Author Topic: calibration time: is it faster in Exx series? (compared to Ex series)  (Read 400 times)

0 Members and 1 Guest are viewing this topic.

Offline calel

  • Regular Contributor
  • *
  • Posts: 87
  • Country: ch
so with the Ex calibration time is about 1.5 seconds where image is frozen during that  time (eg. when quickly switching from one object to another far hotter or colder object)

and I was told that this whole calibration thing is necessary in IR cams

so is it faster in the Exx series?
 

Offline tisher

  • Newbie
  • Posts: 3
  • Country: pl
Re: calibration time: is it faster in Exx series? (compared to Ex series)
« Reply #1 on: October 18, 2020, 09:40:09 pm »
Rec on E30(E60 mod), you can see frozen, also take ~1.5sec
« Last Edit: October 18, 2020, 09:41:47 pm by tisher »
 

Offline calel

  • Regular Contributor
  • *
  • Posts: 87
  • Country: ch
Re: calibration time: is it faster in Exx series? (compared to Ex series)
« Reply #2 on: October 18, 2020, 09:53:35 pm »
ok thx good to know

looks closer to 1s so maybe a wee bit faster

still a long time though. is it theoretically possible to have a "real-time calibration" IR cam? (ie. no noticeable freeze)
 

Online Vipitis

  • Frequent Contributor
  • **
  • Posts: 706
  • Country: de
  • aspiring thermal photography enthusiast
Re: calibration time: is it faster in Exx series? (compared to Ex series)
« Reply #3 on: October 18, 2020, 10:13:34 pm »
the "calibration" here is the FFC, which is a shutter action followed by some maths.
shutter
There are thermal cameras that don't use any shutter, for example BST, but also microbolometers like found in Thermal Expert. You will find some products that claim some other technology. But ask yourself why FLIR and others are still on shutter.
 

Offline calel

  • Regular Contributor
  • *
  • Posts: 87
  • Country: ch
Re: calibration time: is it faster in Exx series? (compared to Ex series)
« Reply #4 on: October 18, 2020, 10:45:18 pm »
But ask yourself why FLIR and others are still on shutter.
cause a cam with shutter method is cheaper I reckon. has to be the only explanation cause there's no advantage to the shutter method right? image freezing all the time can get quite annoying
 

Offline Bill W

  • Frequent Contributor
  • **
  • Posts: 541
  • Country: gb
    • Fire TICS
Re: calibration time: is it faster in Exx series? (compared to Ex series)
« Reply #5 on: October 19, 2020, 12:13:29 am »
There is no reason for it to take 1.5 seconds other than poor design.     :box:

An Argus camera is ~300ms and has to do more than a FLIR does in that time.

Arguably a BST (or pevicon tube for that matter) was being 'calibrated' this way every video field by the rotating chopper wheel.....

Bill

Offline DaJMasta

  • Super Contributor
  • ***
  • Posts: 1955
  • Country: us
    • medpants.com
Re: calibration time: is it faster in Exx series? (compared to Ex series)
« Reply #6 on: October 19, 2020, 06:31:44 am »
I think my e40 (hacked to e60) does it in around a second, maybe just under, so that would be faster than the 1.5s of an e8.  This calibration can be turned off on some cameras, but the longer you go without it the less linearity the sensor has and the farther it can drift from the spot.

I have a Tau2 core that's controllable over USB and you have controls for the FFC calibration settings, the default is time based (technically frame count with the Tau2) and I believe this is what the eX and eXX series both use, but the second is based on drift temperature of the sensor itself.  Since I have it setup in a fixed place indoors and because the core isn't radiometric anyways (I'm not getting absolute temperature values out of it, just the image), I have it configured to do a FFC every 0.3 degree C change of the core, which means after half an hour or so, it will only recalibrate every hour or so and visually it looks fine.  This would not be suitable for precise temperature measurement, though.
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 9909
  • Country: gb
Re: calibration time: is it faster in Exx series? (compared to Ex series)
« Reply #7 on: October 19, 2020, 12:16:29 pm »
FLIR use a comprehensive FFC process in their cameras which may explain the time taken to complete the event.

We have to remember that FLIR are in the business of producing thermal cameras that meet, or exceed their claimed specification. That is why they are one of the big names in the field of thermography. Radiometric measurement accuracy is often the higher priority, rather than ‘pretty pictures’ or uninterrupted imaging.

An FFC event is part of the microbolometer imaging process where accurate temperature measurements are required. The FFC flag not only provides a flat field to level the pixels but it also provides a temperature reference against which to calibrate measurements. It is possible to operate a Microbolometer based thermal imaging system without an FFC event routine but measurement accuracy suffers and some minor variance in pixel outputs when viewing a flat scene is to be expected. Operating in a non FFC event mode is often required for videography where pretty pictures take priority over measurement accuracy. Image processing algorithms provide dynamic ‘virtual’ FFC to maintain an acceptable image quality. The latest microbolometers do not tend to use Peltier temperature stabilisation due to its inherent high power consumption and vacuum contamination issues. Peltier module equipped temperature stabilised microbolometers would cope better without the FFC event than the latest non temperature stabilised microbolometers but the Virtual FFC algorithm does its best in the circumstances.

If a user demands the best possible thermal imaging without interruption, such as required for wildlife videography for broadcast, then a cooled thermal imaging camera may be used as this does not require an FFC event due to its inherent temperature stability. As has been stated, the BST based cameras used a rotating chopper wheel that levelled the sensors view after each frame was read out so no FFC event was visible to the user. BST was not well suited to radiometric measurements however. That may have changed if development of the technology had been continued. A sad loss to the thermal imaging community.

To professional users of Microbolomter based thermal imaging cameras, the FFC event is expected and tolerated. Sure they would prefer to not have the short break in imaging every 2 minutes or so, but they often need the measurement accuracy that the FFC event helps to achieve. In consumer circles where the public are used to high quality imaging from camcorders and phones etc, the FFC event can appear to be an awful imposition on their viewing pleasure. That is because they do not understand the amazing technology that is at work to provide them with affordable thermal imaging. Such is life and FLIR will not lose any sleep over it.

The following link takes you to the FLIR FFC process patent that explains the process and may help the reader to understand what actually happens while the FFC event is taking place. It is not as simple as might first be thought  :)

https://patentimages.storage.googleapis.com/ba/c9/a6/94221fac22ca2f/US8373757.pdf

Fraser

« Last Edit: October 19, 2020, 12:29:18 pm by Fraser »
Cogito, ergo sum
 


Offline calel

  • Regular Contributor
  • *
  • Posts: 87
  • Country: ch
Re: calibration time: is it faster in Exx series? (compared to Ex series)
« Reply #9 on: October 19, 2020, 02:39:21 pm »
ah so FCC is what it's called will check it out thx   8)



so to answer the topic higher end cams do have faster FCC so even if it's indispensable at least it's possible to make it faster (fast enough for our viewing comfort)

wish there was a setting to make it faster for the Ex  :/
« Last Edit: October 19, 2020, 02:43:46 pm by calel »
 

Online Fraser

  • Super Contributor
  • ***
  • Posts: 9909
  • Country: gb
« Last Edit: October 19, 2020, 03:23:13 pm by Fraser »
Cogito, ergo sum
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf