#40980 - 10/18/02 06:54 PM
Optimum Analog Input level for DSP?
|
Deputy Gunslinger
Registered: 10/18/02
Posts: 4
Loc: Long Beach, CA., U.S.A.
|
When using any DSP mode with true analog inputs (cassette, R-R, phono, tuner, etc.)it is my understanding that those inputs are digitized at 48KHz sampling rate and 24bits PCM.
Obviously the 950 has some sort of AGC (automatic gain control) since other A/D units have had manual input level controls which had to be set by the user. Anyone who has ever tape recorded knows how critical recording levels are for maximum S/N level.
Since my compliment of analog equipment outputs nominal 0 dB levels ranging from 250Mv to 2.0 V (which represents an 18dB voltage level difference). It seems reasonable that some optimum input voltage range would produce the best S/N ratio and best use of the dynamic range available with the 24bit A/D converters?
Does anyone have a clue as to what the true situation is regarding this issue? That is, what analog input levels best 'feeds' the 950's A/D converters for DSP processing of analog inputs?
TIA, John Whittaker
|
Top
|
|
|
|
|
#40981 - 10/18/02 07:38 PM
Re: Optimum Analog Input level for DSP?
|
Desperado
Registered: 04/10/02
Posts: 1857
Loc: Gusev Crater, Mars
|
I very seriously doubt that there is any form of automatic gain control on the input of the 950s analog to digital stage (yes, it is 48Khz and 24 bit). They are undoubtedly relying on the maximum output of any device connected to the DSP inputs being in the neighborhood of 2.0 volts. As you might know, in the analog DSP mode, the noise (hiss) level of the 950 is at a worst case level. Perhaps being a bit conservative on the sensitivity of the ADC is why this mode is hissier than the other modes. Of course _theoritically_ a 24 bit digital word has a S/N ratio of 144db - clearly this is bullshit in the real world, as no possible device on the planet Earth, lacking perhaps liquid nitrogen cooling to near absolute zero degrees, could be this quiet. At least the lower 4 bits are recording nothing more than random electron noise, especially in an ADC that cost only a few dollars. If course the marketing department would disagree [This message has been edited by soundhound (edited October 18, 2002).]
|
Top
|
|
|
|
|
#40982 - 10/18/02 09:02 PM
Re: Optimum Analog Input level for DSP?
|
Deputy Gunslinger
Registered: 10/18/02
Posts: 4
Loc: Long Beach, CA., U.S.A.
|
Soundhound,
Sure, and since we don't have liquid nitrogen cooling, then there is going to be some input voltage (from and analog input) that will provide optimum S/N and dynamic range from the A/D conversion. The question is - what is that maximum voltage?
As to an AGC, I've been wondering about all the hiss issues which seem to be somewhat associated with analog inputs (if I've read correctly)? An AGC process boosting low voltage analog inputs could account for some increase in noise - but as you note, who knows?
Time maybe to ask the Outlaws?
John
|
Top
|
|
|
|
|
#40983 - 10/19/02 12:01 AM
Re: Optimum Analog Input level for DSP?
|
Desperado
Registered: 04/10/02
Posts: 1857
Loc: Gusev Crater, Mars
|
Originally posted by John Whittaker: The question is - what is that maximum voltage?
The maximum input voltage is 2.63174821 Volts Actually, I can only make an educated guess that it is somewhere around the 2 volts that a standard CD player puts out. This would be easiest from an engineering point of view. I would think that an AGC would be kinda messy to engineer, one obvious problem being 'how long of a time constant do you build in?' If it wasn't right, the volume of the music would be rising and falling at most inappropriate times as the AGC hunts down the optimum input voltage. And what about classical music where there is one very quiet movement that goes on for minutes, followed by a sudden loud fortissimo? I just don't think any kind of AGC circuit would work well enough without messing with the dynamics of all the types of music and movie soundtracks out there. Of course, they have 144 db of dynamic range to play around with, so they don't need no stinking AGCs! If my 950 wasn't in a very inaccessable rack, I would bench test this parameter; it would be easy to do. Of course, if they made the maximum input voltage really high, like 5 or 6 volts, the result would be that the analog to digital and back to analog conversion would be noisier than optimum. This _could_ be one reason the 950 is hissy in that mode. I have horn speakers that have 106db / 1 watt efficiency, and I can hear the hiss about 5 feet out - I consider myself lucky, as it could be a lot worse!
|
Top
|
|
|
|
|
#40984 - 10/19/02 05:07 AM
Re: Optimum Analog Input level for DSP?
|
Desperado
Registered: 12/11/01
Posts: 1054
Loc: Santa Clara, CA
|
I can probably educated guess that the 950 has no AGC...
AGC IMO of any kind stinks. Had a portable cassette recorder with it a loooooong time ago.
_________________________
If it's not worth waiting until the last minute to do, then it's not worth doing.
KevinVision 7.1 ... New and Improved !!
|
Top
|
|
|
|
|
#40985 - 10/21/02 11:17 AM
Re: Optimum Analog Input level for DSP?
|
Desperado
Registered: 11/29/01
Posts: 1434
Loc: Mount Laurel, NJ
|
No, maybe not, but MGC (manual gain control) is kind of nice. There should be more of them. They should be everywhere.
------------------ Matthew J. Hill matt@idsi.net
_________________________
Matthew J. Hill matt@idsi.net
|
Top
|
|
|
|
|
|
0 registered (),
653
Guests and
1
Spider online. |
Key:
Admin,
Global Mod,
Mod
|
|
|
8,717 Registered Members
88 Forums
11,331 Topics
98,708 Posts
Most users ever online: 900 @ Today at 03:23 PM
|
|
|
|