recording 174 Mb file size with 3 or 4-minute tracks

I appreciate very much your taking the time to help me get started. I have read the Audacity manual, downloaded Audacity 2.1.0 .dmg file (32.8 MB) on both my iMac OS X 10.9 and MacBook Pro Yosemite 10.10.2. I tried recording and got a 20-minute file with a number of “tracks” that were “analysed” and compiled into what appears to be a huge file way too big for the Amazon upload requirements – and the “tracks” were defined by the stops I made to perfect the recording. It took several minutes for this to load.

What I need is file size less than 170 Mb in monaural 192 kbs MP3 at 44.1 kHz (which the Audacity manual has told me is standard when I choose “narration”) and with tracks WITHOUT silence that flow from one to the next of 3 or 4 minutes in time.

I have found no topic in the manual or here in the forum for file size by VIEW of both Mbs AND time. I need to watch as the recording approaches 170 Mb in size so that I may end at a natural pause in the narrative than SAVE by chapter title and section number each file. I also need to make the tracks by TIME so that the listener may find the same place with not too many minutes repetition, as when one is driving and listening to the CD the listener has made from the digital download of my book from my Amazon web page. And when I stop to re-record, I don’t want an automated insertion of “track.” This means that the 170 Mb files will then be compiled by chapters and sections for customer download to make a standard hour-long (or so) CD. Essentially, this question is how to record a standard audiobook we are all familiar with as listeners. Many thanks to all who read this request for an addition to the Audacity manual. Michelle (jungsoul)

Koz will probably answer you and knows more about audio books than I.

It would be good to have information in one place but I expect the answers are already elsewhere in this board of the Forum.

Anyway, just to state one thing - you can’t use MP3 if you want one file with chapters because MP3 doesn’t officially support chapter marks within one file. Almost no MP3 players can see the chapter marks in an MP3. You need to make a zip (or a data CD) containing multiple MP3’s.

To export multiple MP3’s, label the chapter breaks and Export Multiple. You can reduce the MP3 bit rate to reduce the file size (click “Options…” when you select “MP3 Files” in the Export Multiple window).

You can make a single long MP4 (or M4B audio book format) with chapter marks (but you will have to make the chapter marks in other software).

To record where you left off, SHIFT + R records at the end of the selected track.


Generally I’m very unclear what your target format is - an MP3 file, multiple MP3 files, an MP3 data CD or what? How many minutes do you want on the CD?


Thank you, Gale, for your reply. The Amazon requirements are for files no larger than 170 Megs. When you listen to commercially produced audiobooks, the reader simply announces the chapter number and title; this is how I plan to label the files so that they can be transcribed to a CD after downloading from my Amazon book page.

Since 170 Mb is not a large file in terms of time for reading aloud, there will be several for each chapter by number, and my ordering of each 170 Meg file will be by the numbering of sections, such as, Chapter One, Section 1; Chapter One, Section 2, Chapter One, Section 3, Track 1 to 4. In this example, the next file of 170 Megs would begin Chapter One, Section 3, Track 5 and continue for these 170 Meg files to be compiled for download and recording a CD of approximately one hour in length.

I am asking how to produce on Audacity what I have experienced as a listener to a standard CD in an audiobook, which is about an hour. On any audiobook CD, all the listener sees is the time for each track – usually between 3 and 5 minutes at the most, with seconds only. There may be 16 to 20 tracks and the total recording time for all of them is also given. From the producer’s point of view, which is what I am asking help for, there may be three or four “files” of 170 Mbs that are compiled into one CD.

The listener may hear the chapter number only once, and in order to make it possible to stop listening at the end of a chapter, I, the producer, will want to watch for either the end of a file of 170 Mbs or simply the end of a 3 to 5-minute “track” and then INSERT a silence (pause, not noiseless) and start a new chapter reading the number and title. If a chapter should start at the end of a CD, then I might simply group the 170 Meg files by total time of less than one hour and begin a new file or track for the next grouping so that the download can be easily copied to CD. This is why I need to know how to set Audacity to VIEW both file size in Megs and time for setting tracks of 3 to 5 minutes that flow with NO silences.

What all this allows in commercially produced audiobooks is for the listener in a car, for example, to turn off the ignition when arriving at a destination, and to start the engine with only a short repetition of a track of less then 3 to 5 minutes, which may also be the beginning of a chapter, as announced.

170 MB is nearly three hours at Audacity’s default 128 kbps MP3 bit rate, and longer at lower bit rates (1 minute takes about 1 MB at 128 kbps). At 192 kbps MP3, 1 minute takes about 1.5 MB. So 170 MB at either 128 kbps or 192 kbps would produce a file that is already much longer than an audio CD (up to 80 minutes) which is what you need for playing in a car.

Burn the audio CD from WAV files not from MP3 - some CD burning software may not accept MP3. If you burn the CD from a WAV file at 44100 Hz 16-bit stereo, then 3 minutes is 30 MB of WAV file, and 5 minutes is 50 MB (1 minute takes about 10 MB). Note that a stereo file may be required by your CD burner - to export a mono recording as stereo, click Tracks > Add New > Stereo Track. It does not matter that this track is empty.

You cannot have marker points within an MP3 and expect everyone to be able to skip between the points. If you want to split the MP3 book into sections of 3 or 5 minutes, you must make separate MP3 files.

You can split the audio CD into as many tracks as you like, up to maximum 99 (which is less than a minute per track). Each track can be burned from a separate WAV file, exported using Export Multiple. See Splitting a recording into separate tracks - Audacity Manual.

Can you place the three to five minute track breaks at the end of a sentence? If so then all you should need to do if you want the CD to play across the tracks without the customary 2 second gap between tracks is to turn off that gap in your CD burning software.

People making music CD’s who need to be absolutely sure the CD will play across tracks without even the slightest extra pause will export one WAV file up to 80 minutes long, and use a “cue sheet” to tell the burner where to place the tracks. Audacity does not directly support cue sheets but you can label each track break in Audacity then export the label track as a text file and make the cue sheet from that file. See


Again, I thank you for your time, Gale. I’ll have to spend tomorrow attempting to implement what you have posted here last. I do not understand why I had such a large file from my first recording that seemed to be timed at a total of 20 minutes (this timing itself was confusing because of the compiling of successive times when I stopped and re-recorded). I know I was recording at 44.1 kHz, but the bit rates and other specifications were not immediately obvious to me from my settings. I need to familiarize myself with Audacity now before I can tell you that this has answered my need for instructions on producing an audiobook. I cannot use WAV files or change the parameters. Please understand that the requirements are given by Amazon publishing of audio files for sale on the author’s (my) book page. As a long-time audiobook listener, I am trying to translate my experience to production ====>>>>
192Kbps MP3 with each file containing only one chapter or section. Mono files are strongly preferred. Stereo files may be submitted but your audiobook may not contain both mono and stereo files. Stereo files must not be “joint stereo”. [Elsewhere in the instructions, acx (which is an Amazon subsidiary for publishing) indicates that stereo files are used for dialogue with more than one reader/actor]
Audiobooks uploaded to ACX must adhere to the following requirements. The ACX Quality Assurance team may reject titles that do not meet these standards, and their retail release may be delayed. The following requirements help ensure customers get a great listen.

Your submitted audiobook must:
be consistent in overall sound and formatting and be comprised of all mono or all stereo files
include opening and closing credits and a retail audio sample that is between one and five minutes long
Each uploaded audio file must:
contain only one chapter/section that’s shorter than 120 minutes and the section header must be read aloud
have room tone at the head and at the tail and be free of extraneous sounds
measure between -23dB and -18dB RMS and have -3dB peak values and a maximum -60dB noise floor
be a 192kbps or higher MP3, Constant Bit Rate (CBR) at 44.1 kHz

monaural 192 kbs MP3 at 44.1 kHz

Actually, I believe 192 is the ACX AudioBook standard for submission. It’s wildly better than required for a simple MP3. You can get just passable mono voice tracks at 32. ACX requires this terrifically high quality because they are planning on resampling the work into multiple lower quality products as required for corporate production and sales. If you’re not going to ACX, you don’t need any of that.

So traditionally, you can produce a stand-alone block of work (say, one chapter) in WAV (44100, 16-bit, mono) and save that forever as your master archive. Then you can export it in MP3 at whatever quality is required to get your target file size. I don’t think there is any fixed relationship between file size and MP3 quality setting because it depends to some degree on content. MP3 pays attention to content. WAV doesn’t.

After the first few, you get used to the length of presentation and quality levels required. If your goal is straight listening and for some people to “convert up” to Audio CD, then the MP3 can be any quality you wish over the minimum. Audacity default quality is 128 for a stereo show which means mono will probably sail by at 64. Again 32 is the point most people start to notice honky, wine-glass quality problems.

That’s all under those conditions. If someone wants to sample down to get more show into a Personal Music Device or post on-line, the rules change. Downsampling produces quality a combination of the original and new quality rates. Someone resampling a 64 mono voice to 32 is not going to get 32 quality. It will be much worse and probably unlistenable. This is why ACX requires a sky-high delivery rate of 128 (or higher). They want to downsample their products with little or no quality hit.


Yes, and as above, you said “Some People” will want to burn the work to a CD. If you want to do it, you should certainly be using your WAV archive, not any MP3s. MP3 always creates sound damage. Your recording, archive and Audio CD should all be the same super high quality.


I haven’t experienced the track jump problem. If I submit a bunch of sound clips to my CD Authoring Program and tell it not to put gaps in, I don’t get any gaps. It’s true I probably couldn’t do it in the middle of a musical note, but it would certainly fit in the breath you take between chapter one and chapter two.


Koz, thanks for your input; this expands the audiobook recording topic to other people’s needs in the forum. For me to record an audiobook to sell on my Amazon book publishing webpage, I have to do what acx tells me because they are the dragon at the gate. I don’t need to know the fine points you bring up for music recording but I do need to know how to set Audacity to the parameters Amazon and ACX specify on their instructions for which I gave the URLs in my last post. I will spend another day recording and trying to find the file size and time in order to meet the requirements so customers can use whatever format their devices require and still copy to a CD, if they need to. Audiobook listeners (and I am one) want simply to download and listen.

Already, I have learned that the enormous file size (750 Megs) I got after reading a 20-page (typed double-space) chapter in 20 minutes must mean that I was not using the MP3 bit rates in the instructions despite having chosen “narration.” Is there no way to set Audacity for simple 192 kbps, MP3 monaural narrative recording at 44.1 kHz and SEE BOTH TIME AND SIZE? I can figure out from the manual how to do the “room noise” silence in order to create a track, but not how to stop the program from making a track every time I stop or re-record.

You can always pick up where you left off with Append Record, Shift-R. This will not start a new track.

Audacity always works internally at super high quality, uncompressed rates. That is required against the requirements of post production, editing, effects and filters. Those all fall apart very rapidly if you tried it in MP3 or any lower quality.

I think your’e misinterpreting the requirements. The MP3 quality standard should be no less than 192 (and can be higher) and the filesize can’t go over 170M. If you Export one chapter of your story at MP3 192, what’s the filesize?

Here’s a thing they published specifically about creating the MP3s in another program. They’re clearly not expecting each chapter to go over 170MB.


The only way your users can make a CD of the same length and quality as an Amazon 192 kbps MP3 lasting more than 80 minutes is to make a data CD containing that identical MP3 file. Or they can split into multiple MP3’s at 192 kbps (using lossless MP3 splitting tools) and burn those MP3’s to data CD.

If they do that, such a data CD will play in an MP3 CD player or in a computer or in some DVD players, but it will not play in a car because a car will expect an audio CD (maximum 80 minutes).

If you want to record in mono set Audacity’s Device Toolbar (third box) to 1 (mono) recording channel.

I have answered the rest of it already in my first reply. When you have recorded, File > Export Audio… (if you are exporting one file) or File > Export Multiple… (if you want to export one file for every three or five minutes of the book in order to make an audio CD for yourself).

Set the file type in either export dialogue to MP3 Files, then click the Options… button to change the bit rate to 192 kbps.

192 kbps is actually a little under 1.5 MB file size per minute. You can calculate how large the exported MP3 will be from that. If you export exactly two hours (the maximum length that ACX permit) at 192 kbps, the file will be about 165 MB (just under the ACX maximum of 170 MB).


The novel-length audiobooks from the public library are all on multiple Audio CDs.