Flir E4 Thermal imaging camera teardown

[dazler]

How does the hacked e4 compare to the therm-app as far as reading temp of walls, ceiling and to see the studs.
As in the video of mjlorton it was easy to see the studs and hand heat signature with the therm-app but not so much with the regular not hacked e4.

[encryptededdy]

Not the same environment but if you look in the Therm-App thread you'll find the TA has better sensitivity at detecting smaller temperature difference than the E4/8, mainly due to the lens on the E4/8 being terrible.

[Bud]

You can see the studs with the doctored E4.

[-jeffB]

Handprints are easy with the hacked E4, too. This is from two seconds of contact on a heavy hardwood table top. (Fingers were cold, as you can see from the shot of my hand.  )

[böjebba]

I just want to say thanks to everyone that have made it possible to make a useable thermal camera from the Flir E4. Without the possibility to increase the resolution, I had not even considered to buy a Flir.
I bought the Flir E4 in January that came with Firmware 2.8. Followed the instructions and the links in the first post to downgrade to 2.3 and install the hack and the extra tools.
Used Virtualbox with Win 7 32 bit all way.
My only issue was the USB-ports in Virtualbox.  It worked a lot better with the Virtualbox extension pack 

Once again, THANKS!

[targit]

Hi, I have a FLiR E30 which I imported from a U.S. Supplier and its the full 30 fps version.
I would like to know if anyone else has done the E30 to E60 "conversion" on a 30 fps unit without it being then restricted to 9 fps. I can't find an explicit reference to fps in the config files ( I could have missed it ).
I don't know if the frame rate is hardware or software limited, but I would assume (don't like doing that) if it is a security requirement that it would be set by the hardware. Does anyone know this for sure?

Thanks

[Chanc3]

I have fiddled about with both the older and newer generation E60s, both 30Hz models and have not reduced it at all. As long as you have a backup, you should be fine.

Sent from my ONE A2003 using Tapatalk

[targit]

Thanks for that!
I'll give it a go with a bit more confidence now 
And thanks to the all guys who worked this out in the first place.

[targit]

Hi, upgrade done and all working, even seem to have the wifi module and lamp ( can just see the lamp through the plastic cover + have connected to my phone over wi-fi ) I am thinking there is probably a camera under there too, if anyone has managed to remove that cover, could they share how they got it out, I could not find a teardown on the E30 - E60 series, and I think I would have to bugger up the bottom rubber on the handle to remove the screen/keypad/battery cover part ( last 2 screws are under a glued on section of rubber ) which I may do if its absolutely necessary to get the visible camera cover off.
Thanks again

[Chanc3]

Could you confirm what firmware and model number the e30 is? I think the cover around the digital camera and lamp is just glued on...

Sent from my ONE A2003 using Tapatalk

[targit]

Hi, Part number is 49001-1601 model is E30 0.10 "software" (firmware I think) version is 2.23.14
Thanks

[dazler]

thank for the photos, it does look way better than the regular e4.

this is the video I was talking about

at around min 26 you can see the therm-app finding stud on the wall
at 28:30 the flir e4
at 30:00 the handprint.

I wonder how would a hacked e4 performed in that same scenario.

As the e4 goes for around 1k and the therm-app for 1.6k I might not need that much detail.
but don't know if I will later regret not spending that extra $ and get the therm-app.

[man_of_stars]

From what I can see, enhanced E4 outperforms the therm-app significantly!

[targit]

Hi, just spent a few hours stripping the whole camera down in order to find out how that lens was held in, and you are right, it's just bloody glued on
I had convinced myself there was a screw holding it in place, and didn't want to trash the cover by snapping it, so I dismantled the whole thing ( have lots of photos if anyone else wants to see whats inside ) to discover there were no screws under the camera/LED/WiFi boards. So with glue being the only possibility, I pried the cover off with a jewelers screwdriver and that was all I needed to do in the first place ( just like Chanc3 said ).

[Chanc3]

Sorry! I should have checked ours, but both of them were out of the office on jobs. Glad you worked it out though. Will make future repairs nice and easy Haha

Sent from my ONE A2003 using Tapatalk

[targit]

Hi, No worries Chanc3, it was an interesting way to kill a few hours anyway 
If anyone else is doing the E30 upgrade, here are some pics of whats under the cover and one way to make the holes and refit it. after drilling the holes out I filed with a fine needle file and the results look OK without getting  too fussy over the finish inside the holes.

[Bud]

This is a sample of a handprint viewed with the doctored E4

26.avi

[encryptededdy]

From what I can see, enhanced E4 outperforms the therm-app significantly!

mjlortons video seems to have a ~3 degree temp delta, whereas the image by Bud has a ~8 degree temp delta.

Same with the handprint, jeffB's image has a 8 degree (celcius) delta whereas mjlortons video shows a 2 degree delta.

Here's a comparison using very small temperature deltas.

Therm-App:

Noise and resolution test

FLIR Hacked E4, Hacked E40, FLIR One (G1):

a sensitivity test
Flir E40 (hacked 320x240), Germanium (?) lens


Flir E4 (hacked 320x240),  Chalcogenide Glass lens


Flir One => Lens FOL 2 (80x60)

As you can see at low temperature deltas the Therm-App does much better than the hacked E4 and is near the hacked E40.

Still not the same target temperatures though (only the same delta), so still not a super accurate comparison.

Please note that I used the Therm-App and the included lens (with f/1.1 aperture), whereas mjlorton used the Therm-App TH and it's included lens (with f/1.4 aperture, much darker).

[Bill W]

Those seem to show up that that E4 hack has only sorted out the resolution and not the 'added' noise, while the E40 hack may have also sorted out the noise.
Standard specs for NETD (which are at f/1.0, at the sensor, under sensor test conditions of course, so not like it is used in the camera with an f 1.3 lens !):

E4 = 150mK
E8 = 60mK
E40 = 70mK
E60 = 50mK

Just being a Chalcogenide lens is not the issue, f number could be or it is image 'dishancement'.  Indeed the E4 noise is very blotchy suggesting a low resolution itself which contributes to what is for a 320 a rather horrid image. 

Bill

[Fraser]

As Bill W has stated, the noise generator destroys the E4 performance and that was FLIR's intention for that budget model.

I manually edited my noise generator setting in the configuration file of my FW 1.19.8 E4 camera and the noise was reduced dramatically. I think I set the injected noise to 5mk in the belief that some noise was possibly needed in the image processing stages. IIRC the E60 also has its noise generator set to 5mk when shipped. I have to upgrade my second E4 soon so will try a setting of zero for the noise generator.

Fraser

[bookaboo]

5mK on the E60? Can you advise where I can check that figure?

On the E60 I use for work I think the noise is set to 35mK as I can see this line in the conf.cfg file:
.caps.config.image.targetNoise.targetNoiseMk int32 35

It's been my intention to modify that for a while, it's on the long "todo" list. In some respects the stock E60 looks noisier than the modified E4.

[Fraser]

Sorry this is way back somewhere in this thread when we were looking at the differences between the low end and high end units. I may even be mistaken, but I know there was much discussion about whether a noise generator was needed for video processing algorithms and it was decided that a low setting of 5mk was appropriate. Others set the noise generator to zero without adverse effect. I have been meaning to do the same.

Fraser

[Fraser]

Ok I just used google to search this thread for 'noise generator' and all the info is there on the topic. I was in error, the E60 does not have its noise generator set to 5. It is 35 as you say. I think we were all a bit surprised that the noise generator was active on the most expensive models, hence the concern about turning it off completely.

Just enter this search string into Google to read more on the topic......

https://www.eevblog.com/forum/thermal-imaging/flir-e4-thermal-imaging-camera-teardown/  noise generator

[encryptededdy]

I wonder how high in the range you have to go to get no noise generator... T460? T460sc?

[encryptededdy]

Indeed the E4 noise is very blotchy suggesting a low resolution itself which contributes to what is for a 320 a rather horrid image. 

IMO what's happening is that the noise is being generated THEN noise reduction is applied to the generated noise. Due to the high amount of noise the noise reduction is working overtime thereby generating the blotchy image.