RIAA encoding Options

[doctorcilantro]

Discussing some of the various digital RIAA decoding methods, someone asked,

"BTW, if digital RIAA decoding is more accurate than any analog EQ, then I assume the original LPs' RIAA encoding is less accurate. This may be a case of the playback equipment's having more accuracy than the source is capable of delivering."

Thoughts? Thanks, and have a great weekend.

[Soap]

Discussing some of the various digital RIAA decoding methods, someone asked,

"BTW, if digital RIAA decoding is more accurate than any analog EQ, then I assume the original LPs' RIAA encoding is less accurate. This may be a case of the playback equipment's having more accuracy than the source is capable of delivering."

Thoughts? Thanks, and have a great weekend.

Unless you know what equipment was used to apply the RIAA encoding in the first place you only have one option: To apply the RIAA decoding in as accurate a manner as possible.

Without the knowledge of any (potential) flaws (deviation from the design curve) used during "encoding" any straying on your part from the ideal is only going to add more deviation.

EDIT: I assume your goal is to attempt as accurate a reproduction of the artistic intent as possible. If you're trying to achieve your own point of maximum sonic euphoria than this is all moot - color as you desire.

This post has been edited by Soap: Sep 3 2010, 16:57

[pdq]

Discussing some of the various digital RIAA decoding methods, someone asked,

"BTW, if digital RIAA decoding is more accurate than any analog EQ, then I assume the original LPs' RIAA encoding is less accurate. This may be a case of the playback equipment's having more accuracy than the source is capable of delivering."

Thoughts? Thanks, and have a great weekend.

I assume that these days RIAA encoding is probably done digitally. Of course, the analog process of translating this to grooves in a record surely adds some inaccuracy.

Also, the bulk of RIAA decoding has to do with the pickup being a velocity-sensing device rather than an amplitude-sensing device, so RIAA encoding applies a much smaller correction than decoding.

[2Bdecided]

I assume that these days RIAA encoding is probably done digitally.

Why would you assume that? Just asking - it doesn't seem to make much sense.

Analogue RIAA circuits are hardly the weak point in the chain.

Someone once made a big deal out of designing a "phase correct" RIAA circuit, but I'm not convinced that the bog standard circuits are phase incorrect.

De-clicking should work better before you "correct" the phase of the music, but all declickers are designed to work well enough with the phase of the clicks messed with by an RIAA circuit - they just treat it as part of the impulse response of the system (or they should do!).

Cheers,
David.

[doctorcilantro]

I assume your goal is to attempt as accurate a reproduction of the artistic intent as possible.

yes

[DVDdoug]

Analogue RIAA circuits are hardly the weak point in the chain.

Exactly what I was thinking... The curve was designed in the "analog days" the match something easily achievable with analog components. It's just a matter of using precision resistors & capacitors, and perhaps a trim-pot or two to make-up for any remaining deviations. I'd guess that the recording curve was typically accurate within 1dB.

My feeling is that the weakest link was not the equipment, but rather the humans that didn't care as long as the sound was "good enough for AM radio". It seems like it was possible to make high-quality records because once in a while, you'd find a gem that sounded clear & clean, and many classical records were carefully produced to have good sound. But, most of the rock records that I bought sounded somewhat "dull".

[Paul Sanders]

declickers are designed to work well enough with the phase of the clicks messed with by an RIAA circuit - they just treat it as part of the impulse response of the system (or they should do!).

Being mostly just impulses, clicks don't really have a phase. Their duration is too short for that. All the declickers I know of use a statistical method for detecting clicks, based on a short-term analysis (just a few samples) of the high-frequency content of the signal relative to a longer term rolling average. An FFT-based approach is also possible but very slow. I don't know in detail how it works.

[Arnold B. Kruege...]

I assume that these days RIAA encoding is probably done digitally.

Why would you assume that? Just asking - it doesn't seem to make much sense.

Analogue RIAA circuits are hardly the weak point in the chain.

Developing a correct analog RIAA network or its inverse is non-trivial because the component values interact, and was frquently done incorrectly back in the day. Lipshitz tried to set things right with a landmark article that was published in the JAES: S. Lipshitz, “On RIAA Equalization Networks”, JAES 1979.

Article about analog RIAA and inverse RIAA networks

Yes, back in the day it was not uncommon to have an RIAA network that introduced audible errors.