Hi, I'm home brewing recording software. I know it's a log scale, but what
is the calculation to convert from a16 bit sample value to decibal?

Thanks in advance,
Aaron Anodide

"Bob Smith" wrote in message
...
Aaron Anodide wrote:

Hi, I'm home brewing recording software. I know it's a log scale, but
what
is the calculation to convert from a16 bit sample value to decibal?

Thanks in advance,
Aaron Anodide

dBA, dBB, dBC, dBSPL, dBu, dBV, dBFS, dBm, dBW, dBwtf?

What does cubase use on it's graphical meters?

Aaron


--
bobs
we organize chaos

Bob Smith - BS Studios
http://www.bsstudios.com/

Aaron Anodide wrote:

Hi, I'm home brewing recording software. I know it's a log scale, but what
is the calculation to convert from a16 bit sample value to decibal?

Thanks in advance,
Aaron Anodide

20*log10(|S|/32768)

where log10 means log base 10 and |S| is the magnitude of
the sample value.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein

"Mike Rivers" wrote in message
news:znr1058234293k@trad...

In article
writes:

Hi, I'm home brewing recording software. I know it's a log scale, but
what
is the calculation to convert from a16 bit sample value to decibal?

It would be nice if you learned how to spell decibel first.

geez... saw-ee.




--
I'm really Mike Rivers - )

In article ,
"Aaron Anodide" wrote:

What does cubase use on it's graphical meters?

I think they just slap the linear magnitude of the signal on them. That
usually fools most people and that is what it looks like.

You don't really think that they are doing true rms, ballistics and
calibrated dB calculations do you? There is far too much gratuitous
graphics to be done in one sample period...

"Graham Hinton" wrote in message
...
In article ,
"Aaron Anodide" wrote:

What does cubase use on it's graphical meters?

I think they just slap the linear magnitude of the signal on them. That
usually fools most people and that is what it looks like.

You don't really think that they are doing true rms, ballistics and
calibrated dB calculations do you? There is far too much gratuitous
graphics to be done in one sample period...

Granted, but as you move the fader in, and it reports such and such
attenuation in decibels, that couldn't be linear, because I don't think it
would fool anyone.

Aaron

In article ,
Aaron Anodide wrote:

"Graham Hinton" wrote in message
...
In article ,
"Aaron Anodide" wrote:

What does cubase use on it's graphical meters?

I think they just slap the linear magnitude of the signal on them. That
usually fools most people and that is what it looks like.

You don't really think that they are doing true rms, ballistics and
calibrated dB calculations do you? There is far too much gratuitous
graphics to be done in one sample period...

Granted, but as you move the fader in, and it reports such and such
attenuation in decibels, that couldn't be linear, because I don't think it
would fool anyone.

It might be logarithmic then, but it's probably got bizarre ballistics.
Then again, even if it just has peak hold for a while it might still be
useful.

You should know that since base 2 logs are usually faster to do that you
will gain some speed by translating that algorithm into a base 2 log and
a multiplication by a constant. And you should know that what you are
calculating is dBFS, that is decibels with respect to digital full scale.

Incidentally it's spelled "bel" because it's named after Alexander Graham Bell.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Graham Hinton wrote:

In article ,
"Aaron Anodide" wrote:

What does cubase use on it's graphical meters?

I think they just slap the linear magnitude of the signal on them. That
usually fools most people and that is what it looks like.

You don't really think that they are doing true rms, ballistics and
calibrated dB calculations do you? There is far too much gratuitous
graphics to be done in one sample period...

I've been digging thru the VST plugin SDK, and there are *some* ballistics
to 'em. I'd be skeptical that they're proper VU ballistics, but they made an
effort, at least.

--
Les Cargill

In article writes:

Incidentally it's spelled "bel" because it's named after Alexander Graham Bell.

Gee. All this time I thought it was named after Lucille Ball.



--
I'm really Mike Rivers - )

In article znr1058303074k@trad, Mike Rivers wrote:
In article writes:

Incidentally it's spelled "bel" because it's named after Alexander Graham Bell.

Gee. All this time I thought it was named after Lucille Ball.

Ay, Babalu!
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

huh. i never thought about DAW meters having "ballistics", and always
assumed they were peak, as in they just showed the actual magnitude
over a given buffer length, whatever it might be -- and that things
like buffer lenght, os draw functions, video refresh rate, persistance
of vision, etc., had as much to do with the ballistics as anything
else.

that is, i always assumed they kinda wired 'em up to show actual
magnitude in real time, albeit on a log scale.

by watching the meter on a DAW, there's the peak fluttery activity,
the peak hold value, and then the "persistant" level that hangs around
the rms value of the program (in sound forge, at least). i've gotten
pretty good at guessing rms from the meters . . .

cd /..




It might be logarithmic then, but it's probably got bizarre ballistics.
Then again, even if it just has peak hold for a while it might still be
useful.

You should know that since base 2 logs are usually faster to do that you
will gain some speed by translating that algorithm into a base 2 log and
a multiplication by a constant. And you should know that what you are
calculating is dBFS, that is decibels with respect to digital full scale.

Incidentally it's spelled "bel" because it's named after Alexander Graham Bell.
--scott

"Scott Dorsey" wrote in message
...
In article ,
Aaron Anodide wrote:

"Graham Hinton" wrote in message
...
In article ,
"Aaron Anodide" wrote:

What does cubase use on it's graphical meters?

I think they just slap the linear magnitude of the signal on them. That
usually fools most people and that is what it looks like.

You don't really think that they are doing true rms, ballistics and
calibrated dB calculations do you? There is far too much gratuitous
graphics to be done in one sample period...

Granted, but as you move the fader in, and it reports such and such
attenuation in decibels, that couldn't be linear, because I don't think
it
would fool anyone.

It might be logarithmic then, but it's probably got bizarre ballistics.
Then again, even if it just has peak hold for a while it might still be
useful.

You should know that since base 2 logs are usually faster to do that you
will gain some speed by translating that algorithm into a base 2 log and
a multiplication by a constant. And you should know that what you are

Thanks for the tip. Actually, it doesn't matter how fast the calculation
is, because I'm precalculating all the values and putting them in a table
(memory is cheap these days).

calculating is dBFS, that is decibels with respect to digital full scale.

Incidentally it's spelled "bel" because it's named after Alexander Graham
Bell.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

I have a related question (I think). I have an old porta-studio that has bit
the dust, ... but it's got nice VU meters. I'm wondering what would be
involved in salvaging them into some sort of "insert" for some of my other
mixers/recorders that only have digital meters. Is this possible, and can
they be calibrated?

chris b.


"Scott Dorsey" wrote in message
...
Mike wrote:
(Graham Hinton) wrote in message
...
In article ,
"Aaron Anodide" wrote:

What does cubase use on it's graphical meters?

I think they just slap the linear magnitude of the signal on them. That
usually fools most people and that is what it looks like.

You don't really think that they are doing true rms, ballistics and
calibrated dB calculations do you? There is far too much gratuitous
graphics to be done in one sample period...

Software level meters suck big time in my book. They are just useful
as a loose idea of where the signal is.

There is no excuse for this, though, because real level meters that have
controlled and known ballistics are not too hard to write.

Sonic has real meters (and even a phase meter that is trustworthy). A lot
of other DAW systems have plug-ins with real meters available.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

On 17 Jul 2003 13:56:52 -0400, (Scott Dorsey) wrote:

Mike wrote:
(Graham Hinton) wrote in message ...
In article ,
"Aaron Anodide" wrote:

What does cubase use on it's graphical meters?

I think they just slap the linear magnitude of the signal on them. That
usually fools most people and that is what it looks like.

You don't really think that they are doing true rms, ballistics and
calibrated dB calculations do you? There is far too much gratuitous
graphics to be done in one sample period...

Software level meters suck big time in my book. They are just useful
as a loose idea of where the signal is.

There is no excuse for this, though, because real level meters that have
controlled and known ballistics are not too hard to write.

Sonic has real meters (and even a phase meter that is trustworthy). A lot
of other DAW systems have plug-ins with real meters available.
--scott

Spectra Foo has a great metering system that allows you to adjust the
ballistics from VU to anything that you would like it to be.

In article ,
"chris berbaum" wrote:

I have a related question (I think). I have an old porta-studio that has bit
the dust, ... but it's got nice VU meters. I'm wondering what would be
involved in salvaging them into some sort of "insert" for some of my other
mixers/recorders that only have digital meters. Is this possible, and can
they be calibrated?

They are unlikely to be "real" VU meters which are moving iron, measure ac
directly and can be placed straight across a 600 ohm line, rather they will
be moving coil meters with a VU scale printed on them.

If you have the original manual with the circuit diagram see if you can
work out how Tascam drive them. You will need something like a balanced
line input, a full wave rectifier charging a capacitor with a bleed
resistor to discharge it and a scaling circuit to get the full scale
deflection. This may be as small as 1mA so be careful not to blow the meter
coil. Put in some trimmers so that you can adjust the meter against a known
signal.

It is a bit of work, but you could built all that around one quad op amp
and have it on a small card mounted on the back of the meter.
A decent new VU meter movement will cost over 50 pounds, or $75, so four
from a broken portastudio are worth salvaging. You might find better ones
from a more up market broken tape machine, but even a portastudio had
better meters than the crapp on cheap mixers and DAWs now.

IIRC, the 144 meters were hard to replace the bulbs.

In article ,
(Scott Dorsey) wrote:

There is no excuse for this, though, because real level meters that have
controlled and known ballistics are not too hard to write.

Not if you are RTW or DK Audio, but we are talking about Steinberg here...

Sonic has real meters (and even a phase meter that is trustworthy). A lot
of other DAW systems have plug-ins with real meters available.

You can't really expect Cubase to conform to professional standards and
have the same capability as systems based on cards of DSPs written by
people like Andy Moorer. Even if the people that buy/crack it do.