Experiments with CD-4 and soundcard

[Malcolm2010]

Hi,
I tried an experiment today, basically boredom set in convalescing after a visit to hospital. I set about thinking what I could see if I directly connected my turntable to a standard mic input on my sound card. OK I know the impedances and EQ are not correct, but it has shown some interesting results.
I ran the CD-4 tracks from Quadrafile into my sound card, recording at 96Khz 24 bit, please see the graphic for the results. The 30Khz carrier can clearly be seen, and superimposed on this is the centre back announcement showing the sub carrier modulation. I use an Audio Technica AT-120, a cartridge not rated for CD-4 so using a proper one should be better. This may be a (complicated) way to eventually demodulate CD-4, although I really dont think it is worth the hassle.
Amazing what you do when bored !!

Malcolm

 

[Q-Eight]

Tried the same thing.... I couldn't get it to work. How'd you amplify the signal enough?

 

[Malcolm2010]

I built a small external amplifier with a gain of about 20, actually it was a microphone pre amp design, modified slightly to give a higher input impedance. Running the turntable directly into your sound card line inputs will not work, it needs additional amplification.


Malcolm

 

[Quadchuck]

This weekend, I get my new Pioneer DVD-a player, model ...588 something.
I am on to the next phase of my experiment.
I have playable/decodable files on computer for CD-4,(they play fine) and downloaded test version of Cirlinca and made test DVD-A.
I will soon find out if file remains intact in burning and playable on Pioneer.
If all goes well, I am going to dub them "QuadraDVDisc-A" .
I'm hoping the line level / headphone jacks or both output the carrier
and activate decoder with no degrading artifacts.
If all goes well, I next plan to create 4 channel mono files,separate
L and R composite into 2 files and L and R carriers into 2 files
in both the recording and playback process.
Analagous to the "Component" video process where colour is processed separately from luminance and sync.
I can then edit and clean up files(maybe filter them) and it could hold
some promise of improved sound.
It could also fail,due to the fact the sound overlays/mixes on each sum and difference ,canceling appropriately.
Noise could improve, but at the loss of separation.
TBC

 

[Quadchuck]

Behold the Bride Of Quadrastein

It's Alive!!! ...Alive !!! Ah HA HA HA HA Ha.......!!!!!
(No offense Carly)

I've been working in my laboratory all weekend and finally got
my new Pioneer dv-588a.
I must admit I didn't research the player enough, but no matter,
it does work!
Played album from AT15s into Panasonic SE405,tapped
unit at Riaa preamp out(composite signal).
Captured with M-audio 1814 Firewire into computer.
Downloaded test Cirlinca and burned DVD-a
Played back thru Pioneer DV-588a and fed L/R out to same point
as captured signal.
DVD-A disc plays back with CD-4 carrier intact and decoder
plays as if record was played.
Several issues(as always) remain.
1-Since I capture entire side I of record,Imust edit wav file with index points
for each song (How)?24/96 editor or split each song up HMMM
2-Cirlinca gives me a persistant error about a non-existant temp file
hard to get past that point.
3-Pioneer does not have headphone out or any way to vary output
like I can do with computer,therefore to obtain optimum separation
I would have to calibrate SE405 to Pioneer instead of standalone
setting.
Check out the pix

 

[Q-Eight]

That is SOO Cool. Somehow I knew this were possible. Now, if only a practical application can be done

In regards to your questions, depending on which program you're using to burn the disc, I know in Nero, you can edit the .cue file for the different track changes.

I've never used Cirlinca.

And thirdly, can you pass the signal thru a SQ decoder first before passing it on to the CD-4 Demod? The reason I mention this is that both my Lafayette SQ boxes have a variable volume control on them. Could be a simple solution for now.

 

[Quadchuck]

I tried to use DC-art to edit but even though it has many features and a
vast selection of noise reduction for audio files,I still like Nero
better.
Nero 7 does 24/96 and192 now, so I may have to cough up the bucks and get it,though I'm not sure if it burns DVD-a format.
I have tried to separate carrier from audio and clean up each
then recombine.
Since I have no software to clean supersonic freq. I can only
pre-filter and try to remove as much unneeded audio from carrier.
Then I change audio sample back to 16/44.1(that removes carrier)
cleaned with Nero 6 and resampled back to 24/96.
Recombined both in to one .wav.
Too much work, but results did improve overall sound,though not as
much as I would have liked.As they say garbage in,garbage out..
it must be a good source to begin with.

I either need to read Cirlinca manual thoroughly or it may be
too simplistic for this app.

One other strange anamoly,varing the carrier level up or down does
not effect the separation as much as varing the overall combined
level,therefore a simple level control(at least for my decoder)
is sufficient to obtain best separation.
Running it through a SQ decoder even if only for level control
could skew phase for CD-4 difference matrix.
Also,everything in chain must maintain upto 50khz ability, fortunately
it's harder to kill supersonic carrier (30khz) than you would think.
TBC

 

[par4ken]

I might try this someday, I did a short test recording one time and the modulated carrier was clearly present. That particular disc had several large clicks on it, just wonder what effect cleaning it up would have on the sub-carrier. I would first try playing it back from the sound-card through the demodulator , before bothering to make a DVDA disc, though.

 

[quadsearcher]

IMHO, clean up after decode. Is there a multichannel edit/declick solution? Or is the problem that the clicks will interrupt the decode itself..... hmmm. Interesting stuff.

 

[FM_Quad]

I'm wondering if I could get a copy of this, Maybe a single track if the file is too large? I'm wanting to do some experimenting on software decoding. While I have a good TT and cart, I don't have something like the 1814 to digitize it.

Regarding the comments about F/B crosstalk. Chaning the carrier amplitude should not change the L-R decoded. Therefore, proper separation is dependent only on the L+R. The reason why the adjustment is even necessary in a normal playback setup is that every cartridge has different gain. The L+R changes directly with this gain but the L-R (since it is on an FM carrier that is amplitude insensitive) does not. The adjustment simply changes the L+R to match the known, constant L-R.