On Wednesday, March 10, 2010 8:46 AM, Ted Kendall wrote:
> I'm a little disturbed by some of the assumptions here...
AES Convention Paper 5534, S.W. Davies
Equalisation for Archival Transfer: In the Analogue or Digital
See page 3, "Disadvantages" [of equalisation in the digital
AES Convention Paper 7185, R.S. Robinson
Filter Reconstruction and Program Material Characteristics
Mitigating Word Length Loss in Digital Signal Processing-Based
Compensation Curves used for Playback of Analog Recordings
See page 2, Section 2, "Word Length Truncation"
> And as for loss of dynamic range - I think we would be hard put
> to exceed 90dB between the Johnson noise of the pickup cartridge
> (never mind disc noise) and the highest reproducible peak off
> the disc.
If we look at just the bass region for RIAA, the RIAA playback
EQ will apply up to a 19 dB boost @ 20 Hz. If you quantize
the bass region, that 19 dB is represented by 3 bits (6.02 dB
per bit, whether you use 8 bits, 16 bits or 24 bits in your
Now repeat the exercise by making a FLAT transfer. The 19 dB
of dynamic range in the bass is no longer there in playback.
What happens to the bass now that it gets *quantized* with a
FLAT transfer? What just happened to that 19 dB of dynamic
range - how is it now represented once quantized by the ADC?
On the other end of the spectrum, you need an extra 19 dB of
headroom to accomodate the treble emphasis that occurs in a
> The last time I looked, 24 bits gave a theoretical dynamic
> range of 144dB, and commercially available convertors are
> getting within spitting distance. The drop from one end of
> the spectrum to the other on RIAA is 40dB.
The dynamic range of the ADC is not the issue at hand. The
issue is that a FLAT transfer reduces the dynamic range in
the bass and requires increased headroom in the treble.
This is the source for word length loss.
See the above AES references.
The Audio Archive, Inc.
mailto:[log in to unmask]
Disc and Tape Audio Transfer Services and Preservation Consulting
----- Original Message -----
From: "Fred Thal" <[log in to unmask]>fdshg7
To: <[log in to unmask]>
Sent: Wednesday, March 10, 2010 2:26 PM
Subject: Re: [ARSCLIST] Disc EQ in the digital domain
> Hello Eric,
>> Transfering FLAT will cost you about 6-7 bits of dynamic
>> range - an audible loss. EQ in the digital domain will not
>> recover that lost dynamic range.
> Can you describe the mechanism by which the dynamic range is lost in
> this process? Or can you point me to any relevant literature that
> fully explains this?
> Thanks in advance.
> Fred Thal
> ATAE / Audio Transfer Laboratory