(TAU/BOSON PN ID inside) Analog output from Flir Scion?

[hap2001]

Hi. I'm asking this for a friend.
Does Flir Scion provide analog/composite output?
Thank you!

[Fraser]

I am not surprised that your friend is confused !

The earlier LS and PS series offered analogue video output from either the bottom connectors or via a pin on the USB socket. FLIR provided the required video cable with the unit and offered it as a spare part.

I have not used the Scion but just researched the specifications etc and have come to the conclusion that the SCION is not capable of providing analogue video output. There is no mention of such in the specifications except a cryptic “Video” mentioned in the I/O spec. That “Video” would appear to refer to streaming video as a UVC device. This as opposed to downloading saved still images and video that are also provided.

I searched for video cables for the SCION and I only found the cable used for the LS and PS scopes. No mention of a video cable for the SCiON anywhere. The fact that FLIR did not provide a video cable with the SCION, or offer it as an accessory, re-enforces my belief that this BOSON based scope provides only USB based digital I/O.

Given a choice, I would, avoid products that use the BOSON core.The LS and PS used TAU/TAU2 cores and performed well. A good condition used LS/PS unit may be a better idea if analogue video is needed. There will be a difference in specifications however so worth checking that it meets any other needs.

Fraser

[hap2001]

Thank you Fraser.
He already have one, that's why he asked me this.
I always buy thermal cameras with "direct" video output, no matter it's analog or HDMI or what ever.
He just want to view it without a PC or sticking his eye to the tiny viewfinder. (And Wifi streaming isn't an option because no one like to click on a stupidphone in the field)
I found several "UVC to HDMI converter" products, but most of them are Chinese made, I'm sure he won't use them. But this is a good direction to go, maybe I should dig deeper and just find him a good quality signal converter.
Thank you again for your good answer!

[Bill W]

Given a choice, I would, avoid products that use the BOSON core.The LS and PS used TAU/TAU2 cores and performed well. A good condition used LS/PS unit may be a better idea if analogue video is needed. There will be a difference in specifications however so worth checking that it meets any other needs.

Fraser

Hi Fraser,

I was visiting the local RNLI last week and got talking.  They bought a SCION direct, and it is well received by them and seems to have become the accepted unit for RNLI use.  While they (Clacton) got theirs direct, HQ are supporting purchases for boats that are used a lot in cliff / beach search work.

Did not get to play though.

My main gripe with Bosons is the high temperature behaviour, but as an 'ambient' core maybe it is OK now, plus the advantage of 12µm pixels for narrower field use.
That said the RNLI folks are comparing thermal to searchlight / II.

Bill

[Fraser]

Thanks Bill 

My issue with the BOSON relates to the amount of ‘noise’ I found in the image. It reminded me very much of the Lepton in this respect. As we know, 12um pixels can be a challenge on the signal to noise front, yet other manufacturers have managed to cope with this challenge. The DRS Tamarisk used the novel approach of attaching ‘thermal collectors’ to the pixels to gain the best possible thermal capture efficiency from such small pixels. Some clever noise processing was still needed to make the 12um pixel core come up to the imaging performance of 17um and even 25um pixels. When I had access to a BOSON core I was really disappointed with the amount of noise I saw in the produced image compared to a Tamarisk core. Maybe it was an unfair comparison but I thought that FLIR’s use of that clever and powerful video processor IC would have been more evident in the image quality. Maybe the BOSON has improved since my test though ?

I am so pleased that the RNLI are now deploying more thermal imaging equipment. When watching the BBC RNLI documentaries I often wondered why there was not more use of thermal imaging and assumed it was cost related. That said a hyperthermic human in less than ideal seas, with only the head above water, can make a tough target for a thermal imaging system. I still think thermal imaging has much to offer though. Spotting casualties on cliffs, beaches or on top of buoyancy aids is definitely an area where thermal imaging can assist, be it handheld, boat mounted or on a Drone platform. The correct IFOV is important for obvious reasons and whilst a nice wide angle lens may cover more search area, it can seriously degrade the detection distance of the casualty. I hope the SCION is a good fit for the RNLI and other SAR applications.

Fraser

[Fraser]

I have just been advised by a fellow forum member that their current experience with the FLIR BOSON 640 has been very positive and it is providing good imagery. Maybe FLIR sorted out it’s issues ? I have not tested a BOSON recently so cannot confirm the imageing performance of current BOSON builds.

Fraser

[Bill W]

Yes, it is 'stranded on the beach / under the pier' where they said it was most useful, but surprisingly handy even for people in the water.


Back with e2v we did try the RNLI with the Argus TT and P type cameras around 2013, but did not get anywhere.

Info for those less familiar....
https://www.tequipment.net/ArgusTT320C.html
https://www.tequipment.net/ArgusPG320.html

Bill

[Bill W]

I have just been advised by a fellow forum member that their current experience with the FLIR BOSON 640 has been very positive and it is providing good imagery. Maybe FLIR sorted out it’s issues ? I have not tested a BOSON recently so cannot confirm the image ping performance of current BOSON builds.

Fraser

I am not so surprised that FLIR have been able to improve things on the sensor side, maybe not so much fixing anything software related on the Boson.

[Fraser]

With The BOSON being FLIR’s “new baby” I am also not surprised that they have put the effort into developing the core further. This is a wise move when faced with the new cores coming out of Asia that will likely continue to develop and challenge FLIR core sales.

Fraser

[ArsenioDev]

Boson+ has an entirely new ROIC stack descended from the TAU2, from what I've seen far better noise performance to boot.

[Fraser]

Very interested to hear this 

Fraser

[Fraser]

I just took a look at the BOSON+ 320 core specifications. Impressive NETD specifications and an Industrial version with a stated NETD of <20mK. That sounds like a very different core to the early model I tested 

https://www.oemcameras.com/thermal-imaging-cameras/thermal-imaging-cores/flir-boson-series.htm/22320H092.htm

Fraser

[hap2001]

Boson+ has an entirely new ROIC stack descended from the TAU2, from what I've seen far better noise performance to boot.

How can we distinguish Boson+/TAU2+ from their non-plus versions?

[hap2001]

Presumably by the model number/name

Problem is the PN decode section in the datasheet does not mention the + versions.

Edit: Per the latest documents, Boson can indeed be checked by PN, but TAU doesn't mention.

[hap2001]

I just found latest documents (tau/tau+/boson/boson+) have some updates, but each PN decode section are not complete. So I stitch them together and made 2 pictures, I'm posting them here to make you guys ID your core better.
I also noted Boson now have 320 + versions, and radiometry + versions will also release in 2024.

[Fraser]

Very useful. Thank you.

Interestingly, there is no detail of the TAU 80 (80 x 60 pixels) core that was used in the Bullard Eclipse 80 fire fighting camera. I have long suspected that the core is actually a TAU 160 x 120 pixel model with some tricks to restrict the produced image resolution to 80 x 60. It is not in a X2 eZoom configuration so likely pixel output combining to reduce output resolution.

Fraser

[Bill W]

Very useful. Thank you.

Interestingly, there is no detail of the TAU 80 (80 x 60 pixels) core that was used in the Bullard Eclipse 80 fire fighting camera. I have long suspected that the core is actually a TAU 160 x 120 pixel model with some tickers to restrict the produced image resolution to 80 x 60. It is not in a X2 eZoom configuration so likely pixel output combining to reduce output resolution.

Fraser

So exactly the same 'model adjustment' as the E4/E5/E6/E8, and no doubt the K33 - K65 fire camera series for that matter ?

[Fraser]

It could well be the same technique used in the TAU 80. Sadly I know of no way to release the cores full resolution and it would likely involve replacing an fpga .bin file with one from a TAU 160 to do it. Other configuration settings might also need to be changed but the TAU OS is not as modification friendly as that of the Ex series.

I may spend some time investigating the TAU 80 or I could just install a spare TAU 160 that I have.

Fraser

[hap2001]

Hmm, note that 168-34u and 162-50u marked as "effective pixel pitch"? Those have factory combined pixel groups to reduce resolution (to waste sensor).
And 80x60 is really a strange decision, for both Flir and Bullard. It's hardly usable.