Cutting of SQ and QS discs

[kfbkfb]

The trick is that both channels must be limited or compressed exactly the same, with the two devices linked. I have some Behringer stereo compressors with link switches for this purpose.

Ever try some matrix encoded content processed through limiters/compressors (set to high compression) - is the matrix decoded imaging affected a lot?


Kirk Bayne

 

[lukpac]

A question. Cutting a stereo disc involves the cutting ‘needle’ moving left and right in the groove of the master on the lathe. But did cutting SQ and QS discs require new equipment for encoding the vertical modulations? I presume CD4 needed totally new lathes.

Actually, stereo masters are cut with the 'mids' (stuff in the center) moving the stylus laterally, and the stereo 'sides' moving the stylus vertically. For the life of me, I can't imagine getting 2 more channels represented somehow!

That's effectively what gets cut, but in terms of physical cutting and reproduction, it's simply 2 coils 90 degrees from each other, each 45 degrees off vertical, one for each channel. They are wired so that an in-phase signal (i.e., same signal in both channels; mono) moves the stylus laterally, while a completely out of phase signal (i.e., mono but with the polarity flipped in one channel) moves the vertically. If there's a signal in one channel but not the other, one groove wall has modulations and the other doesn't, and the effective stylus motion is at a 45 degree angle.

Matrixed formats are just a stereo signal, so nothing changes. CD-4 used an ultrasonic FM modulated signal to encode the rear channel difference information (with both the front and rear channels being cut in stereo normally). In a most basic sense, cutting and playback is no different from the basic stereo cutting described above, except the cutting and playback systems have to be able to handle frequencies up to 45kHz to preserve the FM modulated rear difference signal. Of course, when cutting at half speed, the cutting equipment would only have to be able to handle frequencies up to around 22.5kHz. But as far as the cutting and playback styli are concerned, they are just dealing with stereo signals that happen to have very wide bandwidth. It's the encoder and decoder that combine/split the audible and ultrasonic content.

 

[austinsignal]

I don't think it's all that different for QS records. The important thing to remember is that, unlike CD-4, any cutting lathe that can cut stereo masters, can cut matrix quad (SQ, QS, EV, et.al.) with no retooling needed.

,

That's effectively what gets cut, but in terms of physical cutting and reproduction, it's simply 2 coils 90 degrees from each other, each 45 degrees off vertical, one for each channel. They are wired so that an in-phase signal (i.e., same signal in both channels; mono) moves the stylus laterally, while a completely out of phase signal (i.e., mono but with the polarity flipped in one channel) moves the vertically. If there's a signal in one channel but not the other, one groove wall has modulations and the other doesn't, and the effective stylus motion is at a 45 degree angle.

Matrixed formats are just a stereo signal, so nothing changes. CD-4 used an ultrasonic FM modulated signal to encode the rear channel difference information (with both the front and rear channels being cut in stereo normally). In a most basic sense, cutting and playback is no different from the basic stereo cutting described above, except the cutting and playback systems have to be able to handle frequencies up to 45kHz to preserve the FM modulated rear difference signal. Of course, when cutting at half speed, the cutting equipment would only have to be able to handle frequencies up to around 22.5kHz. But as far as the cutting and playback styli are concerned, they are just dealing with stereo signals that happen to have very wide bandwidth. It's the encoder and decoder that combine/split the audible and ultrasonic content.

wow thanks!

 

[MidiMagic]

Ever try some matrix encoded content processed through limiters/compressors (set to high compression) - is the matrix decoded imaging affected a lot?


Kirk Bayne

Yes I have.
Without linked compressors, the imaging shifts.
With linked compressors, the imaging is unaffected.

 

[MidiMagic]

https://www.worldradiohistory.com/A...ng-Engineer/70s/Recording-1973-12.pdf#page=23
^^^
CBS claims that their SQ disc can be cut 2.3 dB higher in level than the RM (in other words QS) encoded equivalent


I haven' t read anywhere that Sansui [QS] discouraged using limiters (potential phase shifts) in the QS master tape to master lacquer copying process.


Kirk Bayne

That CBS claim precludes using the space between the back channels in both systems.

There is no difference between QS and SQ levels if center back signals are allowed.

Note that with a record lathe, low frequency signals at higher levels must be cut lateral to avoid overcutting the groove.

This means that even in stereo, the bass guitar, kick drum, and string bass need to be panned to center for louder sound.

 

[kfbkfb]

Yes I have.
Without linked compressors, the imaging shifts.
With linked compressors, the imaging is unaffected.

Matrix quad was promoted as a way to "convert" an FM stereo radio station to an FM quad radio station - during the quad era of the 1970s, many FM stereo radio stations were using severe compression/dynamic range reduction, I've never read about how this level of compression affected the matrix decoding (SQ and QS).


Kirk Bayne

 

[Owen Smith]

Matrix quad was promoted as a way to "convert" an FM stereo radio station to an FM quad radio station - during the quad era of the 1970s, many FM stereo radio stations were using severe compression/dynamic range reduction, I've never read about how this level of compression affected the matrix decoding (SQ and QS).

The BBC did extensive FM quad matrix trials, hence Matrix H etc. But in all the documents I read from their trials, I don't recall any mention of compression. Of course the BBC weren't compressing as heavily as commercial stations often do, but they would likely have been using some.

 

[MidiMagic]

As long as they used a linked compressor, the imaging is retained.

 

[kfbkfb]

https://www.psaudio.com/blogs/copper/a-look-at-ravel-s-works-for-orchestra

QS encoded LP mastering (see Recording data note: "no lat or vert limiting") - So...I guess Sansui did recommend that limiting be limited for QS LP mastering.


Kirk Bayne

 

[lukpac]

https://www.psaudio.com/blogs/copper/a-look-at-ravel-s-works-for-orchestra

QS encoded LP mastering (see Recording data note: "no lat or vert limiting") - So...I guess Sansui did recommend that limiting be limited for QS LP mastering.


Kirk Bayne

I’m not 100% certain, but I believe that refers to sum (lateral) and difference (vertical) limiting. Those would cause differences in the stereo image. Straight stereo limiting would not, and shouldn’t have been an issue.

 

[Electric Moo]

I believe most vinyl pressing prep sums everything below 60Hz or so to mono.

Has anyone here ever pressed vinyl? Would be good to hear about the technical process & how it could differ with matrix encoded material.

Those docs @kfbkfb posted above look good, reading thanks!

 

[barfle]

The trick is that both channels must be limited or compressed exactly the same, with the two devices linked. I have some Behringer stereo compressors with link switches for this purpose.

Hmmm. I have a handfulof QS dbx encoded discs. IIRC, dbx is a compression scheme, and this is intended to reduce surface noise through their decoder.

I can’t say if the dbx affects the QS or not. Interesting thought, though. Hopefully, they figured it out correctly.

 

[lukpac]

I believe most vinyl pressing prep sums everything below 60Hz or so to mono.

The use of elliptical EQ to sum the lower frequencies and the specific turnover frequency varies. As a good example, a relatively high frequency was used for early Beatles LPs, while the MFSL LPs use a lower frequency or no elliptical EQ at all.

It’s not a requirement but it makes cutting at higher levels possible.

 

[Electric Moo]

Probably allows more music to fit onto a side, too?

60Hz seems really low. I must be confusing a highpass filter freq with the elliptical. The conversation I had with a cutting engineer was a long time ago. I bet it's closer to 200Hz & like you say, adjustable.

( I picked up an old copy of Deep Purple's 1st album & it sounds absolutely terrible. There's 27 mins of music on side 1. I wonder if that's how they fit it all on, and why it sounds so bad. The music level is very low & there's no bass ).

 

[lukpac]

Probably allows more music to fit onto a side, too?

60Hz seems really low. I must be confusing a highpass filter freq with the elliptical. The conversation I had with a cutting engineer was a long time ago. I bet it's closer to 200Hz & like you say, adjustable.

( I picked up an old copy of Deep Purple's 1st album & it sounds absolutely terrible. There's 27 mins of music on side 1. I wonder if that's how they fit it all on, and why it sounds so bad. The music level is very low & there's no bass ).

I believe 150 and 300 are (were?) common frequencies. I can’t find the Neumann documentation at the moment, but The Mastering Lab had 40, 50, 75, 110, and 200. The specific frequencies depended on the specific EQ being used.

I don’t recall what my friends who cut records typically use, but I think the frequency generally generally went down over time, presumably because playback equipment got better.

There’s a lot about disc cutting here:

https://pearl-hifi.com/06_Lit_Archive/07_Misc_Downloads/088_Basic_Disc_Mastering.pdf

 

[par4ken]

I believe 150 and 300 are (were?) common frequencies. I can’t find the Neumann documentation at the moment, but The Mastering Lab had 40, 50, 75, 110, and 200. The specific frequencies depended on the specific EQ being used.

I don’t recall what my friends who cut records typically use, but I think the frequency generally generally went down over time, presumably because playback equipment got better.

There’s a lot about disc cutting here:

https://pearl-hifi.com/06_Lit_Archive/07_Misc_Downloads/088_Basic_Disc_Mastering.pdf

Growing up my main program source was always vinyl. Usually played back with some bass boost and a bit of treble as well. To me that is the standard, I suppose that I've fine tuned my system for the sound of vinyl. Now when I listen to a high rez source such as the Quadios, I have to rush to crank down the bass! The Quadios are the worst offenders in that regard. I find that CD's generally sound similar to vinyl but perhaps with a bit more solid bass. Remastered, compressed CD versions on the other hand generally sound like crap!

Summing the bass to mono makes sense especially for vinyl. Summed bass will sound about the same (as if not summed) because bass is generally non directional, if it wasn't then use of a single subwoofer would be useless. Obviously in addition to summed bass vinyl usually employs bass roll off as well. It has been my desire/dream to find some magic equalisation setting to make my vinyl rips sound more like modern releases. That idea (a standard setting) appears to be impossible if different settings were used to begin with. It would be cool to see those Neumann documentation details!

One of my early audio projects was a bass summing amplifier. I've posted about it before. It was an article in an issue of Popular Electronics. In any case it was very cool to watch the woofer cones move slowly in and out in opposite directions while playing a warped record. With the bass summing amp switched in that speaker cone movement stopped completely and the bass still sounded exactly the same. I never understood why manufactures didn't use that design instead of the typical high pass rumble filter!

This reminds me of a story that I once heard about Motown. They had a Canadian company press thier 45's. They wanted them louder than the competition (sound familiar) the solution was to simply cut the bass more, then they could be pressed louder! Motown was happy!

As for Deep Purple twenty one minutes is about average for an LP side. I don't know which pressing you have but I think that the EMI releases sound much better than the US Tetragrammaton Records version. The Canadian Polydor releases "Best of Deep Purple" and "Early Purple" both sounded bad; apparently they were released after Tetragrammaton went bankrupt and were culled from vinyl rather than the master tapes! It was a revelation to get EMI LP's and then the CD's of the Mark 1 band. That lineup remains by far my favourite of all the Deep Purple band variations!

​

 

[Owen Smith]

Talking of length, Queen Greatest Hits on LP is 60 minutes in total. Ouch.

 

[austinsignal]

I believe 150 and 300 are (were?) common frequencies. I can’t find the Neumann documentation at the moment, but The Mastering Lab had 40, 50, 75, 110, and 200. The specific frequencies depended on the specific EQ being used.

I don’t recall what my friends who cut records typically use, but I think the frequency generally generally went down over time, presumably because playback equipment got better.

There’s a lot about disc cutting here:

https://pearl-hifi.com/06_Lit_Archive/07_Misc_Downloads/088_Basic_Disc_Mastering.pdf

Generally, I prefer to lower the level a bit and leave the bass alone if possible. 0db or -3 vs +3 is just fine in most cases and I can get away with not having to roll off the sides at all many times. Regarding modern playback - the ‘modern’ Crosley-esque record players are the most problematic. My first rule is, ‘the record has to play properly’, and that means on all turntables, including the Crosleys. Even with decent groove depth, sudden, loud low frequency stuff (even mono) can cause tracking problems on cheapo gear. Cutting quieter can help a lot. For long sides, cutting quieter is definitely required