A properly calibrated reference display costs more than most independent setups cost in total. Almost nobody working alone has one, and a lot of grading advice quietly assumes you do.
Here is what I do instead. It is not equivalent and it is considerably better than guessing.
Calibrate what you have
An ordinary monitor calibrated with an inexpensive hardware probe is enormously better than an ordinary monitor out of the box. This is the single highest-value step and it costs less than a lens filter.
Manufacturer presets are generally set to look impressive in a shop — oversaturated, over-bright, cool white point. Grading on that means you will pull saturation down to compensate and the result will look flat everywhere else.
A probe measures what the screen is actually doing and builds a correction. It takes about fifteen minutes. Do it every month or two, because displays drift.
This will not make a consumer panel into a reference display. It will make it predictable, and predictable is most of what you need.
Control the room
Almost as important and completely free.
Your perception of what is on screen depends on what is around it. Grading in a room with a window behind you, or with warm lamps, or with a bright coloured wall, will push you consistently in one direction.
What I do: neutral grey wall behind the monitor if possible, no direct light on the screen, consistent room lighting that does not change through the day, and no window in my field of view while grading.
The day-to-day consistency is the point. If your room looks different at ten in the morning and at nine at night, your grades will be different too and you will not know which one was right.
Use the scopes and believe them
This is the part that substitutes most directly for a reference display, and it is the habit that took me longest to build.
The waveform tells you where your luminance actually sits. The vectorscope tells you what your colour is doing. The parade shows the individual channels. None of them care what your monitor is doing.
Practical uses. Set your black point and white point on the waveform rather than by eye. Check skin tones against the skin tone line on the vectorscope, which is a genuine reference that holds across skin colours. Look at the parade to spot a colour cast in the shadows that your eye has adapted to and stopped seeing.
The discipline is to make the decision on the scope first and then look at the picture, rather than the other way round. Your eyes adapt within seconds. The scope does not.
Check on multiple ordinary screens
Once I have a grade I think is right, I look at it on a phone, on a laptop I did not grade on, and on a television.
Not to match them — they will never match. To find out whether the grade survives them.
The failure modes are consistent. Shadow detail that disappears on a phone in daylight. Highlights that clip on a television with a different gamma. A subtle warm grade that becomes an orange cast on a cool panel.
A grade that holds up reasonably across those three is a safe grade. A grade that is perfect on your monitor and falls apart on a phone is a grade you cannot deliver.
Grade less than you want to
The practical consequence of an uncertain display is that heavy grades are risky. If you do not know exactly what your screen is doing, a strong stylised look might be considerably stronger than you think on someone else's.
Subtle grades are more forgiving. Correct the exposure, neutralise the white balance, set the contrast, add a modest amount of whatever look you want, and stop.
This is frustrating if what you want is a distinctive look. It is also why heavily graded work from people without proper displays often looks wrong — not because the taste is bad but because the display lied to them about how far they had pushed it.
Reference images help more than reference videos
Something I picked up from a colourist and have used since. Keep a folder of stills from work whose colour you admire, and put one on screen next to your grade.
Comparing side by side on the same display cancels out most of the display's error. If your image looks warmer than the reference on your screen, it is warmer, regardless of what your screen is doing to both.
This is a relative measurement rather than an absolute one, which is exactly what an uncalibrated environment can still give you reliably.
Where this stops working
Being honest about the limits.
Anything for broadcast or cinema delivery has technical standards you cannot verify without proper equipment, and guessing is not acceptable.
High dynamic range work genuinely requires a display capable of showing it. There is no workaround.
And matching shots from different cameras under difficult lighting is much harder without a display you trust, because the differences you are chasing are small enough to be inside your monitor's error.
For everything else — online delivery, client work, documentary, the vast majority of what most people make — a calibrated ordinary monitor, a controlled room, disciplined use of scopes and a habit of checking elsewhere will get you to a result nobody will complain about.