PDA

View Full Version : "Line level"



jim henderson
01-06-2004, 10:09 AM
My understanding is that for home audio equipment, “line level” is nominally 1 volt = 2 volts P-P = .707 RMS, full-scale. Is that correct?

I made a CD of full-scale test tones: 16-bit samples at 44.1 KHz sampling rate with values range from a low of x8000 to a high of xFFFF. When I play it on my CD and DVD players, I measure approx. 0.17 volts RMS output. Why isn’t it approx. .707 volts RMS?

I’m test driving an Ashly 4.24G DSP with 24-bit A/Ds and D/As, and a 48KHz sampling rate. The manual has this paragraph:

“There are no analog gain trim adjustments on the Protea processors, therefore all the processing (including gain) is done in the digital domain. The Protea processors were designed this way so that when digital inputs/outputsbecome available, there will not be any gain stages left out of the all digital signal path. As a consequence of this design philosophy, it is important to feed the Protea processor with the proper nominal signal level to achieve good signal to noise performance as well as headroom before clipping. The Protea processors were designed to clip at signal levels above +20dBu = 7.75Vrms which places the noise floor lower than -90dBu with the current 24-bit A/D and D/A converters. The optimum input signal level which should be fed into the Protea processor is 0dBu = .775Vrms. This input level will allow 20dB of headroom while giving a nominal signal that is 90dB above the noise floor.”

If my source equipment provides .707 volts RMS full-scale, isn’t the Ashly DSP wasting 20 dB (approx. 3 bits) of input resolution?

Oldmics
01-06-2004, 09:00 PM
Hi Jim,I will attempt to answer your questions.I sense that you are looking for answers to other questions not yet asked.


My understanding is that for home audio equipment, “line level” is nominally 1 volt = 2 volts P-P = .707 RMS, full-scale. Is that correct?

"Line level" is a loose term generally refered at +4 volts.This is measured at 1000 Hertz terminated across a 600 ohm static load.

I made a CD of full-scale test tones: 16-bit samples at 44.1 KHz sampling rate with values range from a low of x8000 to a high of xFFFF. When I play it on my CD and DVD players, I measure approx. 0.17 volts RMS output. Why isn’t it approx. .707 volts RMS?

Compression circuits come into factor here-depending upon the CD burner.


The Protea processors were designed to clip at signal levels above +20dBu = 7.75Vrms which places the noise floor lower than -90dBu with the current 24-bit A/D and D/A converters. The optimum input signal level which should be fed into the Protea processor is 0dBu = .775Vrms. This input level will allow 20dB of headroom while giving a nominal signal that is 90dB above the noise floor.”

If my source equipment provides .707 volts RMS full-scale, isn’t the Ashly DSP wasting 20 dB (approx. 3 bits) of input resolution?


The source should be calibrated at .775.Let us however not argue over this .The 20 db of headroom is close enough to consider correct for this conversation. Wasted headroom ??? Not really when you consider the application that the Protea was intended for.That would be live reinforcement.This is where the 20 db of headroom would get utilised.Wild transients of snare drum and explosive vocal dynamics easily absorb the 20 db safety margin on the front end of that unit.Remember that + 20 db means the unit is clipping!

There are many different opinions about digital preformance in the audio chain throughout the members of this forum.I do not know your intended application for this unit-pro sound or hi fi ??
I will comment and say listen to all of the other digital items in that price range before you make a final decision on the Ashly.

Best regards,Oldmics