making the sound louder on files

[IKu]

is there a way to make the sound louder on mp3's and video files in the actual files rather than manually adjusting the volume? i don't like how some of my sound and video files are louder than others...

[ghost2003]

In iTunes if you right click a file and click get info, in the options tab there is a thing to change the default volume but I dont know if it will change it for all players of just iTunes.

[bailey]

not after there burned

[IKu]

yea, i'm just talking about files on my computer.

[TwoRails]

Is it always just one type of file? If so, try going into your mixer and making sure the volumn levels are all set to about 3/4 of the way up for starters.

[IKu]

no, it's various files, like maybe a video my friend put together vs something i made, one will be significantly louder our quieter i suppose because of the way we made it.

[TwoRails]

Then I don't know of a way without editing the file itself, and just crank up the volume, and / or when recorded, record it at a higher volume.

[bailey]

ok I'm sure gonna like this

how is a digital bit determe how loud its gonna be
as far as I know its eather a 1 or a 0, so where toes the voulme of its recording come in ?

[bailey]

double post
locked up

[ghost2003]

Well there has to be some info on how loud it is becaus they use frequency equalizers(I think thats what they are called) all the time, what would be the point if all ends up the same volume?

[bailey]

hey I am just asking , I don't know. but if it can be recorded at a grater volume level, what is changing ?

[TwoRails]

Originally posted by bailey ok I'm sure gonna like this how is a digital bit determe how loud its gonna be as far as I know its eather a 1 or a 0, so where toes the voulme of its recording come in ?

Because low volume bits are only a half-bit! LOL

My first instinct was to say it's in the file header, but then, when you increase the amplitude using an editor, it does a recoding of the file or the section you're editing. That would tell me that, like a photo, a music file is read as bytes or words, or even maybe a nibble, but not bits. That would indeed yield amplatude and other info.

TwoRails

[bailey]

great answere there
something is a half bit
LOL

I was just wondering what the response would be.

I could look it up or even try it out myself.

[TwoRails]

If you check it out, let me know if I was right, OK? don't see how else it would work.

[IKu]

so... the only way is to re-record it at a higher volume...?

[TwoRails]

That's the only way I know.

[SonicVanguard]

Volume of any give media file is dependant on the creator of the file. There is no way of normalizing all your files because there would be no base reference.

And bailey - you're right, audio data is nothing but 0 and 1, but those are relative to amplitude as it relates to unity gain (0 dB).

Dave.

[TwoRails]

Originally posted by SonicVanguard ... And bailey - you're right, audio data is nothing but 0 and 1, but those are relative to amplitude as it relates to unity gain (0 dB). Dave.

Please excuse my lack of understanding. Everything stored in a computer is 1s and 0s, so I'm still not clear as to where audio files get the reference to be relative to for amplitude. If the reference was in the file header, then changing the amplitude would take a split second. My guess is that, since you have to recode to change the amplitude, this reference point is stored in the data part of the file, not the header. That suggests that the data part is larger than just one bit because as bailey points out, it can be only 1 or 0. Just like a scanner can be, for example, 24 bit or as high as 48 bit, each chuck of data is larger than one bit. I haven't a clue to the data sizes in any type of audio file, but it seems that it has to be greater than one bit to have any reference points at all. Again, I am just taking a WAG.

[SonicVanguard]

Well, I put it as simply as I could. Unity gain is a set standard (0 dB). The very first sound in a sound file has a given amplitude - yes it's a 1 or 0, at a given marker. Volume then throughout the song/video/animation or whatever is based from that first reference. So you could say the reference is in the header - although audio files in and of themselves do not have headers.

Dave.

[TwoRails]

Thanks for the input. Sometimes simply put things go over the head of simple people, like me. This is what I don't understand: If the reference is in the "header," then when changing the "level," why does the whole (or the selected part) need to be recoded?

Here's an example for a photo: If I change the file size, it is "recoded." If I change just the DPI size, that's in the heading, and there is no recoding; just the size is reset.

Sorry to be a pain, I'm just trying to understand.

[SonicVanguard]

You've got to get away from comparing a digital photo to digital audio. It's comparing apples to oranges.

If I have two audio files of an identical song, one recorded with a nominal level at unity gain and the other at -6 dB, the files sizes are the same (nominal being it's loudest point). Although a audio file is binary data, that binary data represents a waveform (it must or it would not be audio). That waveform contains the quantized information that correlates directly to amplitude.

Let's take a look at these two waveforms: waveform1 and waveform2. They are identical waveforms from the same audio file - but one is normalized at -6dB while the other is at unity gain. I think it's pretty easy to tell which one is which. If you look closely, you can see the entire waveform of this particular song in a smaller window above the highlighed waveform. Comparing the two, you would think waveform1 (nomalized at unity gain) would be the larger file. Yet both audio files take up exactly 6,111,232 bytes.

Dave.

[TwoRails]

Thanks for the details and JPEGs, SonicVanguard. I used digital photo as an example because I was trying to relate to something I'm familiar with, as it helps me understand

[SonicVanguard]

Originally posted by TwoRails Thanks for the details and JPEGs, SonicVanguard. I used digital photo as an example because I was trying to relate to something I'm familiar with, as it helps me understand

No harm, no foul. I looked at what I posted and realized it came across somewhat brutal. Sorry about that.

But for the real question: Are you understanding the concept? If you're not and you want me to try and define it differently, let me know.

Dave.

[TwoRails]

Like you said, no harm no foul -- It's hard to see when someone is smiling or not with the written word, and I'm not a great writer and often get misunderstood / mistaken as a result. No offense meant or taken.

I did some research and been to about 8 - 9 different sites on the file structure. Wave files (and apparently all sound files) do indeed have what I call a file “header,” but it’s called “chunks” instead.

The header holds such info as to the type of sound file (Wave, PCM, etc.), number of channels, bit and sample rates, and so on. It can even hold artist info and the like. There is no mention at any site about amplitude or gain in the file header.

The actual “data” section of the file is, of course, the actual sound info. Finding info on headers was easy, but finding the structure of the data section was not very fruitful. I found only one side that discussed midi file structure and there it did state that the gain / amplitude for the sample is in the data “sub-chunk.”

Since gain is not included in the header of a Wave file, and it is included in at least the midi structure, it also must be in the data section of a Wave file. This correlates with the fact that to change the gain of a file, it is recoded. Since it is a change in a bit(s) of the data byte(s), then the file size would – and does, as you have shown – stay the same.

Also, the data is indeed in byte (not bit) sized chunks and can be a byte, a word, a double word, and so on. If the info is not an even byte’s worth of info, it is rounded up to the next byte boundary.

TwoRails

Here’s just one link showing a Wave PCM header. I picked this one simple because it has a nice layout without going into too much detail:

http://ccrma.stanford.edu/CCRMA/Cour...ts/WaveFormat/