High speed thermal imaging cameras - When 60fps is just not enough !

[Vipitis]

wait. you are not just using one... you own it? That is crazy. I would be heavily interested to see how the very high thermal resolution and high speed capture allows you to see indirect body functions(reflexes) in blood vessels and muscle retractions. Certainly something special to call your own. I hope to work with this tech one day.

[Fraser]

Yes, I broke my rule about not buying any more cameras. I just could not resist it as the price was right 

Having Quickly tested it this evening, it is cooling very quickly to 76.1K and providing a thermal image
It is working perfectly The Stirling Cooler is so quiet that I did not think it was running when I first switched it on. The two little forced air cooling fans for the Stirling cooler heat sink make more noise than it does ! Lovely technology 

Upon checking the Pelicase contents I found what look like two high emissivity plates of dimensions approx 100mm x 100mm. Not sure what they are for yet as the camera has internal black bodies for NUC and dead pixel checks. I have seen them in other pictures of these camera kits though so they are a FLIR provided item. Nothing in the user manual about them that I can see. Maybe I missed it though.

The camera is pretty heavy at around 10lbs as it is an all metal casing, chassis and cooler. Built like a tank !

Time for bed now. I will have another play with the camera tomorrow.

Fraser

[Fraser]

The FLIR SCxxxx cameras find uses in all manner of specialist fields, including the military. The attached pictures show an SC6000 attached to a beautiful 10” telescope for long rand work

The images were borrowed from here :

https://www.rcopticalsystems.com/telescopes/10military.html

[Fraser]

The two thick metal plates that are covered in high emissivity paint are calibration plates. The SC4000 is not like your average thermal camera. It has its own multiple NUC tables and can also have additional NUC tables stored in the PC software. As an InSb camera it does not have the FFC tables found in a microbolometer based camera. The user can set up an experiment and use unique NUC tables to ‘normalise’ the scene. That means a scene that contains a temperature range or profit e can be ‘flattened’ to show any change in temperature during the experiment. This ability can take account of optical path characteristics and effectively ‘zero out’ imperfections in the optical path to the camera. The camera contains its own NUC tables that were created at the factory on the  alibration bench. Apparently, user created NUC tables can be even better as they are a true reflection of the whole optical path in an experiment. The two emission plates are used to provide a uniform surface for creating user NUC tables.

This cameras advanced configuration GUI utility reads like the control system for a Space Shuttle ...... there is so much that can be ‘tweaked’ and adjusted to provide the best imaging possible. These cameras truly are scientific instruments and not ‘just’ a thermal imaging camera. Thank heavens it also has default settings and an auto configuration capability ! FLIR said they do not normally supply the camera GUI to ‘average end users’ as it is too much ‘power’ to trust them with. The ‘basic’ camera configuration tool in the Researcher software is considered adequate. Only heavyweight research establishments normally get the GUI to enable maximum performance from the camera. I have the Advanced GUI utility CD in the kit but may leave that well alone until I master driving the camera in the ‘basic’ mode of configuration.

My brain is overloading  with all the information in the camera user manuals and software manuals !

Fraser

[Fraser]

In case anyone at FLIR is reading this thread and panicking ...... fear not, this SC4000 and associated documentation could not be in safer hands I am a British Citizen with no intention of moving this camera out of the United Kingdom. You can message me if you have any concerns

Hopefully no black helicopters will be descending upon my location. If I go silent, you will know why

Fraser

[Fraser]

Some pretty poor pictures I just took of the GUI section of the SC series user manual. Sorry about the poor clarity but you will get the idea of what can be adjusted in the standard GUI.

I have to be careful what I share from the manual as I am aware that some(many) technical details are sensitive. Please do not ask for copies of the manuals as they are a restricted circulation document and I cannot share them.

Fraser

[eKretz]

That is wicked cool. All those adjustable options are about like a thermal imaging tinkerer's wet dream. Very jealous ovah hyeah! Can't wait to see some images.

[Fraser]

I am still learning how to drive the unit so no promises on when I can upload some images, hopefully soon though.

[Hyper_Spectral]

I see higher resolutions mentioned in the software but not in your post. Is this a camera with a 640x512 core that's shrunk for higher frame rates or?

[Fraser]

No I think not. The technical details provided for the various InSb sensor arrays show different pixel sizes between the the two resolutions available and  the lens FOV figures do not tally for a higher resolution array that is windowed in the lower resolution camera.

Fraser

[Fraser]

I heard from a fellow forum member today that the bayonet mount MWIR lenses on my Amber Radiance 1 may/should fit my SC4000 

I will dig out the Radiance 1 from the spare room and see if they are fully compatible. There seems to be a “standard bayonet mount” in the world of thermal cameras which came as a pleasant surprise. I have both the 25mm and 50mm lenses for the Radiance 1. They would make a useful option for the SC4000 

Fraser

[TooQik]

Some awesome videos posted in this thread.   

[Ultrapurple]

How about thermal imaging at 100,000 fps and X-Ray imaging at 1,000,000 fps? Argonne National Laboratory has a new toy too...

[eKretz]

Now that right there is cool! 

I like the laser weld - the weld is totally completed before any heat even gets down into the parent metal!

[Fraser]

Sorry there has not been any more about the SC4000 camera. I have been ill for a few days so have not touched it since initial testing. I will likely need to build a WinXP laptop that has Gigabit Ethernet in order to use the Researcher 2.8 software and the GUI. I just do not feel in the mood to do that at the moment. Maybe over the next few days. For the moment, the SC4000 is back in its nice cosy and safe pelicase

Fraser

[Hyper_Spectral]

Can't you use a VM with XP?

[Fraser]

My past experience with VM builds to create XP on a Win7 platform has not been good. FLIR do warn about how fussy some of their software is about running on Virtual Machines. In the past I think that was because of differences in the way the operating systems handled the I/O ports and some interfaces. Drivers for some bespoke interface cards could also be very fussy about the OS.

I have many good quality Core2Duo, i5 and i7 laptops, including ruggedised Military types that can be brought into service with a quick loading of WinXP on the drive. Thankfully spare laptop platforms are not something I have a lack of 

In the case of these high speed cameras, FLIR recommend a laptop or Desktop computer that is using an Intel PRO/1000 GigE NIC and is dedicated to the role. The CPU needs to be given the best possible chance to keep up with the high data rate coming in on the GigE port in order to avoid dropped frames. I suspect a VM based XP install may suffer issues on that front.

It is very early days for me with this high speed camera. I am still learning about it. Even installing it to run on a standard PC has lead to some head scratching for me    I will get it running properly on a very simple Win XP platform before testing other options like ResearchIR 4 MAX that runs under Win7. The camera uses a Pleora iPort GigE interface and I have not used one of these before. The Pleora driver needs to be ‘happy’ to get imagery out of the cameras Ethernet port.

Fraser

[Hyper_Spectral]

Very good information! Your posts will be invaluable to others following your steps in the future 

[Fraser]

Hyper-Spectral,

I will be documenting my ‘journey’ with this camera in a new thread. I have read the various user and technical manuals that come with it and I am about to embark on the voyage of discovery that will hopefully enable me to get the best out of this amazing piece of equipment.

If I am honest, I am still getting used to the whole idea of owning such an advanced thermal camera. It is more sophisticated than any that I have used previously as it offers so much control over the image capture and processing stages. As I have previously stated, this is a scientific piece of equipment as opposed to a semi-pro or even pro thermal camera. It has a totally different ‘feel’ to it and the menu’s are a little daunting when first viewed. I thought I was very lucky to own a FLIR A40 and SC3000 static cameras. This SC4000 is not the sort of thermal camera I ever expected to own or have the opportunity to buy .... it is all a little surreal to be honest.

The SC4000/6000 series were developed by FLIR-Indigo for use on US Military weapons ranges but they were permitted to sell them to certain approved customers as well. Hence the tight controls on both the camera hardware and even the highly detailed user and technical manuals. These cameras cost in the range of $85K to $140K 

I am pleased that this camera contains a linear Stirling Cooler unit. These have been carefully developed to ensure a long and reliable operational life. The life is stated as >10,000 hours but testing showed them to last more than 250,000 hours ! FLIR cannot guarantee that of course. The down side of the linear cooler is its size, weight and current consumption. Hence why the SC4000 weighs around 10lb and is basically a ‘box’ shape to accommodate the bulky cooler and the PCB’s are mounted around it. The unit draws up to 4A at 24V so it is not exactly environmentally friendly in the power consumption department    I only care about operational life though so these drawbacks are of no concern. The linear cooler primary piston housing also has the advantage of no wires entering through the sealed casing for a rotary type motor. Such feed-thru connection seals are known to sometimes leak, causing loss of the Helium fill. The piston drive is magnetic with the drive coils surrounding the primary piston cylinder and the piston contains a strong permanent magnet. The piston is attached to a precision spring and operated at resonance by the driver circuit.

I own the x15 microscope lens assembly from an AGA Thermovision 680 MWIR thermal microscope. It will be interesting to see whether I can adapt that lens assembly to work with the SC4000  I will also be looking at making some bayonet mount spacer rings for close focus work using the standard 25mm and 50mm lenses. There may even be a photographic bayonet mount that can be adapted. We shall see. My Inframetrics Close-up lenses are dual band so I can also use them with this MWIR camera.

So the next steps for me now will be to create a nice stable laptop running XP and installed with Pleora iPort and the SC4000 GUI for configuration testing and changes. I will then install FLIR Researcher 2.8 SR3 that came with the camera. That should work perfectly as the camera is specifically configured to work with Researcher 2.8 without any tweaking needed. If all goes well I will move the testing onto ResearchIR 4 MAX running on a Win7 platform. ResearchIR 4 MAX is the latest R&D thermal analysis software from FLIR and it supports the SC4000 and 6000 series. Hopefully all will go to plan and I will be able to run the camera on either a WinXP Researcher 2.8 or Win7 ResearchIR 4 MAX platform 

As I say, I will likely begin a new thread to document the background to the SC4000/6000 series and my experiences with it.

Fraser

[Fraser]

I have just completed a WinXP SP3 build on one of my laptops 

The laptop uses a Broadcom GigE port so I will not achieve the best performance but that is not an issue at the moment. I just want to test the GUI and Researcher 2.8. That will hopefully happen tomorrow 

I did some playing with ResearchIR 4 MAX today on my Win7 laptop. Whilst I could see the camera and obtain information from it, the Pleora ‘Camlink to Ethernet’ interface driver did not want to play ball. I did not have the time to sort out Pleora, but I am already beginning to dislike it ! The composite video output from the camera is working fine and this is how I first tested the camera.

Tomorrow I will take my time installing the drivers and software and see if I can get the Pleora driver working properly.

Fraser

[Ultrapurple]

Sounds like you are having a lot of fun! Money well spent indeed - but you're pretty good at finding the bargains the rest of us don't see. I wish I knew your secret...

[Fraser]

Well it has been a frustrating day !

The camera is working perfectly, it is the FLIR software and Pleora drivers that have been giving me problems.

The images that the camera produces a very nice indeed. They have the look of a monochrome CCTV image and there is absolutely no need fro anything like MSX with this camera. It detects such small Delta T and emissivity differences within a scene that the  contrast is excellent. She is a lovely bit of kit, but the software set up ........ Hmmmmm not going so well for me at the moment.

More when I get a moment.

Fraser

[Fraser]

I was not going to post this first image but it may be of interest to some.

This image has just been captured from the SC4000 using ResearchIR 4 MAX with the camera running at 320 x 240 pixels 30fps.

This picture is a very long way from good. I did not fine tune the focus and the software is running in full auto mode with no effort to fine tune the image before capture. I did save it as a PNG file in preference to BMP or JPG. I think upscaling and smoothing are likely active as I have not disabled them.

I took a quick look at my arm and without any fine tuning I could see the veins in my arm. Even in full auto span and centre, this camera produces a respectably balanced image.

I still need to master how to create new NUC, Dead Pixel and Calibration files. I was just happy to get the image up on the software for the moment. I literally loaded the software, started it , loaded the default calibration file and set the focus roughly before hitting the snapshot button with my hand in front of the camera ! My hand is resting on a CD storage box that is a composite of plastic coated wood panels with aluminium edging.

Once I have a better handle on driving and fine tuning the camera/software settings I will see what images I can capture and upload.

Fraser

[Ultrapurple]

That's ... pretty ... good ... 

I am seriously jealous. As, I suspect, is pretty much everyone else. But fortune couldn't have smiled on a better candidate.

  Teardown coming soon?

[Fraser]

@Ultraviolet,

Thank you. It is a pretty crude and hastily captured picture of an uninteresting target. Sorry about that.

Yesterday was quite a frustrating day in some respects. I was very happy to get the image coming up on the computer rather than just on the composite output but I am still struggling with the darned FLIR softwares.

FLIR Researcher 2.8 has a very old fashioned user interface but I did get it running with the camera. The later FLIR ResearchIR 4 MAX is a more modern user interface but there are little icon buttons scattered all over the place and it takes time to navigate the softwares menu’s. It is not the most user friendly of software applications but I have found this with other specialist software. I find the user manuals for both Researcher and ResearchIR to be somewhat lacking when it comes to the integrated camera control menu. The various menu’s are quite confusing and scary when warnings appear saying that you are about to overwrite the cameras internal calibration ! Remember, these scientific cameras are intended to have their internal calibration accessible and updateable by the user. Not something you want to do by accident though !

I am still struggling with how to carry out a NUC and Dead Pixel map update on the camera without erasing the temperature calibration as well. The camera contains multiple NUC tables that are stored in flash memory. The user can set up specific NUC tables for a particular resolution and temperature range. This gives the best possible imaging performance. I just need to work out how to create new NUC tables and how to back up the factory installed tables and calibration before ‘playing’ with them. Nothing is very simple when controlling the camera via Researcher or ResearchIR !

So why do I not use the dedicated SC4000/6000 camera controller GUI that came with the camera kit ? Yes that would be a great utility as it makes controlling the camera so much easier with its logical menu structure. That is the sort of camera configuration utility that I am used to using. Just one problem.....
I have installed the GUI on both a WinXP and a Win7 platform and I get the same problem. The GUI software runs fine but as soon as I try to connect to the camera I get an Exception error saying that file ISC_Camera.dll cannot be located. The software then recommends exiting but the warning can be ignored. Ignoring it achieves nothing as the camera cannot be found when it should be seen on the Ethernet port. Somewhat frustrating. Checking the GUI installation guide in the user manual provides absolutely no help as it is supposed to be a simple install, followed by camera detection and access ! Checking a readme file on the GUI CD provides version detail for the GUI, but also mentions a dependency...... the Indigo “AccessorySDK”. I am thinking that I am missing this essential SDK   The missing “ISC_Camera.dll” file is likely the Indigo SC series Camera dll so without it the GUI cannot talk to the camera or even find it. That file is nowhere to be found on the CD’s or installed software

So at the moment I have a working camera, working software to drive it, a non working advanced configuration GUI and a missing “AccessorySDK” program. The good news is that the GUI is not essential for use of the camera. The Researcher and ResearchIR softwares are intended to provide full configuration and control of the SC series cameras. They are just not as user friendly in doing so and some unusual camera settings that can adversely effect the cameras calibration are deliberately absent. I have not given up on the GUI and will likely contact FLIR requesting the required programs for direct configuration of the camera. That will be a very interesting conversation with FLIR me thinks !

I have worked out the full history of the camera now. I discovered the original owners company name and found all the details of its closure on the Companies House web site. They went into voluntary liquidation in 2013.

Fraser