@nshapiro and I are working on a card upon which you can place a colorimetric tube for detecting formaldehyde, and take a photo of it to record data. Nick sent me the above sketch, and I doodled a design idea below. I figured we could prototype this idea out and further develop it in this research note.
Nick - do you have a typed list of all the info we'd want to include? Here's a diagram of the tube itself which is quite nice, from the company which makes it, Kitagawa:
And the PDF where I found the graphic: 710_1.pdf
38 Comments
What about color and white balance?
Your example image shows an awfully strong red/yellow back light which will change hue of the colorimeter.
It might help to build this photo card around a colorchecker card or on top of a colorchecker card, with instructions to focus the camera at a particular point.
This might all be moot if the color changes along the graduated marks. In that case, even a grayscale image would suffice to demonstrate where the gradient changes color, although then you'd want to make sure you have contrast set appropriately. Contrast can shift gradients. Might want to build the photo card on a gray card to help with that.
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
Secondary thought:
Is the gradient blue and black, or white and gold?
lulz
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
Good questions Bryan. @nshapiro, I'd put those black boxes along the top of my sketch to provide color balance help, but Bryan's right; the tube itself could be enough in theory. Do you know what's needed? But they could still help to align the image, or even include a QR code.
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
We were talking about relative humidity being important (but accuracy not being a big deal) so including not just a thermometer but also a hygrometer would be useful. Something like this one (7$)
Reply to this comment...
Log in to comment
Sorry for the slow reply! I think my spam filter may have been hoarding emails from PL.
The color changes from a waxy yellowish orange to pink and it is not the gradient of the color that matters but how far it goes along a long graduated marks like @btbonval mentions.
I don't know anything about color theory. What color background would make the distinction between yellowish orange and pink most dramatic? I think we may need a photo of the card and a close up of the tube. So that would be four photos total. 1 of the tube and its PL specific serial number from a smart phone to prove that it was un-opened at the time stamp. 1 of the pump on a non-formaldehyde emitting surface (top of the fridge is a good one). 1 of the card and the tube and 1 close up of the tube. Is that too much?
I bought a cheap humidity gauge like the one attached to that thermometer and it has already crapped out. Mine was from petco and for reptile cages.
Here is a list a preliminary list of the entry fields we need on the card (I tried to break it up to look like @warren 's much better design):
A
B
C
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
@mathew I just ordered the temp/humidity meter you recommended. In terms of design maybe it can fit horizontally under the A, B, C columns?
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
What about a thermochromic temperature strip adhered to the card? https://www.google.com/search?tbm=shop&q=thermochromic+thermometer&gws_rd=ssl or isn't that good enough resolution?
Nick, I heard you found the therm. and humidity meters were too big? Are they not the same as shown in the initial photo?
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
I'm seeing 2-degree accuracy on some of those strips. are they cheaper than an alcohol thermometer? do they photograph better? buy some and lets see!
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
Here is what the thermo and humidity meter look like next to the card. We could still use an alcohol thermo, this one is just a bit clunky. Also I worry about the metal coil humidity gauge. The pecto reptile one i bought stopped working after a few days and I think they are vulnerable to jostling damage.
Reply to this comment...
Log in to comment
we would probably want 1 degree accuracy as the visual read out of the tube has to be multiplied by a temperature coefficient on a 1 degree scale. it would only be a 3% difference between degrees, but that is added to 10% variation in the entire method so we could chop that 3 percent out by having one degree accuracy on the thermo.
Reply to this comment...
Log in to comment
Hey Nick - can you post the dimensions, or just a photo of the thermometers you have with a ruler in the photo?
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
Hey Jeff, The card mock up is 5x8. The alcohol thermo and humidity meter is 9x 2.5 at its thickest point. I have 2 inch diameter coil spring humidity and thermomoters and one 2x4 inch digital meter with danglely probes. I wonder if we will be able to get actual 1 degree accuracy with the alchohol or coil spring. so maybe the thermochromic won't be reducing our accuracy. the other way we could go is to more expensive, accurate digital. The photo uploader isn't working for me now but will upload a photo as soon as i can get it working.
Reply to this comment...
Log in to comment
There we go! working now.
Reply to this comment...
Log in to comment
we should check the accuracy of the digital thermometer-- sometimes they give 3 sig figs but actually only do 2.
if we switched to a vertical orientation for the thermometer it would fit on an 8.5x11 sheet of paper.
Reply to this comment...
Log in to comment
OK, cool - i agree vertical orientation will probably work better, and OK, going to go for a letter-sized sheet; Nick, could you try to select final choices for thermometers (please feel free to order as many as you like, I'm assuming you're doing this with our research supplies budget?) balancing:
I'll start designing now based on the idea that the humidity gauge will be circular and the thermometer linear, and try to do a design that is a bit flexible. Which of the pictured ones are your preferences currently? I'm going to go with the combined therm/humidity sensor, and assume we can swap in smaller ones if we find them.
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
One confounding issue is where the card goes in library shipment. I'm looking at foam-lined reusable boxes we can package everything easily and minimize breakage or loss.
the Seahorse 300 case is the cheapest US-made box I can find that I think is the right dimensions. at 9.5 x 7.5 it is a little smaller than an 8.5x11 card.
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
I think the card could be folded in half, print outward, so the stuff pushes it flat onto the table.
Here's a draft, but I do think we could make it smaller if we can source smaller parts:
test-photo-card.pdf
Reply to this comment...
Log in to comment
This looks! great! what do you think the QR code would be? We might need there to be a solid color that the tube is placed on top of for consistency and contrast.
What color background would make the distinction between yellowish orange and pink most dramatic?
I've just been spending out of pocket but good to know there is a research supplies budget!
I'll get on the themo/humidity meters tomorrow AM.
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
Yes, maybe ask Shannon about the budget. The QR code just points at wherewebreathe.org/qr - we can hook that up later to redirect anywhere we like.
Good question about yellow/pink - one way to do it would just be to print very clear red and yellow in a very consistent pigment which is fade-proof. @mathew, do you have an idea for this based on what we've done for the Infragram card? I guess a fade-proof ink may be best; should we ask Charles?
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
It looks like most of the cheap hygrometers are of the coil and dial variety which I worry about the robustness of.
This one is temp and humidity for 11.39 (free shipping). The lab thermometers I saw were all very long.
http://www.tmart.com/HC520-Household-Digital-Thermometer-Hygrometer-with-Probe-Gray-Silver_p291407.html?cc=USD&fixed_price=hk_intl&gclid=CJ6WkK_2ssUCFcmPHwodQqoA2Q
The bigger the card the farther the camera has to be from the tube and the less accurately we can verify the reading. so compactness might be a priority for this one.
also my tube kept rolling off the photo card. in the original i had cut a sliver of paper to fold back into the open lip of the tube that held it in place with no added glue, but we could also just put a spot of two sited tape on there.
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
The tube ought to be taped to the card, I think, so you don't mix it up if you are storing the cards and tubes together.
+1 compactness for varying reasons.
Hiking thermometers are also very compact; i used to have one like this one:
http://www.amazon.com/Campmor-44082-Weather-Gauge/dp/
Also:
http://www.amazon.com/Coghlans-9712-Zipper-Pull-Thermometer/dp/B000ECQ9KY/
And a more compact combined one: http://www.amazon.com/Zoo-Med-Economy-Thermometer-Humidity/dp/B004PBB1JE/
And for digital ones, these look kind of promising! <5cm largest dimension:
$4.29: http://www.amazon.com/Digital-Thermometer-Temperature-Humidity-Hygrometer/dp/B00OFOD874/
$6.99: http://www.amazon.com/Vktech-Celsius-Digital-Thermometer-Humidity/dp/B00E4BOQR4/
Temperature Measuring Range: -20℃-+70℃ Humidity Measuring Range: 10%-99%RH Temperature Measurement Accuracy: ±℃ Humidity Measurement Accuracy: ±5%RH Display Resolution: 0.1℃(temperature), 1%RH(humidity) Powered by: 1 x 1.5V LR44 button battery ( NOT included) Size: 45*15mm (D*T)
Reply to this comment...
Log in to comment
we can put double sided tape down on the spot-- they remove the cover and stick the tube down. we already have a ton of 1/8" double sided tape.
Reply to this comment...
Log in to comment
double sided foam tape, maybe? any formaldehyde concerns in the tape itself?
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
Great. I've ordered the last two as they have the accuracy we need--will report back by the end of the week. way cheaper than the one i had found.
Sticky tape should be fine as we won't be pulling air through the tube when its on the card. it would take hours or maybe days for the formaldehyde to passively get into the tube. In an early iteration with a toluene tube i had some tape in the air intake system which yielded wildly high results but as long as we're not sucking exposed glue directly into the tube we should be in the clear.
Reply to this comment...
Log in to comment
if, for the library, we're going to mount everything in a box with a face plate, lets look at how we can integrate the tube's scale and the thermometer and hyrometer into the faceplate. also seen in: http://publiclab.org/notes/nshapiro/11-03-2014/diy-formaldehyde-test-kit
Reply to this comment...
Log in to comment
Discussed on the call today having a pad of note pages attached to the lid of the box in the illustration above. The last two thermometer/hygrometers slot into square/round holes and sit flush, and are battery powered, so they wouldn't need to be on the card in this boxed version. Cool!
Reply to this comment...
Log in to comment
Small edits to the card:
we can delete the "within 2 degrees" helper text.
we can reduce size (thus increasing detail of tube) of the card
zip code is missing (or some sort of location)
Have we figured out the color scheme that would most make the transition from waxy orange to pink most apparent?
I found a black background to be better than white or grey. maybe we could have the nice detail of the tube with an arrow pointing to a blacked out section of the card that would have the sticky tape on it? or we could try a black silhouette of the tube? but we might need something bigger than that so that there is contrast of the whole visual vicinity
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
Nick: I am not disputing your background observation but why do you think the vendor says a light background and compare it to an unused tube?
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
I'm not sure I can tackle the complete redesign on a smaller card just yet, as I am just on my laptop this week. But I can change the text.
Nick, if the vendor says light background, maybe we should stick with that? Can you upload some comparison images? Perhaps a black background helps with contrast, but could we place it on a white background with a block of black nearby so that the camera exposes well?
How transparent is the tube? Is it a problem that I've reproduced writing on the tube; will it show through the tube and cause confusion? I could make the details very very light grey so they are just a guide...
Re: location, how anonymous do we want to encourage people to be? A photo of this card from a smartphone will already have pretty exact GPS location, which If we encourage people to upload, we may want to scrub or de-res by removing significant digits of GPS (or follow best practices for de-res'ing location -- this is a good question for @jlev, actually).
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
Here are the test photos on an old tube. Its not a perfect test because the rest of the tube may not be the same substance as the rest of the tube as it is not changing color after a month of being open but this should still be pretty close:
Reply to this comment...
Log in to comment
Maybe I was wrong! The glare seems to be worse on the back background. Which looks best to y'all?
only the outer two mm or so is transparent so maybe best if there is no writing under that part, but I think color matters more than writing. Light grey as a guide would probably work.
For water quality monitoring rounding down to three decimals in lat/long is the standard i think. but we could perhaps just wipe that field altogether? Or people could have the option of leaving the gps in as long as we explain it well enough.
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
I see the most contrast in color in the first two images. I'll see about a new version after the staff call in 20 mins.
Reply to this comment...
Log in to comment
Hi, I wasn't able to edit the tube graphic very well on my laptop but here are some of the other changes; I hope it's good enough for your immediate purposes?
test-photo-card-2.pdf
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
This looks great! Thanks! The thermometers arrived today. Will bring them in and photograph them tomorrow.
Reply to this comment...
Log in to comment
The round meter appears to be inaccurate. here are some test photos:
Reply to this comment...
Log in to comment
I think the sensor is at the back so we might want it to be popped out of the case so that we don't just pick up the heat from the pump.
Reply to this comment...
Log in to comment
Can you test filling one out and taping a tube to it, and photographing it in landscape? For the video. Maybe on a carpeted floor or a wood table, in diffuse (morning) light? Thanks!
Is this a question? Click here to post it to the Questions page.
Reply to this comment...
Log in to comment
forgot to add a link-back here: I made a smaller version of the card on the formaldehyde kit for borrowing:![card](https://i.publiclab.org/system/images/photos/000/010/430/original/IMG_20150624_221017.jpg)
download pdf
Reply to this comment...
Log in to comment
Login to comment.