CD-4 demodulation

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

ingresman

300 Club - QQ All-Star
Joined
Apr 24, 2006
Messages
321
Location
Manchester UK
Just an ideol thought....
If you had a turntable with a CD cartridge and a PC with a good enough sound card to sample up ro 48khz (if such a thing exists ?) would it be possible to capture the CD-4 track and demodulate in software ?
 
Yes and no.
Yes for the first part (record the CD4), No for the second (software demod).
 
Just an ideol thought....
If you had a turntable with a CD cartridge and a PC with a good enough sound card to sample up ro 48khz (if such a thing exists ?) would it be possible to capture the CD-4 track and demodulate in software ?

In order to capture a 45khz frequency digitally, the sound card would have to sample at twice that ... 90khz ... so you would have to record it at 96khz ... but you cannot demodulate with software. Get a demodulator on eBay for $50. If you have a suitable turntable and cartridge, you are almost there anyway. Mike.
 
thanks for that guys,
Just of out interest why could you not demodulate in software?
I had a quick look at wikipedia (http://en.wikipedia.org/wiki/Quadraphonic#CD-4_.2F_Compatible_Discrete_4_.2F_Quadradisc) and it confirms my own (30 yr old now) that CD-4 was based around summation and difference and so should be fairly easiliy decoded with some algerbra.
E.g Stero radio is brodcast as L+R in the normal audio range to give a mono signal and L-R is broadcast somewhere above 15k and the 19khz for the stereo indicator.
At the receiver the L+R and L-R signals are added for the Left channel ((L+R)+(L-R)=2L) and subtracted for the right channel ((L+R)-(L-R)=2R).
Would it not be a case of find the maths and off we go ?
 
thanks for that guys,
Just of out interest why could you not demodulate in software?
I had a quick look at wikipedia (http://en.wikipedia.org/wiki/Quadraphonic#CD-4_.2F_Compatible_Discrete_4_.2F_Quadradisc) and it confirms my own (30 yr old now) that CD-4 was based around summation and difference and so should be fairly easiliy decoded with some algerbra.
E.g Stero radio is brodcast as L+R in the normal audio range to give a mono signal and L-R is broadcast somewhere above 15k and the 19khz for the stereo indicator.
At the receiver the L+R and L-R signals are added for the Left channel ((L+R)+(L-R)=2L) and subtracted for the right channel ((L+R)-(L-R)=2R).
Would it not be a case of find the maths and off we go ?

It really is not that easy, simplistically the process is similar to a standard FM signal, but in normal FM the subcarrier is amplitude modulated. CD-4 uses an FM modulated subcarrier with both phase and frequency modulation, in addition to ANRS noise reduction. It is not just a simple matter of addition/subtraction as in a normal FM transmission. The subcarrier has to be demodulated first, then ANRS applied and only then can it be added/subtracted to the main signal. I and others on this forum have had some success in "recording" the signal direct from the lp and playing it back through a demodulator, a half way house.

Do a search and you will find it

Malcolm
 
Do you have technical documents about the modullation. I would think once you have the cd-4 captured in a .wav, demodulating should be possible.
I must admit I didn't know that ANRS was used. I recall a script on this site for doing dolby decodes, ANRS is as close to dolby as makes no difference (JVC in the 70's)
 
Been having thoughts about this and about the question of cartridge quality to play CD-4, I thought I might try initialy to simplly capture the CD-4 signal into a 2ch wav.
I've found a sound card http://www.scan.co.uk/Products/ProductInfo.asp?WebProductID=332488 which looks like it might do the job (it has 192k sample rate) and also does 4-in 4-out which would be usefull for other captures.
Does any one have an comments on this card e.g. it's great, it's rubbish that won't do the job etc?
All comments greatfully recieved.
 
A soudcard with 96K sample rate is enough to record a undecoded cd4, no need to go up to 192.
 
A soudcard with 96K sample rate is enough to record a undecoded cd4, no need to go up to 192.

That is correct. The highest frequency that can be recorded is one-half the sampling rate. 96K will record a 48kHz signal. Mike.

P.S. I still don't think it will work.
 
I know but a man needs a hobby !
To be honest it's just an excuse to research the technology and play about a bit
 
Ignoring the fact that there is no software available that can do this kind of demodulation (you would have to write your own) you would still have to find a phono-amp that would pass through the full bandwidth audio signal. I seem to remember discussion about modifying a CD-4 demodulator to do the job, but again. its another piece of the chain where no off-the-shelf solution exists. You'd have to roll your own.

So, to make this work, you have to be an electrical engineer and a software engineer... or know some talented folks willing to lend your "hobby" a hand.
 
A simple mic preamplifier will do the job, without problem of frequency range, then you do RIAA in software.
Another possibility is to use a external RIAA preamplifier.
 
If you search through the postings you will find details of what has been done so far, My approach was to linearly amplify the phono signal ( no RIAA eq.) then record at line level into my sound card at 96Khz 24 bit sampling. On playback I simply attenuated the signal and fed it back into the demodulator phono input. Someone else, and I forget who, modified a demod and took the signal from a later stage just after the 1/2 eq and recorded that. Both approaches worked, but to carry this out in software is MUCH more difficult. Please try, but as Cai mentioned you need specialist software skills in DSP, nothing off the shelf will do, you are really on your own.

Search on old posts, a lot of answers are there

Malcolm
 
Thanks for all the help guys.
I wrote an SQ decoder based on the fine work that went on with Audition, I got a good intro to DSP doing that. I've got the code and the .wav classes and I've got half a web site done http://www.geocities.com/ingresman1961/ to let people have it, but I'm so busy with other stuff and christmas its untrue.
The RIAA shoud be ok to do, ANRS close to dolby probably doable, demodulation should be doable with some research, I'll see how far I get but it wont be my lifes work
 
Back
Top