iampoor1 said:
https://www.youtube.com/watch?v=E5XCvsNUkmI
OK that's 15 minutes I'll never get back.
I spot the classic lm13700 OTA in the circuit, looks like it is used as an expander. .
Yes, they used inexpensive OTA for playback gain manipulation. The encode is a below threshold 2:1 compressor, and playback uses a symmetrical 1:2 expander. Of course if the playback levels are not calibrated properly wrt threshold it will mistrack, similar to Dolby B NR.
Anyone have any stories of this being forced upon them? I am too young to remember much before the demise of cassettes!
I can tell you more than you want to know... Back in the early 80's I got a phone call from the editor of Popular Electronics, and he offered me the Christmas issue cover of the magazine (then 500k circulation) if I designed a kit CX decoder. It was how you say an offer I could not refuse.
In short order I was made a CX licensee by CBS and received the full CX documentation package along with a stack of CX albums. I only had a few months to crank out the design so I put the CX decoder into an existing tape NR package from another kit of mine.
The CBS documentation included two proforma decoder designs, that apparently the other licensees copied verbatim. I didn't and discovered a mistake in the playback time constants (about 10% error), so I corrected that for my design, and notified my contact at CBS.
I would be very curious to see the gear that was used at the mastering stage for this.
The encoder was designed and fabricated by URIE.
Literally weeks before my kit article hit the street, I was attending the AES show in NYC and decided to visit the URIE booth to chew the fat with them. When I mentioned that I found the mistake in the CX documentation their eyes lit up and they said, so you're the guy.
... It turns out that there were already tens of thousands of decoders out in the wild with the wrong time constants in them, so CBS decided to change the time constant in the encoder to agree with the mistake. Of course those MF'rs at CBS neglected to tell me, so I was about to publish the only correct CX decoder design, that they were making wrong by arbitrarily changing the system specifications, without telling me.
Luckily for me I was able to make the value tweak to my design, and change my kit article before it printed, so we were back in sync... BUT all the records already encoded and pressed would never agree with all the decoders out in the market, and only new pressings after that change would be completely correct.
I probably over engineered my decoder... I went to such lengths to deliver accurate decoding that I even specified a tantalum cap for the side chain time constant circuit because the URIE encoder used tantalum in their compressor design. This is a subtle esoteric point, but in a side chain circuit when charged and discharged with different impedances the capacitor's dielectric absorption can influence the capacitor's waveshape. Clearly over-engineered in light of tens of thousands of decoders with time constants wrong by 10% from the start, and a catalog of existing CX records cut with the earlier time constants that were now wrong. You could say it was a major cluster f__k.
I included a clever threshold level trim using a bi-color LED fed with two legs of a LTP... when the audio level was matched to the threshold both LED colors would just light up. All the rest of the time it provided a nice light show alternating between green and red.
IMO CBS was pretty arrogant expecting listeners to not notice the difference of 2:1 compression. It didn't sound as bad as un-decoded DBX because DBX also used about 12 dB of HF pre-emphasis so sounded tinny too. CX undecoded was not unlike listening to Dolby B cassettes without NR decoding switched on. This was not the first time CBS had a conflict with audiophiles. They promoted a copy protection scheme that involved a very narrow notch filter, that they also argued nobody would be able to hear. :
:
For $500 each maybe I could scrape up an old kit or two, I even have CX albums... Nah let sleeping dogs lie. 8)
JR
PS: FWIW my car CD player has a simple 2:1 compressor built in (switchable) to improve the audibility of quiet passages inside a noisy car, it doesn't suck.
