Author Topic: why E6 costs 3x more than TG267? (they're basically same camera)  (Read 4209 times)

0 Members and 1 Guest are viewing this topic.

Offline calelTopic starter

  • Regular Contributor
  • *
  • Posts: 97
  • Country: ch
why E6 costs 3x more than TG267? (they're basically same camera)
« on: September 15, 2020, 07:55:49 pm »
both have same resolution 160x120
same thermal sensitivity
MSX
usb connectivity
IP54

they're basically the same camera (same capabilities). that would make the E6 a total scam right? it's like 3x more expensive than the TG267  ???
 

Offline Bill W

  • Super Contributor
  • ***
  • Posts: 1103
  • Country: gb
    • Fire TICS
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #1 on: September 15, 2020, 09:00:13 pm »
They are far from identical, E6 is a TAU core and TG267 is a lepton.

So E6 is portrait format, larger screen, better NeTD (60/70) and better / narrower lens.

TG267 temperature accuracy looks a bit suspect as you'd expect on a lepton, and the NeTD is likely suspect too.  It may well not carry forward into actual image quality.

Whether this is worth 3x is a valid ask, but in hand anyone would take the E6

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 13170
  • Country: gb
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #2 on: September 15, 2020, 09:03:52 pm »
+1

The TG267 and E6 are like comparing chalk and cheese.

Fraser
« Last Edit: September 15, 2020, 09:18:07 pm by Fraser »
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline calelTopic starter

  • Regular Contributor
  • *
  • Posts: 97
  • Country: ch
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #3 on: September 15, 2020, 09:45:01 pm »
They are far from identical, E6 is a TAU core and TG267 is a lepton.
andn that means...? (I'm new to this whole IR thing tbh)

what are those cores? physical thing or just a software thing? is core same thing as sensor?
if it's physical then how is Tau core better? more sensitive? or the pixels in a tau sensor are better quality then pixels in a lepton sensor? (like the difference between the pixels of a classic TNT monitor & those of an IPS monitor?)
or does it make the IR image update faster in real time when moving camera? I noticed that the e4 is always "calibrating" every 30s or so & it freezes the image for quite a few seconds when doing so, that don't sound like very impressive performance (what exactly is it calibrating anyway?) and that's supposed to be high-end TAU sensor, this means the lepton sensor would perform even worse then that?


also is screen size important if IR resolution's the same?  if you just increase display size without increasing resolution then logic says the whole image will just get blurrier right?  ???

+1

The TG267 and E6 are like comparing chalk and cheese.

Fraser
but the resolution (IR pixel count) is the same & for IR cameras that's the most important criterion right?
« Last Edit: September 15, 2020, 10:22:43 pm by calel »
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 13170
  • Country: gb
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #4 on: September 15, 2020, 11:10:22 pm »
There is much to be understood when it comes to comparing two thermal cameras. There are cameras costing a few hundred Dollars and others that cost tens of thousands of Dollars. They are equal only in that they can detect heat and display it to the human user.

In simple terms, the E6 is part of the Ex series of cameras that were developed as an entry level thermal imaging camera that meets the needs of less demanding users of the technology. The series uses a cost reduced microbolometer sensor that was developed for use in automotive night vision Thermal cameras and the lens is quite basic. The camera is fixed focus but achieves decent performance. The physical resolution of the cameras microbolometer sensor is 320 x 240 and this is what was released in the E4 upgrade  :) The Ex series is a decent product thanks to the quality of the microbolometer, lens and firmware that processes the thermal scene coming from the sensor array. In terms of value for money, the Ex series WERE reasonable but since its release, other, more affordable cameras have been produced by other competing companies. BUT, and it is a big BUT, the E4 camera remains the darling of the series thanks to the E4 upgrade that has been created on this forum. The upgrade means a user can buy a bottom of the range E4 and upgrade its resolution to the full 320 x 240, switch off the artificial noise generator and add functionality that goes beyond that offered by even the top of the series E8  :-+ An upgraded E4 is a thing of beauty that costs less than $1000? In terms of value for money and the image that it can produce, there is not much to touch it, if anything. Dongle based thermal cameras are cheaper but you need a phone and some people want the all-in-one solution rather than using a phone with something hanging off of it.

The TG series are bottom end thermal imaging intended for users who are very limited in their needs. They are “budget budget” thermal cameras. What is amazing about these TG series cameras is that they are relatively inexpensive. This is all thanks to FLIR developing the Lepton Core. This is a thermal imaging sensor array with 80 x 60 or 160 x 120 pixels that integrates significant image processing within its diminutive physical dimensions. The whole core is smaller than a sugar cube ! Sadly to create such a marvel of engineering, FLIR had to accept some compromises in terms of the cores performance. The sensor array uses the 12 micrometer size pixels that are still relatively insensitive and so the noise content of the produced images is relatively high. They used a silicon ‘printed’ diffractive lens structure that further reduces sensitivity and imaging performance. The result is an amazingly compact thermal imaging core that produces OK images, but not great images. FLIR use a lot of image processing to improve the images that come from the Lepton core, but there is only so much that can be done with sub optimal data as the source.

Basically the FLIR Lepton is intended to be a very affordable thermal imaging core that may be easily integrated into products that do not need to provide impressive thermal imaging output. The scene just needs to be ‘adequate’ for the expected users needs. The FLIR One series is an example, as is the Thermal dampness meter and the TG series thermal imagers..... they produce an adequate, if a little disappointing thermal image. To improve that image would require the Lepton core to be completely redesigned from the ground up as the current design has reached its limits. That is exactly what FLIR did when they embarked on the FLIR BOSON core Project. That Core has proved to be disappointing however !

With thermal imaging, you often get what you pay for and if you need high quality imagining with accurate temperature measurement that can be trusted, it does not come cheap. There are some nice affordable options available though that would easily out perform any Lepton based camera.

If you need to know exactly what makes one thermal camera ‘OK’ and another ‘Amazing’ in terms of the images it produces, you will need to study IR lens design, microbolometers, NeTD and image processing algorithms. These dictate the quality of the image, measurement accuracy and the cost of the product to manufacture.

It is not just about the pixel count.

Fraser

« Last Edit: September 17, 2020, 12:47:30 pm by Fraser »
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 
The following users thanked this post: bdeianov

Offline calelTopic starter

  • Regular Contributor
  • *
  • Posts: 97
  • Country: ch
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #5 on: September 15, 2020, 11:32:30 pm »
ok so if IO understand correctly even the ex series & their Tau core is entry level (and the TG series & lepton core is entry-entry level)
and the main difference is thermal sensitivity? (which means the advertised 0.07°C sensitivity could be BS?)



if so, what's the next level above eX & their Tau core?



in fact what's the best core they got? as in, any cameras with IR images so good they dont even need any processing?
 

Offline Vipitis

  • Frequent Contributor
  • **
  • Posts: 867
  • Country: de
  • aspiring thermal photography enthusiast
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #6 on: September 16, 2020, 12:45:15 am »
the Exx line is a jump up in lens quality and also framerate.

Txxx or even T1K is using even bigger versions of the cores, bigger lenses and far higher level of calibration. Beyond that is scientific cameras, mostly cooled and far different technology to LWIR microbolometers. All cameras need processing,  you won't get by that. While the Tau and Tau2 aren't the most modern cores, they certainly get the better image quality compared to the modern Boson.

I don't have a Tau core, but two different Lepton generations as well as big cores with bigger pixels and better "sensitivity" and various lenes and lens elements. You can see the difference in the texture of a hand already. but I might get to the point where I can do some side by side comparison, as there are many already.
 

Offline calelTopic starter

  • Regular Contributor
  • *
  • Posts: 97
  • Country: ch
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #7 on: September 16, 2020, 01:36:32 am »
so does Exx use superior core to Ex? or same (Tau)?

(also I've an old E4 dating back to 2014 does it have Tau or Tau 2?)
 

Offline CatalinaWOW

  • Super Contributor
  • ***
  • Posts: 5236
  • Country: us
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #8 on: September 16, 2020, 02:57:54 am »
A reminder that "better" is only meaningful relative to the application.

If you are looking for hot bearings, or ICs which haven't been connected to their heatsink there will be little or no difference between those cameras.

If you are looking for pretty pictures in a hot desert the temperature varies by 6-20 degrees even a .1 deg NET gives you dozens or hundreds of grey scale levels and the difference in cameras will be small.

But in an indoor scene the total temperature range is likely to be only a couple of degrees (excepting heaters and other isolated cases).  Every grey level you can get is important and fixed noise becomes important.

Trying to get usable images at long range on a cold cloudy day where scene temperature range is small and atmospheric attenuation further reduces contrast challenges all but the very best cameras.

You need to know what you want to decide what will make you happy. 
 

Offline calelTopic starter

  • Regular Contributor
  • *
  • Posts: 97
  • Country: ch
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #9 on: September 16, 2020, 03:30:23 am »
You need to know what you want to decide what will make you happy.

this level of detail would make me very happy

(does FLIR even sell cams capable of something like this? looks more like military grade stuff I imagine this would cost hundreds of thousands if not millions)
 

Offline CatalinaWOW

  • Super Contributor
  • ***
  • Posts: 5236
  • Country: us
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #10 on: September 16, 2020, 03:58:06 am »
You can work out the rough specs for the camera yourself.  Count pixels in both dimensions.  Looks greater than 640x480 to me.  You can also get rough estimate of NET.  Assuming 6 degrees C for dynamic range and well over 100 grey scale levels gives an estimate well below 0.060.  You could firm up these numbers by taking the image and looking in detail, though I am sure it is 8 bit video so you can only confirm 64 grey levels.  The camera appears to have a fairly high frame rate also.

Such cameras have come down in cost.  The last I checked you could get them for a few tens of thousands.
 

Offline calelTopic starter

  • Regular Contributor
  • *
  • Posts: 97
  • Country: ch
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #11 on: September 16, 2020, 04:18:59 am »
yeah youtube allows viewing this vid in 1080p though the vid itself doesn't appear to be 1080p but definitely way above 480p

so are they even commercially available? do common companies like FLIR make them? can private citizens buy them or is this only sold to government & corporates?

FLIR's 480p products like the e95 cost around $10000 so wouldnt a cam like the one used by National Geographic cost far far more


ps. what's "NET"? the only NET I'm familiar with is Microsoft's .NET :/
« Last Edit: September 16, 2020, 04:22:38 am by calel »
 

Offline CatalinaWOW

  • Super Contributor
  • ***
  • Posts: 5236
  • Country: us
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #12 on: September 16, 2020, 10:21:36 am »
An abbreviation to an acronym.  NETD.  Noise Equivelant Temperature Difference.

 

Offline Vipitis

  • Frequent Contributor
  • **
  • Posts: 867
  • Country: de
  • aspiring thermal photography enthusiast
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #13 on: September 16, 2020, 10:46:14 am »
Check my reply in this post, I tried to get the real information there: https://www.eevblog.com/forum/thermal-imaging/thermal-imaging-in-mainstream-media/msg3158370/#new

In short: you can barely get you hands on XGA+ cooled cameras, in working condition. But it's not impossible and see recent posts on this forum as an example. There is a difference in LWIR vs MWIR and from the freebooting reupload you send I can only guess. But I'd say it is the Merlin footage as BBC has access to that, there are some more links to be found once you research into it.

LWIR VGA is available and affordable and could be made into a filmmaking platform, however you have to find an eBay bargain - which many have shown on this forum - is possible. Older cores mean bigger pixels and that also entails bigger lenses, which to me makes the most difference, but be aware that DIY is not an easy task. Maybe try to convince Sierra Nevada to give you a quote on the Vayu HD, as I intended a while ago.
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 13170
  • Country: gb
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #14 on: September 16, 2020, 10:47:01 am »
Calel,

Likely a Leonardo Horizon camera as that is popular with wildlife videographers. It is often used and is either loaned to videographers, leased or, if they are rich, sold. It is a case of if you need to ask the price, you likely cannot afford it. Over $100K anyway.

The Horizon uses a high resolution cooled imaging core, hence the crisp, low noise imagery that it produces.

https://www.leonardocompany.com/documents/20142/3150491/Horizon_IR_camera_LQ_mm07756_.pdf?t=1538987622892

Fraser
« Last Edit: September 16, 2020, 10:48:45 am by Fraser »
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 13170
  • Country: gb
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #15 on: September 16, 2020, 11:24:23 am »
Just to clarify something regarding the Ex series.....

The Ex series do not actually contain a TAU or TAU2 core. The TAU is a complete core and video processing module in its own self contained aluminium housing. It can form the basis of an excellent thermal camera or be used ‘naked’ on its own. I can understand confusion regarding TAU as it is used in some other FLIR cameras.

The Ex series contains a microbolometer that was developed for use in vehicle night vision systems, such as the Autoliv product line. The aim of the project was to reduce the manufacturing cost of the microbolometer. The resultant microbolometer found its way into the Ex series design. The Ex series cameras comprise a microbolometer and ROIC module connected to a main PCB that deals with video processing, user interface and battery management. As such it is not actually a TAU based design, even if the technology used is similar.

Fraser
« Last Edit: September 16, 2020, 09:19:36 pm by Fraser »
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline calelTopic starter

  • Regular Contributor
  • *
  • Posts: 97
  • Country: ch
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #16 on: September 16, 2020, 06:27:43 pm »
so if it's not a Tau do we know if the E6 sensor's any better then the TG267's Lepton then? maybe its the same quality for all we know
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 13170
  • Country: gb
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #17 on: September 16, 2020, 07:07:26 pm »
No, the E6 microbolometer is far superior to that of the Lepton. I never said I did not know the identity of the microbolometer in the E6, just that a TAU core is not used  ;)

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

Offline calelTopic starter

  • Regular Contributor
  • *
  • Posts: 97
  • Country: ch
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #18 on: September 16, 2020, 07:59:17 pm »
 ok ok so the...core/sensor/thingamajig used in the e6, what's it called? gotta have its own reference right?  ???

and how good is it quality wise? above Lepton but below Tau?
« Last Edit: September 16, 2020, 08:01:23 pm by calel »
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 13170
  • Country: gb
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #19 on: September 16, 2020, 08:34:24 pm »
FLIR Microbolometer Die's :

FLIR ISC0601B (320 x 240 25um pixels)

Used in:

Exx Series
TAU 320
Autoliv FLIR NV2
ETS 320
A320
ix Series

FLIR ISC0901B (336 x 256 17um pixels)

Used in:

Ex series
TAU2 336
Autoliv FLIR NV3
Vue 336
Quark 336
Muon 336
Trafisense BPL317

Performance of the two microbolometers is very similar to the point that it makes little difference which one is used. It is often the lens and video processing stages that follow the Microbolometer that dictate the imaging performance. For example, TAU2 has more refined video processing than the earlier TAU core. The Exx series has a decent quality manual focus lens that easily outperforms the Ex series smaller fixed focus lens. So even though similar image processing is used, the Exx produces a superior image.

I think that is enough from me. Unlike Google I have a life   ;)
« Last Edit: September 16, 2020, 09:48:52 pm by Fraser »
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 
The following users thanked this post: bdeianov, nikitasius

Offline calelTopic starter

  • Regular Contributor
  • *
  • Posts: 97
  • Country: ch
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #20 on: September 16, 2020, 09:06:35 pm »
I also noticed the Exx has better image than Ex @ same resolution

eevblog.com/forum/testgear/flir-e4-thermal-imaging-camera-teardown/?action=dlattach;attach=182634;image
eevblog.com/forum/testgear/flir-e4-thermal-imaging-camera-teardown/?action=dlattach;attach=182636;image

the e60 (hacked e40?) image also has lower brightness for some reason 



>> I reckon no hack can give the e4 same image quality as the e60 or any Exx?  :-[

« Last Edit: September 16, 2020, 09:13:40 pm by calel »
 

Offline Bill W

  • Super Contributor
  • ***
  • Posts: 1103
  • Country: gb
    • Fire TICS
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #21 on: September 16, 2020, 09:34:52 pm »


what are those cores? physical thing or just a software thing? is core same thing as sensor?
if it's physical then how is Tau core better? more sensitive? or the pixels in a tau sensor are better quality then pixels in a lepton sensor? (like the difference between the pixels of a classic TNT monitor & those of an IPS monitor?)
or does it make the IR image update faster in real time when moving camera? I noticed that the e4 is always "calibrating" every 30s or so & it freezes the image for quite a few seconds when doing so, that don't sound like very impressive performance (what exactly is it calibrating anyway?) and that's supposed to be high-end TAU sensor, this means the lepton sensor would perform even worse then that?


also is screen size important if IR resolution's the same?  if you just increase display size without increasing resolution then logic says the whole image will just get blurrier right?  ???


but the resolution (IR pixel count) is the same & for IR cameras that's the most important criterion right?

Core = sensor plus some electronics and software.  Usually reused in several end product cameras or even sold to other camera builders as this is where the development costs are.

I'd compare a Tau vs lepton as more like a Ford and a Trabant.  Both have 4 wheels so whats' the difference  ;D

The screens do matter as the E series (and similar K series) is a very nice screen and even if the sensor is less you can upsample quite well with a reasonable powered core (or in the lepton based FLIR-One using the attached mobile phone).  TG267 is not so well appointed.

Calibration is needed as the sensors do drift in offset, mainly by column brightness, for very small temperature changes.  The calibration is telling the camera 'here is a blank scene of xx°C - so make your output to match'.  An array of offsets is then applied to the sensor output until the next calibration when a new array of offsets are calculated.
The TG267 / Lepton is as bad or worse for calibrations.

Pixel count and noise should be the most important criteria, but a good 160x120 upsampled can equal a grim 320x240 where too many corners have been cut on lens optics, screen and processing.

Bill


Offline calelTopic starter

  • Regular Contributor
  • *
  • Posts: 97
  • Country: ch
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #22 on: September 16, 2020, 10:45:29 pm »
Calibration is needed as the sensors do drift in offset, mainly by column brightness, for very small temperature changes.  The calibration is telling the camera 'here is a blank scene of xx°C - so make your output to match'.  An array of offsets is then applied to the sensor output until the next calibration when a new array of offsets are calculated.
The TG267 / Lepton is as bad or worse for calibrations.
do the Exx cams calibrate faster than the Ex? 

(real time? any IR cam capable of this?)



if I come across a 2nd hand E40 I'd be tempted to buy that too (if it's an old one & assuming those can be hacked to E60)
 

Offline Vipitis

  • Frequent Contributor
  • **
  • Posts: 867
  • Country: de
  • aspiring thermal photography enthusiast
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #23 on: September 17, 2020, 12:37:43 am »
there is different "calibration" steps. Cameras are calibrated in the factory to give accurate readings within spec, this is expensive and time consuming to get done again, but needed in some fields to ensure accuracy.

There is a FFC or flat field calibration done during camera operation that is basically just a metal flag being held in front of the sensor every few minutes to correct camera drift, essentially as the sensor itself is heating up. There are shutterless cameras with different ways to minimize those effects.
 

Offline Bill W

  • Super Contributor
  • ***
  • Posts: 1103
  • Country: gb
    • Fire TICS
Re: why E6 costs 3x more than TG267? (they're basically same camera)
« Reply #24 on: September 17, 2020, 08:32:19 pm »
do the Exx cams calibrate faster than the Ex? 
(real time? any IR cam capable of this?)

I have not measured the FLIR E model times, the FLIR-one is certainly slow.
There has (probably...#) to be a minimum time as you freeze frame, take a few fields to close the shutter, a couple for the blank field, a few to open again while calculating the new offsets and possibly some field if making other adjustments like gain mode.

Can you measure your cameras ?
Get an analogue video feed out onto a scope, display a good dozen frames, wave a soldering iron in front while it calibrates, and see how many frozen fields you get.
Argus4 cameras are 14 fields

#
A rotating chopper (as in Raytheon BST / Pevicon cameras) is an every field flat calibration but I do not know anyone who has done it on a microbolometer.  It would make the camera larger and less accurate in temperature, but would eliminate any freeze up time.
One for the home-brew software folks to try?

Bill



Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf