thermometer calibration?

[Mike Anderson]

I'm ready to process film for the first time. I have a new thermometer. Does it need to be calibrated?

I've worked as a butcher, and we calibrated our thermometers by putting them in a cup of ice, and ensured they read 32 degrees. Is that how you do it in photography? Edited: can't do it that way with this thermometer, it doesn't go down to 32F.

...Mike

[Eric Woodbury]

No. Your thermometer could be fine at 32F and way off at 70F.

I assume you are doing BW and as such, it need not be perfect, just consistent, altho it should be within a degree F. I just bought a new dial thermo and compared it to all my others, some alcohol, my old dial, and a mercury thermometer I have left from college chemistry. The mercury has been my standard of comparison for 30 years. They all matched, even the cheapo alcohol. That's my calibration. So they could be off, but all off together. The new dial thermo was about $30.

[BarryS]

Hi Mike--If it's for black&white work, I wouldn't worry. Precision (repeatability) is more important than absolute accuracy. If you're doing color work, you'll need some sort of reference standard.

[Heroique]

Yes, a consistent temperature from job to job is what's key.

For example, if you like results when your thermometer reads "68 F," then you want to be sure that each time your thermometer says "68 F," you're getting a consistent temperature, whatever the "real" temperature happens to be.

What makes inferior thermometers so exasperating is that w/ each indication of, say, "68 F," you may actually be getting varying temperatures. Maybe 67 for one job, and 71 for another, even though your thermometer always says 68...

I used to use analog dial thermometers, because, well, I just like them the best. But then I noticed they had a tendency to vary w/ barometric pressure changes! Maybe the better dial thermometers don't do this.

[Doremus Scudder]

The standard way of calibrating your various thermometers is to have a standard, factory-calibrated, mercury thermometer and then compare your process thermometers to it. The usual base-line photo thermometer is the Kodak Process Thermometer. They are factory checked to be accurate to 1/4° F.

FWIW, if you have a thermometer that reacts linearly, like a mercury thermometer with etched scale or even a non-mercury glass thermometer from a reputable firm, the calibration method you mention is still valid, i.e., using a container of distilled water and crushed ice and seeing if the thermometer reads 32°F or 0°C. Using boiling water is less accurate, since the actual boiling temperature is strongly influenced by altitude and air pressure. If your thermometer is off at freezing point, it may still read linearly, in which case it simply needs an adjustment factor. This, however, would have to be determined against a known, accurate, thermometer.

Mercury glass thermometers are very linear as long as the mercury column has not separated and the scale is etched accurately at the factory. A spirit thermometer with an accurate scale and tight manufacturing controls is also fairly accurate (e.g., Kodak Color Thermometer). That's why a good thermometer from a reputable firm is needed for a base-line.

Unfortunately, metal probe/dial thermometers are notoriously inaccurate and non-linear. What I do is use a base-line thermometer that is reliable and has been checked with the above method (I have both Process and Color glass thermometers) and then bring a volume of water to processing temp. (e.g., 20°C/68°F) using the base-line thermometer. I then immerse all my other thermometers and either adjust them (some have rotating dials for calibration) or use stickers to indicate the error at processing temp (e.g., "reads 2°F over"). I do this once every 6 mos/yr.

This way, I'm assured of being within margin. As mentioned above, you only need to be really accurate for color work.

That said, for my B&W processing, I use a Zone VI compensating developing timer, which adjusts the "seconds" according to the actual temperature; faster "seconds" for warmer temps, slower for lower. This allows me to process at room temperature the whole year, as long as I heat a bit in the winter and keep the darkroom under 80°F or so in the summer. I just then have to make sure that freshly-mixed solutions and the wash water are at ambient temp before processing.

The main thing you need to be careful of with B&W processing is that the solution temperatures are close. Reticulation and other problems can occur if there is a too-large disparity in temps, say more than 2°C or so (I might be a bit conservative here...)

Hope this helps,

Doremus Scudder

[Jon Shiu]

Can thermometers be calibrated against a digital fever themometer in a warm glass of water?

Jon

[Doremus Scudder]

Jon,

The only issue in your suggestion is the accuracy and linearity of the digital fever thermometer that you are using. If these type of thermometers are accurate in the range they are designed to work in, i.e., around body temperature, then you could calibrate against them in that range. Whether such a thermometer reads accurately enough at 68°F or not is another question. Such thermometers may not be linear enough to be accurate at temperatures they are not designed for.

Maybe someone who is more familiar with digital thermometers can comment on the accuracy and linearity of fever thermometers.

Best,

Doremus Scudder

[Lenny Eiger]

Everyone else is correct that repeatability is the key vs accuracy. This is a good thing, every thermometer at the photographic level is garbage. I'll repeat: crap. I have half a dozen glass ones, including a Kodak Process, very old one, and they are all inaccurate - up to 6 degrees off (yes, including the Kodak fancy one). When attempting to purchase a real thermometer I was willing to spend even $80-100. I tried abut 6 of them , they were are supposed to be accurate to within .2 or something, but oh, that only meant at "points" which means 32 and 212. Near 70 they could be a couple of degrees off. A lot of false advertising. I spent a lot over time to get good thermometers and these folks piss me off.

I complained about this last year and a fellow sent me one that is perfectly accurate, probably about $450 or so new, I think. It's a sorry state of affairs. You can not imagine for a minute that the thermometer you are using is accurate, unless you've spent a lot of money.

Be happy about repeatability. At least that's there, and have your own development times for your film.


Lenny

[Bruce Watson]

Can thermometers be calibrated against a digital fever themometer in a warm glass of water?

There is no reason to think a fever thermometer is an accurate standard to calibrate against.

[Bruce Watson]

Everyone else is correct that repeatability is the key vs accuracy.

No, it's not. Repeatability and accuracy are two different and independent concepts.

A thermometer can exhibit excellent precision (repeatability) and horrible accuracy (read 20C when the actual temperature is something else, like 22C or 17C). Such a thermometer would be thought precise, but not accurate. It would work fine for photography once you calibrate your workflow like you would with the Zone System or other method.

On the other hand, you might find a thermometer that you can test over and over and over and find that on average it's very accurate. But it's not precise -- it takes many measurements (that all fall in a range of say 17 - 22C) to get enough data points to average together to arrive at the correct temperature. Such a thermometer would be thought accurate, but imprecise. It would be a nightmare for photography because it would be nearly impossible to correctly calibrate your workflow with a method like the Zone System -- every run would give you different results. Talk about frustration!

I'll take the first one over the second one any day for photography. Precision is my friend. Of course, what one really wants is an accurate and precise thermometer. And as you correctly point out, those can be quite expensive and hard to find.

Edit: Darn -- such a nice rant, and all because I couldn't read that Lenny and I are in violent agreement. Sigh... Sorry Lenny.

[Heroique]

Bruce,

I wish I’d read your distinction between “precision” and “accuracy” way back in HS chemistry class. My teacher only confused me about the two terms. You’re crystal clear.

(BTW, I think Lenny is making the same distinction. He says, “repeatability is the key vs accuracy.” I don’t think he’s equating the two…)

[Bruce Watson]

I think Lenny is making the same distinction. He says, “repeatability is the key vs accuracy.” I don’t think he’s equating the two…)

I think you're right. I didn't read that well -- my bad. Now where did I put that optometrist's phone number? :rolleyes:

[ic-racer]

I pitched my 1974 model year Spiratone dial thermometer last week for failing to be either precise or accurate :)

I keep a mercury thermometer as a standard but I need a new dial thermometer 'beater' for routine use.

Any dial thermometer currently available from B&H or Freestyle that is any good?

[Scott Knowles]

Lots of suggestions. All I can say is from my USGS days where we calibrated all field thermometers in the lab against a NBS thermometer (which I still have one) using a 3-5 point graph from an ice bath to near boiling water. This was done routinely with digital thermometers as they are the type most prone to error and drift, and when you're measuring tenths of degrees C, it's important to be accurate within 0.05 degrees C (half the measuring limit).

This isn't much help, except to take at least a three point measurement of a large range against some accurate source. And don't assume ice water is 32 degrees or boiing water 212 degrees. Good luck.

[jim kitchen]

I have two calibrated thermometers that are NIST traceable with certificates, and their cost was significant a few years ago, if not astronomical, where they are calibrated to 0.01 degrees Fahrenheit. One unit is digital, complete with matching calibrated thermistors, and the other is calibrated mercury in glass. I use these calibrated thermometers for scientific calibrations within an oil facility, a natural gas liquid facility, and a natural gas processing facility that require this level of precision.

One day I decided to test my black and white dial thermometer, to see whether the twenty-degree Celsius reading was accurate, and the calibrated thermometer read sixteen degrees Celsius.

Damn, and I was having such a great day...

Cheap thermistors can be plagued by inconsistent, lowest bid manufacturing, as would a cheap dial thermometer, where you get what you pay for…

That said, I wrote measurement manuals, and technical facility operations manuals for several major oil and gas companies around the globe, whether they were natural gas, natural gas liquids, or crude oil facilities, once we commissioned the facilities, and where I published a chapter that happens to be directly related to calibrating a thermometer, whether it is a thermistor or mercury in glass. If you wish to read a chapter that happens to be extremely boring, such as an economic journal in the bathroom, I will send the chapter to you via an email. I clearly state the required calibration procedure regarding a thermometer's calibration, but it is buried within the written chapter complete with a few images, and I will remove the company's corporate name and logo from the chapter's footer. The chapter also addresses measurement uncertainty to satisfy the truly anal. The calibration document is the standard within the industry. Contact me directly by email, if you so desire.

jim k

[neil poulsen]

I use an electronic thermometer meant for photographic use. Very consistent, it can be calibrated, it reads in tenth degrees, and it's quick to come to temperature. After using this one, I would never use anything else.

I have a Kodak mercury thermometer that's supposed to be accurate to within a half-degree, and I use this to calibrate the digital. I calibrate at about 70 degrees, because I notice that the slope on the digital thermometer can be off a bit. Within a few degrees of 70, it stays accurate to within a 10th degree. The digital also does a good job of maintaining calibration over a long period of time. I rarely use the Kodak. It stays in it's case in the drawer to reduce the risk of being broken.

I recently encountered three Weston dial thermometers; they were all off by by 3-5 degrees, and not necessarily in the same direction.

[ki6mf]

Make sure your thermometer is the kind that can be adjusted. These usually have a nut on the back that can be turned with a wrench. I use 3 thermometers. My main device is a waterproof digital that I use most of the time. I check all against a glass color thermometer and my Dial analog mostly stays on the shelf. I do regularly check each against the other using the Ice Cube and water method. I do use a water bath to control temperature for chemistry.

[Tom Monego]

The big SS encase Kodak process thermometer I have found to be the best. But my wife dye thermometer was good too and about 1/2 the price. One of my college teachers used to say buy three thermometer and use the one in the middle, same with lenses. The school would keep a kodak Process Thermometer in the "cage" so students could test their thermometers against it. Surprising how off some were.

Tom

[d.s.]

An old saying...

A man with one watch knows exactly what time it is.
A man with more than one watch is never sure of the correct time.

d.s.

[Mike Anderson]

Thanks for all the responses.

It seems the bottom line is you can calibrate a thermometer by comparing against a calibrated thermometer. You'd think photo stores would have some sort of calibrating service, not to any scientific level but just to know it's within 1 degree would ease my mind. With all the variables involved with film exposure and processing it would be nice to eliminate one area of slop.

Anyway, my fist film development seems to have been more or less successful.

Thanks again,

...Mike

[MelanieP]

Can thermometers be calibrated against a digital fever themometer in a warm glass of water?

Jon

--->You should instead calibrate it against the standard temperature for calibration which is 212 deg F. I think it's a more accurate way. You can check this post on calibrating thermometer (http://www.jesrestaurantequipment.com/jesrestaurantequipmentblog/calibrate-your-restaurant-thermometer/) as a guide. :)

[Brian Ellis]

. . . The main thing you need to be careful of with B&W processing is that the solution temperatures are close. Reticulation and other problems can occur if there is a too-large disparity in temps, say more than 2°C or so (I might be a bit conservative here...)

Hope this helps,

Doremus Scudder

Hi Dory - If you're saying that developer, stop, and fix should all be kept within 2 degress of each other, that hasn't been my experience and trying to do that can drive one crazy. When processing film I kept the developer at 75 degrees, which was the temperature at which my zone system film tests were run (no need to worry about 68, any temperature within reason is fine and 75 was easier for me to maintain in my Florida darkroom). Stop and fix could be anything from about 68 if the air conditioner was on to as high as about 85 if it wasn't. I never had a reticulation or other problem with those disparities.

As for thermometers, I used the Kodak Process thermometer you mention and it always worked fine for me. Of course at a cost of $100 or so it should have. I also used the Zone VI compensating developer timer for prints but not for film except during the brief time I tray processed.

[Ole Tjugen]

For all B&W work, I use a digital thermometer: My finger. I know exactly how 20 C feels on my finger, unless it's too cold or too hot to work anyway. It's accurate enough and repeatable enough that I get repeatable result from year to year, and that's good enough for me.

For colour processing, I use a mercury thermometer which was sent to me by mistake many years ago. I know I didn't order it, as it would be illegal to ship by air, and for the same reason I couldn't send it back. I haven't checked it against any other thermometer, but it does read within a fraction of a degree when my finger says it's 20C.

[Brian K]

--->You should instead calibrate it against the standard temperature for calibration which is 212 deg F. I think it's a more accurate way. You can check this post on calibrating thermometer (http://www.jesrestaurantequipment.com/jesrestaurantequipmentblog/calibrate-your-restaurant-thermometer/) as a guide. :)

I would not recommend using 212 degrees as the calibration point for a thermometer requiring accuracy at 68 degrees. Thermometers are not always linear in their response.
I would instead invest on a seriously accurate mercury photography thermometer, like the kodak color one, use it as a reference only thermometer to calibrate your working thermometers and then using that as the standard for all your processing. What is more critical to film processing is consistency over accuracy, assuming that you have taken the time to calibrate your processing based on your thermometer. Even if your thermometer is not dead on accurate, if you have calibrated your processing to it's inaccuracy, using test film and densitometry, you can still process dead on.

[al olson]

--->You should instead calibrate it against the standard temperature for calibration which is 212 deg F. I think it's a more accurate way. You can check this post on calibrating thermometer (http://www.jesrestaurantequipment.com/jesrestaurantequipmentblog/calibrate-your-restaurant-thermometer/) as a guide. :)

I presume that you are suggesting 212 F because that is the boiling point of water at STP. However, the boiling point changes with respect to changes in barametric pressure. If using boiling water it would be necessary to compute the boiling temperature based on corrections to STP. Not the most convenient reference point for most of us.

Further, barametric pressure changes according to elevation. For example, I live at 7500 feet and the boiling point is around 198 degrees, which could be a calibration point, but it is way outside the range of photo processing even for color chemicals.

[Drew Wiley]

Heck - up in the high country water will boil and you can put your finger in it. For the
darkroom I use the Kodak Process Thermometer exclusively. It's way more responsive
than any of the digital ones, a probably more accurate too. A wise investment.

[BetterSense]

I'm lucky to have a kodak process thermometer. I don't worry about calibrating it; I consider it The Standard. I also don't use it. I use two of these, which I calibrate against the process thermometer:

http://www.amazon.com/Taylor-9842-Commercial-Waterproof-Thermometer/dp/B00009WE45

I love these thermometers by the way. They are very fast responding, waterproof enough to dunk in trays, and were within a fraction of a degree of my Kodak process thermometer upon delivery.

[Grif]

[QUOTE=BetterSense;737508]I'm lucky to have a kodak process thermometer.
http://www.amazon.com/Taylor-9842-Commercial-Waterproof-Thermometer/dp/B00009WE45
QUOTE]

Wow,,, how do they do that level for $10? Anyway, we've used Taylor alcohol minimum reading thermometers for years in the orchard during frost season, they're etched with a "bug" in the alcohol column. We got them calibrated by the local weather service every few years, unless the column separated, I've never seen one a full degree off at 32F. They used distilled water/ice in a styrofoam ice chest, and if my grey hair is right, they actually corrected the freezing point for pressure. I knew those things once :-(

Agree 100% on the repeatability. I'd use one of the taylor alcohol's I've got (0.2F graduations) calibrated at 32 as a sanity check at 68 any day. I would not expect it to be accurate to less than a degree,,, but I'd bet lunch on reproducable to 0.2F at 68F.

[ki6mf]

Ditto on the Glass Kodak color thermometer. I have one just for the purpose of calibration. In my travels I also found a digital waterproof cooking thermometer which when checked against the Kodak thermometer was spot on. I now use the digital most of the time and ofter check it against the glass Kodak. While I am doing this i always check the calibration of my analog B&W thermometer.

[Doremus Scudder]

Hi Brian,

I think in °C, not °F, so my 2°-worth in Celsius is really 3.6°F. Still, as I mentioned, it might be a bit conservative. However, I have reticulated BPF 200 a time or too when the wash-water temperature changed, and I don't think the difference was much more than 2-3°C... I imagine Kodak, Ilford and Fuji films are hardened at the factory and more resistant to reticulation. Still, it's a good idea to keep the processing temperatures close.

It is nice to resurrect this thread every so often just to remind me to calibrate my thermometers again :-) I use my Kodak process thermometer as a standard, and they are supposed to be accurate and linear... but after what Lenny complains about, I'm now not so sure. Oh well, at least it is precise.

As for using the boiling point to calibrate with, as mentioned above, it is much to variable to use as a standard. Crushed ice in distilled water, however, is always withing a fraction of a degree of 0°C (or 32°F if you prefer). This is great if you know your thermometer is linear. If it's not, you are only calibrating at way to cold for processing. Better to use a factory-calibrated thermometer as a standard even if it is not completely accurate.

There must be a way of reading IF or something that would allow a reading of the radiation directly from the solution... or?

Best,

Doremus Scudder

[Lynn Jones]

If you have a dial thermometer, they are notoriously inaccurate. What i do is decide what temmp I want to work at (often I would choose 70F) then I would place a liquid thermometer at 70 degrees, put the dial in there a note its actual reading. With a 3 cornered file, I would file a notch at that point. However, nothing beats a liquid column.

AND whoever said that temperature accuracy was not important in B/W couldn't be more wrong, or doesn't know anything about quality control. My liquid column thermometers are guaranteed to be +/- 0.15 degrees F, I wish they were even better.

Lynn

[Mike Anderson]

I just googled "Kodak Process Thermometer" - none on ebay, one on NY craigslist $35 (June 8, 2011).

If I found a used one that looked undamaged, and it told me icy water was 32 degrees F, would it be a good idea to assume it's accurate enough around 60-80 degrees F?

...Mike

[cyrus]

I just googled "Kodak Process Thermometer" - none on ebay,

Cuz I just bought it! :p
Seriously, why not use a certified precision scientific thermometer that also comes with metal cases so they won't break?
http://www.novatech-usa.com/Products/Glass-SAMA-Thermometers

[neil poulsen]

. . . I'll repeat: crap. I have half a dozen glass ones, including a Kodak Process, very old one, and they are all inaccurate - up to 6 degrees off (yes, including the Kodak fancy one). . . .

Interesting. I've been using a Kodak mercury thermometer as my "standard" to check my digital thermometers.

But I agree, repeatability is important. Even though I may be off a degree or so, it's important that my temperature be repeatable from my calibrations to actual developments.

[Jeff Bannow]

How about something like this: http://www.thermoworks.com/products/thermapen/splashproof_thermapen.html

Along with a calibrated mercury thermo to test it against?