Posted by Smarty on 10/07/05 07:16

Steve,

The original post specifically asked about "a little hand held digital
camera", not the PD150 camera style with low impedance balanced inputs, and
switchable AGC. I don't disagree with your solution for a balanced output,
since 5 dB of 'disagreement' in our opinions is well within the nominal
range of line level inputs which the original poster is likely to encounter.

The unbalanced input of a typical consumer camcorder is, however,
considerably more susceptible to hum and noise than the balanced XLR inputs
of the pro and some prosumer camcorders, and moreover has (almost always) a
non-defeatable AGC. Thus, a bit less attenuation, and putting a bit higher
signal into this high impedance unbalanced input will result in better
signal to noise performance than driving it with a smaller signal and thus
relying on the AGC to a greater extent to bring up the gain. It is a
trade-off of distortion for SNR, and the specific camcorder's
characteristics should really be accounted for, particularly its' AGC, since
those simple AGC circuits which tend to overshoot or pump or "breathe" can
and will sometimes add more objectionable effects than the hiss and hum and
other background noise which arise when a weaker (more highly attenuated) is
used.

Smarty




"Steve King" <steveSPAMBLOCK@stevekingSPAMBLOCK.net> wrote in message
news:7sGdnSmB0rMeBtjeRVn-pw@comcast.com...
> "Smarty" <nobody@nobody.com> wrote in message
> news:iv-dnTLi_8s_CNjenZ2dnUVZ_sidnZ2d@adelphia.com...
>> Most if not all of the camcorders do AGC on the audio inputs, and have at
>> least 60 to 80 dB of dynamic range. Virtually all the camcorders use
>> unbalanced inputs. I have build T-pad attenuators with a simple
>> 3-resistor T-network to do precisely what the original post asked for,
>> and the attenuators which do around 30 or so dB seem to work just fine.
>> Putting in too much attenuation forces the AGC to "gain-up" and crushes
>> the SNR by raising the noise floor.
>>
>> Smarty
>>
> I have found that Jay Rose is right on, when his tests pointed to an input
> signal of -30 db as a sweet spot for VX2000/PD150 type cameras. If I
> recall correctly, his test set-up was +4 db out of a balanced mixer with a
> 35 db attenuator following.
>
> Steve King
>

[Back to original message]