With the shift to digital audio and streaming, the Digital to Analog conversion is one of the most important steps in your signal chain, but it’s also among the hardest to understand. Most modern listeners understand at least the basics of why a higher resolution audio file sounds better, and get the general gist of a good amplifier providing power and volume without distortion, but the DAC feels like it just shouldn’t matter that much, since there should only be one right way to convert the digital signal into an analog one – right? While that may be the case on a basic level, there’s a lot more to how a DAC works that can make a difference for how it sounds, and how it will end up working in your chain. We covered a lot of DAC basics in our original What is a DAC and Why Do I Need It? and we're going to dive a little deeper as we ask "Does your DAC really matter?"
What is a DAC?
A DAC is a device that receives a digital signal as input, converts that signal, and outputs it as analog audio. A digital signal is typically input from a source like:
-
USB (B or C)
-
Optical/TosLink
-
Digital Coax/SPDIF
-
Bluetooth
-
HDMI/I2S
-
AES/EBU
-
WiFi
For a device that’s only a DAC, the output will be a line level from an analog output:
-
RCA
-
XLR
-
3.5mm or 4.4mm line output
Some DACs also function as a preamp, which provides a variable volume control for the output – generally still RCA or XLR – and others still have a headphone amp built in, in which case you’ll have a headphone output:
-
3.5mm
-
6.3mm
-
4.4mm
-
4-Pin XLR
There can be a number of mixups on the name here, because sometimes when someone says “DAC” they mean the DAC chip itself, while other times they mean a device with digital inputs and analog outputs – but no amplification – and beyond that often a DAC/Amp combo is also just described as a DAC, especially in the portable space, where many DAC/Amp combos are just described as “dongle DACs” or “portable DACs.” The differences between different DACs can be differences at any of those levels too: different DAC chips, different preamp design, or different headphone amplification. For our purposes, we’re going to focus on the idea of a DAC as a complete standalone device that receives digital input, converts it, and outputs line-level analog audio.
How Does Digital to Analog Conversion Work?
When a digital file is created, the system takes an analog waveform, and stores each point in that waveform, called a sample, as a number representing a point on the waveform at a precise moment. It takes thousands of samples per second and strings them together to create a file that contains the map for recreating the original waveform. CD quality audio is 16-bit and 44.1kHz, which means that each sample is represented as a 16-bit number (between 0 and 65,536), and there are 44,100 samples for each second of playtime. The DAC needs to take each digital sample as an input and generate the analog signal as output.
There are two main methods of encoding digital audio, and subsequently two ways that it’s decoded. Pulse Code Modulation (PCM) is the most common way of encoding audio, and is used for CDs. With PCM, you fully convert and playback each sample as you go, so with a 16-bit 44k.1Hz file, the DAC does exactly what we already described, sequentially decoding 16-bit samples 44,100 times a second. Direct Stream Digital (DSD) takes a different approach, only reading 1 bit at a time, so with a theoretical 16-bit 44.1kHz file, the DSD DAC would read 705,600 1-bit samples a second, rather than 44,100 16-bit samples.
DSD vs PCM is more of an engineering debate than a difference that most people will be able to immediately hear, but oversampling and interpolation are two elements of DAC processing that are more clearly audible. Oversampling is the process of increasing the bitrate of a file by duplicating the existing samples, so a 44kHz CD quality file becomes 88kHz, but no actual new information is added to the file. The goal here is to reduce errors and distortion by providing a greater number of correct samples, and to also help with the process of interpolation.
Regardless of how the conversion is processed, you’re still going to be left with tiny gaps between the samples, as compared to the full analog waveform, and this is where another important element of the DAC comes into play: interpolation. Interpolation is the process of filling in the gaps to create a complete smooth waveform from the provided samples. The filter design on the DAC is going to determine exactly how this interpolation takes place, and typically the listener can perceive some difference between something like a “Sharp/Fast” filter that’s going to sound more clean and precise, and a “Slow Roll-off” filter that’s going to sound a little smoother but maybe not as precise. Exactly how the DAC handles interpolation can make a big difference in the listening experience. The difference also tends to be more pronounced with lower quality files: with MP3s, you might be dealing with something like 12,000 samples per second instead of 44,100 at CD quality, which means that there’s a greater impact to the sound from interpolation.
So for the complete digital to analog conversion process, the DAC receives a digital input from a file, it may or may not oversample the file depending on the design, spaces between the samples are filled in through interpolation, and then final audio is output as an analog waveform.
What’s the Difference Between Different Types of DACs?
While the end goal remains the same, there are a number of different designs used to get there. Typically DACs found in audiophile products have one of these four designs:
-
Delta/Sigma Chip DACs
-
R2R DACs
-
1-Bit/DSD DACs
-
FPGAs
The vast majority of DACs use a Delta-Sigma design with an integrated on chip solution. If you’ve seen a product with an ESS, AKM, Cirrus Logic, or ROHM DAC, it’s one of these. While each DAC chip and its available filters have identifiable characteristics, other elements of the implementation can make two DACs with the same chip sound somewhat different. Delta-Sigma DACs are often implemented with multiple DAC chips in one device, with separate DACs for left and right channel being the most common design, but often devices will use as many as 8 chips in a device. The Delta-Sigma DAC design relies on oversampling of a signal and then comparison of multiple of the same sample to generate output, so having multiple chips processing the same signal simplifies the process and creates a cleaner output.
Delta-Sigma DACs are the most popular, but the oldest design that’s still commonly used is the R2R DAC. R2R DACs require more physical space and engineering, because the design is based on a Resistor Ladder, with each bit getting an individual resistor that receives current as input, and the combined output from the resistor ladder is the final sound. R2R is probably the easiest design to understand, and there are a number of variations that use the same principle, like dCS’s Ring DAC, and most 1-Bit/DSD DAC designs share the same basic architecture as a resistance ladder.
The other major type of DAC design is an FPGA DAC. A Field Programmable Gate Array (FPGS) is a microprocessor that can be configured by a programmer to do whatever they need it to do, including operations like digital to analog conversion. FPGAs were popularized by Chord Electronics and enable manufacturers to create a completely unique DAC design. FPGAs have made their way into other DAC designs as well, often serving as the connecting element that manages signal flow between more advanced Delta-Sigma or R2R configurations.
Do Different DACs actually sound different?
Assuming that all DACs are rendering the provided digital audio in the same way, how is it possible that some DACs sound different without one being more or less colored than another?
The first thing that can change the sound of a DAC is differences in the input design and digital processing. Anything that comes in between the input jack and the DAC chip can impact the exact information that the DAC is processing. The original signal could be changed by deliberate digital processing or from noise or interference in the design. The same holds true for the output: once the signal has been converted to analog there are multiple points in the process of amplification and output that can have a measurable impact on the final signal.
While the differences in the input and output design are likely to provide the most clearly measurable differences, the differences between the DAC design, filtering, and interpolation are what you’re more likely to hear when comparing the sound between different DACs.
When it comes to interpolation, the difference in sound is not so much tuning – in the idea of one DAC adds more treble, while another adds more bass – it’s in how those interpolation filters react to different elements in the sound. One DACs filtering might result in a sharper treble attack in the treble because of the way it fills in fast attacking elements in the upper frequency ranges, while another might sound warm because of the way it translates the decay of bass notes. It’s not that the DAC is increasing the bass or treble volume from the signal that is there, it’s the choices the designers made in how it extends or cuts off notes in the spaces between what the original signal provides. The spaces themselves are tiny and momentary, but the sum of the subtle differences has an impact on how you hear each tone and the music as a whole.
The Bottom Line
That last line essentially boils the whole DAC discussion down to its essence: the sum of subtle differences has an impact on how you hear each tone and the music as a whole. Outside of converting the music accurately and without introducing noise or distortion, there are any number of subtle elements from the circuit design to the filter design and interpolation that’s different between DACs and while each may to designed to provide the most accurate possible listening experience there are absolutely small sonic differences between DACs that can add up to a big difference in the final result of how you hear your music.
