Yet another cheap thermal imager incoming.. Seek Thermal

[mikeselectricstuff]

I saw a mention that it reduces lag

[JcDenton]

I saw a mention that it reduces lag

Yes, it seems to be a bit smoother now.  It also fixed a crash that I used to get when using the recent apps key, and trying to relaunch the app and getting a black screen.  Overall it feels much better.  Still issues with the prompt screen to set as default camera, would be nice not to have to click that every time, but I'm just glad they are still making improvements.

[gammaxy]

I'm a little late to this thread, but have read most of it.  Does anyone understand how Seek fixed the gradient issue?  I watched mikeselectricstuff's video and understood the gradient was due to the lens and probably not the shutter.  Calibrating out the gradient seems like it would be difficult without a manual process where the camera is pointed at a scene of uniform temperature.

This does not seem to be the case.  Any method I can think of to detect and fix gradients could end up filtering out an actual gradient.

Is there only one or a small number of possible gradient patterns and they detect which one it is and filter it out?

[Fraser]

At the moment only SEEK Thermal know exactly how the gradient issue was resolved or should that be concealed    There is an additional image processing file added to remove the gradient issue from the displayed image and this suggests some smoke and mirrors processing that counters the gradient using maths. For me this is disguising an issue rather than dealing with the root cause. A 'band-aid' solution. What the 'processing' actually does is not known but it may or may not have a negative effect on some target scenes. I certainly noticed a mottling effect on the image after the 'fix' was released. A further reduction in the image quality. The correct approach will be for SEEK to solve the root cause of the gradient as I believe it is hardware and not software related. That may take time and money though.
As I have previously stated in this thread, I find it very disappointing that SEEK even released a camera with such a gradient issue. Not what I would expect of a company looking to make their mark on the thermal camera market. it wasn't a minor gradient, it was a socking great one on some cameras. Then again, as has also been stated, what should we expect for $200 ? I am saying nothing 

Aurora

[Seeker]

I haven't had the time yet to ascertain if the update has resolved the flakiness of the app (it deciding (at various points) to be uppity about working contunuously without repeated reboots etc).
I have noticed the "bong" alarm (which only used to sound when it wanted a reboot) now sounds whenever you exit the app, and also when you unplug the unit. Is more "bongs" better? Time will tell.

[eneuro]

I wonder, why did they choose this split screen solution?

This mode is useless.
I wonder, why they didn't made another mode with thermal image at the centre at 1/10 size of whole screen and fill rest of the screen with resized calibrated visual image with adjustable diameter of thermal image  and fast toggle buttons or automatic mode to switch between themal and visual images

[WS-PI]

@Aurora
I used Google to find infos about microbolometers and gradient and there are a lot of papers available. All give the info that non-uniformity is a common problem for not TEC stabilized microbolometers. The only solution is using some clever math. Even FLIR removes the gradient by caculation. Apparently Seek had bad luck to release the software before the gradient was removed completely. Otherwise no own would talk about Seek and gradient.

[Fraser]

WS-PI,

Sorry I cannot agree with you there. I am very familiar with NUC and FFC and their operation. I repair and service thermal imaging cameras for fun 

The key difference is that NUC is set in correction tables at calibration and may be considered a 'course' correction for individual pixel characteristic differences. The Microbolometer is a tricky little array that is hard to keep uniform as its temperature changes. The FFC uses a shutter that passes in front of the microbolometer and then equalises the small differences in individual pixel outputs to produce a flat field. A sort of fine tuning of the microbolometer. It is VERY unusual to have the sort of gradient across the microbolometer that appears on the SEEK. Such a large and block type gradient would normally be cause by poor hardware design and that is not the purpose of NUC or FFC. They are there to correct the natural characteristics of the microbolometer.

If the gradient in the SEEK were a simple NUC issue, it would have been caught during calibration and corrected. Such would have been a simple fix and we would not have seen the number of SEEK cameras with a gradient issue over an extended period of production.

Please do not think that I have a grudge against the SEEK camera. I am positive on the product objective, disappointed with SEEK Thermal as a company, and technically incisive on the SEEK camera design, which is flawed. Engineers analyse designs and comment on such, that is all 

Aurora

[Fraser]

For those interested in thermal imaging equipment in general..... you may like to take a look at my latest acquisition...... a thermal microscope    Cost....GBP35 

https://www.eevblog.com/forum/testgear/thermal-camera-the-aga-680-thermal-15x-microscope-retro-tech-porn/msg616597/#msg616597

Aurora

[efahrenholz]

I know this is the wrong thread for this, and I will post it there too, but this thread has been on and off the rails quite a bit so shoot me. I saw this image posted in the Flir One twitter page. This image looks a bit...high res, for what it is a somewhat low res camera sensor. Even at the new claimed resolution, does this seem possible?

https://twitter.com/FLIR_ONE/status/564862767719403521

[frenky]

There is no way that tha image was made with wlir one (2).

It seems that it's real 640x509 themal image and no MSX needed for more details...
Just compare it to the blobs of colors on the flir one:

[PA0PBZ]

Apart from it being posted on the Flir One twitter account there is no suggestion at all that it was made with a Flir One. It's just a nice shared picture.

[efahrenholz]

So the jurys finding is that this image did not come from the next generation flir one, but a professional camera and the image was posted under the flir one twitter account in error. And this post is off topic.

I keep trying to decipher which photos flir have posted came from the next gen imager and not the older one. Is there such little visual difference that it's practically indistinguishable?  I thought doubling the resolution would make a major difference. This is a fairly significant increase in resolution.

[miguelvp]

So the jurys finding is that this image did not come from the next generation flir one, but a professional camera and the image was posted under the flir one twitter account in error. And this post is off topic.

I keep trying to decipher which photos flir have posted came from the next gen imager and not the older one. Is there such little visual difference that it's practically indistinguishable?  I thought doubling the resolution would make a major difference. This is a fairly significant increase in resolution.

The same can be said for thermal.com (seek) so it's not really off topic that much
on their web page, not all of their images come from a seek thermal unit and look like a higher resolution unit.

[JcDenton]

So the jurys finding is that this image did not come from the next generation flir one, but a professional camera and the image was posted under the flir one twitter account in error. And this post is off topic.

I keep trying to decipher which photos flir have posted came from the next gen imager and not the older one. Is there such little visual difference that it's practically indistinguishable?  I thought doubling the resolution would make a major difference. This is a fairly significant increase in resolution.

I saw this image -
Isolated the image below.

[nabla]

1) Software (v.1.4.02) seems really unstable on Android 5.0 Lollipop running on a stock Nexus 5. Usually it will immediately crash the first time I run it, but once relaunched it will stay running.

So the app work on Lollipop? Seek says it won't!?
I'm about to buy a Lollipop Phone and, of course, want to use the Seek Cam with the new phone

[frenky]

I have Android 5.0 on LG G3 and app works fine...

[nabla]

Okay thank you. Seems as if Seek-FAQ is outdated...

[Seeker]

I haven't had the time yet to ascertain if the update has resolved the flakiness of the app (it deciding (at various points) to be uppity about working contunuously without repeated reboots etc).
I have noticed the "bong" alarm (which only used to sound when it wanted a reboot) now sounds whenever you exit the app, and also when you unplug the unit. Is more "bongs" better? Time will tell.

Flakiness not resolved. Completely helpless with it until it gets in the mood to pay along. Final assessment: SHIT.

[ixfd64]

For those who aren't following the Mu Optics developments, there are good news and bad news.

The bad news: Mu Optics is no longer delivering their product. I can't say I'm surprised.
The good news: Seek Thermal and FLIR are offering a 10% discount to people who have contributed to the Mu Optics campaign. The Seek Thermal discount code is: MUOPT15.

More details here: https://www.indiegogo.com/projects/mu-thermal-camera-a-great-tool-to-save-on-energy-costs#activity

[miguelvp]

The good news: Seek Thermal and FLIR are offering a 10% discount to people who have contributed to the Mu Optics campaign.

The bad news: about that is that Seek Thermal used to be $199 and it's now $249, so they were better off purchasing it at full price a month ago before the price change.
The good news: you can get the new XR with the 10% discount since that wasn't available before, but I'm not so sure the focus lens is worth the extra $50, or really  $100 from the introductory Seek Thermal original price.

[cynfab]

And the other bad news is that the "coupon" for the Seek discount expired about 10 days after it was announced.

From the comments page:
"Coupon code “MUOPT15” is not valid. The coupon expired on Fri Feb 13 00:00:00 PST 2015."

Chances are that the Flir "coupon"  has expired as well, or will not be applicable to the "new" Flir One for android.

Mu ==> bad situation all around.

[c4wd]

Has anyone else had dead pixel issues with the last few updates? Starting with version 1.5.x I had a bottom eight of the screen containing dead pixels and now with the most recent update nearly 75% of the screen contains dead pixels and makes it unusable. Attached is a video of the Seek Thermal module on the 1.8.4 update, interesting enough when using the module on 1.3.1 it works perfectly fine without any issues. I've been in contact with the developers and the most recent update sounded promising, yet it turned out disastrous.





https://drive.google.com/file/d/0B9Zl4KisxaP8cENFN0I3XzFnU0U/view?usp=sharing

[eneuro]

I've been in contact with the developers and the most recent update sounded promising, yet it turned out disastrous.

Are there developers in this team-really? 
http://en.wikipedia.org/wiki/Test_case
A test case, in software engineering, is a set of conditions under which a tester will determine whether an application, software system or one of its features is working as it was originally established for it to do.

Too many mistakes, too often 

Hopefully, it is easy write own software for Seek toy...

[c4wd]

I've been in contact with the developers and the most recent update sounded promising, yet it turned out disastrous.

Are there developers in this team-really? 
http://en.wikipedia.org/wiki/Test_case
A test case, in software engineering, is a set of conditions under which a tester will determine whether an application, software system or one of its features is working as it was originally established for it to do.

Too many mistakes, too often 

Hopefully, it is easy write own software for Seek toy...

Agreed. I cannot wait until the SDK is released, if it ever is. It's a disappointment my camera has been nearly unusable for the past few months. I'm just confused as to why it worked perfectly fine on the 1.3 update, yet each update after seems to make it worse. Perhaps they contracted the original development to someone other than themselves and an inhouse developer isn't up to par.