A method for improving Milky Way exposures in light pollution

What I've learned over the years with digital cameras -- you can't trust the preview image. The histogram tells much more of the story (if not the whole truth). My photographs have become much better (in a technical sense) after I learned to expose almost to the point of blowing highlights, making sure there is little blank space at the right (where the highlights are displayed) of the histogram as possible (hence the term Expose To The Right). As someone downthread stated, this improves the signal-to-noise ratio of the captured data. There's an unwritten assumption that the camera's built-in auto-exposure capability isn't sophisticated enough to expose the scene properly, and the photographer has to take control. This is often the case.

And speaking of SNR, there are a few tricks astro-photographers have been using for years that occasionally find their way into mainstream photography. One is image stacking, where you combine multiple photographs of the same scene (useful for landscapes if a little blur isn't a problem for you) into a single image. This pushes the shot noise (the random distribution of noise) down even further. I've gotten clean images out of cheap cameras, and even from cheap cell phone cameras (given enough images).

The other technique is dark frame subtraction, where you take an photo with the lens covered at the same ISO/shutter/aperture as the rest of the set. Digital sensors have readout noise and heat noise which tends to be very predicable. Subtracting this noise gives you a cleaner shot as well.

On a side note, I do wish when talking about ETTR and even digital photography in general that the term "overexpose" wasn't used when explaining the technique of overriding the cameras default metering with exposure compensation. Overexposure would be blowing the highlights, or at least the highlights of details that the photographer considers important (e.g. blowing the highlights of street lights is usually okay). The "over" portion makes it sound like the photographer is doing too much of something, when in reality ETTR is a technique of giving enough exposure, but not too much.

That is the right technique in general when overexposing would clip highlights. Here's that's not really a risk, so the other direction works better.

What article doesn't explain is the physics behind why ETTR captures more light data: our perception of light scales logarithmically, while a CCD captures it linearly. Each stop is a doubling (or halving) of the amount of light captured. CCDs work by increased voltage based on captured photons.

Imagine back to when digital photography just started, when we were stuck with 10-bit RAW images (before processing them to 8 bit JPGs). That's 1024 different values per channel, not a whole lot of room to begin with. Because we're converting a linear signal to a logarithmic one, that means that highest stop in the image takes up 512 of those values.

EDIT: As rightfully pointed out in a reply, 8-bit JPGs are not in linear colour space any more - so this logic does not apply there (however, JPG compression is more harsh on shadow details, since it's lossy, and it tries to move that loss of information to the parts of the image where you're less likely to notice, like the shadows, but we're digressing)

By that logic, the low light data should have a lot more banding than the highly exposed data. Try it with level tools on a RAW file and see for yourself.

So you can imagine that when the dynamic range of (part of) the scene that you want to capture is less than the dynamic range of your sensor, you want to be on the right edge of the histogram (all else being equal - if this causes motion blur due to long exposure it's pretty useless)

Caveat: this is potentially outdated knowledge, perhaps the analog/digital converter in modern cameras actually converts the voltage of a CCD to a (pseudo)logarithmic scale for the digital output, which might partially circumvent this issue (although the sensor is still linear). And nowadays many of the better cameras have 14-bit A/D converters - the usefulness of which also largely depends on the quality of the sensor - so there's a lot more detail in the shadows than before.

I'd rather see a comparison between two 30-second exposures at 1600 and 6400 than what was presented here. That would give a much better idea of how much detail you're losing from sensor gain.

If the scene has a lot of variation in brightness, then I'll always opt to save the highlights over the shadows. Shadow detail recovers remarkably well, while clipped highlights are ghastly.

ETTR is a real technique to maximize SNR. Underexposing and then boosting shadows is just a good way to guarantee that your image has no contrast, and comes out looking flat and grey all over.

I think it's time to introduce fines for office buildings that are unoccupied overnight, but have lights on. Certainly in high rise towers this would be tens of thousands of lights.

More stars, less energy consumed. Win, win.

In my small city we are a few people interested in night observation. We are pondering about lobbying the city hall to have a "dark night" once a month, maybe something from 1am to 3am, where public lighting would be off.

Considering the mayor comes at every smallish local events just to remind people of his name, he may be interested in something that would make a bunch of people praise him for little effort.

"Then an image sensor with a sufficiently fast global shutter could disable itself during every brief pulse of light"

all that light scattered in the atmosphere and reflecting off other things would still be around when the LEDs are out.

I had this idea too but...

This article uses 20 seconds exposure, so it wouldnt be helped by PWMing the lights.

Also, you'd have to do it much faster than 100Hz. Wave Christmas LED lights around and you'll see a strobe effect.

People claim to feel (not see) 100-120 Hz flicker (I wouldn't dismiss them off hand).

Finally, you'd have to have a powerful, lighting, LED able to toggle on/off at fast enough speeds. That isn't obviously available, if for no other reason large internal capacitances.

As for synchronizing it, that's not too bad. I don't think you'd need a GPS clock. Use an accurate crystal clock and the mains as a 60*3 = 180 Hz reference. Turn the light off every 0, 120 and 240 degrees to cover all three phases.

Fun! Solving the wrong problem though - problem with light pollution is the effect on the environment and human psychology, not on astrophotography.

Or we could simply go back to sodium vapor lamps for streetlights, because there are simple sodium vapor filters for astronomy that can entirely filter it out without losing any of the light required for astronomy.

This entire issue is already solved, and works fine.

It is an interesting idea, but most of the lighting industry is interested in reducing flicker, not increasing possible sources of it.

You'd be fighting the current.

It's somewhat of a known problem space. Tackling light pollution boils down to easy steps, you just have to be mindful in design...

- Only be on when needed

- Only light the area that needs it

- Be no brighter than necessary

- Minimize blue light emissions

- Be fully shielded (pointing downward)

Light pollution is (IMO) driven by lack of knowledge of the problem & solutions, a perception that lights = wealth/luxury, and the low cost of electricity.

This might be enough for photography, but the speed of light would probably push this above 1 millisecond for the area that needs to be dark for large telescopes.

But engineering the light sources to be astronomy-friendly and nature-friendly in general seems like a good idea. Basically limit things to the red-orange part of the spectrum. It worked for sodium lights.

Fix the duty cycle to a particular slice of the phase of AC voltage (e.g. the first 0-pi/2) and you accomplish the same thing much more practically, at least within a single power network.

I heard once that it is easier to filter out (of the photos) light that comes from LED, but I can't find a good resource on it. Anyone know if this is true?

All you need to do is make a copy of your image, blur that copy enough to get rid of the stars, and then subtract it from the other image. It's going to affect larger structures like the milky way, though.

There are light pollution reduction filters made specifically for astrophotography. I was shocked at how good these photos were w/o using such a filter. I don't think you ever could have recovered so much of milky way using film w/o a filter in such a polluted sky. (I was certainly never able to.)

Googling for an example, amusingly enough, the first link is to the same site:

https://www.lonelyspeck.com/hoya-intensifier-review-an-affor...

That filter is not specifically for astrophotography, just one that happens to work well. Here's a discussion of other options:

https://www.cloudynights.com/topic/494162-light-pollution-fi...

Adobe Lightroom can also merge bracketed RAW (DNG) images. That is, you take N DNG images and you get 1 DNG image out, which then preserves more image information that allows you to non-destructively edit the final image. It's available under "Photo Merge > HDR". Unfortunately, it doesn't provide a lot of settings to tweak the merging algorithm, unfortunately.

ISO is not just software gain. On a lot of higher-end cameras (read: just about any ILC) there's a definite analog gain stage prior to ADC.

Since there's a fixed level of sensor noise, when you increase the analog gain you decrease the level of noise in the shadows relative to the sensor noise, essentially increasing SNR. This isn't true for all cameras, however.

This chart [http://www.photonstophotos.net/Charts/PDR_Shadow.htm#Canon%2... ] demonstrates the difference between three cameras, one of which is known to be largely "ISO invariant" (the Nikon D750), one which is not (the Canon 6D), and one which looks like it has a 'native' ISO of 800 (the Sony A7R II)

As you can see, the D750 doesn't gain much (heh) from increasing ISO -- the total image noise in the shadows isn't improved by increasing the ISO through the range. If you're shooting a 6D in a low-light situation, however, you see improvements in shadow quality up to about ISO 3200, after which it turns out that you will be fine if you just raise the shadows in post. If you're shooting an A7R II you might as well just leave the ISO dial on 800 all the time.

That's not all, though, since it turns out when you increase ISO you basically lose 1EV of highlight off the top of your image, as you allude to, and you also lose around 1EV of DR. (Chart here: http://www.photonstophotos.net/Charts/PDR.htm#Canon%20EOS%20...)

It is, like all things in life, a balance.

It differs between the different camera sensors. With CCD sensors, there were switchable amplifiers, but sometimes camera companies added software-pushed sensitivities. The most modern CMOS sensors are called ISO-less as the higher ISOs are implemented not by changing the amplification, but by digital post-processing. So for a wide range of exposures, there is little difference between higher ISO settings and pushing in post-processing. There might be some slight differences after all, as the in-camera processing might vary as well as some low bits skipped at low ISO values.

No the ISO setting works in the analog domain, amplifying the signal prior to digitisation, so it is different from digital postprocessing. Edit: apparently my knowledge on this is outdated.

Imagine a vinyl record, and that you are sitting on the record, near, but not at, the outer edge of the record. South of you, but close to the horizon, is the centre of the record - containing a massive black hole. North of you are all the stars in the galaxy between you and the edge of the galaxy. Most of the stars in the galaxy are scattered along the disc, but with a fairly large variation. The closer you get to the centre, the greater the number of stars - that is the brightest part of our night sky.

Some ancient clusters of stars have drifted a fair bit away from the disc, but are still gravitationally bound to it, and its dark matter. If you were on another galaxy looking from above, our galaxy would look like a spiral with arms. If you looked edge on, it would look more like a cigar.

My understanding is that based on the position of the Earth, most of the time what we see is the next band over, from our position on the edge of a neighboring band. But like you I've had trouble finding a detailed explanation of our exact POV within the galaxy.

Some quick digging found

The PureNight filter is coincidentally a crowdfunded product by Lonely Speck, the people (Ian Norman and Diana Southern) who contributed this story article.

my town has just gone to LED now, so its a bit harder to do :/

Still I'm sure one could filter out the peaks from the LEDs (they are phosphor based still) but it's not going to be effective as taking out the sodium notch.

Quite recently I lived in a small town far from cities with almost no street lights.

I thoroughly enjoyed being able to simply look up and gaze at the glorious Milky Way unaided.

I'm curious if in many decades self driving vehicles will eliminate the need for most night time street lighting.

The answer is the 500 rule. To avoid star trails, exposures should be less than 500/focal-length-in-135-equivalent.

In other words, if he shot this on his 'favourite' Rokinon 24mm, an exposure time of 20 seconds is reasonable to avoid all star trails.