than with Flir One G1 in this youtube video:
(the resolution is horrible)
start from 01:00
 

[encryptededdy]

Thanks, I'll try that. But I don't understand how this is necessary, because when I plug it in directly, without any cable, then the pins are not shorted either (or are they shorted internally in the flir one? but then a straight cable should work as well?)

Because the cable leaves the ID pin unconnected. Standard USB only has 4pins, MicroUSB has 5 (5th pin is ID). Inside the FLIR One ID is shorted to ground, but if your extension cable doesn't connect the ID pin to the FLIR One, then the phone can't detect that it's shorted.

[Spike101]

Because the cable leaves the ID pin unconnected. Standard USB only has 4pins, MicroUSB has 5 (5th pin is ID). Inside the FLIR One ID is shorted to ground, but if your extension cable doesn't connect the ID pin to the FLIR One, then the phone can't detect that it's shorted.

Ehm yeah, pretty obvious now thanks

[schlafli]

Just received my Flir one for Android and thought I'd share a few observations I haven't seen mentioned on the forum.

I ordered on the 10th of August and it arrived today (10th September).

I have tried it on a YotaPhone 2 (snapdragon 800) and a Galaxy S5 and both work without issue. So for anyone worried about performance, I'd say any 2 year old flagship phone should be able to run it fine at full FPS (9hz).

There is also support for android wear but the delay is about 2-3 seconds, so it's really just a gimmick. Speaking of:

Themalception  !




Regarding the resolution, I'm pretty sure it's 160x120 thermal resolution. I took a picture of some nichrome wire and played with the ranges in Flir Tools to bring out the details:



Dividing the number of pixels/step by the vertical resolution gives 160px. See post #320 for update. Looks like it might only be 128x96. https://www.eevblog.com/forum/testgear/new-flir-products/msg752153/#msg752153


Finally, sometimes there is a gradient and/or banding (both images are of a wood surface with even temperature):





In this example there is a delta of 1.5C from the centre to the edge. It does seem to only happen when first turning the camera on and disappears after 20-30 seconds and a few calibrations. I guess it's just the lepton sensor heating up.

If I find anything else, I'll post an update 

p.s., The white squares in the images is me bluring out the gps coordinates that flir tools automatically adds...

[schlafli]

Found one more thing:

Whilst the official temp range is -20 to 120C, the camera will go from at least -34 to 150C.

The Android app will simply report any value above 120C as ">120C", but if you view the picture in Flir Tools, that range goes up to 150.

Picture of the sky (it was cloudy, so I don't know the minimum temp yet):


Picture of the hob:


Flir Tools adds a little warning triangle next to any measurement above 120c and will top out at 150.2C.


 

[encryptededdy]

I wonder what the accuracy is like below the official "-20C". We know the accuracy on the FLIR E4 is pretty bad a low temperatures.

[tomas123]

Regarding the resolution, I'm pretty sure it's 160x120 thermal resolution. I took a picture of some nichrome wire and played with the ranges in Flir Tools to bring out the details:



Dividing the number of pixels/step by the vertical resolution gives 160px.

You can't post the original image because your embedded gps coordinates.
But can you post the RAW values from your nichrome wire?
Thanks

Code: [Select]

exiftool -b -RawThermalImage FLIR0080.jpg > t1.pngsee https://www.eevblog.com/forum/testgear/flir-e4-thermal-imaging-camera-teardown/msg348398/#msg348398

[schlafli]

But can you post the RAW values from your nichrome wire?

Here you go (I think t9 is probably the best):

[schlafli]

I just had a look at t9.png after doing the byteswap/auto-level from the other post.

It looks like the sensor is only 128*96 

There are 19 points over a 47 pixel range giving a physical pixel -> image pixel ratio of 2.47.

Dividing the image resolution by 2.5 gives 128x96.

[mct75]

I just had a look at t9.png after doing the byteswap/auto-level from the other post.

It looks like the sensor is only 128*96 

There are 19 points over a 47 pixel range giving a physical pixel -> image pixel ratio of 2.47.

Dividing the image resolution by 2.5 gives 128x96.

Someone here prodded the crap out of a FLIR CS rep and eventually got them to admit it was 160x120.  If that's not true it will be a mess.

[tomas123]

I just had a look at t9.png after doing the byteswap/auto-level from the other post.

It looks like the sensor is only 128*96 

I used your original 16 bit image with a gamma adjustment to avoid overexposed areas.
some Imagemagick steps
see here for details:
https://www.eevblog.com/forum/testgear/flir-one-thermal-imaging-camera-teardown-and-hacks/msg551882/#msg551882

Code: [Select]

>convert  -define png:swap-bytes=on t9.png -auto-level -scale x1280 -gamma 0.6 -depth 8 t9swap.png
>convert -size 8x8 xc:none -stroke gray -strokewidth 0  -draw "line 0,0 0,7"  -draw "line 0,0 7,0"  8x8.png
>convert -size 960x1280 tile:8x8.png 960x1280.png
>convert t9swap.png 960x1280.png -compose overlay -composite overlay.png


there are four fault locations:
1. your image is unsharpen (wrong distance)
2. the Flir One lens is a bottle bottom
3. the wire is to hot (optical crosstalk)
4. the Flir One G2 resolution of 160x120 is a fake

one square = one sensor pixel of Flir One


in this crop you can guess the real resolution of 160x120


not bad, but compare with a Flir Exx  :

one square = one sensor pixel of Flir Exx

the result is really sharp (thanks to good optics of the Flir Exx)

[NathanFowler]

Can MSX not be turned off?  It seems like more of an annoyance than anything else, because I would use a Flir One for macro-type imaging than anything else.

Using the FLIR Tools application on iOS, not FLIR ONE, on the G1 I am able to turn the MSX blending off by selecting the circular icon to the bottom left and selecting IR only with no MSX; I've not noticed any gain issues.  Is this not an option on the G2?

[cynfab]

Nathan,
Can you see the GPS location with your IOS Flir Tools?
On the Android Flir Tools Mobile I see my GPS location with the wrong E/W designation ie. it's placing me somewhere in China.
The samples that came with the Flir App show all E Longitude, and I'm wondering if there is a bug in the Flir One Android app, or Flir Tools Mobile.
I read out the exif data with exiftools and I think there may be some new/changed Tags in the Flir One files.

GPS tags from the exif data from my Flir One image with exiftool:

GPS Version ID                  : 2.2.0.0
GPS Latitude Ref                : North
GPS Longitude Ref               : West
GPS Altitude Ref                : Above Sea Level
GPS Map Datum                   : WGS84
GPS Latitude                    : 38 deg 53' xx.yy" N
Camera Software                 : 1.0.0

sample images that came with the Flir One app,  I see the following GPS tags:

GPS Version ID                  : 2.2.0.0
GPS Latitude Ref                : North
GPS Longitude Ref               : East
GPS Altitude Ref                : Above Sea Level
GPS Map Datum                   : WGS84
GPS Latitude                    : 59 deg 27' 19.02" N
GPS Longitude                   : 18 deg 7' 50.04" E
GPS Position                    : 59 deg 27' 19.02" N, 18 deg 7' 50.04" E
Camera Software                 : 1.0.0



   ...ken...

[Fraser]

I was the one who plagued the life out of the FLIR CS Rep.

If you check in the FLIR one true resolution thread you will see the responses.

You will also see that I had to work hard in order to get a straight answer on the resolution question. Eventually I gave the CS Rep no choice except a "160x120 pixel ... Yes or no" answer. She actually covered herself by saying that according to the information she had gleaned, it is 160x120 pixels.

Let us just park that for a moment though.

What does 160x120 pixels actually mean in the real world ?

You can have the highest resolution sensor in the world and still destroy its performance with poor optics and/or lossy image processing. We know from the SEEK camera that the 12um pixel microbolometer is a noisy beast. Why should FLIR's version be any better. Remember FLIR took SEEK Thermal to court over the 12um technology usage. We do not know what FLIR is doing in its noise suppression algorithms so such could affect apparent resolution ?

What I do know is that the LEPTON appears impossibly small to my more conventional technology familiar eyes. Such miniaturisation may produce a tiny affordable core but I have to wonder if it is fair to expect it to perform as well as a more conventional larger core?

I am not defending FLIR in any way but I remain pretty amazed at what SEEK Thermal and FLIR have achieved. The lenses will be far from optimal and we may be witnessing the down side of the miniaturisation and cost cutting efforts. I am still very pleased that FLIR made this higher resolution version of the core though. Even if it does not perform as well as a conventional 160x120 microbolometer, it surely must be better than 80x60........ Or is it ?

I noted a comment on some FLIR blurb that the LEPTON 3, with 4x the pixels, provided imaging "as good as, or better than" the previous LEPTON core.

Such statements do make me wonder what FLIR knows about the new core that we do not.

When more of us have the new FLIR ONE GII, we may be able to assess its performance better.

At the end of the day, it is not the specs that are so important as the real world performance. When I compare the F1G2 with the TESTO and NEC 160x120 cameras we will have a better idea on the new 12um technologies performance.

Aurora

[encryptededdy]

I don't think this is a actual limitation of 12 micron pixels by the way - BAE's MicroIR T640 is a 640x480@60Hz 12um thermal imaging core currently found in some "IR Patrol" handheld products, where they are able to deliver resolution on par with conventional 640x480 sensors.

[NathanFowler]

Nathan,
Can you see the GPS location with your IOS Flir Tools?
On the Android Flir Tools Mobile I see my GPS location with the wrong E/W designation ie. it's placing me somewhere in China.
The samples that came with the Flir App show all E Longitude, and I'm wondering if there is a bug in the Flir One Android app, or Flir Tools Mobile.
I read out the exif data with exiftools and I think there may be some new/changed Tags in the Flir One files.

Ken, thanks so much for the response.  You'll have to forgive me for my lack of experience with the Android Ecosystem.  On iOS I have two FLIR-enabled tools available; "FLIR Tools" and "FLIR ONE"; with FLIR ONE I cannot disable MSX blending.  With FLIR Tools I can disable this so that I am only presented an IR image with no MSX blending.  The below homescreen image shows the respective icons which I assume would be similar on Android.

GPS tags from the exif data from my Flir One image with exiftool:

GPS Version ID                  : 2.2.0.0
GPS Latitude Ref                : North
GPS Longitude Ref               : West
GPS Altitude Ref                : Above Sea Level
GPS Map Datum                   : WGS84
GPS Latitude                    : 38 deg 53' xx.yy" N
Camera Software                 : 1.0.0

sample images that came with the Flir One app,  I see the following GPS tags:

GPS Version ID                  : 2.2.0.0
GPS Latitude Ref                : North
GPS Longitude Ref               : East
GPS Altitude Ref                : Above Sea Level
GPS Map Datum                   : WGS84
GPS Latitude                    : 59 deg 27' 19.02" N
GPS Longitude                   : 18 deg 7' 50.04" E
GPS Position                    : 59 deg 27' 19.02" N, 18 deg 7' 50.04" E
Camera Software                 : 1.0.0

Ken, forgive me if I am telling you something you're already aware of, but on iOS there is a centralized "Location Services" daemon and configuration panel that allows me to selectively permit which applications may and may not talk with AGPS/GPS to determine location.  FLIR Tools does not have such an option to grant permissions whereas FLIR ONE does.  I took a FLIR ONE image without this enabled, and one with it enabled in the "Location Services" master panel as well as with-in the Settings of the FLIR ONE application itself.  'locationd' (Location Services) did not report this setting being used by FLIR ONE through historical audit (it will show applications which have used location services in the last 24 hours)

First image, no location services enabled.  I use jhead but the EXIF output should be similar:

Code: [Select]

$ jhead IMG_0223.JPG
File name    : IMG_0223.JPG
File size    : 271324 bytes
File date    : 2015:09:11 19:46:42
Camera make  : FLIR Systems AB
Camera model : *
Date/Time    : 2015:09:11 19:33:22
Resolution   : 480 x 640
Focal length :  1.6mm
Focus dist.  : 1.00m

Now, with location services enabled:

Code: [Select]

$ jhead IMG_0225.JPG
File name    : IMG_0225.JPG
File size    : 260735 bytes
File date    : 2015:09:11 19:46:45
Camera make  : FLIR Systems AB
Camera model : *
Date/Time    : 2015:09:11 19:35:51
Resolution   : 480 x 640
Focal length :  1.6mm
Focus dist.  : 1.00m

As you can see above the location stamping for EXIF is devoid, which is quite interesting.  I could be an outlier because I have an application firewall enabled as well as a strict pf (pfctl) firewall enabled that could be breaking it because I do not allow these applciations to communicate with the IPv4/IPv6 Internet except FLIR Tools over an RFC 1918 CIDR /32 for FTP.

Hopefully this is helpful and if not, I apologize in advance.

[cynfab]

Thanks Nathan,

It looks like there are a lot of differences between the IOS app and the Andriod app.
Well... Flir has had over a year of Flir One G1 on IOS to fiddle with the app. and there are a lot of things not present in the Android app. as it is relatively new.
I'd like to see some of those features on Android and IMHO the GPS data shoulld be included in the IOS images too.

But my real question is:
Is the problem I'm seeing in Flir Tools Mobile related to the Flir One Android app, or is it a bug in the Flir Tools Mobile app.

I've asked Flir One support and not received any answer but:
"This could be a result of the fact that the FLIR ONE was not designed or tested with FLIR Tools"

How can Flir develop a mobile app and then a brand new shiny Android Thermal Imaging product and not test it???

The second question is about the exif tags in the Flir One file: They seem to be different between the sample image and my image. Are there new Tags? does exiftool properly parse the Flir One tags in Android... how about IOS image files.

The potential privacy issue prevents lots of folks from posting raw Flir One images since they contain serial #'s as well as location data..... me too.
It would be nice to be able to scrub this data, either in the Flir One app, Flir Tools Mobile, or exiftool.

As another point of interest, I was able to get the Flir Tools Mobile app running under Archon: http://archon-runtime.github.io/
which allows Android runtime .apk's to be installed and run under the Chrome Browser.

    ...ken...

[Bud]

In Windows you can right-click on the image, go to Properties and there you can erase many image properties. Should retest with perhaps Exiftool to check what is erased what is not.