Mp3table
Dominic Richens
news:3d521e20$1...@news.comindico.com.au...
> For all you dudes that don't get it, the original post was a troll.
> So don't take it seriously.
Really?
Given that a 100k mp3 of a sample will be of higher quality than a 100k wav
of the same sample, doesn't Vikram have a point? Also, if you only have 32M
of sample RAM, couldn't you do a better grandpiano with mp3 encoded
SoundFonts, since you could have more samples with better cross-fading and
velocity-fading?
Bryce Fischer
<dominic...@sympatico.ca> wrote:
>"js" <j...@hotmail.com> wrote in message
>news:3d521e20$1...@news.comindico.com.au...
>> For all you dudes that don't get it, the original post was a troll.
>> So don't take it seriously.
>
>Really?
>Given that a 100k mp3 of a sample will be of higher quality than a 100k wav
>of the same sample, doesn't Vikram have a point?
explain this please... I don't see how a 100k mp3 of a sample will be
of higher quality than a 100k wav.
Please provide supporting documentation along with citations. Because
I'm not taking the fact that what you say above is "given"
--
Bryce
I don't like having disks crammed into me...
unless they're Oreos, and then only in the mouth
Jeffery S. Jones
<dominic...@sympatico.ca> wrote:
Both mp3 and wave as the file size gets smaller compromises on
frequency response. At the point where the rough sonic
characteristics are similarly useful, the size difference isn't that
much. High data rate mp3s vs. high compression wav (note that samples
tend to use some method *other* than uncompressed raw data) may be no
better than 2-5x. At the low end, you have to ask yourself whether
doubling the sample space is worth compromizing on acoustic clarity.
Plus, if you're talking about designing a new sampler to use mp3 or
some superior lossy compression format, I think you're up against the
fact that more RAM comes with new tech. If you are talking a PC based
sampler, we're already in the gigabyte sample range -- what benefits
would you get from the small savings that using mp3 would offer?
I mean, other than reducing the effective quality of the samples to
shoehorn in more simultaneous sounds? I can't see where a gigabyte
piano sample set is going to somehow sound much better just because
you've included a few more levels of sample in mp3 vs. higher quality
unlossy sounds. Esp. for stereo samples, where lossy compression
loses clarity of location.
--
*-__Jeffery Jones__________| *Starfire* |____________________-*
** Muskego WI Access Channel 14/25 <http://www.execpc.com/~jeffsj/mach7/>
*Starfire Design Studio* <http://www.starfiredesign.com/>
*Graphic Reflections and Websites* <http://www.execpc.com/~jeffsj/>
Laurence Payne
halving the sample rate or halving the bit depth. Or by compressing
as a mp3. Which would sound better?
>explain this please... I don't see how a 100k mp3 of a sample will be
>of higher quality than a 100k wav.
--
My Cubase FAQ page is
www.laurencepayne.co.uk/CubaseFAQ.htm
Feedback welcome.
Christof Pflumm
Very bad idea IMO. Mp3 uses psychoacoustic effects like masking of
soft sounds by louder ones. Mp3 throws away information. That's
o.k. for songs that you want to listen to, because you can't hear the
discarded information anyway (that's the whole point of the
compression).
But if you use it for samples, you can't be sure what happens when you
mix those samples. It could be that you have thrown away a lot of
information that you need in a mix with other samples.
Bye,
Christof
Dominic Richens
> <dominic...@sympatico.ca> wrote:
> >
> >Given that a 100k mp3 of a sample will be of higher quality than a 100k
wav
> >of the same sample, doesn't Vikram have a point?
>
> explain this please... I don't see how a 100k mp3 of a sample will be
> of higher quality than a 100k wav.
See one of the other posts. To paraphrase what they said, MP3 compressess
by throwing away information our ears apparently don't need. WAV file is
(according to some) 9/10 wasted information. Given my sample is 3 seconds
long, I'll need to encode to 256kbps. For wave, this would be 8bits at
32kHz or 16bits at 16kHz. Given the typical 10:1 compression ratio of MP3,
the sampling could be 24bit at 96kHz, then compressed to 256kbps.
It might sound better to sample at only 16bits/44.1kHz and compress less, I
dunno.
> Please provide supporting documentation along with citations. Because
> I'm not taking the fact that what you say above is "given"
You're kidding, this USENET, not "Nature".
Seriously, though, I'm not saying it is a good idea, I just wanted to get
the discussion on track, instead of it ending with "it's a troll".
Rob
>
> "Dominic Richens" <dominic...@sympatico.ca> writes:
>
> > "js" <j...@hotmail.com> wrote in message
> soft sounds by louder ones. Mp3 throws away information. That's
> o.k. for songs that you want to listen to, because you can't hear the
> discarded information anyway (that's the whole point of the
> compression).
>
> But if you use it for samples, you can't be sure what happens when you
> mix those samples. It could be that you have thrown away a lot of
> information that you need in a mix with other samples.
>
professional skill!) depends on careful working of the interplay of ALL
frequency components of the tracks being mixed. So, e.g. the sound of a
guitar will have an effect of the sound of a piano mixed with it. Also,
all sounds interact,
it's called "colouration". You get constructive and destructive
interference. Remove frequencies and you will lose this effect and the
overall sound will be dreadful. It matters not that you are removing
"sounds the ears can't hear, or the brain puts back", the fact is you
ARE removing frequency components and these are missing from the AUDIO
signal.
If you have tried to take an mp3, render to .wav, record and mix in a
new sound, you will know that it does not work as well as using the raw
audio.
Also, re-encoding just leads to more destruction of the audio.
Therefore, a NON-STARTER for any serious audio work.
My first and last word.
rob.
3rd Album
Jeffery S. Jones <jef...@execpc.com> wrote:
> On Tue, 13 Aug 2002 13:31:52 -0400, "Dominic Richens"
> <dominic...@sympatico.ca> wrote:
>
> >"js" <j...@hotmail.com> wrote in message
> >news:3d521e20$1...@news.comindico.com.au...
> >> For all you dudes that don't get it, the original post was a troll.
> >> So don't take it seriously.
> >
> >Really?
> >
> >Given that a 100k mp3 of a sample will be of higher quality than a 100k wav
> >of the same sample, doesn't Vikram have a point?
No, since you can't have uncompressed MP3s. So WAV's highest quality
will be much better than MP3's highest quality.