Limitations of Fluorescent Lighting

[Curtis N]

Note: The images in this post were replaced in December 2011, so some of the early comments relating to the original images are now irrelevant.

Occasionally a POTN member will ask about an exposure or white balance problem where fluorescent lighting is eventually determined to be the culprit. This thread will hopefully explain the difficulties associated with photographing under fluorescent light and advise on how best to deal with the problem.

Fluorescent light bulbs emit photons as electricity excites the atoms inside them. Our homes and businesses are powered by alternating current (AC). This type of current is continually reversing polarity in a cycle known as a sine wave, illustrated in this graph.
http://www.sprags.com/images/mainpower_sine_wave.jpg

As the current goes from zero to positive to zero to negative and back, the changing voltage causes the excited atoms in the fluorescent tubes to emit light of different intensity and different wavelengths. This results in a light source of continually changing brightness and color temperature.

In the US and Canada, the electrical current alternates at 60 cycles per second (60hz), or one cycle every 1/60 second. During each cycle, two pulses of light are emitted, one during the positive portion of the cycle and one during the negative portion.

If your camera's shutter is open for less than 1/60 second, light from only part of the cycle will be recorded by the image sensor, and you never know what part of the cycle you'll catch. At 1/240 shutter speed, for example, you may catch a pulse of light at its peak intensity and highest color temperature, resulting in an overexposed, blue-toned image. Or you might catch the middle of the cycle where the light output is dim, resulting in an underexposed, orange-toned image.

"No worries!" you say. "I shoot RAW, so I can adjust the exposure and white balance with my RAW converter to make it right." Sorry, but it ain't that easy. The problem is that at fast shutter speeds, the second shutter curtain begins to close before the first curtain is fully open (this is why your camera has a maximum shutter speed for flash). As the shutter curtains follow each other across the frame, different parts of the sensor are exposed at different times. So different parts of the image will be illuminated by light of varying intensity and color temperature. This makes for some truly ugly images, and even if you're a Photoshop master, it's no easy fix! Adding more fluorescent fixtures won't help, either, since they are all powered by the same current and therefore synchronized with each other.

These four images of a sheet of white paper are from a four-shot burst at 1/320 shutter.
http://performancephoto.smugmug.com/photos/i-GsKNHft/0/XL/i-GsKNHft-XL.jpg

So when you're faced with shooting under fluorescent lighting, here are your options:
1) The fastest shutter speed you can use is 1/120 (1/125 is close enough). Theoretically this will catch precisely 1/2 of a cycle, with all parts of the image illuminated by the full spectrum of output by the lamps.
2) The second fastest shutter speed you can use is 1/60, to catch a full cycle. This should give consistent results. Use Tv or Manual mode. You want to avoid 1/100 or 1/80.
3) Multiples of 1/2 cycle, i.e. 1/40 (1.5 cycles), 1/30 (2 cycles), 1/20 (3 cycles), etc. We're now in tripod territory, but anything slower than this should also work fine.
4) Overpower the fluorescent light with flash if you can. Set your shutter at flash sync speed, get as close to your subject as possible, and stop down until the maximum flash distance is not much more than your subject distance, forcing the E-TTL flash metering to maximize the light output.
5) Since fluorescent tubes vary in color temperature, a custom white balance is recommended, using the same shutter speed guidelines listed above.

Here is a four-shot burst of the same sheet of paper at 1/60 shutter.
http://performancephoto.smugmug.com/photos/i-sKCc4g6/0/XL/i-sKCc4g6-XL.jpg

A caveat to non-US readers: In Europe, the current alternates at 1/50 second. The above speeds would need to be adjusted accordingly (1/100, 1/50, 1/25, etc). If you are familiar with AC frequency standards in various parts of the world, feel free to add to this thread with that information.

We can't always control the lighting conditions we shoot in. But by understanding how both cameras and fluorescent lights work, we can maximize the quality of the images we capture in those conditions.

[robertwgross]

To make fluorescent lighting matters trickier, some tubes have "good phosphors" (the tube coatings that glow) and the spectrum is full. Other tubes have "bad phosphors" and the spectrum is not full. Some tubes are worn out and getting dim, or else the control circuit allows them to flicker visibly. For some strange reason, you tend to find all of these in churches where weddings are held, so the wedding photographer must check this stuff out in advance. If fluorescent lighting is there, then the slow shutter techniques are relevant and overpowering the fluorescent light is generally the way to go.

The graph at the top has a confusing vertical scale.

---Bob Gross---

[Curtis N]

The graph at the top has a confusing vertical scale.You're right. I'll fix it.

[SkipD]

Curtis, what is the cause of the gradient in brightness in your example shots? I fully understand the color and brightness differences but not the gradients in each frame. I suspect is has to do with the angle at which the paper was lit.

Thanks.

[Curtis N]

what is the cause of the gradient in brightness in your example shots?Maybe I should have explained it, or maybe I should have setup the shots with more even lighting. The sheet of paper was taped to the edge of a fluorescent shop light and hanging below it. So the high side of the paper was closer to the light and therefore brighter.

Edit 9/22/05: The original images that Skip referred to have been replaced by shots taken under even lighting. Any variation in brightness across the frames is indeed caused by the fluorescent light phenomenon referred to in the original post.

[PhotosGuy]

Good thread, Curtis!
"4) Overpower the fluorescent light with flash if you can." On a big room, I'd try to just fill the light on the important area & let the rest of the room go a touch darker (& greener). It helped to isolate the subject. Other times I'd use two lights, like this overexposed (cull) example:
Simple "every-day-emergency" location lighting (http://photography-on-the.net/forum/showthread.php?t=66358)

[Curtis N]

Bob, Skip, Frank, thanks for your insights and input.

I replaced the original images with new ones taken with more even lighting. Hopefully that will help avoid confusion.

I'm still hoping some of our members in the other hemisphere will provide information on AC frequencies in other parts of the world.

[SkipD]

This thread should be a STICKY thread. Pictures say far more than words can, and so many folks getting into photography don't understand flourescent lighting at all. Good job, Curtis.

[maderito]

Curtis -

Excellent post.

We had a recent POTN query on halogen lighting which uses tungsten filaments but delivers a brighter light with a continuous spectrum. The same type of variation was noted at high shutter speeds. I think the problem is generic to any lighting source powered by AC, including your humble incadescent bulbs.

Since you're already set up, perhaps you could check standard incadescent light bulbs also!

This is the post I'm referring to: http://photography-on-the.net/forum/showthread.php?t=99429

My guess is that color temperature varies over the AC power cycle when using other AC driven light sources. With fluorescent, the problem is simply uglier and much harder to correct for because of the underlying discontinuous spectrum.

[SkipD]

We had a recent POTN query on halogen lighting which uses tungsten filaments but delivers a brighter light with a continuous spectrum. The same type of variation was noted at high shutter speeds. I think the problem is generic to any lighting source powered by AC, including your humble incadescent bulbs.I don't think that lamps which emit the light by heating a filament are going to exhibit the same problem as flourescents because the heat of the filament (and thus the emitted light) can't change nearly as fast as the fluctuating voltage in the AC power. Flourescent light isn't generated by heating, and that light source definitely does change at twice the AC power cycle rate.

I think the problem that was described in the other thread was really caused by the mixing of flourescent light with that from the halogen lights.

[Curtis N]

This thread should be a STICKY thread.Well, I did add it to my post in the Photography Tips & Tutorials List (http://photography-on-the.net/forum/showthread.php?t=52418). So hopefully we can direct people to it when the need arises.

I agree with Skip's opinion on tungsten lighting. When you turn an incandescent or halogen bulb on or off, you can tell that the filament takes a second or two to warm up and cool down. Larger, higher wattage filaments are slower. It's unlikely that the temperature changes significantly within a 1/60 second cycle.

[maderito]

SkipD - You certainly could be very right. Curtis could prove the point one way or the other! I've read that incadescent light does vary its temperature color over time. What you say makes a lot of sense too. So now I'm very curious to know. Remember - it would be an unsual shot that uses high shutter speeds under available incadescent light. So it's not a problem that would come up very often.

Addendum:
I just did a high speed burst of a white sheet of paper illuminated by a halogen light (tungsten filament) with shutter speed = 1/400. On review, the image and the histogram changes across the burst.

[SkipD]

I had to try it too..... I just did a series with my 20D at 1/400 shutter speed under a halogen desk lamp (small 50-watt halogen bulb). Absolutely no difference in brightness, color, or appearance of the histogram in a half dozen exposures.

[PhotosGuy]

I've read that incadescent light does vary its temperature color over time. "over time" is the key. Generally, it does as the filaments age & the lamp surface is coated with crap in the air. This can take months.
There's another more important factor, too. 120v & 60 cps is a "non-standard". Government regulations mandate that an average per 24-hour period be met. With varying loads during the day, there can be a wide variation in the power that you're getting in any given hour. The 24-hour average is usually made up in the last hour of the day! Wish I could remember the source for that info.
Do I worry about it? No. I have better things to do, so I use CustomWB & RAW & get on with the job.

[PacAce]

Curtis, what is the cause of the gradient in brightness in your example shots? I fully understand the color and brightness differences but not the gradients in each frame. I suspect is has to do with the angle at which the paper was lit.

Thanks.
The reason for the gradient in brightness when shot at 1/320 is because the full frame is not exposed at the same time. The curtain travels across the frame exposing just part of the frame as it goes from one end to the other. At 1/320th of a second, the frame is exposed to about 1/5 of a complete cycle or just short of a quadrant's equivalent (see Chris' graph). Depending on the timing, the shutter could be opening when the flourescent light is at or near the peak and going towards zero (Chris' first pic in burst), going from just past 0 towards the peak (2nd pic) or going from one side of 0 to the other side of 0 (3rd pic). If the timing were such that the cycle was approaching the peak when the shutter was closed, the frome would be mostly grayish blue with a hint of brown at both side of the frame.

When the shutter speed is at or below sync speed, the frame is fully exposed so the gradient is less dramatic. However, there will still be a slight gradient because the curtains still have to travel from one end of the frame to the other to open up and it takes a finite amount of time to do that and during this time, the flourescent light is continually changing in intensity.

[Croasdail]

Very helpful, I have been experiencing this a lot in the gym that I have been shooting volleyball in. Explains a lot, but unfortunately the cure isn't so simple. I can't slow down the shutter, and I don't have strobe power enough yet...

[Curtis N]

I just did a high speed burst of a white sheet of paper illuminated by a halogen light (tungsten filament) with shutter speed = 1/400. On review, the image and the histogram changes across the burst.I'm curious to know what white balance setting you used. With AWB, the camera could just interpret each shot a bit differently. A test like this requires a consistent white balance, either tungsten or custom should suffice. You might also try a similar test in sunlight to see if that gives you results that are more or less consistent than tungsten.

PacAce, Skip's original question re brightness gradient referred to images that have since been replaced. The original images I posted had a significant difference in brightness from top to bottom, due to poor testing procedures.;) Sorry for the confusion. Your explanation is still valid, though.

With varying loads during the day, there can be a wide variation in the power that you're getting in any given hour.Are you talking about voltage fluctuation or frequency fluctuation? I used to have an optical tachometer designed to measure engine RPM on model airplanes by holding it in front of the prop. Every time I pointed it at a fluorescent fixture, it was dead on 3600 RPM. I'm not an expert on this electrical stuff, but I would think that a motor-driven clock would not keep correct time if the AC frequency wasn't right.

[maderito]

I had to try it too..... I just did a series with my 20D at 1/400 shutter speed under a halogen desk lamp (small 50-watt halogen bulb). Absolutely no difference in brightness, color, or appearance of the histogram in a half dozen exposures. I'll try again when I'm not so much in a rush. Perhaps my handheld shot was not stationary enough. We'll see. I was shooting at 8.5 fps, manual exposure settings, and a set Kelvin temperature to eliminate AWB as a factor.

The posting I referred to by Doc Nickel was shot with 4 halogen flood lights illuminating a still object with a light tent type configuration in the setting of ambient fluorescent lighting. The shooting parameters were ISO 400, f/7.1 and 1/350 sec. Under those conditions, ambient fluorescent lighting is contributing virtually no light to the exposed imaged. 99+ percent of the light is coming from the halogens. If ambient indoor lighting could illuminate subjects properly with those camera settings, we wouldn't need flashes.

From my point of view, the experiment with halogen light fluctuation has already been done - i.e. the images posted in the thread: http://photography-on-the.net/forum/showthread.php?t=99429

I also quite agree with Frank -- this is all very academic. :)

Sorry to post and run -- I'll re-visit this thread later this evening.

[PhotosGuy]

Are you talking about voltage fluctuation or frequency fluctuation? Voltage. All the old clocks would be wrong if it was frequency as you said!

[PacAce]

PacAce, Skip's original question re brightness gradient referred to images that have since been replaced. The original images I posted had a significant difference in brightness from top to bottom, due to poor testing procedures.;) Sorry for the confusion. Your explanation is still valid, though.

Curtis, in light of the fact (sorry, no pun intended ;) ) that what I responded to is for the image that is currently displayed in your post I would say that not only is my response very valid but it is the only explanation that would makes sense and does not confuse other people, like myself, who have not seen the orignial images you were referring to (unless, of course, you go back and repost the original image that Skip was referring to to validate your response to him. :confused: ??? ;) :lol:

[Curtis N]

unless, of course, you go back and repost the original image that Skip was referring to to validate your response to him.Rather than do that, I edited my original response to Skip's original question.

Man, I wish I had done this right the FIRST time!:o

[robertwgross]

All the old clocks would be wrong if it was frequency as you said!

I've been to the place where the whole power grid for the western U.S. is monitored. At different points on a gigantic map, there is a frequency meter that shows the actual power frequency at that moment. Nominal is 60 Hertz, and the meters will read from 59.9 to 60.1. In reality, they typically show a variation from about 59.99 to about 60.01. Also, it is maintained so that it averages to almost perfect 60.00 through the course of 24 hours.

Voltage is not so clearly controlled nor monitored since it fluctuates widely over the distance from the power distribution point or power substation and the time of day.

What this means to the photographer is that power conditions vary, and you want to choose shutter speeds and color balance temperatures carefully, and keep checking to make sure that your color balance hasn't shifted on you.

---Bob Gross---

[UncleDoug]

When the shutter speed is at or below sync speed, the frame is fully exposed so the gradient is less dramatic. However, there will still be a slight gradient because the curtains still have to travel from one end of the frame to the other to open up and it takes a finite amount of time to do that and during this time, the flourescent light is continually changing in intensity.


I vote for your explanation.

This is similar to the effect you see when a Video camera records what is seen on a TV, stepish & jumpy, when the video rates are not the same.

[maderito]

Desktop halogen light illuminating white paper.
Shot with 1D MKII, 8.5 fps, f/8, 1/400 sec, ISO 400, white balance set to daylight (not AWB).

I'd say there is variation in luminance and color balance which is refelected in the histograms.

http://www.pbase.com/image/49655542.jpg
http://www.pbase.com/image/49655542

[Titus213]

Excellent thread - thanks. I generally take Bob's advice and just try to overpower the flourescents. There was this one auditorium I remember though. It was lit by massive banks of flourescent lights, the walls were painted a lovely green and the drapes that blocked all outside light thru the windows were a near perfect match to the green walls.....My pictures always looke weird.

[Curtis N]

Maderito, thanks for sharing your test shots. It appears the luminance varies more so than the color. When I looked at the histograms of my fluorescent shots, the blue would go one direction and the red & green would go the other way.

I'd say that test needs a control. I'd like to see similar shots at 1/60 or 1/30 shutter speed. And when you get a chance, do the same thing with sunlight. I'd be interested to see how the results compare.

If you look closely at my 1/60 shots under fluorescent light, they aren't exactly the same. They're just a lot closer than the ones done at faster shutter speeds. And I confess that I really don't know why they aren't identical.

(Note to self: For the next "demonstration thread" I do, use animated GIFs to really impress people.):)

[PacAce]

Maderito, thanks for sharing your test shots. It appears the luminance varies more so than the color. When I looked at the histograms of my fluorescent shots, the blue would go one direction and the red & green would go the other way.

I'd say that test needs a control. I'd like to see similar shots at 1/60 or 1/30 shutter speed. And when you get a chance, do the same thing with sunlight. I'd be interested to see how the results compare.

If you look closely at my 1/60 shots under fluorescent light, they aren't exactly the same. They're just a lot closer than the ones done at faster shutter speeds. And I confess that I really don't know why they aren't identical.

(Note to self: For the next "demonstration thread" I do, use animated GIFs to really impress people.):)
Go back and re-read my first post in this thread. ;)

[maderito]

Curtis,

Good points.

Agreed, the color balance variation is very minor - despite the dancing histograms. The luminance change is more obvious, but not dramatic. A quick check on a GE incadescent lightbulb gives similar results.

You're right - should have done the control shots. Here are the histograms for a shutter speed of 1/30 sec on 2 conecutives frames of an 8.5 fps burst. (You have to look closely to see the very minor differences). I'm done testing. Thanks for gettting this post started.

[Curtis N]

I'm done testing.:lol:
I second that motion!
It can be time-consuming!

Do you know when the next "Measurebator's Anonymous" meeting is scheduled?

[robertwgross]

Do you know when the next "Measurebator's Anonymous" meeting is scheduled?

At the meeting, the topic for discussion is whether "anal-retentive" should be hyphenated.

---Bob Gross---

[maderito]

At the meeting, the topic for discussion is whether "anal-retentive" should be hyphenated.

---Bob Gross--- Bob - What was the consensus at the meeting?

[PhotosGuy]

"Consensus"? :D Like you'd ever get one!
Re: "anal-retentive" And I think the mods should ban Bob for using the "A" word, too!

[J Rabin]

Curtis. I feel like I am missing something here. I bought daylight balanced FL bulbs for $14.98x3 instead of $500 Bowens Tri-lights and here is what I get by just sticking a Whi-Bal card in the first shot (RAW) and then Clicking-balancing the rest on conversion:

http://aesop.rutgers.edu/~rabin/Insects/slides/Tomato_Hornworm_Parasite.htm

Who would use shop lights for photography? Jack

[Curtis N]

Jack,

You don't say what shutter speed was used for that shot. Fluorescent bulbs will work fine, as long as your shutter speed is long enough. Those are some very nice images in that link. The fact is, most of us don't have the opportunity to install our favorite light bulbs wherever we take pictures, and most of us shoot things that move faster than tobacco hornworms.

Using a Whi-Bal card and the "click" white balance feature of a RAW converter is usually a very good way to set a custom white balance, regardless of lighting. Or you could use basically anything grey or white, as long as it isn't overexposed with one or more colors clipped.

Who would use shop lights for photography?
I'll refer you to the last sentence of my original post: "We can't always control the lighting conditions we shoot in. But by understanding how both cameras and fluorescent lights work, we can maximize the quality of the images we capture in those conditions." Most photographers outside a studio environment need to adapt to the lighting available. Flash has its limitations and in some situations it's not welcome.

A few months ago, a forum member posted images taken at a Karate competition, held in a place with fluorescent lights, where flash wasn't allowed. I felt sorry for him, because he bought a digital SLR specifically for that purpose, and there was no way to solve his problem (he needed a fast shutter speed to catch the action).

[eddarr]

I am bumping this thread to add some additional reading. This question has been coming up a lot lately.

Extreme example of banding (http://photography-on-the.net/forum/showthread.php?t=615803&highlight=light+cycle)

Detailed explanation on Sports Shooter forum by Guy Rhodes (http://www.sportsshooter.com/message_display.html?tid=20873)

Indoor sports video podcast about adding artificial light to counteract cycle and WB issues (http://web.mac.com/dmwierz1/Site/Podcast/Entries/2008/2/21_Artificial_lighting_for_indoor_sports__video_po dcast_episode__3.html)

Soccer pictures showing white balance issues w/discussion (http://photography-on-the.net/forum/showthread.php?t=598824&highlight=light+cycle)

[GBRandy]

Wowzer....this is an old thread that was brought back from the dead.....

I'll add in my example.... This was a sequential shoot done under the fluorescent in my basement..... Pretty obvious what's going on here.
http://www.tessfamily.com/uploaded2008/Lightcycles.jpg

[eddarr]

Yeah I know this is an oldy but goody but there have been a bunch of these questions on the sports forums lately. I just wanted to add a few newer threads and get the link so I can stop typing the same thing over and over again.

[tkoutdoor]

Yeah I know this is an oldy but goody but there have been a bunch of these questions on the sports forums lately. I just wanted to add a few newer threads and get the link so I can stop typing the same thing over and over again.Holy Smoke! You're a grave digger! :cool:

[RandyMN]

With this thread so old, what about the newer energy efficient bulbs that are out? I also have a set of daylight balanced flourescent bulbs that are specifically made for photography.

Seems to me that technology has advanced since 2005 enough to make some types of flourescent lights no longer subject to these problems.

[Curtis N]

Compact Fluorescent Lamps have electronic balasts that cycle much faster than 60 hz and do not suffer the same issues.

But I still don't like fluorescent lights.

[vipantonio]

Hi and thanks for the useful tips. Am I the only one who cannot see pictures in the first post?

[gofer]

Maybe lost in forum updates - the original post is over six years old :D

[AntonLargiader]

The subject could use a bit of an update. For one, most (I think) newer fluorescents have electronic ballasts which shouldn't give cycling issues. Some institutions will change all of the old ballasts in their lamps and some won't. New construction and renovations within the last 10 years or so are more likely to have electronic ballasts.

Depending on the venue, you may be able to establish whether they have magnetic or electronic ballasts. Just because a fixture has T8 bulbs, though, it doesn't necessarily have electronic ballasts. All magnetic ballasts cycle, while electronic ones are so fast they essentially don't.

[tonylong]

Earlier this year I shot an event in a middle-school gym with plenty of flourescent lights and no problems, no matter what shutter speed I was able to use, so I figure that it does vary. It may be helpful to do some test shots!

[Curtis N]

Ok I just replaced the images in the original post. The lights used were 48" fluorescent fixtures purchased in 2010. While electronic ballasts are becoming more common in new commercial construction, there are still plenty of 40-year-old (or older) fixtures in use.

[mdvaden]

Seeing this old thread revival, what I'm gathering is that though florescent may not be the optimum choice for photographers, that the newer daylight florescent lighting of 2011 is better than back in 2006.

?

:cool:

[Channel One]

Seeing this old thread revival, what I'm gathering is that though florescent may not be the optimum choice for photographers, that the newer daylight florescent lighting of 2011 is better than back in 2006.

?

:cool:


Part of that may be due to the EPA’s LEED program which requires new and retrofitted commercial buildings to meet certain energy saving requirements, and in order for florescent fixtures to meet those standards the use of an electronic ballast is required.

Due to this requirement, most new florescent fixtures on the market are equipped with solid state ballasts and since those ballasts operate at high frequencies 15Khz+ the flicker problem has been eliminated.

Now that has still not solved the problem of a low CRI exhibited by most commercial flo’s, but with the flicker gone that can be easily solved by placing ½ negative green gel over the tube or fixture and then white balancing out the blue spike.

However, as a consequence of the green movement LED lighting is making headway and it is now being discovered that LED’s, especially when dimmed, introduce flicker and to make matters worse in some designs it varies in frequency so there is no “rule of thumb” shutter setting to work around it. L

Wayne

[airfrogusmc]

Seeing this old thread revival, what I'm gathering is that though florescent may not be the optimum choice for photographers, that the newer daylight florescent lighting of 2011 is better than back in 2006.

?

:cool:

All shot in fluorescent
http://i4.photobucket.com/albums/y118/airfrogusmc/IMG_6597-1.jpg

http://i4.photobucket.com/albums/y118/airfrogusmc/IMG_7178.jpg

http://i4.photobucket.com/albums/y118/airfrogusmc/IMG_8094-1.jpg

http://i4.photobucket.com/albums/y118/airfrogusmc/IMG_5550.jpg

http://i4.photobucket.com/albums/y118/airfrogusmc/IMG_2875.jpg

http://i4.photobucket.com/albums/y118/airfrogusmc/IMG_8452.jpg

[AntonLargiader]

What's that got to do with anything? Do you know what kind of ballasts were installed in each case?

[airfrogusmc]

What's that got to do with anything? Do you know what kind of ballasts were installed in each case?

Has everything to do with it. Proofs always in the work. The rest is just talk. These were all old light fixtures and I always assume they are old and shoot at even shutter speeds of 1/15, 1/30, 1/60 & 1/125 but no higher than 1/125 and get reasonable results. I work in this type of lighting all the time.

[Curtis N]

Airfrog -
Did you do any color adjustments in post besides setting the white balance?

My experience has been that in general, fluorescent does a poor job with reds. The skin tones look pretty good in the first four, not so good in the last two.

[airfrogusmc]

Airfrog -
Did you do any color adjustments in post besides setting the white balance?

My experience has been that in general, fluorescent does a poor job with reds. The skin tones look pretty good in the first four, not so good in the last two.

I usually shoot a gray card in the same light as my subject and then use that as my base when converting from raw. Adjust to get the gray right. Get that right and usually everything else is pretty close.

I've been working in it for 20 + years ya kinda get the hang of it eventually.

In the old days of transparency film it was usually 30 magenta filter with ektachrome 400 iso daylight. I would go in a couple days early, shoot a gray card in the light I would be shooting the day of the shoot and then adjust filtration as needed after looking at the chromes. Much easier now.