Poll

Has the hackabiliy of the E4 made you buy one :  

Yes, I was already looking at the competition at a similar price, but the hack swung it to E4
274 (27.9%)
Yes, I'd not considered buying a TIC before, but 320x240 resolution at this price justifies it (as either tool or toy!)
444 (45.3%)
Yes, I was going to buy an E5/6/8 class of unit but will now get the E4
49 (5%)
No, but am looking out for a cheap i3 to hack
50 (5.1%)
Not yet, but probably will if now that a closed-box hack becomes is possible
164 (16.7%)

Total Members Voted: 803

Author Topic: Flir E4 Thermal imaging camera teardown  (Read 3769475 times)

0 Members and 9 Guests are viewing this topic.

Online Fraser

  • Super Contributor
  • ***
  • Posts: 13148
  • Country: gb
Re: Flir E4 Thermal imaging camera teardown
« Reply #7000 on: May 22, 2015, 06:46:41 pm »
Indeed you have more noise on your image.

Sadly I cannot offer an explanation for this  :(

Aurora
 
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline OrBy

  • Regular Contributor
  • *
  • Posts: 220
Re: Flir E4 Thermal imaging camera teardown
« Reply #7001 on: May 22, 2015, 06:48:18 pm »
A brief comparison... recently on this Thread, a user on my request kindly provided his hand shot with smallest span possible - 2C. Just now I repeated the same with my arm.

The attached are jpegs (mine is radiometric I believe, FLIR006)

Enlarge and you'll see finely grained pixel noise o mine.

Check your config for the following section and tell us the number that is in place of the #:
.caps.config.image.targetNoise entry
.caps.config.image.targetNoise.enabled bool true
.caps.config.image.targetNoise.targetNoiseMk int32 #
 

Offline Solare

  • Regular Contributor
  • *
  • Posts: 134
  • Country: se
Re: Flir E4 Thermal imaging camera teardown
« Reply #7002 on: May 22, 2015, 06:59:35 pm »
To start, I still have no confidence I do things in the right way... help please if I am wrong,

I use cfccfg.py script to decrypt my conf.cfc file, and then I can find the strings you requested:

.caps.config.image.targetNoise entry
.caps.config.image.targetNoise.enabled bool true
.caps.config.image.targetNoise.targetNoiseMk int32 0

From this I would say, there should be no noise, but there is noise exactly at level ~100 mK
And you know what this noise looks artificial! I ran entropy calculator over the high frequency filter, and it is uniformly distributed!!!!!!! According Physics, should be Gaussian! But Gaussian noise is much more expensive in terms of CPU cycles to create! :)))

The user posted the first palm I believe, operate sub- 2.3.0 FW. Could it be they added 'tiny-piny offset' in dll so 0 is not zero anymore :) Just to keep us busy so to say?

That's why I would like to try different noise settings... but please tell me:

Am I using the proper tool? Since it decrypts with my SUID, it should work but I don't know, if it encrypts well. I am afraid to blindly overwrite the hacked conf.cfc in my device. Because obviously after I make changes, the CRC will be altered...

I just have no experience and need your support. Do I use the proper tool and method?

PS: I was wondering before, why the hack they removed web-server Service applet from the 2.3.0 firmware? Assume you have lost your "zero" of noise, and want to find it. What tool would help you best? Right - the Noise Measure from Service menu!

And now, it is gone... and with the old web-server this particular applet doesn't work too...
« Last Edit: May 22, 2015, 07:09:50 pm by Solare »
 

Offline Solare

  • Regular Contributor
  • *
  • Posts: 134
  • Country: se
Re: Flir E4 Thermal imaging camera teardown
« Reply #7003 on: May 22, 2015, 07:20:00 pm »
... if there was a friendly E8 user to share his conf.cfc, things would become clear at once...
 

Offline Solare

  • Regular Contributor
  • *
  • Posts: 134
  • Country: se
Re: Flir E4 Thermal imaging camera teardown
« Reply #7004 on: May 22, 2015, 08:53:32 pm »
...ahaha,

just discovered the same type of noise on pure Digital Camera View! This by all means cannot be natural noise!
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 6877
  • Country: ca
Re: Flir E4 Thermal imaging camera teardown
« Reply #7005 on: May 22, 2015, 10:28:49 pm »
Turn off MSX , looks you had MSX on when made your palm picture
Facebook-free life and Rigol-free shack.
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 6877
  • Country: ca
Re: Flir E4 Thermal imaging camera teardown
« Reply #7006 on: May 22, 2015, 10:34:21 pm »
Am I using the proper tool?
Yes but it is only for the cfc. You should have installed all other files as part of the usual reso and menu updates.
Facebook-free life and Rigol-free shack.
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 6877
  • Country: ca
Re: Flir E4 Thermal imaging camera teardown
« Reply #7007 on: May 22, 2015, 10:36:20 pm »
Forget about worrying for CRC, it is bypassed.

The first palm pic was mine, made on a doctored 2.3.0
« Last Edit: May 22, 2015, 10:39:07 pm by Bud »
Facebook-free life and Rigol-free shack.
 

Offline Solare

  • Regular Contributor
  • *
  • Posts: 134
  • Country: se
Re: Flir E4 Thermal imaging camera teardown
« Reply #7008 on: May 23, 2015, 12:04:27 am »
hey Bud,

thanks...

I am too tired now so I cannot say for sure, but things went better since I did very stupid action:

- I downloaded cfccfg.py V2 instead of V1.
- I applied the script over the conf.cfc file I found on doubly hacked (res and menu) device of mine
- I got the conf.cfg file with, among others, Target NoiseMk set to exactly 0

now, without even touched this string, I re-enabled Circle section of the conf.cfc (set the values same as for the Box),
re-encrypted it back with the same script (through temporary file ofc),
 
- and uploaded it to the original directory with overwrite existing old file.
- cold reboot at this point with battery pulled off the unit

Now.. why the hack it did worked I don't know... I accepted reality that I have the worst sensor on this Thread and even didn't try to play with Noise level... obviously I was too tired after days of non-sleeping reading this because somebody suffered reading this too before so they mostly keep calm and want me to suffer too :) joking

Well, I didn't alter the NoiseMk value = 0.

But after reboot, all random pixels dancing gone!

Well I still have drifting and aliasing and banding but this is Nature of microbolometers, right?

I still have vertical stripes - but this is a Nature of manufacturing process and poor calibration (poor according our high standards ofc!) :)

But there is large part of noise which wasn't removed or influenced by, this operation.

Clearly I can see now that this is either artificial or lack of EM design, but I see the horizontal stripes running up and down like semi ghosts but not on one place like vertical stripes... you know what it reminds me?

Old analogue TV sets when operating with weak signal when there is humming 50Hz or so from nearby power mains!!!

I have 50 Hz mains here and 60Hz is actual frame rate... shit... couldn't it be just simple VIDEO issue?

Mike ! Can you here me!? Hypothetically, where and how such the humming could enter the circuit? Maybe it need Earth? :) Or it is poor screening or some screw holding wire is weak?

If I only could remove the running horizontal lines, I would be finally happy...

Because now I can see temperature temporal noise is low. But still it is not possible to benefit from this good achievement, because horizontal line interference is strong...

Mike ... what is especially interesting, the same horizontal running line is on MSX and even pure Digital Camera View!!!!
But this is absolutely different and separate device!!!! How two devices could share the same exact glitch?!!!

Maybe guys from Flir decided to introduce second noise generator :) or there is loose video cable inside... Mike what do you say?

Bud - thanks but I didn't use MSX or anything but Infrared on this photo.

Thank you guys!

P.S> The funnies thing is, that the Circles never appeared in the Menu :) Oh, screw them... BUT WHY THIS WORKED?!!! :-DD
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 6877
  • Country: ca
Re: Flir E4 Thermal imaging camera teardown
« Reply #7009 on: May 23, 2015, 12:29:21 am »
Solare,

You doing really well. Welcome to the club
 :)

It worked possibly because you cycled cfc encryption. That may have removed some errors caused by patching.
Facebook-free life and Rigol-free shack.
 

Offline Solare

  • Regular Contributor
  • *
  • Posts: 134
  • Country: se
Re: Flir E4 Thermal imaging camera teardown
« Reply #7010 on: May 23, 2015, 01:46:15 am »
Jee, thanks!

 O0 I am one of the Club!

Hurray... I can safely sleep now... in my dreams, I will have it working...

Really... I 've made some additional tests and compared to this

What I have at the moment is central device, but with 320x240 and menu enabled and without pixelation... exactly. You see, all the picture is not steady still as on the right case. MSX is very noisy too, these blue horizontal lines appear spontaneously and heavy pixelation on dark areas... looks awful overall.

Without MSX it even looks better now, but this 'unsteadyness' really pissing me off...

I think now, once I know how to alter the conf.cfc, I will try to play with NoiseMk setting. I remember there was report that noise decrease as value decrease up to zero but then suddenly at high values like 1100 it again smooth.

Thinking of this I may assume the noise generator has several blocks each responsible for own type of image distortion - pixelization, vertical stripes (why not?), horizontal lines etc., etc. And then, IF there is a bitmask, not a single value, so overall noise is something like this

LSB0-4 - Pixelation + LSB5-8 - Vertical Stripes + MSB - 9-16 - Horizontal Lines whatever,

So the sum may be actually 5 or 45376, and it is not the fact that 5 is better especially if they applied some unknown offset...

Can't hold it any longer... sleeping...

 :o
« Last Edit: May 23, 2015, 01:48:25 am by Solare »
 

Offline bookaboo

  • Frequent Contributor
  • **
  • Posts: 725
  • Country: ie
Re: Flir E4 Thermal imaging camera teardown
« Reply #7011 on: May 23, 2015, 08:23:07 am »
Solare I think you have a faulty camera or you have made an error during the upgrade. I'd uninstall the hack and if the problem remains return your camera.
 

Offline HighVoltage

  • Super Contributor
  • ***
  • Posts: 5453
  • Country: de
Re: Flir E4 Thermal imaging camera teardown
« Reply #7012 on: May 23, 2015, 10:52:13 am »
If I would buy an E4 now, do I need to look for any specifics to be able to use the hack or do I need to buy an older model?
Could someone advise me, which one to buy?
And where in Germany is the best place to buy an E4?
Thanks
There are 3 kinds of people in this world, those who can count and those who can not.
 

Offline Solare

  • Regular Contributor
  • *
  • Posts: 134
  • Country: se
Re: Flir E4 Thermal imaging camera teardown
« Reply #7013 on: May 23, 2015, 11:06:36 am »
Read my experience :)

You may be lucky, and maybe not.
 

Offline HighVoltage

  • Super Contributor
  • ***
  • Posts: 5453
  • Country: de
Re: Flir E4 Thermal imaging camera teardown
« Reply #7014 on: May 23, 2015, 04:25:00 pm »
OK, Thank you,
I will just order one next week and see what FW is installed.
May be I will have to return it.
There are 3 kinds of people in this world, those who can count and those who can not.
 

Offline Solare

  • Regular Contributor
  • *
  • Posts: 134
  • Country: se
Re: Flir E4 Thermal imaging camera teardown
« Reply #7015 on: May 23, 2015, 06:50:36 pm »
Maybe it's a good idea to ask what is the Firmware version before placing your order?

There is nothing special in 2.3.0 what could make it better than 1.19 t ex., but both the menu and the resolution hack were in our hands for a longer while, to less problems and less risk..  for instance, files are encrypted and signed on 2.3.0, unlike 1.19. And also, old Firmware has Noise measurement wen server utility, which among others, were excluded from 2.3.0...

However there is little chance you'll find new 1.19 E4 by now... maybe refurbished would be a better (and cheaper) option?
 

Offline HighVoltage

  • Super Contributor
  • ***
  • Posts: 5453
  • Country: de
Re: Flir E4 Thermal imaging camera teardown
« Reply #7016 on: May 23, 2015, 08:55:31 pm »
I looked around for a while and asked about FW versions.
Either they don't know, or it is FW v 2.3.0

But if I understand the information in this tread correctly, FW v2.30 also can be hacked, it is just a little more difficult, right?
Or are there some models of E4 with FW 2.3.0. that can not be hacked?
Thanks for any advise
There are 3 kinds of people in this world, those who can count and those who can not.
 

Offline Solare

  • Regular Contributor
  • *
  • Posts: 134
  • Country: se
Re: Flir E4 Thermal imaging camera teardown
« Reply #7017 on: May 23, 2015, 11:03:06 pm »
@HiVoiltage:

Yes, it is pretty hackable... just be aware that particular sensor in your E4 may differ a lot, as I learned... if you are lucky, you get sensor from E8, then you'll get clear, nice picture. If you are not lucky, it will be bad sensor, noisy... yes, the hack will increase resolution to 320x240, but it is questionable benefit, unless your sensor is good... this is lottery :) no way to know before you buy. As I said, Flir removed possibility to measure noise in 2.3.0, so you are helpless mostly in fine tuning :(

@Mike and the Fathers Founders:

Do you know anything about ADC reference chip used actually in E4? It may be so they just started to use noisy reference, so hypothetically, if this one is located and replaced to a good one, things may improve considerably?

@All:

After setting Target NoiseMk to 135 and uploading, I had a messgaebox: ResMon.App crashed. I closed the window, disconnected USB and put the device to sleep. Then, removed the battery (my usual sequence).

After reboot, I had screen full of coloured pizels and VERY slow bootup (10 min), the device parsed some XML files and loaded something from factory.d. It looked almost bricked :)

Then, it showed temperatures 220 -270C :) crazy

Another reboot, without removing the battery, showed no problem and the noise was really changed to 135 mK. This artificial noise masked almost raw and column noise, maybe this is the purpose of the procedure for not really good sensors?

 

 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 13148
  • Country: gb
Re: Flir E4 Thermal imaging camera teardown
« Reply #7018 on: May 23, 2015, 11:48:53 pm »
The possibility of second grade nicrobolomters has been looked into.

To date there is no evidence that this is done with the E4. My sources suggest that FLIRs image processing can cope with all manner of microbolometer issues. None of the E4 cameras that were tested failed the functional pixel specification for a FLIR core.

The noise that you describe is far more likely to come from a source other than the microbolometer and its integral ROIC. It is known that interference patterns appear on a HW1.1 E4 when the battery is getting low.IIRC the. Noise is most visible on the visible light camera. One of the DC to DC converters feeding a noise sensitive stage may be the cause of such. If it is a boost converter, life for it gets harder as battery potential drops.

I do not know how easily you can exchange your camera but if I were you  I would return it to stock build and get a refund or exchange unit.

Aurora
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline Solare

  • Regular Contributor
  • *
  • Posts: 134
  • Country: se
Re: Flir E4 Thermal imaging camera teardown
« Reply #7019 on: May 24, 2015, 01:05:11 am »
Aurora, thanks!

Yours are the second voice for return & refund... however, my realities are such that this refund will be indeed, pain in my .... so I would rather do anything to cope with the problem, including soldering :)

Well really, I was thinking about either a better adc reference or power supply :)

Idea! Big tantalum capacitor in parallel with battery would help maybe? :)
« Last Edit: May 24, 2015, 01:23:48 am by Solare »
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 6877
  • Country: ca
Re: Flir E4 Thermal imaging camera teardown
« Reply #7020 on: May 24, 2015, 03:24:22 am »
Solare,

try recording a video in USB mode and see if any of these noise goes away, i.e. caused by the LCD and not by microbolometer. I used VirtualDub but probably there exist many other programs for recording USB video.
Make sure you use a codec that does not add much pixelation/noise.

EDIT: Here is a sample
http://wikisend.com/download/138290/test2.avi
« Last Edit: May 24, 2015, 03:35:35 am by Bud »
Facebook-free life and Rigol-free shack.
 

Offline Solare

  • Regular Contributor
  • *
  • Posts: 134
  • Country: se
Re: Flir E4 Thermal imaging camera teardown
« Reply #7021 on: May 24, 2015, 12:57:27 pm »
@Bud,

thanks, it's very curious...

What VLC version do you use for streaming? I failed to stream with the latest version 2.2.1  :-//
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 6877
  • Country: ca
Re: Flir E4 Thermal imaging camera teardown
« Reply #7022 on: May 24, 2015, 01:04:01 pm »
What is VLC ?
Facebook-free life and Rigol-free shack.
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16272
  • Country: za
Re: Flir E4 Thermal imaging camera teardown
« Reply #7023 on: May 24, 2015, 01:28:03 pm »
www.videolan.org One of the best media streamers/players/converters around. If it is in a digital format then you can almost always be sure VLC will play it. Even works on those horrid RM files that the player always wants to phone home about. Can even fix corrupted files somewhat so they will play.
 

Offline Solare

  • Regular Contributor
  • *
  • Posts: 134
  • Country: se
Re: Flir E4 Thermal imaging camera teardown
« Reply #7024 on: May 24, 2015, 01:35:47 pm »
Guys,

I have so far tried two VLC versions... 2.2.1 (latest) and 2.010 (2014)

Both failed... Bud, can you tell exactly what software and version are you using to stream?

@SeanB

Did you actually try the 2.2.1 version with streaming video from E4? Any success report?

The guide I have found on Flir site, also doesn't work... they also do not tell the exact software and version number... it looks different from both the 2.2.1 and 2.0.10, I couldn't recognize it and interface seems different...

UPDATE:

Managed to stream from VideoUSB device, but still not able to record (lack some M4 encoder?)
It is streaming true DirecShow Filter at 3 fps, quality is bad,  spatial and temporal noise.

Maybe it is because the unit got too warm (it shows 44C on the inside sensor while ambient temperature is over 30C), but I remember clearly in the night when it was 10 degrees below, there wasn't that many spatial pixelization (not removed by calibration proc).

Maybe put it into fridge for a while ? :0

Joking :) DO NOT DO THAT!!! it will condensate water vapor on sensor... :(




« Last Edit: May 24, 2015, 02:54:02 pm by Solare »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf