"MP3Gain: How can it be possible?", It 's indicated that the gain adjustments are lossless Options

[2Bdecided]

Interesting how, on a generally friendly board, a conversation can go so wrong.

There's actually often friction with newbies here. Get off my lawn type of reactions.

Yeah, well, I bought a shed last week. So I can be as middle-aged as I want now!

Seriously, maybe we need to be more friendly?

There are at least two things that provoke a bad response though. The anti-DBT/ABX brigade (which, being generous, is "just" a different viewpoint, albeit one fundamentally opposed to the founding principles of HA), and people who think they know something when they don't, and refuse to learn from the people who invented the stuff! There's often overlap between the two of course.

Cheers,
David.

[Destroid]

Seriously, maybe we need to be more friendly?

There are at least two things that provoke a bad response though. The anti-DBT/ABX brigade (which, being generous, is "just" a different viewpoint, albeit one fundamentally opposed to the founding principles of HA), and people who think they know something when they don't, and refuse to learn from the people who invented the stuff! There's often overlap between the two of course.

I forgot to mention your name in my previous post. My apologies. There are literally dozens of others I didn't remember to mention. So, may I convey a direct thank-you as well

As far as the ABX issue, I think it's OK to take a hard-line. This is me being the lossless-oriented person

In regards to the informative stance, it might be taken into account that persons finding this forum are likely searching "high-quality MP3" and "LAME," and that the understanding of the technology is about HQ audio with a catch but not really sure what the "catch" is all about (meaning lossy encoding).

I thought MP3Gain was a scale-factor thing (which I recently got another crash-course regarding CBR).

[bandpass]

The anti-DBT/ABX brigade (which, being generous, is "just" a different viewpoint, albeit one fundamentally opposed to the founding principles of HA)

Yes, if they decide HA is not for them, ideally they would do so understanding that it's "just a different viewpoint" from theirs, and not go off spreading the word that we're a bunch of self-righteous bastards (as I'm sure some do).

[greynol]

Allow me to throw another reason into the mix as to why this one went sideways. It is insulting to be told that something that took you years to learn can be trivialized in a few short sentences. This is especially true here as the questions asked require a broad background to answer.

I steered clear of the flamebait this time around. Last time I showed this poster where he went wrong and explained what was right I was called a douche. I suppose I should not have addressed his fallacy of arguing from authority in the way that I did. While here it was "I'm smardt" there it was "I hasses golden earz".

This post has been edited by greynol: Aug 3 2012, 16:26

[alanofoz]

You're saying full scale is not maximum amplitude?

Yes. To be specific, I'm saying that for the 32 bit floating point format 0dBFS is << the maximum possible amplitude.

[[JAZ]]

You're saying full scale is not maximum amplitude?

Yes. To be specific, I'm saying that for the 32 bit floating point format 0dBFS is << the maximum possible amplitude.

Ok, now that is clearer, and you are correct:

Max: 3.4028234 × 10^38
Min non denormal: 1.18 × 10^−38
absolute non-denormal range: 20 * log10(1.18 × 10^−38/3.4028234 × 10^38) = -1526.18dB.


But your initial sentence was misleading, since you seemed to imply that, when talking about audio, the full amplitude was used, which is not true. So my value of theoric SNR is correct because what matters is 0dbFS, not max amplitude.

[alanofoz]

Hmmm... I re-read my post and didn't think it misleading at all - unless you take that sentence out of context. But I think when you consider that I then said that I artificially created a file with a 1500dB dynamic range, and followed that with Not much practical use though..., there's no implication of full amplitude being used in practice. (I have occasionally dealt with 32 bit files with > 0dBFS, but that's another story.)

What I do think misleading though (sorry), was

...Difference from max value to min value is: -758.56dBFS...

and that's what I was reacting to because, as I said, I think it's more like 1500dB, in fact the figure you have now calculated. If you had said

...Difference from 0dBFS to min value is: -758.56dB...

I wouldn't have batted an eyelid.

OTOH if you had said that my post was pedantic, you'd be right (ask my wife, although I call it attention to detail...).