dBi vs dBd: Antenna Gain and Decibels Explained

Two antenna advertisements can quote gains that differ by more than two decibels for what is, in practice, the same antenna. The reason is usually dBi vs dBd: the two units use different references, and the larger number is not a better antenna. Once you understand what each reference means, you can compare any two antennas fairly, add up the gain and the losses in your own station, and read the claims on a data sheet without being misled.

This guide explains decibels from scratch, defines dBi and dBd, shows how to convert between them, and then follows the numbers through a real station, from transmitter to feedline to antenna and out into the air. It also covers the related terms ERP and EIRP, which appear in many national regulations.

Decibels in Five Minutes

A decibel is not a unit of power. It is a way of expressing a ratio between two powers on a logarithmic scale. If you transmit with power P2 and compare it to a reference P1, the difference in decibels is ten times the base-10 logarithm of P2 divided by P1. The logarithm is what makes decibels so convenient for radio, because very large and very small ratios become short, manageable numbers and because chains of gains and losses can be added instead of multiplied.

A few values are worth memorising, because they let you do most station arithmetic in your head:

Two points about the table cause a lot of confusion. First, power and voltage ratios are not the same for a given number of decibels, because power is proportional to the square of voltage. The rules above are for power unless the column says otherwise. Second, a gain of 3 dB does not mean the signal sounds twice as loud. It means twice the power, and the effect on the other station is modest, which is why people say that a six decibel improvement is roughly one S-unit. Our article on RST signal reports and the S-meter explains how S-units relate to signal strength.

What dBi and dBd Actually Mean

Antenna gain is always a comparison. An antenna cannot create power, it can only redistribute it, so gain in a particular direction is measured against something else. The two common references are:

• dBi is gain relative to an isotropic radiator. An isotropic radiator is a theoretical point source that radiates equally in every direction. It cannot be built, but it makes a convenient mathematical baseline. By definition its gain is 0 dBi in all directions.
• dBd is gain relative to a half-wave dipole. A dipole is a real antenna that can be built, tested and compared. It concentrates energy a little, so it already has some gain over the theoretical point source.

The standard figure for the dipole’s gain over an isotropic radiator is 2.15 dBi, as described in the general explanation of antenna gain. That is the whole difference between the two units:

• gain in dBi = gain in dBd + 2.15
• gain in dBd = gain in dBi – 2.15

So an antenna described as having 8.15 dBi has 6 dBd, and an antenna with 9 dBd has 11.15 dBi. They are the same antenna described two ways. Always check which unit a manufacturer is using. Some quote the larger dBi figure because it looks better in advertising, and some products give a number with no unit at all. If the unit is missing and the figure looks generous, assume dBi and subtract 2.15 before comparing with a dipole.

Comparing Antennas Fairly

When you compare two antennas, both numbers must use the same reference. A simple rule is to convert everything to dBd when you are comparing against a dipole you know well, or to dBi when you are comparing against published specifications and modelling results, which are usually in dBi. Antenna modelling software typically reports dBi, so if you use a program such as the ones described in our antenna modelling walkthrough, you will see numbers that are 2.15 higher than the same antenna expressed relative to a dipole.

Three further cautions apply when you read any gain figure:

1. Gain is in a direction. A quoted gain is normally the maximum, at the peak of the main lobe. In other directions the gain is lower, sometimes much lower. A beam with high forward gain has deep nulls to the sides and rear, which is why it needs a rotator.
2. Free space is not the real world. Models and data sheets often give free-space gain. Over real ground, reflections add to or subtract from the signal at the takeoff angle you care about. A low horizontal antenna over ground can have a very different pattern from its free-space description. Our guide to antenna height and takeoff angle shows why the ground matters so much on HF.
3. Gain is not efficiency. A short antenna with loaded coils can have a respectable pattern and still waste much of the power as heat. Gain figures that ignore loss will flatter such antennas.

For vertical antennas the usual reference is different again. The ideal quarter-wave monopole over a perfect ground plane has a defined gain, often stated in dBi, and in real installations the ground losses reduce it. A figure quoted for a vertical without saying what ground it assumes is of limited use.

Gain Is a Trade, Not a Gift

It helps to picture what gain really does. A fixed amount of power leaves the antenna. A higher-gain antenna squeezes the same power into a narrower beam, so more of it goes in the wanted direction and less in the others. It is like adjusting the nozzle of a garden hose: the volume is unchanged, the jet is stronger. Because the same effect works in reverse, a high-gain antenna also hears better in its main direction, and the same gain applies on receive and transmit.

That has three practical consequences:

• Narrow beams need aiming. A high-gain beam is useful only if it points at the other station. See our comparison of a Yagi and a hex beam for how designers trade gain for size and bandwidth.
• Receive gain is not always helpful. If the noise is arriving from the same direction as the signal, a higher-gain antenna raises both. What matters on receive is the signal-to-noise ratio, not the signal strength alone.
• A little gain is cheap, a lot is expensive. Each extra 3 dB roughly doubles the physical size or the number of elements, so the cost and mechanical load rise quickly.

Following the Numbers Through a Station

The great advantage of decibels is that a whole station can be added up on paper. Take a transmitter, a feedline and an antenna.

1. Transmitter power. Suppose the radio produces 100 watts. In decibels relative to one milliwatt, that is 50 dBm. Relative to one watt, it is 20 dBW.
2. Feedline loss. A length of coax that loses 2 dB at the operating frequency takes the signal from 50 dBm to 48 dBm at the antenna, which is about 63 watts. The loss is largest at high frequencies and in thin cable. Our article on coaxial cable for ham radio gives realistic loss figures and shows how much a poor cable choice can cost.
3. Mismatch loss. If the antenna does not present a good match, a further small amount is reflected back. With a modest standing wave ratio, this is usually well under a decibel, but it is worth knowing it exists.
4. Antenna gain. A 6 dBi antenna adds 6 dB in the best direction, so the effective figure in that direction is 54 dBm relative to an isotropic radiator, which is 251 watts.

In this example, 100 watts at the radio becomes the equivalent of about 251 watts radiated by an isotropic source, in the best direction of the antenna. If you replaced the coax with a lower-loss type that lost only 0.5 dB, you would gain 1.5 dB, about a 41 percent increase in power at the antenna. In many stations, the cheapest improvement is not a better antenna but a shorter or better feedline.

The same arithmetic works on receive. A feedline loss of 2 dB reduces both the signal and, depending on the situation, may add noise relative to what the receiver could hear. On the quiet upper bands, losses ahead of the receiver matter more than on the low bands where atmospheric noise dominates.

ERP and EIRP

Regulations in many countries limit power in terms of the radiated power in a particular direction rather than the power at the transmitter. Two terms appear:

• EIRP is equivalent isotropically radiated power. It is the power an isotropic radiator would need to be fed to produce the same signal strength in the direction of the antenna’s maximum. It equals transmitter power, minus feedline loss, plus antenna gain in dBi.
• ERP is effective radiated power, and in common usage it uses a half-wave dipole as the reference, so it equals transmitter power, minus feedline loss, plus antenna gain in dBd.

Because the references differ by 2.15 dB, EIRP is always 2.15 dB, or about 64 percent, higher than the ERP for the same signal. A rule that sets a limit in ERP is therefore stricter than a rule that sets the same number in EIRP. When you read a regulation, check which one it uses and which direction it refers to. The rules for amateur stations vary between countries and between bands, so always consult the official regulator for the country in which you are operating. We deliberately do not quote numeric limits here: they differ and they change.

Common Mistakes With Gain Figures

• Adding dBi and dBd figures. If one antenna is quoted in dBi and another in dBd, convert one of them first, or the comparison is off by 2.15 dB.
• Forgetting the feedline. A high-gain antenna on a long run of thin coax can be worse in practice than a plain dipole on short, good cable.
• Treating decibels as percentages. A loss of 3 dB is not a 3 percent loss. It is half the power.
• Ignoring the pattern. Peak gain says nothing about the nulls. A beam can lose signal strongly outside the main lobe.
• Believing receive gain equals better copy. A change in S-meter reading from a higher-gain antenna is not always an improvement in readability if the noise has risen with it.

If you want to check what your own antenna and cable are actually doing, an inexpensive analyser can measure return loss and cable loss directly. Our guide to the NanoVNA antenna analyser explains how, and the numbers it gives are decibels, so everything above applies.

The Bottom Line

The difference between dBi and dBd is exactly 2.15 dB, the gain of a half-wave dipole over an isotropic radiator. Convert to a common reference before comparing, remember that gain is directional and measured under assumptions that rarely match your ground, and subtract feedline loss before you count any benefit. If you remember the three anchors, 3 dB is double, 6 dB is four times, 10 dB is ten times, you can check any manufacturer claim in your head and decide whether a new antenna, a better cable or simply a change of height will do more for your signal.

Frequently Asked Questions

Is dBi or dBd the larger number?

For the same antenna, the dBi figure is always 2.15 higher than the dBd figure. A gain of 5 dBd is 7.15 dBi. A bigger number does not mean a better antenna, only a different reference.

How do I convert dBd to dBi?

Add 2.15 to the dBd value to get dBi. To go the other way, subtract 2.15 from the dBi value. The offset is the gain of a half-wave dipole over an isotropic radiator.

Which unit should I use when comparing HF antennas?

Either, as long as you use the same one for every antenna. Many amateurs think in dBd because the dipole is the familiar benchmark, while modelling software and manufacturer data often use dBi. Convert before comparing.

Does a higher-gain antenna always give a stronger signal?

Only in the direction where its gain is highest, and only if the antenna is efficient and mounted well. Gain in other directions may be lower than a simple dipole, and losses in coils, ground or cable can cancel the benefit.

What is the difference between ERP and EIRP?

EIRP uses an isotropic radiator as its reference and ERP uses a half-wave dipole. For the same signal, EIRP is 2.15 dB higher than ERP. Check which one a regulation uses and which direction it refers to.

How much does feedline loss matter compared with antenna gain?

Often more than people expect. Every decibel lost in the cable is a decibel taken away from the gain. On higher bands or with long, thin coax, the loss can equal or exceed the gain of the antenna, so improving the cable can be the cheapest upgrade.