Introduction To Audio Technology, maybe helpful

Hello all,

it does not entirely match the topic description, but it is possibly useful for the people in this forum.

If you want some background about the technology used to record things, you may be interested in my free (CC) text .

Introduction To Audio Technology

I certainly refer to Ardour in it!

I am interested in feedback.

1 Like

Ill take a look at it when I get a moment, is the source published on Github or similar by chance to provide you with pull requests(As a quick glanec suggested even basic editing issues like carful instead of careful, unless that was a pun)?

    Seablade

Hi,

the source is currently not published. It is way more than text.It inlcudes xfig graphics, scripts for Gnuplot and even shell skripts creating gnuplot scripts, etc. I found it too hard to export all in a meaningful way, although that may change some day.

For now, I will manually fold back any improvement that is provided. During the writing of the text I had problems with the typo check for two languages (the original is German) in Texmaker. I am afraid that shows in the result, although I did an overall check in Emacs in the (preliminary) end.

oh, and carful was not a pun. :frowning:

I changed carful to careful. A lot better, thanks!

“Error failed to load PDF document”

First let me say thank you for putting this out there, I apologize that a lot of what you are about to read will be more critical, but at least you are trying and put forth a significant amount of time and energy into this.

So glancing through quickly there is a fair share of language translation issues I think making it more difficult to read, not bad, but in a technical document these become much more of an issue.

That being said, you are kind of all over the place for an ‘introductory’ document, jumping into more advanced math (As in math I never even studied and would have a hard time understanding) at the start, then going back to cover absolute basics of how we hear or what type of microphone is what, etc. I wouldn’t worry about trying to explain acoustics for instance, until I described how sound acted in a space and how we hear it. The math behind a fourier transform is not necessary to understand enough of how analog to digital conversion works for 90% of audio work for instance, not really an introductory topic for sound technology, more DSP introduction topic there, and DSPs themselves can be a more advanced topic for sound.

I also have to mention you have some lack of understanding in some areas, for instance I can jump to areas I know pretty well, and point out some errors in what you did write, which is likely more detail than needed, while skipping more basic topics. ie.:

The classical form of radio wave transmission uses frequency modulation (FM). To improve the signal to noise ratio it uses a 1:2 compression in the transmitter and a reverse expander in the receiver. This may lead to artefacts that may disturb in extremely demanding applications.

Companders in analog FM transmission are usually not quite as simple as a 1:2 compressor and expander, which is why when you use systems from one manufacturer with another manufacturer, the sound quality will be very wrong. In fact Lectrosonics, a higher end manufacturer of wireless microphones, has created multiple emulations to allow for use of other manufacturer’s microphones with their receivers to account for the differences. Honestly you don’t need to go into how the expansion and compression works here, other than to identify it exists, again identifying the ‘basic’ level of introduction needed to cover a topic. Of course for a basic level of introduction on the topic of wireless microphones at all you get into topics you didn’t address such as intermodulation (Or existing/other users of common spectrum, or differences between spectrum, etc.) so it may be better to just remove this section all together, especially because in order to truly understand it requires a knowledge of the laws of the country you are in, instead of just understanding the physics behind it.

Glancing through the live sound section, your advice on the whole isn’t bad, but it is just advice, and missing basic information. Rather than advice people to use DIs and emulation(Which almost never flies professionally), you need to think about why that amp or cabinet should be mic’d, or ways that sound can be controlled from it (You allude to this earlier with acrylic sheets, but don’t go into much detail with examples of why, which this can be a great one).

Another example of giving examples without touching to much on the thought process behind them:

10.5.1 Minimal Solution, no Bass
A very basic setup for signals with little bass (speech, small acoustic instruments) can be done with two small loudspeakers on stands on each side of the stage. One single central loudspeaker would either stand in the way between artist(s) and audience or behind the stage at the worst place for the danger of feedback.

Most of the time if you are doing a smaller setup, it is actually preferable to use a single speaker. While you are correct in that a single speaker behind the performer can lead to feedback or be visually disruptive, very often one speaker set to the side of the performer is enough to provide coverage needed, and since most of a smaller setup is going to be in mono anyways, actually provides a clearer coverage as producing a mono mix in two speakers both covering the same area will lead to phase cancellation and a loss of clarity compared to a single reproduction source.

You description of arrays for live sound is primarily incorrect honestly. The important aspect to them is using the fact the LF drivers are arranged to be mounted very close together, not that they can be mounted with a 5 deg angle. This allows for summing, and through the fact that audio has a travel time, the delay creates a natural phase cancellation and summation at frequencies dependant on the length of the array to provide more directionality than would be experienced by a single driver, and to replicate in a horn requires a very large horn. However due to physics this often doesn’t translate to HF sound, which is in turn guided by other means (Waveguides etc.) in arrays, which is why the hybrid curved array became popular as opposed to the constant curvature described in your document, as you could focus HF sound to where it was needed better than a constant curve. Of course all of this results in phase smearing in the HF compared to single carefully controlled drivers, but the tradeoff is often considered worth it for many cases (Though not all).

Similar problem with your description of cardioid subs, which are generally arrays of drivers, not a single driver with a hole in the rear of the box (Which sounds more like a typical ported sub honestly). Time delay and phase/polarity is used to create zones of summation and cancellation given the travel time of audio at specific frequencies.

I haven’t had time to read the entire document, and again let me state that I am appreciative that you are willing to do this. I had something similar on a Drupal site once for my classes but lost it when the database was corrupted, and haven’t ever gone back to recreate it since, I just don’t have time, so I do understand how much an investment of time and effort this was. I just think that some focus on specific introductory topics may bring out strength to your document, and leave some of the other topics out for a different document.

           Seablade
2 Likes

It loads and displays just fine here. It opens in Firefox, Chromium and also as download. 135 pages. PDF 1.5

$ curl -I http://www.baggytrousers.de/AudioTechnology
HTTP/1.1 200 OK
Date: Wed, 27 Feb 2019 00:33:12 GMT
Server: Apache/1.3.41
Content-Location: AudioTechnology.pdf
Vary: negotiate
TCN: choice
Last-Modified: Tue, 26 Feb 2019 17:02:50 GMT
ETag: "1b528db-48548f-5c75713a;59f5fab1"
Accept-Ranges: bytes
Content-Length: 4740239
Content-Type: application/pdf

$ curl -s http://www.baggytrousers.de/AudioTechnology | sha1sum 
e722573e16ee738009e99f1ed78e679a37d937c4  -

Do you get the same checksum, if not, I could mirror it somewhere.

Strange - the checksum matches - but so far I can only open it using Chrome on Windows (which might be a little ironic) - Chrome on Ubuntu fails as shown, latest version of Safari on macOS 10.14.3 also fails with just a blank document, Firefox also fails with an empty document, downloading the pdf and opening in Adobe reader 2019.010 fails with “This file is corrupted and cannot be repaired”. Foxit reader similarly. I don’t have this issue with any other pdfs, on or offline.

Please try again. There has been a problem with the URL.

After the initial remarks by Seablade I re-checked for typos and fixed more than I thought. Please reload the document.

For the record, it works fine here on both Chrome on Windows and ChromeOS. (And I think I tried it on OS X earlier but can’t remember to be certain)

     Seablade

After going again through typos (there seems to be a problem with the spellchecker in emacs, I found errors I did not see the last time this morning), I revised the section about FM transmission. I definitely overgeneralized because after hearing about the compander artefacts I checked some schematics I found on the internet and there was a 1:2 compressor/expander in what I found. It was not a good idea to generalize this.

I already wrote about the problems with interferences and dual use of frequencies. I added the problem of using frequencies that are too close together.

I will look into the other points later.

It’s funny. I could open it before in Firefox, now I can’t. Even if “wget” it, I cannot open it in Okular. But it does open in Chromium…

Anyway, I can’t take a very good look right now, but noticed a couples of things (all of it are just nitpicking). Did you use LaTeX? If so:

You missed the \ in some \log's, \sin's, \cos's… Also, I would recommend using \text for some subscripts that are abbreviations, like P_{\text{ref}} instead of P_{ref}.

Also, I’d suggest not to use * for products in formulas (like on pg 12). If you need a multiplication symbol, I’d go with \cdot.

In integrals we usually use some space (usually \, – a third of space) before the \mathrm{d}x.

Again, just nitpicking. Thanks for sharing it!

For anyone looking for an excellent well written guide to all aspects of DSP and audio processing in general, I would recommend:

The Scientist and Engineer’s Guide to Digital Signal Processing

2 Likes

Well this is likely going more into detail than might be appropriate for an introductory text, use of RF transmission (And this can apply to certain types of digital as well as analog transmission) is subject to interference from more than just existing use of frequencies. Intermodulation is a completely different topic, and more often the cause of issues when dealing with multiple channels as people don’t understand it much.

But my main point of all that was to suggest that some benefit could be had by not trying to cover everything, but instead cover a more focused introduction with better clarity. So for example if RF transmission of audio is covered it should be in a bit more depth than what you have I believe, but that just opens a door so there is an argument for it not to be talked about at all. I would also suggest that focusing on Live, Studio, or that technology that overlaps between the two only (One of those three options) would be appropriate for an introductory text.

Another example to use for comparison would be the text ‘Digital Sound and Music’ written by a few folks on my side of the pond. I think it suffers a bit from the same, cover to many topics but is a bit better I think about focusing on the introductory in most cases.

http://digitalsoundandmusic.com/curriculum/

Take a look and see if it might give some inspiration.

I have seen that book recommended a lot Mike, I am actually recommending it for my wife as she is looking to possibly put her math degrees to use other than teaching high school and I think it might be a good intro to other possibilities for her. If I remember right that book however does depend on a working knowledge of calculus, which I never really studied though she has offered to teach me, I am curious if you know a good reference that doesn’t depend on calculus for DSP or even if that is possible? More for myself there, as I always like to expand my knowledge beyond the existing limits.

It depends on what you want to do - I think if you want to gain a detailed theoretical understanding, or build your own DSP solutions, emulations etc ‘from the ground up’ - then you’ll almost certainly run into it at some point, there’s no getting away from it - but, there are also plenty of ‘cookbook’ examples out there which can be used (and adequately appreciated) without getting too deep into the maths.

Yes it is LaTeX, using a Koma Script format. I do approve your kind of nitpicking :slight_smile:
I changed everything you mentioned except the format of the subscripts for now - time for my breakfast ran out after doing the change to the German version. I just could not find a good way to let

Meta-X-query-replace-regexp

to do the work completely. I had to do it by hand :frowning: