{"id":19543,"date":"2014-11-03T17:01:13","date_gmt":"2014-11-03T21:01:13","guid":{"rendered":"https:\/\/noelborthwick.com\/cakewalk\/?p=19543"},"modified":"2014-11-03T17:01:13","modified_gmt":"2014-11-03T21:01:13","slug":"audio-specs","status":"publish","type":"post","link":"https:\/\/noelborthwick.com\/cakewalk\/2014\/11\/03\/audio-specs\/","title":{"rendered":"Basics: Five Questions about Audio Specs"},"content":{"rendered":"

By Craig Anderton\u00a0<\/span><\/p>\n

Specifications don\u2019t have to be the domain of geeks\u2014they\u2019re not that hard to understand, and can guide you when choosing audio gear. Let\u2019s look at five important specs, and provide a real-world context by referencing them to TASCAM\u2019s new US-2×2 and US-4×4 audio interfaces.\u00a0<\/span><\/p>\n

First, we need to understand the decibel (dB). This is a unit of measurement for audio levels (like an inch or meter is a unit of measurement for length). A 1 dB change is approximately the smallest audio level difference a human can hear. A dB spec can also have a \u2013 or + sign. For example, a signal with a level of -20 dB sounds softer than one with a level of -10 dB, but both are softer than one with a level of +2 dB.\u00a0<\/span><\/p>\n

1. What\u2019s frequency response?<\/strong> Ideally, audio gear designed for maximum accuracy should reproduce all audible frequencies equally\u2014bass shouldn\u2019t be louder than treble, or vice-versa. A frequency response graph measures what happens if you feed test frequencies with the same level into a device\u2019s input, then measure the output to see if there are any variations. You want a response that\u2019s flat (even) from 20 Hz to 20 kHz, because that\u2019s the audible range for humans with good hearing. Here\u2019s the frequency response graph for TASCAM\u2019s US-2×2 interface (in all examples, the US-4×4 has the same specs).<\/p>\n

\"\"<\/p>\n

This shows the response is essentially \u201cflat\u201d from 50 Hz to 20 kHz, and down 1 dB at 20 Hz. Response typically goes down even further below 20 Hz; this is deliberate, because there\u2019s no need to reproduce signals we can\u2019t really hear. The bottom line is this graph shows that the interface reproduces everything from the lowest note on a bass guitar to a cymbal\u2019s high frequencies equally well.\u00a0<\/span><\/p>\n

2. What\u2019s Signal-to-Noise Ratio?<\/strong> All electronic circuits generate some noise, so you want the lowest possible noise level. Noise increases as you turn up the gain. For example, here\u2019s the US-2×2\u2019s mic preamp noise with the volume turned up a fifth of the way.<\/p>\n

\"\"<\/p>\n

The noise is less than -130 dB\u2014in other words, compared to a signal at full level the noise is over 130 dB softer. This means the signal-to-noise ratio (the ratio of the full-level signal to the noise) is 130 dB, which is extraordinarily quiet. But what happens if we turn up the mic preamp gain two-thirds of the way (about right for recording a really quiet vocalist)?<\/p>\n

\"\"<\/p>\n

Now the noise level is under -110 dB. This is still extremely quiet, and backs up TASCAM\u2019s claims regarding their mic preamps\u2019 performance. Consider that Compact Discs can at best reproduce levels no lower than about -90 dB\u2014so essentially, the noise is so low a CD can\u2019t even reproduce it.\u00a0<\/span><\/p>\n

These two graphs also hint that specs have the potential to mislead rather than enlighten. Company \u201cA\u201d might use a graph like the upper one for their marketing, while Company \u201cB\u201d might choose a more \u201creal-world\u201d graph like the lower one. Company \u201cA\u201d could imply their mic preamp is quieter\u2014\u201cjust look at the graphs!\u201d So it\u2019s important to know the conditions under which specs are taken. In this article, all the specs (except for the first noise graph and second crosstalk graph below) are with the mic preamps turned up two-thirds of the way to reflect real-world, not theoretical, conditions. This underscores that even when pushed, these mic preamps have excellent specs.\u00a0<\/span><\/p>\n

3. What\u2019s Total Harmonic Distortion?<\/strong> As all guitarists know, distorting a signal adds harmonics\u2014and we love them! But just as all circuits generate some degree of hiss, they also generate some degree of distortion. Total Harmonic Distortion graphs show the level of the harmonics generated by distortion.\u00a0<\/span><\/p>\n

This test feeds in a 1 kHz signal at maximum level. In theory, the output should consist only of that 1 kHz signal. Any other signals represent distortion.<\/p>\n

\"\"<\/p>\n

\u00a0<\/span>The graph shows a little bit of 3<\/span>rd<\/sup> harmonic distortion at 3 kHz and 5<\/span>th<\/sup> harmonic distortion at 5 kHz. Both are below -100 dB, and if there\u2019s any other harmonic distortion, it\u2019s so low it\u2019s below the noise. The fact that there are only two visible harmonics, and that they\u2019re so low-level, speaks to the interface\u2019s audio quality.<\/span>\u00a0<\/span><\/p>\n

4. What\u2019s Intermodulation Distortion?<\/strong> This type of distortion occurs when two signals interact with each other to produce additional signals. These additional signals also represent distortion, and many people consider intermodulation more objectionable than harmonic distortion. This test feeds in two signals at maximum level, one at 60 Hz and one at 7 kHz. Any output signals other than these two frequencies represent distortion.<\/p>\n

\"\"<\/p>\n

Unfortunately (or actually, fortunately for those who have these interfaces) the intermodulation distortion is so low it\u2019s very difficult to see any additional signals. But if you look really<\/em> closely, you\u2019ll see a tiny little spike (circled in red so you don\u2019t miss it) poking its head up from the noise. As you\u2019ve probably guessed, this represents superlative performance.\u00a0<\/span><\/p>\n

5. What\u2019s Crosstalk?<\/strong> Crosstalk occurs when one channel picks up some of the signal from another channel. This happens because some circuit elements radiate signals, while other elements can pick up those signals. Careful mechanical design and signal isolation can reduce crosstalk, but you can\u2019t get rid of it entirely. Crosstalk is more likely with high gain, high frequencies, and often, low frequencies.<\/p>\n

\"\"<\/p>\n

This crosstalk spec, taken with a high gain setting, is comparable to similarly priced devices, although other interfaces sometimes have a low-frequency crosstalk increase that\u2019s not present here. When you turn down the gain to a level suitable for recording an instrument like steel-string guitar, crosstalk goes way down.<\/p>\n

\"\"<\/p>\n

This helps illustrate an important point: Specs can be a \u201csnapshot\u201d of a gear\u2019s particular state, or be represented by a range of numbers. So, use specs as a guide, not a judge. Granted, the specs for these TASCAM interfaces are excellent. But what made me curious to check them out is I wanted to know why the interfaces sound<\/em> good\u2014because that\u2019s the ultimate spec.<\/p>\n

 <\/p>\n

 <\/p>\n","protected":false},"excerpt":{"rendered":"

By Craig Anderton\u00a0 Specifications don\u2019t have to be the domain of geeks\u2014they\u2019re not that hard to understand, and can guide you when choosing audio gear. Let\u2019s look at five important specs, and provide a real-world context by referencing them to TASCAM\u2019s new US-2×2 and US-4×4 audio interfaces.\u00a0 First, we need to understand the decibel (dB). … <\/p>\n

Continue reading “Basics: Five Questions about Audio Specs”<\/span><\/a><\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[45,92,1591,73,111,90,1,53,54],"tags":[1898,2469,2470,2471,2472,2473,2474,2475,2476],"class_list":["post-19543","post","type-post","status-publish","format-standard","hentry","category-audio-analysis","category-daw-best-practices","category-guest-blogger","category-music-production-tips","category-studio-recording","category-tascam-news-2","category-uncategorized","category-windows-7","category-windows-8","tag-audio-interfaces","tag-audio-specs","tag-crosstalk","tag-decibel","tag-frequency-response","tag-intermodulation-distortion","tag-signal-to-noise-ratio","tag-tascam-interface","tag-total-harmonic-distortion"],"_links":{"self":[{"href":"https:\/\/noelborthwick.com\/cakewalk\/wp-json\/wp\/v2\/posts\/19543","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/noelborthwick.com\/cakewalk\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/noelborthwick.com\/cakewalk\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/noelborthwick.com\/cakewalk\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/noelborthwick.com\/cakewalk\/wp-json\/wp\/v2\/comments?post=19543"}],"version-history":[{"count":0,"href":"https:\/\/noelborthwick.com\/cakewalk\/wp-json\/wp\/v2\/posts\/19543\/revisions"}],"wp:attachment":[{"href":"https:\/\/noelborthwick.com\/cakewalk\/wp-json\/wp\/v2\/media?parent=19543"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/noelborthwick.com\/cakewalk\/wp-json\/wp\/v2\/categories?post=19543"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/noelborthwick.com\/cakewalk\/wp-json\/wp\/v2\/tags?post=19543"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}