CD-4 adjustment question

QuadraphonicQuad

Help Support QuadraphonicQuad:

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

Larry might have something more specific in mind but when I read a comment like that I think it refers to the problem of side imaging for all quad/surround formats that use Pair Wise Mixing as the source. It only works for center front.

Using the front speakers or stereo only speakers as a pair if Left = Right then you should get a very distinct center front image. As its been remarked on the forum by myself & others there are times listening to 4.0 that the phantom center front is so specific you have to get up & put your ear to center front speaker to make sure it's not on! A friend of mine with really nice Thiel speakers the imaging is even better. A vocalist not only comes from the middle but is so specific it seems to have a height & finite distance away. Like you could get up & touch the singers nose.

If you try PWM for center left where Lf=Lb you get a very strong left side impression but it is very broad with no really specific localized sound from the middle of the speakers. Same problem with center back where Lb=Rb. You get a definite "behind you" effect but again it is broad left/right not specific center back.

That doesn't mean you can't hear center side or rear. If you were facing forward & someone was standing exactly to your left side talking, it would be quite precise. Same thing if they stood directly behind you & were talking. So the problem isn't with our hearing just PWM 4.0 doesn't work perfect for all directions.

The only solution I know of is to use real speaker locations to the sides such as in 7.1 ch theater systems and with a dedicated center back. ATMOS might bring something of a solution to side imaging as well. Of course Ambisonics purported to have all this corner/side/height thing figured out years ago.
 
Last edited:
I mean Frank Zappa, The Doors, Jethro Tull... this should be some dynamic exciting stuff. Most of the mixes are just plenty tame & fail to excite.

You should really check out the remastered Best Of The Doors quad mix on Blu-Ray - the fidelity is excellent (maybe the best I've ever hear these songs sound) and the mix is not tame at all. Just a friendly recommendation :)
 
Larry might have something more specific in mind but when I read a comment like that I think it refers to the problem of side imaging for all quad/surround formats that use Pair Wise Mixing as the source. It only works for center front.

Using the front speakers or stereo only speakers as a pair if Left = Right then you should get a very distinct center front image. As its been remarked on the forum by myself & others there are times listening to 4.0 that the phantom center front is so specific you have to get up & put your ear to center front speaker to make sure it's not on! A friend of mine with really nice Thiel speakers the imaging is even better. A vocalist not only comes from the middle but is so specific it seems to have a height & finite distance away. Like you could get up & touch the singers nose.

If you try PWM for center left where Lf=Lb you get a very strong left side impression but it is very broad with no really specific localized sound from the middle of the speakers. Same problem with center back where Lb=Rb. You get a definite "behind you" effect but again it is broad left/right not specific center back.

That doesn't mean you can't hear center side or rear. If you were facing forward & someone was standing exactly to your left side talking, it would be quite precise. Same thing if they stood directly behind you & were talking. So the problem isn't with our hearing just PWM 4.0 doesn't work perfect for all directions.

The only solution I know of is to use real speaker locations to the sides such as in 7.1 ch theater systems and with a dedicated center back. ATMOS might bring something of a solution to side imaging as well. Of course Ambisonics purported to have all this corner/side/height thing figured out years ago.
So let me get this situation straight. If I'm listening to a Quad conversion of a mix that was not channel correct of say some old QS LP, as in the fronts were rotated 90 deg. one direction or the other, and it's being played back digitally (not through say a SM) then the Phantom middle should show up on the sides?
edit: pardon me CD-4 we're talking about, but still?
 
So let me get this situation straight. If I'm listening to a Quad conversion of a mix that was not channel correct of say some old QS LP, as in the fronts were rotated 90 deg. one direction or the other, and it's being played back digitally (not through say a SM) then the Phantom middle should show up on the sides?
edit: pardon me CD-4 we're talking about, but still?
Hmmmm not quite sure I understand your hypothetical situation you're describing but heck, that doesn't stop me from making a reply. My Magic 8 Ball says: yes definitely.

I'm not aware of any ch mislocations like that in real life but if the fronts were rotated incorrectly 90 deg to the right then what should have been left front is now right front and what should have been right front is actually placed at right back. The earliest Sansui decoder included a very strange 1/4 turn & 1/2 turn controls on their units. Never ever did it seem to help anything. I think it was just some kinda added value front panel junk when quad was just a hatchling.
 
You should really check out the remastered Best Of The Doors quad mix on Blu-Ray - the fidelity is excellent (maybe the best I've ever hear these songs sound) and the mix is not tame at all. Just a friendly recommendation :)
The only CD-4 I have of the Doors is Best of & the only digital discrete I have is the DVD-A of L.A. Woman. Also the entire discography in stereo from HD Tracks. With out a doubt I should check the usual places for Best of on Blu-Ray!

I'm not a purist or traditionalist. For the most part all of the remix / updated format sounds much better than the quad classics. Best example I've heard is comparing the original 4.0 quad Aqualung to the 5.1 both included in the super duper edition.

If CD-4 could do some crazy wild things on demo discs, why was so much pop/rock pretty meh? I think they were much less likely to experiment on big time money makers for the label. You know, give 'em some quad but not too much....
 
Side imaging problem?

The side-imaging problem is the problem that, when a part is panned between the side speakers on one side, the human hearing system locates the speakers instead of interpolating a position between them. This is because the close ear gets a signal from each speaker and the far ear gets a delayed and diminished signal from each speaker. The delays are correct to locate both speakers, but are incorrect to locate the side image between the speakers.

In order to hear the part in the correct place, the listener must turn his head.

If a system does not have the side-imaging problem, the part can be panned smoothly around the listener with no breaks in continuity caused by a jump in the image when the part goes to either side.

The problem does not appear in front of or behind the listener.

Most systems have the side imaging problem, including all discrete quadraphonic systems. Special recording tricks would need to be used to prevent the side-imaging problem. QS, EV, DQ, SQ, BMX, and H also have this problem.

Systems that do not have the side imaging problem include DD (Dynaco diamond), Denon QX, Dolby surround, Pro Logic, Pro Logic II, and 7.1 (I haven't heard that yet). I also made an octophonic system that does not have it. In all of these, panning a part all the way around the listener is done very smoothly. I have done it many times.

I designed a microphone system that eliminates the side-imaging problem. I tested it by having someone walk around the mics and talking (He kept saying "can you hear me now" like the ad). When I played it back through a QS decoder, I heard him all the way around without a jump, and at the distance he was from the mics.
 
The side-imaging problem is the problem that, when a part is panned between the side speakers on one side, the human hearing system locates the speakers instead of interpolating a position between them. This is because the close ear gets a signal from each speaker and the far ear gets a delayed and diminished signal from each speaker. The delays are correct to locate both speakers, but are incorrect to locate the side image between the speakers.

In order to hear the part in the correct place, the listener must turn his head.

If a system does not have the side-imaging problem, the part can be panned smoothly around the listener with no breaks in continuity caused by a jump in the image when the part goes to either side.

The problem does not appear in front of or behind the listener.

Most systems have the side imaging problem, including all discrete quadraphonic systems. Special recording tricks would need to be used to prevent the side-imaging problem. QS, EV, DQ, SQ, BMX, and H also have this problem.

Systems that do not have the side imaging problem include DD (Dynaco diamond), Denon QX, Dolby surround, Pro Logic, Pro Logic II, and 7.1 (I haven't heard that yet). I also made an octophonic system that does not have it. In all of these, panning a part all the way around the listener is done very smoothly. I have done it many times.

I designed a microphone system that eliminates the side-imaging problem. I tested it by having someone walk around the mics and talking (He kept saying "can you hear me now" like the ad). When I played it back through a QS decoder, I heard him all the way around without a jump, and at the distance he was from the mics.
I love it when folks talk mics 🎙 and surround; not enough info on here about that relationship as far as I’m concerned!
 
I designed a microphone system that eliminates the side-imaging problem. I tested it by having someone walk around the mics and talking (He kept saying "can you hear me now" like the ad). When I played it back through a QS decoder, I heard him all the way around without a jump, and at the distance he was from the mics.
With speakers in the corners?

If so, impressive.
 
Hmmmm not quite sure I understand your hypothetical situation you're describing but heck, that doesn't stop me from making a reply. My Magic 8 Ball says: yes definitely.

I'm not aware of any ch mislocations like that in real life but if the fronts were rotated incorrectly 90 deg to the right then what should have been left front is now right front and what should have been right front is actually placed at right back. The earliest Sansui decoder included a very strange 1/4 turn & 1/2 turn controls on their units. Never ever did it seem to help anything. I think it was just some kinda added value front panel junk when quad was just a hatchling.
That might of been useful if your speakers were arranged roughly in a square and you wanted to sit facing a different direction, I don't see any other reason for it.
 
Larry might have something more specific in mind but when I read a comment like that I think it refers to the problem of side imaging for all quad/surround formats that use Pair Wise Mixing as the source. It only works for center front.

Using the front speakers or stereo only speakers as a pair if Left = Right then you should get a very distinct center front image. As its been remarked on the forum by myself & others there are times listening to 4.0 that the phantom center front is so specific you have to get up & put your ear to center front speaker to make sure it's not on! A friend of mine with really nice Thiel speakers the imaging is even better. A vocalist not only comes from the middle but is so specific it seems to have a height & finite distance away. Like you could get up & touch the singers nose.

If you try PWM for center left where Lf=Lb you get a very strong left side impression but it is very broad with no really specific localized sound from the middle of the speakers. Same problem with center back where Lb=Rb. You get a definite "behind you" effect but again it is broad left/right not specific center back.

That doesn't mean you can't hear center side or rear. If you were facing forward & someone was standing exactly to your left side talking, it would be quite precise. Same thing if they stood directly behind you & were talking. So the problem isn't with our hearing just PWM 4.0 doesn't work perfect for all directions.

The only solution I know of is to use real speaker locations to the sides such as in 7.1 ch theater systems and with a dedicated center back. ATMOS might bring something of a solution to side imaging as well. Of course Ambisonics purported to have all this corner/side/height thing figured out years ago.
While PWM is not ideal for side images it does sort of work, likewise center back imaging will not be as precise as front imaging, but still works better than side imaging. This makes a point that I keep harping on, straight QS is not a good way to enhance stereo. QS places a hard left equally to the left side but with a 90 degree phase shift between front and back, making side imaging even worse. Same goes for the right side. Surround mode is much better as hard panned signals go to the back speakers only, anything panned between the those extreme positions image from the front, with this mode of operation it's rare to have anything attempt to image from the side.
 
This makes a point that I keep harping on, straight QS is not a good way to enhance stereo. QS places a hard left equally to the left side but with a 90 degree phase shift between front and back, making side imaging even worse. Same goes for the right side.

Good point & I am aware of that. Better would be Variomatrix applied to RM or Scheiber coefficients. Indeed Sansui QS decoding does not have 90 deg phase shifts integral to the basic decoding (I think the SM does) but does all the decoding with L, R, L+R & L-R. The phase shift is applied only at the output after the decoding. So this is a case where the phase shifts are good for QS, not so much for stereo enhancement. A reasonable approach would be to remove the Sansui phase shift circuits, invert the right rear ch and then any center L/R output would be matching in phase between front & rear.

Surround mode is much better as hard panned signals go to the back speakers only, anything panned between the those extreme positions image from the front, with this mode of operation it's rare to have anything attempt to image from the side.

The surround synth is my preferred mode. Under those conditions, as you say, it's rare to have anything image from the sides, except for dynamically panned sounds which is not uncommon in pop/rock. It's all about give 'n take. Operating in the surround synth mode reduces some of those side phase shift problems but it also causes corner emphasis & your still left with how to produce a good side center image which I think is important.
 
Scan-200529-0004.jpg
Scan-200529-0005.jpg
Adjustment record? What adjustment record? Grado says not to use it! The phono cartridge that I used for years was the Grado F-1+. I loved this cartidge for stereo, previously I had used the Shure M91ED and the lowly M3D before that. Anyway according to Grado's hype low capacitance cables weren't required for CD-4, with their CD-4 cartridges, due to the low inductance of the Super-Flux Bridger design.
Anyway I always had problems with the innermost grooves no mater what I did, or what demodulator I used. I think that the advice that low capacitance cables were not necessary was false.
Recently I picked up an almost new F-1+ for less than the cost of a replacement stylus. Reading the set up note that came with it, it says a few interesting and astonishing things, one being "DO NOT USE THE CD-4 ADJUSTMENT DISC!", adjust the carrier level to maximum and set anti skate to 0. Adjust separation control 1/2 to 3/4 open.
I'm now using a moving coil (Sony XL-MC3) for CD-4 and stereo, it works well! I also have an Audio Technica AT14Sa that works very well but is not as detailed as the Sony XL-MC3.
Why bother with another Grado, well my daughter developed an interest in vinyl, I installed a Grado cartridge for her, the stylus of most models are interchangeable but even the cheapest ones are pricey, so why not?
 
Last edited:
Reading the set up note that came with it, it says a few interesting and astonishing things, one being "DO NOT USE THE CD-4 ADJUSTMENT DISC!", adjust the carrier level to maximum and set anti skate to 0. Adjust separation control 1/2 to 3/4 open.

But also:

Copy of Scan-200529-0005.jpg


What is special about Pioneer demodulators?
 
BTW, there are a lot of internal adjustment potentiometers in JVC 4DD5 demodulators. I believe my 405 Technics only a pair.

Would love comments about the Output Matrix Balance section. TIA.
 

Attachments

  • JVC_4DD5_mods.pdf
    769.2 KB · Views: 469
With speakers in the corners?

If so, impressive.

Yes!

The information to restore the side image is collected by the mics during the recording process. It works on any RM system and in stereo headphones.
 
A reasonable approach would be to remove the Sansui phase shift circuits, invert the right rear ch and then any center L/R output would be matching in phase between front & rear.
That's what I like about the Photolume decoder, no all-pass phase shift circuit. The left rear should probably be phase inverted, the only thing is a center back signal will now image out of phase, which is probably OK as that condition doesn't exist with regular stereo. I briefly tried flipping the speaker leads to the left back speaker to see if it made much difference and didn't really notice any, I had to switch back for normal listening. Phase reversal should help with side imaging as without it left front and left back are out of phase with each other.
 
Back
Top