Resolution is not everything (Comparison Infiray P2 pro vs. Tooltop ET692C)

[Phil1977]

Hi,

due to the other Tooltop thread I was intrigued to compare the Tooltop ET692 with the Infiray P2 pro. The ET692 is a stand alone device in the usual gun-grip shape and pretends to have 256*192 thermal pixels. The P2 pro seems to be well known here.

Let the comparison speak for itself:

https://youtu.be/tV7L-qe1aR0

PS: Beside the side-by-side effect there was no processing on the videos.

[jackobo]

Always been wondered why this happens. is that hardware or software processing? Had the same effect with more expensive Hikmicro mini 2 Plus camera,with manual focus lens. Terrible lack of contrast whatever you do. Very similar to this tooltop, everything in the background messy and lit with this pinkish color.(((

[Phil1977]

I´m quite sure it´s the same main reason why there are good and bad visual cameras: The lens takes the image, the sensor is only a readout.

And far-IR-lenses can get very complicated and expensive. I'm quite surprised that the Infiray P2 is so good with it´s small lens assembly - but size is obviously not everything.

[Fraser]

There are many factors that can effect the image quality produced by a thermal camera. I will list some below.

If we consider the actual imaging cores to be identical in terms of build and core firmware.....

1. A basic setting can seriously affect imaging performance ..... FOCUS. Poorly set focus has been seen in other cores and minor adjustment can improve image clarity. The lens is supposed to be set at the hyperfocal distance for best focus over the greatest depth of field from half the hyperfocal distance distance to Infinity. In budget cores and cameras, the setting of focus distance can vary so it is worth checking.

2. A thermal core module is intended to be supplied with the required low noise power supplies. If poor quality power supplies are used, then noise can get into the cores analogue stages and degrade the image quality. A classic error would be to use the same power supply rail for both the digital and analogue parts of the core as digital noise can get into the analogue stages via this path. I cannot recall whether the analogue/digital power supply separation occurs inside the Tiny1C or is required externally.

3. Core manufacturers often offer noise reduction settings that an OEM may tune to their needs. This is an area where an EM may sacrifice image crispness for lower visible noise or there can be multi frame noise reduction that results in a lower frame rate and potential image smear to achieve a lower noise image. there are many ways to deal with the image noise that is inerant in microbolometer based imaging cores. This can mean that two cameras that use the same core produce images of differing appearance.

4. LCD panel quality and specification. There are many sources of LCD panels available to an OEM. The image quality produced by panels can vary greatly from one LCD panel manufacturer to another. FLUKE had an issue with this exact situation when one of their camera series suffered from a washed out, low contrast image on the LCD panel. A new panel supplier was found and the side by side comparison between the old LCD panel and the new one was "night and day". I will see if I can locate the comparison picture and add it of I do. Basically, not all LCD panels are necessarily the same in terms of image quality.

5. Saved image data compression can affect the quality of images that are saved for later review. It is not uncommon to look at a saved image on a laptop or tablet and think that it is not as clear as how you remembered it when you viewed the scene in real time. Despite the low cost of flash memory these days, some manufacturers still employ high levels of data compression when saving the image data. This can really spoil the image and is so unfortunate as the quality was present but literally thrown away in the save process. Viewing the relatively low resolution thermal images on a larger display such as on a laptop or tablet can also reveal deficiencies that were not obvious on the smaller screen of the thermal camera. be aware that many thermal camera manufacturers upscale/interpolate the thermal image data to fill a higher resolution LCD display. When saved to the Flash memory this upscaling/interpolation is not present.

6. Faulty production processes can also lead to issues with a cameras image quality. I have seen Cyanoacrylate glue (superglue) used near to the imaging core and the out-gasing of the adhesive creates a cloud on the cores lens, degrading its performance. SEEK Thermal had this issue when bonding their lens into the lens barrel on the 1st generation SEEK Thermal dongle. I also found it in a prototype camera that extensively used Cyanoacrylate within the casing... the out-gasing created a cloud over both the lens and microbolometer !

Well that covers the most common causes of image issues between cameras that use the same imaging core.

Fraser

[CatalinaWOW]

Fraser deliberately left out the properties of the core in his list, but one obvious difference that could differentiate cores would be the cell fill, the portion of the unit cell actually comprising the bolometer.  I have seen core designs with far less than half of the pixel actually contributing to the image, and others with well over a half.  The higher the percentage the harder to do, so an obvious place to cut corners for cost reasons.

If I were to prioritize the ideas presented here I would rate lens quality and focus at the top, and too close to prioritize one over the other, followed by raw core performance, contamination (Fraser's number 6 expanded to include finger oils on the lens and other such antics) and LCD module quality.

[globoy]

I cannot recall whether the analogue/digital power supply separation occurs inside the Tiny1C or is required externally.

Externally.  A 5V rail spec'd to max 200 uV noise over 1Hz-50KHz.  Digital stuff runs on a separate 3.3V connection.

[Fraser]

Further to CatalinaWOW’s comment on cell fill. I have always been impressed with how DRS approached the problem of the pixel support arms reducing the absorption area. They have what I call an “umbrella” mounted on the pixel that extends over the suspension arms to create the greatest thermal absorption surface area possible within the cell. I attach pictures of the DRS enhanced pixels.

[Bill W]

OK I'll add another similar factor - the vacuum.

As pixels get smaller (and so becomes more difficult to get high fill) a good vacuum is more critical than ever.  Any air in the package and the thermal differences on pixels get conducted away.
Used to be a lot worse with peltier coolers in there and even more stuff to outgas after manufacture.

As a story, at Argus we had a camera that developed a fault - and it got left on over the weekend.  It was run for 60 odd hours with the peltier on full heat !  Not only sorted the outgassing but also activated the getter and cleared it all out.  50% increase in sensitivity - so much so it was not useable in production !