The MTC-MS-II: An Improved Precision Mid Side Matrix

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

The MTC-MS-II: An Improved Precision Mid Side Matrix

Post by mediatechnology »

This is a preview of a design I'm working on for an improved MS Matrix.

Update 3/1/18: Construction information for the MTC-MS-II is located here: https://www.proaudiodesignforum.com/for ... ?f=7&t=941

The MTC-MS-II boards are available in limited quantities now.

MTC-MS-II Product Page: https://www.ka-electronics.com/kaelectr ... Matrix.htm
MTC-MS-II Bare PC Board: https://ka-electronics.com/shop/index.p ... duct_id=85
MTC-MS-II Bare PC Board and ICs: https://ka-electronics.com/shop/index.p ... duct_id=88
MTC-MS-II Assembled and Tested: https://ka-electronics.com/shop/index.p ... duct_id=76

Image
The KA-Electronics MTC-MS-II Improved Precision Mid Side Matrix for Audio Mastering

Large format schematic for the MTC-MS-II: http://www.ka-electronics.com/images/pn ... heet_1.png

The MTC-MS-II board includes a number of new features to make MS work easier.

MTC-MS-II Unique Insert Gain Offset:

Side levels are often significantly lower than Mid. With highly-correlated elements Mid often has "mono build-up" producing level increases of 3 to 6 dB.

The MTC-MS-II addresses MS level challenges by applying Insert Gain Offset between Mid and Side. The Encoder can operate with 0, 6 and 12 dB offset between Mid and Side. When gain offset is applied in the Encoder, the Decode gain is compensated to maintain the original M/S ratio.

Mid Gain Offset: Reduces Mid Encode Gain by -6 dB. Mid Decode Gain is automatically increased by +6 dB to compensate. Headroom is increased by 6 dB.

Side Gain Offset: Increases Side Encode Gain by +6 dB. Side Decode Gain is decreased by -6 dB. Side gain offset allows outboard gear to be driven hotter.

By using Mid and Side Gain offset together, an offset up to 12 dB is available.

Individual Encode Decode Bypass

The MTC-MS-II's Encoder can be bypassed to forward Left and Right into the Insert.

The Decoder can be also be bypassed to forward Mid/Side to the final Left and Right Outputs (J2B) or jumper-configured to forward the Decode inputs directly to the output. (J2A)

By using Encode and Decode bypass together the Insert can be switched from Mid/Side to Left/Right.

Individual Insert Bypass

Mid and Side have individual Insert bypass relays. Need to compress only Mid? Bypass the Side Insert completely.

Processing both Mid and Side? Bypass one at a time to hear each processes' contribution.

On-board relay logic is available to bypass Mid and Side Inserts simultaneously for instant A/B comparisons.

On-board Width Control

The MTC-MS-II has on-board Width control ranging from 0 to 200% by varying Side Level from -∞ to +6 dB relative to an anchored unity gain Mid level.

Control is smooth and predictable with a "feel" similar to the Width Controller using the LR±S method. A potentiometer or stepped make-before-break rotary switch can be used. The Width Control is relay-bypassed.

On-board Tilt Equalizer Provides Frequency-Dependent Width

A gentle tilt equalizer can be inserted into the Side channel to provide frequency-dependent Width. The range at 10 kHz is +/-3dB. As 10 kHz is boosted, 100 Hz is decreased by an identical amount with a "pivot" frequency of 1 kHz. A 0 to 6 dB range in frequency-dependent Width is available from the 100 Hz to 10 kHz with either positive or negative slope. Width can increase with frequency or decrease with frequency.

The Tilt Equalizer follows the overall Width control and is relay-bypassed.

Individual Mid and Side Pre-Decode Mutes

The Mid and Side channels have individual relay mutes to audition the processing of each insert.

When Mid is muted, Side is presented to the Left and Right outputs in their actual perspective similar to an "in place" Solo function.

(Side is a "mono" single-channel signal. When presented "in place" the positive polarity is heard in Left and the negative in Right. This is how we hear it in the stereo image when recombined with Mid.)

The Mid and Side channel mutes are placed in the return path rather than the Insert Send to eliminate "compressor recovery" artifacts from muting sends.

Improved Encoder and Decoder Headroom

Mid Side encoding is well-known for Mid "build-up" on highly-correlated material.

The MTC-MS-II eliminates Mid build-up and the resulting headroom reduction with Mid Gain offset. Internal headroom in the Mid channel is increased by 6 dB to +23 dBu. The external overload point is +27 dBu. (+/-15V supply rails.)

Though not fully-appreciated Mid Side decoding has additional gain build-up during Left and Right reconstruction. When M+S is added to M-S to produce "L" the actual result is 2L.

The MTC-MS-II uses a combination of THAT1240 and THAT1246 line receivers to eliminate encoding and decoding headroom and level bottlenecks. By switching the THAT1240 in a unique way its gain can be varied from -6, to 0 and to +6 dB. This eliminates the requirement for additional gain or attenuation stages that would add a differential phase delay to one process path but not the other. High frequency crosstalk performance is maintained.

Relay and Power Distribution

Image

Large format schematic for the MTC-MS-II Relay and Power Distribution: http://www.ka-electronics.com/images/pn ... heet_2.png

The next steps.

I'm putting the finishing touches on the layout.
Gold
Posts: 677
Joined: Sat Dec 01, 2007 5:20 pm

Re: The MTC-MS-II: An Improved Precision Mid Side Matrix

Post by Gold »

mediatechnology wrote: Wed Jan 31, 2018 3:38 pm By switching the THAT1240 in a unique way its gain can be varied from -6, to 0 and to +6 dB. This eliminates the requirement for additional gain or attenuation stages that would add a differential phase delay to one process path but not the other. High frequency crosstalk performance is maintained.
This looks great Wayne. It's much more product like. Could you expand on the gain switching of the 1240's? I don't think I've seen that before.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: The MTC-MS-II: An Improved Precision Mid Side Matrix

Post by mediatechnology »

Gold wrote: Thu Feb 01, 2018 11:51 am
mediatechnology wrote: Wed Jan 31, 2018 3:38 pm By switching the THAT1240 in a unique way its gain can be varied from -6, to 0 and to +6 dB. This eliminates the requirement for additional gain or attenuation stages that would add a differential phase delay to one process path but not the other. High frequency crosstalk performance is maintained.
This looks great Wayne. It's much more product like. Could you expand on the gain switching of the 1240's? I don't think I've seen that before.
Thanks Paul!

Here's the 1240 information: viewtopic.php?f=6&t=935
emrr
Posts: 570
Joined: Sun Mar 02, 2008 10:27 pm
Location: NC, USA
Contact:

Re: The MTC-MS-II: An Improved Precision Mid Side Matrix

Post by emrr »

Looks like smart refinement of the package.
Best,

Doug Williams
Electromagnetic Radiation Recorders
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: The MTC-MS-II: An Improved Precision Mid Side Matrix

Post by mediatechnology »

Thanks for the feedback everyone.

I'm doing layout now and have managed to get all ten relays on the board.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: The MTC-MS-II: An Improved Precision Mid Side Matrix

Post by mediatechnology »

Some recent updates posted to the original thread-starter.

Update 2/7/18: Added J2 to select the source when the Decoder is bypassed to allow easier Left/Right processing in the insert. Attached relay and power distribution schematic.

Update 2/5/2018: J1 was moved to be before the switchable M/2M stage and a Phoenix connector added for stepped Mid Attenuation. When J1 is linked Mid level is anchored.

The project is currently in layout stage. The layout is almost finished. Currently error-checking.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: The MTC-MS-II: An Improved Precision Mid Side Matrix

Post by mediatechnology »

Boards on order! Should know if they work by next week. If they do I'll have 3 extra.

Image
The MTC-MS-II Precision Mid SIde Matrix Encoder/Decoder with Width and Side Tilt EQ.
User avatar
dagoose
Posts: 15
Joined: Wed Feb 03, 2010 6:41 am
Contact:

Re: The MTC-MS-II: An Improved Precision Mid Side Matrix

Post by dagoose »

wow! this looks great Wayne!
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: The MTC-MS-II: An Improved Precision Mid Side Matrix

Post by mediatechnology »

Thanks Jeffrey.

I wasn't planning on upgrading the original MS but after running some experiments and listening to the result I felt that I had to. Getting 10 pounds of relays on a 5 pound PC board was the challenge.

This is a plot I did to demonstrate the effect of the Tilt EQ on Width.
One input channel is driven. Side gain is 6 dB for 200% Width.

The top pair of traces are the Side response with the Tilt EQ at its extreme settings.

The bottom pair of traces, which are technically a crosstalk curve, are representative of the resulting Width.
(Width is a little difficult to display because the analyzer can't discern the polarity of "crosstalk" - it only sees level - so it doesn't know the difference between crosstalk and gain in Side.)

The red trace shows increasing HF Width along with HF boost in Side.
The violet curve shows increasing LF Width as the Tilt control is moved to LF boost.

The Tilt EQ slope at maximum (800 Hz-2 kHz) is about 1.8 dB per octave.
The resulting Width curve is about 3.4 dB per octave over the same range.
The measure pivot frequency is 1.2 kHz.

In most cases the top half of the control will be used but I felt like having LF boost available would be useful for corrective situations where the LF are more narrow than they have to be.

Image
MTC-MS-II Tilt Equalizer Curves.
Post Reply