Hti HT-301 384 x 288

[RBsonic]

I have question for all who use HT-301 with HTI Image App.

If I see on my phone I see this picture:




But if I make photo, the picture is longer on the right side.




So App shows a shortened picture.




My phone has 1920 x 1080 px  screen resolution with hardware buttons. Can you test and say if you have same situation.
But if you answer, please write your phone model and screen resolution.

[RBsonic]

No one can test it ??!!   

[Nogtail]

My phone has 1920 x 1080 px  screen resolution. Can you test and say if you have same situation.
But if you answer, please write your phone model and screen resolution.

I originally used the app on my now deceased OnePlus 3T (RIP) with a 1920x1080 screen but hardware buttons and didn't notice any problems. I'm now using the app on a Google Pixel 2 with a 1920x1080 screen, but software buttons which are always visible. I get a very noticeably cropped view in the app with half the logo obscured by the UI. Pictures and video come out fine though.

[Nicky]

Hi  we have updated the app of ht-301 , pls download it from our official website . or send it to you , my email  hti201801@hotmail.com

[Ambassador]

program v301-2020.08.18
Nothing new, only now you can hide the logo.
The cross with the highest temperature, twitches like crazy.

[svgurus]

So, noone tried putting  lenses from thermal expert q1 on hti 301 or iray t3s? the sensors seem to be of same size

[RBsonic]

I tried to screw out lens (HT-301) to reduce focus distance, but it seems that lens has a locking inside.

[zorek]

so you couldn't get the lens off? At the beginning of this thread someone posed pics of the lens/mount

https://yadi.sk/d/cKbfsocK7y2jZg

I'm also interested in a wider field of view, but it looks like the et-q1 is the best option, unless there's a way to replace this lens

[kwukduck]

So i've been messing around with the HT-301 for a bit, i love the crisp view and good ability to see small variations in temperature and high frame rate.
However, measuring temperature is highly inconsistent and varies over time. Both with HTI app and Thermviewer.

When i just let the camera stare at one point it will change temperature over time, slowly dropping down 0.1 degree every second or a few. It starts at 28 degrees C and over the coming minutes it will show a drop to around 24. Until it recalibrates with a click (or triggered manually) and goes back to 27 again, starting the whole process again.
The difference with every reset seems to become smaller as well until about 15 minutes continuous operation, at that point recalibrating will set it to around 24 degrees where it will stay.

Is this normal behavior? Is it an effect of the camera having to 'warm up'?

I calculated the 'temperature correction' at the temperature it indicates after 'warming up', this seems most reliable.
But effectively this means the device is useless for any temperature measurements the first 15 minutes or so, is this the case with all thermal imagers? A fluke of this specific model or mine specifically? (I realize thermal imagers are not the best to get an EXACT temperature reading, but a decent estimate would be useful)

[Fraser]

That sounds like normal behaviour for the sort of camera you are using.

When it comes to temperature stability there  are commonly three types of microbolometer used in thermal cameras. The fully temperature stabilised microbolmeter uses an integrated Peltier element to maintain the die operating temperature at the set level. There are microbolometers that contain a small heater instead of a Peltier element and these can heat the die to the set operating temperature but cannot cool it. Finally there are the non temperature stabilised microbolometers that contain neither a Peltier Element nor a heater. Such microbolometers are now common and operate in the “thermal equilibrium” mode. That is to say, the microbolometer die warms up due to self heating from it and the ROIC beneath it. At some point the self heating balances with the natural heat loss of the microbolometer and thermal equilibrium is reached. This is in no way a stable state as the microbolometer is sensitive to changes in ambient temperature around it. The solution to the non stabilised microbolometer temperature drift problem is the careful characterisation of the microbolometer in an environmental chamber to establish the effects of ambient temperature changes on its output. It is possible to create correction tables that counter the effects of ambient temperature change and the Flat Field Correction flag (if one us used) may be used as a calibration reference as its temperature is known.

Why does your camera take 15 minutes to reach thermal equilibrium of the microbolometer ? There can be many reasons for this but commonly the time to stabilise can be a variable dictated by the microbolometer design, its mounting and the Delta T that is present around the microbolometer module in the period after initial switch on from ‘cold’. In the case of the dongle cameras, the support electronics will be producing heat and this will create a Delta T around the microbolometer until thermal equilibrium is reached within the Dongles casing. Only then can the microbolometer also reach thermal equilibrium. As mentioned, it is possible to measure the ambient temperature around the microbolometer and apply corrections to measurements..... but how well that works depends upon how well the camera is designed.

The lower cost thermal cameras are not well known for temperature measurement accuracy and their manufacturers sometimes make it clear that they are not accurate Radiometric thermal cameras. Some models of dongle thermal camera may be more accurate than others depending upon how the ‘warm up’ period is handled by the firmware etc.

To answer your question...... no I do not think your camera us faulty. Yes I think it is behaving like many other such cameras.

Fraser

[bap2703]

All uncooled thermal imagers discussed here are bolometers = a thermometer that is heated by the infrared flux.
Being a thermometer it's just sensitive to all temperature variations such as the one caused by its own operation.
That's a intrinsic feat of thermal imagers.

One way to mitigate that is to calibrate the camera at many operating temperatures, but that's expensive. Another way could be to regulate the temperature of the camera, that's also expensive (and power hungry). Having a temperature feedback loop as close as possible to the sensor is also an option, guess what, that's expensive 

Not all thermal imagers are made equal and that shows in specifications : the temperature uncertainty will range from 1% of reading +- 1°C to 5% of reading +-5°C. The most common being 2% +- 2°C like your HT-301. (and I suppose that's optimistic / applicable only under "normal" conditions)

Burnt by Fraser 

[Fraser]

It is also worthy of note that when I discussed the temperature measurement accuracy of the FLIR E4 with FLIR they confirmed my belief that it is operated in a non temperature stabilised thermal equilibrium mode. The microbolometer die rises in temperature to around +32C through self heating and at that point thermal equilibrium is reached. This can take a few minutes to occur and is in no way a stable state ! The camera monitors the ambient temperature around the microbolometer and applies corrections to measurements. The accuracy of the measurements decreases the further away you go from the ambient temperature at which the camera was calibrated. This is because the correction tables are less than perfect in The E4  FLIR consider the Ex series “accurate enough” for the intended market segment at the most common ambient temperatures.

More expensive thermal cameras either use a temperature stabilised microbolometer or they employ far more sophisticated temperature correction tables that necessitated extensive development and environmental chamber testing. Such efforts are not normally justified for ‘budget’ cameras. The FFC flag can be a very useful tool to the camera designer however. If you know the surface temperature of the FFC flag and use it as a Blackbody reference surface, you can correct measurement errors in the cameras measurement system, at least for a short period after an FFC event. This is why the FLIR Lepton uses an FFC flag to improve its measurement accuracy over the version that has no FFC flag. Any budget thermal imaging camera that does not incorporate an FFC flag will likely struggle with measurement accuracy and even those with such a FFC flag may not be very accurate due to other factors influencing temperature measurement performance, such as the available Emissivity offsets.

Fraser

[kwukduck]

@Fraser
Thanks for the reply, that makes sense and is what i thought is going on too. After this 15 minutes it's quite stable and useable, i've not done many tests yet where accuracy is concerned once it's in 'equilibrium', when i do i'll report back my findings here.

Can you explain how the calibration click sets it back higher again? Why wouldn't it just continue drifting down to the 'real' temperature from before the calibration click as it's getting closer to equilibrium?
I'm not even sure what this calibration click does exactly or what triggers it to do so automatically.

[Fraser]

I regret that without knowing more about how this dongle camera deals with its temperature change compensation I cannot say why it behaves exactly as it does. The 'click' is the FFC flag solenoid bringing the FFC Flag into the optical path in front of the microbolometer.

The FFC event is triggered more often after a cold start in order to manage the output of the microbolometer as it warms up. The FFC events become less frequent after thermal equilibrium has been reached. From that point on the FFC event management works to maintain a flat image field and also to check measurement accuracy against the FFC flags known temperature. Corrections are then applied to provide the best performance possible, or adequate performance in the case of lesser cameras. The FFC flag event management system can activate an FFC event at a fixed interval dictated by the manufacturers product profiling or it can dynamically apply the FFC in response to ambient temperature change limit or detected excess drift in the microbolometer die temperature. Some cameras use a combination of both fixed and dynamic FFC event activation.

Fraser

[Angel1024]

Fraser,kwukduck,bap2703
I have same behavior on T3S, i using ThermViewer app with it and this app can lock colour palette for maximum and minimum readings. Factory app, Xtherm Infrared and HTI Image both working with this camera but no have this function. I think ThermViewer will work with HT-301 too because soft for HT-301 works fine with T3S.

The body for this camera is maded from aluminium and its perfectly conducting heat and cold too. Microbolometer requires a constant temperature whenever opearting, in "serious, professional" measurment devices using thermal regulator such as Peltier element to keep microbolometer temperature constant, but this camera designed for use as cellphone gadget and add Peltier regulator inside it will drain cellphone battery very quck. And camera will be bigger than now. Make thermoisolated case also affect on mobile gadget size.
But we can provide conditions to use this mobile phone toy as measurment instrument by keeping in mind easy "feature" of this camera:
this device very sensitive to ambient temperature changing because it have non thermo isolated aluminium body and no internal regulator inside. Whenever you adjusting focus your fingers warm up lens and body of camera and microbolometer inside camera will change own temperature too. As result - measurment readings will float (Accuracy of 2-3-5% and pro\not pro its just an marketing tricks, this kind of cameras all just are usb cellphone toys with app )
So if ambient temperature is constant and device is warmed up, thermal equilibrium has been reached, this will measure very accurate. But if we touch it with warm hands or cold something or go outdoor in winter it needs to reach equilibrium and need wait some time, more temperature difference - more longer time to wait equilibrium.
And lens i think need to try cover with something not so thermalconductive as aluminium, maybe plastic ring, to adjust focus without fingers temperature interference, or maybe work in winter gloves  it's not bug, it's a feature.

[rasselfast]

Received my copy of HT-301, disappointed with the quality of the sensor. My copy has visible image defects, 2 pixels are damaged. These pixels are very visible and interfere with your device. Each pixel has a secondary illumination around itself, I think it is dust particles directly on the surface of the temperature-sensitive element of the sensor. Dust particles on the surface of the protective window will not give such a bright effect, they will be out of focus. I have schematically depicted the proposed sensor device and the location of the dust particles (picture attached).

The very next day, I returned the device to the seller, yesterday I received a refund for the value of the goods from the trading platform, but the customs payment in the amount of about 80 USD was lost irretrievably. I will re-order, but from a different seller. So the quality of the HT-301 is a lottery?

[Vipitis]

the dead pixels get their halo due to sharpening. But you should normally have them mapped out on a fresh device. Not sure if you can do a dead pixel map by yourself.

[rasselfast]

While I had the device, I connected it to my computer and got a picture without interpolation and found 2 more dead pixels (2 black dots). The dead pixels are invisible when using the official phone app. Probably, when the device is turned on, the application creates a map of dead pixels and they are masked when displayed. The other two pixels with halo effect are clearly visible in the application, the reason is not broken pixels, but dust. Maybe I'm wrong. In the picture without interpolation, the halo effect with a radius of 1 pixel is clearly visible, without interpolation, each pixel is independent of the neighboring pixels, this suggests a parasitic illumination of neighboring pixels from a dust particle. The official application interpolates the original image by increasing the size of the defect to a radius of 7-8 pixels. I was very sorry that I had not filmed the video, it was very clearly visible there.

[zrq]

I don't think <10 bad pixels are something that worth complaining about given the typical normal pixel operability of 99.6%
 for the FLIR cameras on datasheets. You may point the camera to the clear night sky (~seeing 4K cosmos background +some scattered radiation) and you can see some if not all the dead pixels, even masked.
You can try to reach to the manufacturer and I'm sure they have some way to mask the defective pixels in the firmware (maybe by sending some unknown Zoom channel commands, please share if you know). I'm 100 percent sure about this as mine also developed 2 dead pixels in the first 2 months of its life, and I successfully returned it to the manufacturer to get these pixels masked.

[rasselfast]

I have no complaints about broken pixels, I saw them only after I saw the picture without interpolation. They are invisible when using the official app. Dead pixels are most likely calculated automatically when the device starts up, for this it is enough to compare a couple of frames and find identical pixels - the map is ready. A healthy pixel makes noise, while a dead one gives the same value. In this way, you can hide newly appearing dead pixels and there is no need to contact the manufacturer. But this is unreliable information - these are my assumptions.

But there is a problem: these are spots on an already finished picture after all post-processing. In the screenshots, these are spots with a diameter of about 15 pixels that hang motionless and constantly change their color. Located one on the top left, the other on the right below the logo. In post #174, the color screenshot has two orange spots. They are very visible and irritating. This is a clear visible defect that is not hidden, but enhanced. As I wrote earlier, I am inclined to believe that this is a dust particle that is located directly on the temperature-sensitive element of the matrix and has a size comparable to the pixel size. The dust particle "shines" on adjacent pixels to create a halo effect.

[All Seeing Eye]

My dog for T2search

[Odiug]

Hi All,

I received my HT-301 yesterday and also have this rather loose lens. That's annoying! It's so loose that the focus is not stable.

When I try to unscrew I will hit a resistance. I did not dare to apply excessive force. Is it just a "soft" stop and can I safely unscrew further?
Yes, I saw the pictures of a disassembled device, but how was it done?

I was considering to use one or two turns of teflon tape on the thread.

Regards
Guido

[Odiug]

I was curious and had to open the device. 

You need an allen key with a size slightly smaller than 1.5 (metric) to remove the screws on the back. Maybe something imperial. I filed down a 1.5.


You can then carefully wiggle the whole electronics and shutter assembly out of the housing. No need to remove any additional screws. That's how the back of the lens looks like:


What you see is a metal ring with two bend clips hold in place by some sort of glue. It's soft like hotglue.



For now I did not dare to remove this in fear of scratching the lens. 

Regards
Guido

[Skashkash]

Hi All,

I received my HT-301 yesterday and also have this rather loose lens. That's annoying! It's so loose that the focus is not stable.

 The threads on my HT-301's lens were also too loose and rough for my taste.
 
  My lens does unscrew all the way, but it did not occur to me to try Teflon tape.
  So I put a very small amount of silicone/ptfe grease on the threads.
   
    Helped a lot.   
 
  PS.  Thanks for the pics. 
 

[Fraser]

A word of advice..... never use a grease on a focussing thread that can either outgas or migrate. Avoid silicone based greases as once applied they migrate and can be very hard to completely remove. Silicone grease is an excellent lubrication material but it is very badly behaved when it comes to seating where it is initially placed. If silicone grease migrates onto the cameras lens out the microbolometer window, it is very bad news.

The lubrication of optical threads should be done with optical grade greases that do not outgas or easily migrate. For situations where a high viscosity grease is required, look at optical helical greases as used on camera lenses, binoculars and microscopes. Helical greases are commonly available and sound appropriate for this cameras loose threads situation. It is not an inexpensive grease so a group buy may be an idea where a number of people share the contents of one pot of grease.

Make sure to choose a higher viscosity helical grease as you are dealing with a loose thread engagement situation and not a precision helicoid assembly !

https://www.ebay.co.uk/itm/Japan-Hobby-Tool-Helical-Grease-for-Lenses-10-15ml-Helicoid-Repair-Lubricant-/123452055953

https://www.amazon.co.uk/Japan-Hobby-Tool-Helical-Grease/dp/B001500ZCC

https://www.micro-tools.de/en/Greases-Oils/Grease-Helical-10-Light-8ml.html

https://www.micro-tools.de/en/Greases-Oils/Grease-Helical-30-Medium-8ml.html

Google helical grease and you will find plenty of guidance from the photographic forums. I attach a viscosity chart for one helical grease maker. It came from here.....

http://forum.mflenses.com/helicoid-lube-question-t74606,start,25.html

I recommend reading the excellent post by ‘Gerald’ that appears half way down the thread page. He provides good insight into Helical grease behaviour and some good links to expert guidance.

The Japan Hobby Tool #30 Helical grease looks a promising viscosity for the task. There is the issue of focus drag however where a thick grease pulls the lens in the opposite direction to the focussing adjustment force when you release it. If this is likely with #30, go for #10.

Fraser