bplist00_WebMainResource_WebSubresources _WebResourceFrameName^WebResourceURL_WebResourceData_WebResourceMIMEType_WebResourceTextEncodingNameP_Shttps://docs.google.com/document/d/16frWVfB3VtR7952A1W3mCjMUXZrgv_qOeM8fRAsi6GE/pubOs
The Effects Of Analog To Digital Conversion On Audio
Mike Arnoult, Graham Reid, Kyle Kovalik
04/27/2015
When it comes to audio fidelity, analog to digital converters have always been in the forefront of the discussion. Does a more expensive analog to digital converter actually improve your audio fidelity? Over the past two months we have researched and completed an experiment in order to test this theory. We tested two different audio converters, using the same signal path, as well as the same source (Mike Arnoult singing “Baby In Hertz” written by singer/songwriter Scott Hawley). In the end we found that there were more factors at play, rather than just the converters. However, it goes without saying that the converters, like most audio equipment, add coloration to the audio signal. When it comes to the tangible evidence, the “better” converter (Antelope) outputted a much “cleaner” signal than that of the “lesser” converter (Focusrite).
II. Introduction
An analog to digital converter (otherwise known as A/D) is a device that converts a continuous physical quantity to a digital number. A converter is defined by its bandwidth, the range of frequencies that it can measure, and its signal to noise ratio. This is how accurately it can measure a signal relative to the noise that it produces. The bandwidth is defined by the sample rate and the dynamic range is defined as the resolution, the number of output levels it can quantize a signal to (linearity and accuracy), how well the quantization levels match the true analog signal, and jitter. Jitter is small timing errors that introduce additional noise.
The importance of A/D conversion in the audio engineering industry is growing every day due to the digitization of the methods of recording music. As Dennis Bohn wrote, ”We could go so far as to say that data conversion is the art of digital audio while everything else is the science, in that it is data conversion that ultimately determines whether or not the original sound is preserved. (This comment certainly does not negate the enormous and exacting science involved in truly excellent data conversion).” [2] (Bohn, “Digital Dharma of Audio A/D Converters”) The conversion process has become just as important as how the sounds are generated and captured. Every step of the recording process “colors” the sound. If audio engineers are seeking to have more control over how a recording sounds, then they must consider their analog to digital conversion.
III. Methodology
For our experiment, we decided to track a vocal through the same microphone and mic pre-amplifier, using two different converters to compare the analog to digital conversion. We decided to use a vocal because it is usually the instrument that is most “up front” in a mix and is a practical application of our experiment. We used a Neumann TLM 102 through a Neve 511 pre-amplifier using both the Antelope Zen and Focusrite Saffire. We did this to limit the amount of factors. The microphone pre-amplifier and microphone were our control variables. We chose to use the Antelope Zen because it was available to us and is considered a very high quality converter. We chose the Focusrite Saffire because it is considered a “higher end” interface at a decent price point and it is an interface that many Belmont University students use. The Antelope Zen is retailed for $2,500 and the Focusrite is retailed for $900. Below are the specs that the manufacturers provided for both audio interfaces. We recorded the two tracks into Pro Tools 11 at 192kHz at 32bit.
Focusrite Saffire Pro 56: Antelope Zen
[4] Musicians Friend [2] Sweetwater
Dynamic range is the “difference, in decibels, between the loudest and the quietest portion of a program. [3] (Davis & Jones “Sound Reinforcement Handbook”) The THD or total harmonic distortion is “specified” distortion, “for individual harmonics, a composite value representing all the harmonics.” [3] (Davis & Jones “Sound Reinforcement Handbook”) THD is usually accompanied by the “ + N” which takes into consideration the noise specification, meaning that there is a possibility that the harmonic distortion is actually a result of noise voltage.
IV. Experimental Results
When we played back the tracks recorded through two different converters we heard a major difference right away in spectral qualities of the audio. We played them back through KRK Rokit 5 studio monitors which are considered “cheap” studio monitors and even they could show us a difference. The Antelope sounded like it had much more intensity in the higher frequencies, while the Focusrite sounded “muddy” as compared to the Antelope. The Antelope sounded more “crisp” and “clear”. In order to fully know what these differences were, we decided to analyze the audio files through Wavelab and Shaart. A problem we found was that the analysis software could only take 48kHz at 24bit. We had to bounce down the tracks at a lower quality, which could have changed our results drastically. It would have been really interesting to see the difference between the data we received from a 48kHz at 24bit session as compared to the original 192kHz at 32bit session. We ran the Wavelab spectrometer, spectroscope, and loudness meter analysis tools on both the Antelope and Focusrite. We have screenshots of the spectrogram and the waterfall graphs for both the Antelope and Focusrite as our other documentation is in video form. The graphs below indicate that the Antelope is able to capture more detail in the “higher frequencies” than the Focusrite, while the Focusrite has a tendency to capture more in the “lower frequencies”. This explains the audible difference in clarity that we heard through the Antelope.
You can see here that the Antelope has a more uniformly distributed frequency spectrum while the Focusrite has less detail in the “higher frequencies”.
Figure 1: Focusrite Saffire Spectrogram Figure 2: Antelope Zen Spectrogram
Additionally, these waterfall graphs show that the Antelope has a more “even response” in the higher frequency range.
Figure 3: Focusrite Waterfall Figure 4: Antelope Waterfall
V. Conclusion
We concluded that the analog to digital conversion of an audio signal greatly affects the spectral qualities of that signal. The Antelope Zen yielded a more “brilliant” and “clearer” sound. We found it to be accentuating the “higher frequencies” of the spectrum, while taming the “lower frequencies” of the spectrum. The Focusrite Saffire Pro 56 yielded a more “muddy” and “less bright” sound. Accentuating the “low-middle” range frequencies of the spectrum, while taming the “higher frequencies”. One may not be better than the other from an artistic point of view. All audio gear is a tool for the engineer to craft the sounds that he/she is working with. The A/D conversion is one of these tools. The different spectral qualities that can be offered by different A/D converters will result in another artistic decision that the engineer can make. We predict that the study of A/D conversion will become increasingly important in the audio engineering industry.
VI. Ideas For Future Work
An extension of this study would be to analyze clocking and jitter using software that was not available to us. [5] Jim Kaiser, an Audio Engineering Professor at Belmont University in Nashville, Tennessee, informed us that the analysis software for clocking, jitter, and total harmonic distortion analysis costs over $3,000. He said that it would be important to analyze these in that there is a correlation to audio fidelity with clocking, jitter, and total harmonic distortion. At the very least we would have liked to analyze our materials at 192kHz, 32-bit float and compared the differences in the total harmonic distortion (THD). The lower the value of THD, the more likely is it to create an accurate representation of an audio signal. According to specifications of the Focusrite Saffire 56, the total harmonic distortion is as low as -109dB, which is -4dB less than that of the Antelope. However, we cannot detect that with the human ear. Lastly, it would have helped our experiment to track with pink noise. Our initial experiment was meant to be a real world comparison, so we chose to test a human voice.
VI. References
[1] “Antelope Audio Zen Studio” Antelope. Sweetwater.Web. April 2015.
[2] Bohn, Dennis. Digital Dharma of Audio A/D Conversion.Rane Corporation,1997. Web. April 2015.
[3] Davis, Gary, and Ralph Jones. Sound Reinforcement Handbook. 2nd ed. Milwaukee: Hal Leonard, 1989. Print.
[4] “Focusrite Liquid Saffire 56” Focusrite. Musician’s Friend. Web. April 2015.
[5] Kaiser, Jim. “Audio Conversion.” In person interview. 5 Mar. 2015.
[6] Palmaccio, Joe. "A/D Conversion." In person interview. 10 Mar. 2015.