Working CD-4 (software) Demodulator!

[Brice]

Is there a way to view in an oscilloscope (ideally, PC-based and cheap) the carrier signal by itself? With the right tonearm you could micro-adjust the rake angle until the carrier is cleanest and/or loudest (a single, constant signal is easier to monitor).

It's quite easy to judge subcarrier amplitude even from the composite baseband+subcarrier waveform. If you record needle-drops of CD-4 material into a DAW (like the GNU (GPL) Audacity program), you can inspect the subcarrier amplitude with ease. (I've attached a screen shot).



Test recordings of some "dead-wax" (silent grooves) towards the centre of the record are best. If one exists, the inter-track grooves between the last two tracks on a side is very suitable because the subcarrier is always present even when the baseband modulation is silence.

I do wholeheartedly agree that stylus rake angle (SRA) is critical to the best CD-4 replay. Appendix 2 of this page:

http://pspatialaudio.com/vertical_tracking_error.htm
develops a very simple model of the devastating effect upon frequency response if a Shibata/ line-contact stylus is not orthogonal with the groove.

Richard

 

[proufo]

It's quite easy to judge subcarrier amplitude even from the composite baseband+subcarrier waveform.

Thanks for you reply, Brice.

I meant in real time.

If the tonearm allows for it, you should be able to nail the angle by looking at the carrier waveform, checking for maximum amplitude and/or perhaps shape.

 

[Brice]

I meant in real time.

Sorry, my bad. I hadn't understood that.

If you are looking for a really great way to analyse subcarrier quaility in real time, I thoroughly recommend the technique developed by Shure which is very well explained in this reference:

Anderson, C. Roger, "Carrier Crosstalk Considerations in the CD-4 System," J. Audio Eng. Soc. Volume 25, Issue 7/8, pp. 460-465 (August 1977)

This is available in the AES e-Library here; http://www.aes.org/e-lib/browse.cfm?elib=3358

It requires an oscilloscope in X-Y mode, plus you'd need some gain and filtering. I've used it a lot and it tells you a whole heap of stuff in a very simple way. If you have a hardware decoder, it would easy enough to find the relevant places to probe with the 'scope.

Richard

 

[Brice]

All,

I am happy to announce that a new version of the CD-4 decoder has been released in Stereo Lab (from version 3.1.13). This is a free upgrade to anyone who has the original decoder as part of Stereo Lab Audiophile. The Classic CD-4 decoder is selected in the main Quadraphonic Decode mode menu. The new CD-4 decoder is also available in the free Demo version of Stereo Lab, so that you can verify the software and try it.

The new Classic CD-4 decoder in Stereo Lab has been developed in response to requests for a decoder which more closely emulates the hardware decoders of the 1970s. The decoder decodes to the four-loudspeaker regular-polygon layout which was popular in the heyday of "quad". The technical background to this decoder is given on this page:

http://pspatialaudio.com/CD4_classic.htm
Please note that the two available CD-4 decoders in Psptial Audio Stereo Lab should be thought of a complimentary: the initial decoder as a "catch-all" decoder for media in all states, and the new decoder designed to give good results with ONLY the best CD-4 media available. To put it blutly, as much as the original decoder is forgiving, this one isn't!

Best wishes,

Richard

 

[proufo]

Sorry if this is a stupid (perhaps disrespectful) question but have you used (and coded) a gain adjustment using a CD-4 calibration disk to balance (null with the test record) the AM and the FM-derived tracks?

 

[Brice]

Sorry if this is a stupid (perhaps disrespectful) question but have you used (and coded) a gain adjustment using a CD-4 calibration disk to balance (null with the test record) the AM and the FM-derived tracks?

Hi Proufo,

Not disrepectful at all. It's a good question. Have a look at this page:

http://pspatialaudio.com/JVC_CD-4.htm
Towards the bottom of the page, you will see, "Hardware decoders require adjustment of the FM/baseband sensitivity; usually via a couple of potentiometers on the rear of the CD-4 decoder case. These are adjusted using signals from a special test record which was supplied with the unit. In Stereo Lab, no such special material is required and this adjustment is made automatically."

We do this by calculating the average loudness of the baseband and demodulated signals and using this information to match the loudness of the baseband and FM signals (pretty much as a human would do). This was greatly facilitated by the fact that the commercial program already implements loudness measurement to ITU BS.1770-3 so we could re-use this code.

Ironically enough, the only signals which tend to cause problems for this algorithm are test signals! That's because they don't obey statistical equivalent loudness (which music does).

Richard

 

[proufo]

Hello Brice and many thanks for your reply.

We do this by calculating the average loudness of the baseband and demodulated signals and using this information to match the loudness of the baseband and FM signals (pretty much as a human would do). This was greatly facilitated by the fact that the commercial program already implements loudness measurement to ITU BS.1770-3 so we could re-use this code.

Ironically enough, the only signals which tend to cause problems for this algorithm are test signals! That's because they don't obey statistical equivalent loudness (which music does).

Won't this be affected by the particular mix, channel content?

I'd guess that, for example, if you want a separate soloist in a rear channel there won't be "statistical equivalent loudness" in all channels or at least in the same-side channels.

I tend to believe that the test signals should be used for matching.

Thanks in advance.

 

[Brice]

Hello Brice and many thanks for your reply.



Won't this be affected by the particular mix, channel content?

I'd guess that, for example, if you want a separate soloist in a rear channel there won't be "statistical equivalent loudness" in all channels or at least in the same-side channels.

I tend to believe that the test signals should be used for matching.

Thanks in advance.

Remember that we match the loudness over a whole track, or over an entire side (however long is the needle-drop). So statistics are on our side; even if there are some anomalous conditions during the duration of the file.

Remember too that we are matching the loudness of the DIFFERENCE signals, not the rear signals. Imagine the most lopsided mix possible in which the CD-4 was encoded so that it played pure, forward stereo (with the back channels silent). Even in this very unusual situation, the difference signals have to be identical loudness with the baseband signals so that the rears will cancel completely.

Your point is well taken however, I guess few engineers would argue with the contention that test signals are better then relying on the stochastic nature of music. However, the practical problem arises that the test material has to be a record (so that you match the baseband velocity-sensitivity of the cartridge with the FM deviation). A file won't do: only a real test disc.

My experience with CD-4 test discs has not been a happy one: many of them are very badly recorded and in poor condition. If anyone wanted to produce a new CD-4 test disc, I'd be delighted to help collaborate.

Richard

 

[quattro64]

Hi again Richard,

Congrats on the new SW! Will this version have a provision for those of us with the Panasonic SG carts? I guess I need to send an e-mail for the new demo SW.

 

[Brice]

Dear Quattro64

Will this version have a provision for those of us with the Panasonic SG carts? I guess I need to send an e-mail for the new demo SW.

Not this version specifically. In any case, I would want Panansonic cartrdige owners to be able to use either of the two decoding regimes. So it would be better if this function (which you are quite right to prompt me about, because I did promise it) was selectable on the CD-4 Preferences tab.

Here's a question.... The Panasonic cartrdiges are characterised by two differences (compared with a standard MM cartridge):

1) RIAA EQ is not necessary, and
2) One channel needs to be phase-inverted with respect to the other.

Should I include the phase-inversion as part of the support for these cartridges? The hardware decoders did. It's easy for me to do either. So it's really a question of what makes users' lives easier. (Note that the little hardware solution I suggested earlier in this thread does NOT support phase-inversion on one channel.)

As to the second question, if you already have Stereo Lab, the new software version is downloadable. Just install it "on top of" the older version. All your preferences will be inherited.

Best wishes,

Richard

 

[quattro64]

Thanks! Yes I will be building mine with a switched phased channel in the hardware, I think it is up to each person on how they do this. Maybe you could make the software so that each channel can be phase inverted, or not, independently? I think there are several good SG preamp designs out on the webs. I believe to get the best out of these carts one should build their own SG phono amp. I will download this new version to test with my MM carts that might be CD-4 compatible.

 

[kfbkfb]

https://forum.audiogon.com/discussions/benz-cartridge-phase^^^
"The Benz-Micro Lukaschek PP-1 T9 Phonostage does invert polarity..."

I also suggest separate check boxes for L and R to invert the phase.

That would also deal with the scenario of cascaded phono cartridge
amplifiers (probably due to using a Moving Coil phono cartridge
amplifier + another amplifier to bring the signal up to "line level"
for the Computer input) which could result in feeding the
Stereo Lab CD-4 Decoder a Baseband signal of -F-B, which would
swap the Front and Back channels.

Kirk Bayne

 

[Brice]

I will be building mine with a switched phased channel in the hardware, I think it is up to each person on how they do this. Maybe you could make the software so that each channel can be phase inverted, or not, independently?

It makes sense to perform the phase-inversion in the hardware, otherwise the preamplifier is only good for needle-drops into Stereo Lab because you couldn't listen to it (seriously) without the phase inversion. Even record-monitoring would be uncomfortable.

It would be trivial to offer "either or both" phase inversion on the baseband signal. I just need to figure out a way to offer this in the GUI without it being confusing and creating problems for MM users.

For the record, having spent an hour peering at old hardware decoders based on the QSI chip, it the right channel which is identical for MM and for strain-gauge, except for dropping the RIAA EQ. It is the LEFT channel which is inverted in these designs when the switch to strain-gauge is made.

A needle-drop made with a strain-gauge cartridge would a very useful addition to my CD-4 test file collection. If anyone could provide one of these, please let me know.

Richard

 

[quattro64]

Yep, Most will likely build it in, wouldn't hurt to have a phase switch somewhere though, can't have enough switches ☺

 

[Mendemender]

It's taken me a while get around to it but I finally managed to try out the software decoders and compare them with my Marantz CD400b. I tried the first record which came to hand which was Jefferson Starship, Red Octopus in a slightly worn condition.
The hardware demodulator produced a sandpapery sound and wasn't really acceptable. I tried the "fail gracefully" demodulator and whilst there wasn't any sandpaper, there wasn't much four-channel effect either. If that were all that was available then to be honest I wouldn't bother with it.

I was pleased to see that there is now the HQ demodulator available so I gave that a try. The results were much better. Not only was the 4-channel effect like the hardware version but there also seems to be less sandpaper. It was good enough that I invested in a copy of "Audiophile". Here are some files for reference:

http://www.csogb.co.uk/Music/Grunt/BFD1-0999_24bit_JeffersonStarship/SideA_AT/NeedleDrop.wavhttp://www.csogb.co.uk/Music/Grunt/BFD1-0999_24bit_JeffersonStarship/SideA_AT/Marantz.wavhttp://www.csogb.co.uk/Music/Grunt/BFD1-0999_24bit_JeffersonStarship/SideA_AT/FSdemod.wavhttp://www.csogb.co.uk/Music/Grunt/BFD1-0999_24bit_JeffersonStarship/SideA_AT/HQdemod.wav
In due course I will clean a few more records and try those.

 

[kfbkfb]

Have you tried any of the Hugo Montenegro Quadradiscs (with the ping pong ping pong) Quadraphonic mixes
on both the original Pspatial (software) CD-4 decoder and the HQ CD-4 decoder?

Kirk Bayne

 

[Mendemender]

Hi Kirk,
no I haven't tried any Hugo Montenegro yet. I had a look and the only one I have is Hugo in Wonderland which I bought recently and haven't gotten around to playing yet. Is that one worth trying?
At present I'm playing with Aqualung. It's noticeably different to the stereo CD that I have.

 

[kfbkfb]

https://www.americanradiohistory.com/Archive-Audio/70s/Audio-1975-03.pdf^^^
Page 78 (review of Hugo In Wonderland Quadradisc)
(appears to be a ping-pong-ping-pong Quad Mix)

I'm interested to know how the playback imaging compares between the
original (soft fail) CD-4 decoder and the (HQ) Classic CD-4 decoder.

Kirk Bayne

 

[Mendemender]

OK, I'll dig out Wonderland and do a needle drop in the next few days.

 

[Mendemender]

OK, well here's a needle drop of the track that the reviewer found so good. Personally I find "ping-pong sound" makes my head hurt You can try it out and form your own opinion.
Wonderland needledrop

Richard, are you reading this? Is there any way to adjust the parameters of the noise/click removal other than just turning it on or off? I normally use DeNoise and ClickRepair and I think a bit more control would be useful.