Storage requirements for compressed audio

The church I belong to is planning a website which will include 
recorded sermons. The proposal for site design and hosting did 
not explicitly mention how much online storage was included.

A brief search told me a minute of MP3 occupied ~1 MB. No mention 
of fidelity. I assume that would be satisfactory for speech. What 
if they wanted to record choral music. I know essentially nothing 
about the details of MP3 [I date from when modems include an 
acoustic coupler ;].

What search terms should I be using?
Any comments/suggestions/etc ?
TIA

Richard Owlett  <rowlett@cloud85.net> wrote:

>The church I belong to is planning a website which will include 
>recorded sermons. The proposal for site design and hosting did 
>not explicitly mention how much online storage was included.
>
>A brief search told me a minute of MP3 occupied ~1 MB. No mention 
>of fidelity. I assume that would be satisfactory for speech. What 
>if they wanted to record choral music. I know essentially nothing 
>about the details of MP3 [I date from when modems include an 
>acoustic coupler ;].
>
>What search terms should I be using?
>Any comments/suggestions/etc ?

Most listeners consider 320 kbps MP3 to be very acceptable for music.
That comes to 2.4 MB per minute.  You might consider giving listeners
a choice of that, or a slower rate such as 128 kbps, since their
feed may be slow.

S.

On Friday, September 23, 2016 at 9:30:35 AM UTC-4, Steve Pope wrote:
> Richard Owlett  <rowlett@cloud85.net> wrote:
> 
> >The church I belong to is planning a website which will include 
> >recorded sermons. The proposal for site design and hosting did 
> >not explicitly mention how much online storage was included.
> >
> >A brief search told me a minute of MP3 occupied ~1 MB. No mention 
> >of fidelity. I assume that would be satisfactory for speech. What 
> >if they wanted to record choral music. I know essentially nothing 
> >about the details of MP3 [I date from when modems include an 
> >acoustic coupler ;].
> >
> >What search terms should I be using?
> >Any comments/suggestions/etc ?
> 
> Most listeners consider 320 kbps MP3 to be very acceptable for music.
> That comes to 2.4 MB per minute.  You might consider giving listeners
> a choice of that, or a slower rate such as 128 kbps, since their
> feed may be slow.
> 
> S.

unless you are an "audiophile"  an MP3 encoded at 128 kbps will be perfectly acceptable for stereo music.

For sermons, speech in mono,  64 kbps or less will be fine.

There are many MP3s on the internet encoded at various rates that you can listen to and judge for yourself.

 
I listen to music in the car at 128 kbps, and while I can occasionally hear a compression artifact, I  do not find it objectionable.

m

On 23.09.16 14.17, Richard Owlett wrote:
> A brief search told me a minute of MP3 occupied ~1 MB. No mention of
> fidelity. I assume that would be satisfactory for speech.

Depending on the details even less might be enough.

First of all, if you have only one channel, the required bit rate 
reduces by about 30%.
Using another codec, i.e. HE AAC Codec instead of MP3 could save another 
40%. But this one is not free. Using Ogg Vorbis instead saves only about 
15% compared to MP3 at comparable quality.

> What if they
> wanted to record choral music. I know essentially nothing about the
> details of MP3 [I date from when modems include an acoustic coupler ;].

Choral is one of tho most complex things for MP3. Mainly because our 
acousthesia is very sensitive to human voices. Bit rates in the order of 
185 kbps (VBR, lame encoder with --alt-preset standard) should give good 
results. Now you are at about 1,4MB/min.

> What search terms should I be using?
> Any comments/suggestions/etc ?

For speech only I would recommend single channel audio with about 100 
kbps MP3 for high quality results. If quality is subordinate and it is 
just for good understanding you could even use half of that.

You may look for "lame speech settings".


Marcel

<makolber@yahoo.com> wrote:

>For sermons, speech in mono,  64 kbps or less will be fine.
>
>There are many MP3s on the internet encoded at various rates that you
>can listen to and judge for yourself.
>
> 
>I listen to music in the car at 128 kbps, and while I can occasionally
>hear a compression artifact, I  do not find it objectionable.

This is all true, but why under-design if you don't have to.
Something like a choral mass (assuming the vocalists and instrumentalists
are virtuoso) can be much better at the higher rate of 320 kbps.
The goal of this genre of music is to immerse yourself in it, any
artifacts will detract.

Just my opinion.

Steve

On Fri, 23 Sep 2016 14:37:14 +0000 (UTC), spope33@speedymail.org
(Steve Pope) wrote:

><makolber@yahoo.com> wrote:
>
>>For sermons, speech in mono,  64 kbps or less will be fine.
>>
>>There are many MP3s on the internet encoded at various rates that you
>>can listen to and judge for yourself.
>>
>> 
>>I listen to music in the car at 128 kbps, and while I can occasionally
>>hear a compression artifact, I  do not find it objectionable.
>
>This is all true, but why under-design if you don't have to.

For the same reason that not everybody needs a 4k TV or computer
monitor.

>Something like a choral mass (assuming the vocalists and instrumentalists
>are virtuoso) can be much better at the higher rate of 320 kbps.

Why stop at 320kbps?   Isn't more better?

>The goal of this genre of music is to immerse yourself in it, any
>artifacts will detract.

Only if those artifacts are discernible to the listener.   Dropouts
from the any portion of the link not being able to continuously
support the rate are generally far more annoying than compression
artifacts.   Pick your tradeoffs wisely.

>Just my opinion.
>
>Steve

On 23.09.16 17.13, eric.jacobsen@ieee.org wrote:
>> Something like a choral mass (assuming the vocalists and instrumentalists
>> are virtuoso) can be much better at the higher rate of 320 kbps.
>
> Why stop at 320kbps?   Isn't more better?

Well, just because it is the maximum that MP3 can handle. ;-)

But you are right. 384 kbps MP1 sounds even better. MP1 reproduces even 
most waveforms rather exactly.

And 320 kbps MP2 is usually better than 320 kbps MP3 too, because MP3 
cuts some parts of the music regardless of the available bit rate. (No, 
I do not talk about band 21, i.e. >16 kHz)
But, well, people want MP3, so they get MP3; even though any standard 
conform MP3 decoder must be able to decode MP2 as well.


>> The goal of this genre of music is to immerse yourself in it, any
>> artifacts will detract.
>
> Only if those artifacts are discernible to the listener.   Dropouts
> from the any portion of the link not being able to continuously
> support the rate are generally far more annoying than compression
> artifacts.   Pick your tradeoffs wisely.

Fortunately even an old 1,4MB floppy can deal with this bit rate. 
Unfortunately it is full after half a minute. ;-)

But you are right. 320 kbps MP3 is not reasonable within this context. 
Strictly speaking it is not reasonable in any context, because it is 
superseded by MP2 at high bit rates.


Marcel

Notice I've deleted all replies? <chuckle>
Though deleted, they have educated me ;> !!

The "answer" [please note quotation marks] *AS ASKED* 
legitimately range from
less than "1 MB /sec" to ">3 MB /sec".

This leads to multiple questions:
  1. Does a host care how an audio file is encoded?
  2. What encoding can a modern browser handle transparently?
     [The last time I retrieved an audio file was when dial-up 
was SOP ;]

Assume I'm totally ignorant. A certain resident on PLUG might 
cheerfully certify same.

<eric.jacobsen@ieee.org> wrote:

>On Fri, 23 Sep 2016 14:37:14 +0000 (UTC), spope33@speedymail.org

>>The goal of this genre of music is to immerse yourself in it, any
>>artifacts will detract.

>Only if those artifacts are discernible to the listener.   Dropouts
>from the any portion of the link not being able to continuously
>support the rate are generally far more annoying than compression
>artifacts.   Pick your tradeoffs wisely.

Which is why I wrote in my first post to this thread, "You might consider 
giving listeners a choice of that [320 kbps], or a slower rate such as 
128 kbps, since their feed may be slow."

Selecting at semi-random a site I often listen to, KALX that being
U. C. Berkeley's radio station, they have four streams: Ogg at 128 kbps,
and MP3 at 128k, 56k, and 24k mono.

Steve

Richard Owlett  <rowlett@cloud85.net> wrote:

>Notice I've deleted all replies? <chuckle>
>Though deleted, they have educated me ;> !!
>
>The "answer" [please note quotation marks] *AS ASKED* 
>legitimately range from
>less than "1 MB /sec" to ">3 MB /sec".
>
>This leads to multiple questions:
>  1. Does a host care how an audio file is encoded?
>  2. What encoding can a modern browser handle transparently?

Unfortunately, assuming you want to support PC's, MAC's, and
both Apple and Android mobile devices, and you want this 
to stream on all these devices (as opposed to, or in addition to, allowing 
file download), you're going to have to support several filetypes
and possibly embed a player such a jplayer into your website.  

A typical Android phone with no extra apps installed will not
stream audio from a typical audio-streaming website, even when
it's an mp3 stream.  (This deficiency creates an entire business model 
for Tunein, a startup out of San Francisco, which basically is paid by 
broadcasters to create a website/app combination that reliably streams.)

For what you're doing a something like jplayer should handle the
streaming to most HTML5 browsers.

Another option in your case is to upload the content to Youtube
and link to it.  Youtube takes care of transcoding and streaming
to different devices.

Steve