Yet another cheap thermal imager incoming.. Seek Thermal

[Uho]

I painted with black paint. There are no changes.

[WS-PI]

@YewSuck

are you able to capture videos with the galaxy SII ? I tried that before with a cyanogenmoded SII and everything worked fine exept the video function. I tried also the omnirom without success.

[z01z]

@efahrenholz
There's no gradient for me for at lower temps either, see the attached pics.
It's a bit strange, because on one of the pics 5 deg difference is shown but there's no gradient visible. Maybe sensor sensitivity is less at lower temps, causing more noise.

My guess is that they use the sensor readings taken while the shutter is closed for an additional correction value. Say one pixel reads 30 C when it should read 25 C, they will know how much offset is needed to correct that pixel.

[WS-PI]

I did also some measurements. I placed a heatsink for transistors in our freezer for 1 hour. With an IR-thermometer the temperature measured was -18.5 deg Celsius. With the Seek I got only strange measurements. On one picture the high/low detection was enabled and on the other the point measurement. As you can see it gives different readings for the same point on the heatsink. Another strange obersavtion I made is that even if I move the Seek the max value of 0 deg stays stable in the center of the screen while the min value showed some variations and also pointed to different Locations.
For measurements at low measurements the Seek is of no use. With this behaviour the missing emissivity setting is not a Problem.
Now I start thinking what Seek is meaning with "-40C to 330C detection" in their datasheet. Apparently detection and measuring is not nearly the same.

Video of measurement while moving the Seek
http://youtu.be/boTx3c_PlLo

[Fraser]

The FLIR E4 also has problems reading low temperatures. This may be to do with the calibration processes used during manufacture.

Aurora

[efahrenholz]

I found this link that compares the e6 to the seek. The author does some excellent side by side examples, and shows that the higher resolution of the seek (in theory it's higher) still looks awful. Whatever flir is doing to provide clean noise free images, it definitely means that new flir one coming out this year will probably kill the Seek. Gradient or not, the image output looks like early 80's technology. Fyi, it hasn't circulated much but a spokeswoman at the flir booth did say they were shooting for a $200-$300 price range. Id even buy it at $400.

http://www.homesmsprealestateblog.com/2014/12/seek-thermal-infrared-camera-a-199-ir-camera-for-androids-and-iphones.html

[sgken]

Hi folks,

Had to join so I could add my 2 cents worth. Have read the whole thread since I got mine back in mid December.

Golly, it seems that some of you would complain if you were hung with a new rope! https://www.eevblog.com/forum/Smileys/default/grin.gif

I don't suppose you complain that your creator gave you eyes that aren't photometrically calibrated and that in spite of an overall field of view of nearly 180 degrees that we only see clearly for less than 2 degrees of that. And there are many many other limitations of the eye/brain. But, there isn't any technology that can exceed them in all factors. Fact is, they work very well for their intended purpose. So does the Seek.

With an accurate IR thermometer, how long would it take to map the 32136 points that the Seek can do in an instant? Yes, it may be more accurate except in the time it takes you to make those measurements, the first ones have likely changed. With any imaging IR device, you would see those changes as they occur.

That I can see in a glance a hot circuit breaker, leaks around windows and doors, an overheating electronic component for $200 is really amazing. As Aurora mentioned, a lens for one of his other cameras cost more that the Seek.

Does a fireman really care that his view is a few degrees off? No, he needs instant information of where the hot spots are and an imaging device like the Seek fits that requirement. In other words, for precision, yes, spend that 50K. For seeing the unseen- get a Seek!

Was I disappointed that there was a gradient when I first got it? Yes but it still could do the basics I was hoping for. Now that the gradient is gone, it's even better.  Is it perfect? Na. But, again, I can see something I wouldn't have been able to with my budget. Temp wise, I haven't checked it's absolute accuracy but it seems within reason I.E. usable. My room is about 70F, my heater around 425F. Yup, about right.

As I see it, emissivity is something important for accurate measurements. As most scenes are of mixed emissivities, which one do you choose? If you choose Highly Polished Brass at an emissivity of .03 what's that going to do for the rest of the measurements? Most emissivities seem to be lumped at the high end, so is it really that big of a deal for casual use?  You would have to keep changing emissivities to accurately measure everything in the scene. Do you really need to know accurately what the temperature of that bear's nose is or is it good enough to know that he is there?

Fact is, I didn't get it to be a precise scientific instrument, I got it to see the world of IR, looking for leaks, hot spots and just plain fun. It meets all those requirements.

As to noise, I have seen it vary. I don't know the parameters but I would expect more noise on a narrow temp range. I have seen it produce some sharp images and some fuzzy ones. If they improve it, that would be great. If not, I'm still very happy.

Thanks for listening, no flaming intended, just asking folks to lighten up a bit. In spite of it's limitations, this is an amazing thing we hold in our hands!

Ken

[eneuro]

It's a bit strange, because on one of the pics 5 deg difference is shown but there's no gradient visible. Maybe sensor sensitivity is less at lower temps, causing more noise.

Maybe Seek "improvement" is.... adding more noise in their software.
This is easiest way to cancel human voice in open space environments I used while working in a big company a few years ago-simply wrote ocean waves sound simulator in OpenAL and noise created by artifical waves from random positions made human voices present around in open work space  were not possible to understand, so not a problem no more 

Those images from their software looks still really crappy 
We'll see, maybe they will be forced now to go down with price for Seek Thermal and at price below $100 there were no complains while it could be a quite usefull toy easy to connect to PC using own software.

[efahrenholz]

@Ken,

For the non commercial use of the device, it's fine. It does give you a visual image of the scene. There's no arguing that. It's cheap at $200.,it gives you instant digital output, different palettes, saves images and video. It's even about 9hz in frame rate, so it's fast enough to use on moving objects.

However...

This is an electronics engineering forum. People here are the most critical of any test equipment. Yes, many of us have questioned the capability of the camera. We have examined the sensor, lens, software, to determine it's usefulness in a serious manner.

The overall verdict is meh. It's fair. The problem is the sensor design. They had to blank out pixels to be able to even compete with flir patents. So the actual claimed resolution is much lower. Seek came in here and even stated that it wouldn't matter because the wavelength was larger than the pixel size. So, in theory, any pixel that is bad could be getting the same radiation a functioning neighbor has. In sound, yes that's probably somewhat right. However not only do the sensors have blanked pixels, but they are *plagued* with dead/stuck pixels. Mine is horrible. The raw image looks like bad analog tv reception. But hey, we can fix it. Run per pixel averaging. Well that works, until you get groups of bad pixels polluting the average. It also increases the overhead for processing, which slows down the image display time. My camera has to do so much post processing that my frame lag is about 3/4 second. That gets disorienting very quickly.

But it's $200, all these issues should be acceptable. And to a degree they are. But the claimed resolution was deceptive. If a pixel is supposed to give real data, its included in the resolution count. Not imaginary or guessed data. Then there was the image quality. They pulled a trick used by the night vision manufacturers like bushnell or ATN. They show simulated product images which use a very high quality camera, or just image processing to make it look real. That's bad marketing. Just check the very first page. It's totally bologny. These things kind of burned me. If it was $400 I would have still bought it. It wasn't the price point. It was the form factor and stated qualities. And yes, I'm aware of Therm-App. And I'm considering that camera along with the new flir one. I kinda don't like how the therm-app is mounted to the phone; Seek appeared to be the holy grail since it was so small. Sadly, this is an unholy lemon.

I hope you are happy with the purchase, if it suits you. I am just going to call it what it is. I'm not pissed about the money spent. I'm pissed with the time wasted, and the deception. I hope they eventually fix the image problems but it's probably not going to improve much from here. It took them awhile just to fix the gradient, but they removed emissivity in the process.

[kanu]

Hi all, been lurking here following along since i got my seekthermal in mid October order 1200ish or so.

I Had about 4C gradient only after 10 minutes so i guess i was one of the luckier ones and as others i see no more gradient with the latest update.

came to point out this little bit of info on 9 to 5 geek from the 5th of jan from a CES reporter, appears to be a physical body refresh with no circuitry change is happening "soon".

has just shown a second version of the temperature-sensitive camera that’s set to be released in the next few months. Currently unpriced but planned to sell for a small premium over the original model, the Seek Thermal Camera with Zoom will be capable of zooming in up to three times by using a twisting front lens

http://9to5mac.com/2015/01/05/ces-2015-olloclip-cpl/

[ricksastro]

... But the claimed resolution was deceptive. If a pixel is supposed to give real data, its included in the resolution count. Not imaginary or guessed data.

Disagree here...it's exactly analogous as cell phone cameras claiming to be 24 mpix (because that's physically how many pixels there are), but the real resolution pales in comparison to "real" large chip 16 mpix cameras because noise and diffraction impacts those tiny pixels more.   And even full frame cameras don't live up to their "resolution" claims at higher ISOs.   Noise kills detail, that's a fact that won't change.

Seek has tiny pixels.   They are going to be far noisier than the larger sensored thermal imagers, thus reducing the comparative resolution.  It it also cheaper.

As with visible light cameras, those that understand these facts best are those who care most about them and are willing to pay more to mitigate the noise.    If you shop by price and "good enough", you are going to get what you pay for.   No different here.

Now that the gradient effect is no longer visible, I love the device for what I bought it for...comparative visualization of a thermal field.   I can get a good idea of differences in temperature.   I won't use it for accurate absolute temperature, but I honestly didn't expect a lot in that realm since it becomes much more of a cumbersome process (with emissivities and so forth).   

As an engineer, I do understand any frustration that it could be better...that was how I felt about the gradient.   But I also understand productization to a price point involves compromises and Seek hit the nail on the head for someone like me that doesn't necessarily know what I'm missing.

[YewSuck]

@YewSuck

are you able to capture videos with the galaxy SII ? I tried that before with a cyanogenmoded SII and everything worked fine exept the video function. I tried also the omnirom without success.

I can capture videos in the Seek app just fine, but they ARE laggy and a little choppy.  I had not used that function ever till today.   Thanks for making that clear.    My post was really to those that are having lagging issues and are running a faster new generation phone.   They got to much bloat ware running, because it works on the SII just fine on most functions.     Also a aftermarket Rom will probably add back the OTA function for those that do not have it. 
-YS


EDIT:   The video could not be played back on the device or my computer.   Sorry for the bad info.

[eneuro]

has just shown a second version of the temperature-sensitive camera that’s set to be released in the next few months. Currently unpriced but planned to sell for a small premium over the original model, the Seek Thermal Camera with Zoom will be capable of zooming in up to three times by using a twisting front lens

http://9to5mac.com/2015/01/05/ces-2015-olloclip-cpl/

That Seek Thermal zoom capability is interesting in anounced new version 
I wonder if they change also slightly lens holder mounting itself.
Anyway we found quite nity small custom shutter, so we'll try remove oryginal one and see what happends.

When new updated hardware version of Seek Thermal with this Zoom feature might hit the market?

[YewSuck]

I pulled this from another forum and thought it might be relevant to the topic since there were a few asking for mounts.  This a link to someone selling a mount.  Moderators if this seems inappropriate then please delete link.  I have nothing to do with this vendor.

3D Printed Dock for Seek Thermal. 
http://www.shapeways.com/model/2944122/seek-thermal-dock.html?li=shortUrl




-YS

[Seeker]

Quite handy! I don't need one though, I have no intention of dropping anything. The seller didn't mention the current difficulty in finding an extension micro-B OTG cable. Amazon list one on their SeeK page, but I've yet to see one on eBay (among the hundreds of others). The one I tried from there didn't work, and couldn't find anything to bridge in the bared male connector...just four pins showing.
***

[eneuro]

3D Printed Dock for Seek Thermal. 
http://www.shapeways.com/model/2944122/seek-thermal-dock.html?li=shortUrl

Since no plans to use Seek Thermal even in Android version mounted directly into crappy phone/tablet USB port but via extension cable, do not like at all this "Z" shaped zig-zag Seek Thermal.
This thermal camer acould have much more compact design if MPU and USB port were in the same axis with lens holder and Seek sensor back and put in some kind of tube/cylinder which might be more practical for custom designs.
I hope Seek will make such versions of this devices.
For the moment no problem we can cut USBport and solder USB pins directly to Seek PCB and make it more compact.

Zoom feature is much more interesting thing and waiting to test this  in circular motion enhanced setup   

[efahrenholz]

@rickastro,

I agree with you that the pixels are too small. They are literally just large enough to accept the radiation in the wavelength the material is sensitive to. I have a little experience with image intensifier tubes using microchannnel plates. Any excess generated electrons around bright sources of light will bounce off the edges of the channel openings and enter into the wrong channels. This causes a bit of sparkling around bright objects. I believe the same thing could be happening with the seek detector.

Frame averaging is basically the saving grace for the camera. But they appear to be averaging four frames, directly on the processor of the device. This number crunching, and any other tasks the chip has to do, is surely eating up processor time. If they moved the averaging task off to the phone, it might be possible to squeeze a few more frames per stack. As it is now, its possibly dropping frame speed in favor of processing. This would greatly improve the noise reduction, improve the temperature sensitivity, and they could regionally unlock the fps through where the app was downloaded (of course smart people will get around it.)

However, they decided to go the other route. I truly think it's all in the maximum achievable fps of the sensor which will determine if it ever gets a cleaner image.

[rjardina]

Quite handy! I don't need one though, I have no intention of dropping anything. The seller didn't mention the current difficulty in finding an extension micro-B OTG cable. Amazon list one on their SeeK page, but I've yet to see one on eBay (among the hundreds of others). The one I tried from there didn't work, and couldn't find anything to bridge in the bared male connector...just four pins showing.
***

God Damn! When it comes to anything smaller than my thumb I have butter fingers. I dropped mine about 3ft/1m and the case came apart and the lens broke off the PCB. Plugged it in and it no longer worked. My phone still recognized it at the time of dropping it but just got a black screen. replaced the ribbon cable and with no luck. I let it set for about a week until I got a magnifier ended breaking the D- pad off too. I know it's toast. But would it be useful to anyone at all for reverse engineering? Excluding the broken pad it looks fine under a 10X magnifier. I thought I would offering it to someone before offering it to the trash can.

[Fraser]

IMHO Mike is a most deserving recipient for your deceased SEEK as be contributes greatly to the community and will publish his discoveries in a quality video. Just my 2 cents.

At lest one person here has a broken lens so they may request your lens to use on their camera ?

It is very good of you to offer your camera to others. Thanks.

Aurora

[heavybarrel]

I fully agree with Aurora, the communities good friend Mike is very deserving, aside from being extremely qualified.  If he has the time, he is most likely the only person i know of who is busier than i am. Hats off to you for donating your device, admirable sir!!

[efahrenholz]

@rjardina,

If your lens mount broke free from the pcb, the sensor is likely toast. The bond wires are literally a millimeter from the left inside wall of the lens holder. Any slight movement and they get jammed up and short the sensor. I've done it twice and was able to repair it using a microscope. I did it a third time and broke 3 bond wires. Needless to say, it's now a $200 write off. I'm waiting  on the new flir one now, and watching development here of what I could have had, had I not jacked with my camera in the first place.

[PaulF]

I pulled this from another forum and thought it might be relevant to the topic since there were a few asking for mounts.  This a link to someone selling a mount.  Moderators if this seems inappropriate then please delete link.  I have nothing to do with this vendor.

3D Printed Dock for Seek Thermal. 
http://www.shapeways.com/model/2944122/seek-thermal-dock.html?li=shortUrl

-YS

Hi there - I'm the designer of the dock on Shapeways.  The dock was mainly designed to be attached to something else.  The back is completely flat, so it can be attached using adhesive or velcro to a phone or tablet case or something else.  For example, I'm planning to rig up a handle out of some PVC pipe to make it easier to hold.  It could also be attached to a longer pole to see things that would otherwise be difficult to get at.

If anybody has any questions, I'm happy to address them.  I saw one question about extension cables.  There aren't too many out there, but YCS on Amazon has both 6" and 3' OTG cables that I've confirmed work.  I've got the 3' cable attached to my dock, and it's quite handy.  A single packet of Sugru is plenty to attach the cable to the dock, and reinforce the strain relief at both ends.

I'm curious - would anybody like to see any different accessories for the Seek Thermal camera?  For example, I'm considering making an adapter plat that would let you mount the dock to GoPro camera mounts.  They have mounts for many applications, including helmet mounts. 

Thanks,

PF

[mikeselectricstuff]

Quite handy! I don't need one though, I have no intention of dropping anything. The seller didn't mention the current difficulty in finding an extension micro-B OTG cable.

A search for "micro USB extension" on ebay shows hundreds, though I don't know how many of them will be OTG compatible

[mikeselectricstuff]

I thought I would offering it to someone before offering it to the trash can.

Yes please!

[Uho]

efahrenholz @rjardina,
I did the tests and damage lens and casing. Yesterday damaged track. Restored under the microscope. I'd love it handy. I still plan on doing tests. I will pay shipping. I have the address both in Ukraine and in the United States. Please contact me.