I'm building a new mastering chain and master-project template. In there, I use QuadElectra's Stereo Splitter running into two EQs for mid and side. I routed this like the QuadElectra video on the product page, and this works up to a point. The problem is that when I bypass the Stereo Splitter, I lose the signal chain. I want to be able to bypass the splitter, to be able to A-B the effect. There is a Reason mastering video by learnReason, where he does bypass the splitter, but no explanation on the routing. How can I solve this?
The other, less relevant but related, question is about the other M/S splitter REs. I have both the RE-180 Mid-Side convertor and the Anansi Mid Side router, but both split into a single socket mid-signal (logical) and a single socket side-signal, which seems to lose left/right separation for the sides. What am I missing here?
Mid/Side processing in Reason, some questions (SOLVED)
- crimsonwarlock
- Posts: 2327
- Joined: 06 Nov 2021
- Location: Close to the Edge
Last edited by crimsonwarlock on 18 Oct 2022, edited 1 time in total.
-------
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
Mid and side signals are mono, so you only need one jack for each.
For the bypass question, not knowing your specific setup (or the QuadElectra device in question) you could use a Line Mixer to switch between a dry channel vs the M/S channels. Also, it's the EQs you want to bypass right? Bypassing the EQs should leave the stereo signal intact unless you're doing further processing beyond just the EQs you mention.
For the bypass question, not knowing your specific setup (or the QuadElectra device in question) you could use a Line Mixer to switch between a dry channel vs the M/S channels. Also, it's the EQs you want to bypass right? Bypassing the EQs should leave the stereo signal intact unless you're doing further processing beyond just the EQs you mention.
Selig Audio, LLC
- crimsonwarlock
- Posts: 2327
- Joined: 06 Nov 2021
- Location: Close to the Edge
OK, to answer my second (less relevant) question myself, at least in part: I set up a test with Anansi and another with RE 180. They both do, in fact, retain the left/right information for the side channel. However, the mechanics of this totally eludes me, especially as the Anansi documentation specifically states that it is a MONO signal.
-------
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
- crimsonwarlock
- Posts: 2327
- Joined: 06 Nov 2021
- Location: Close to the Edge
I was hoping you would chime in here
I just posted that I found out it works. However, how can I have separate left and right signals over a mono connection? The sides are left and right, so that's stereo... right?
Bypassing the EQs is only part of it, as the Quadelectra RE does a bit more for mid/side management. Only bypassing the mid/side RE gives a clear picture of what is happening there.selig wrote: ↑18 Oct 2022For the bypass question, not knowing your specific setup (or the QuadElectra device in question) you could use a Line Mixer to switch between a dry channel vs the M/S channels. Also, it's the EQs you want to bypass right? Bypassing the EQs should leave the stereo signal intact unless you're doing further processing beyond just the EQs you mention.
-------
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
- integerpoet
- Posts: 832
- Joined: 30 Dec 2020
- Location: East Bay, California
- Contact:
I only have a few seconds here right now, but it occurs to me a Combinator has an effects bypass button.crimsonwarlock wrote: ↑18 Oct 2022Bypassing the EQs is only part of it, as the Quadelectra RE does a bit more for mid/side management. Only bypassing the mid/side RE gives a clear picture of what is happening there.
- Shocker: I have a SoundCloud!
- crimsonwarlock
- Posts: 2327
- Joined: 06 Nov 2021
- Location: Close to the Edge
That one simply set ALL embedded REs to bypass. Not useful, as I have a whole stack of other stuff in the mastering stack (which is actually a combinator). Besides, it will still break my signal path as it will switch the Stereo Splitter to bypass as well, and that is precisely the problem.integerpoet wrote: ↑18 Oct 2022I only have a few seconds here right now, but it occurs to me a Combinator has an effects bypass button.
I have a hunch from testing RE-180 and Anansi, so I'm going to test an assumption. If that not works, Anansi seems to have a routing solution for this built-in, so now that I know its sides keep separated I might be able to use that.
-------
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
The Shallow End
The "S" in mid side stands for SIDE - not SIDES. It is a mono signal, derived from subtracting one channel from the other. A common misconception is that the Side signal is a stereo signal missing the center panned information. This is only conceptional true, and only when the processing the Side signal and THEN decoding the result. As an encoded signal the side is mono.
If you think about it as a math problem it MAY become more clear (I know this helped me understand the concept).
The Deep End
The Mid signal is really just a mono signal, which (ironically?) also contains the sides by definition. It is defined as L+R, which literally means +L summed with +R.
M = L + R
The Side signal is L-R, which literally means +L summed with -R. With me so far?
S = L – R
The 'magic' comes when you decode the side signal, splitting it into two signals, inverting one and panning them hard L/R. This can be expressed as L = L -R, and R = R -L. Or literally L = +L summed with -R, and R = +R summed with -L. While at this stage it appears the signal is stereo, it is not - meaning anything panned hard L or R is actually in BOTH channels, UNTIL you add the Mid signal back into the equation.
And so when you combine those two signals Side signals with the Mid signal panned center, an interesting thing happens. The Mid signal, being in the center, appears in both the final left and right channel output and is therefor combined with each of the two Side signals described above.
Taking each channel one at a time, this means the Left channel has a +L and a -R signal (Side) added to a +L and +R (Mid) which is easy to see the +L summed with +L gives you the left channel on it's own and the -R summed with the +R nulls leaving ONLY the original left channel signal on it's own in the left channel - perfect. Same for the right, summing the L+R (Mid) signal with the R-L causes the L to null, leaving only the original right channel. Signals panned center are unaffected and as such remain in the center.
Thus the exact original stereo sound field is perfectly reproduced!
Now for the important part - changing ANYTHING with either signal will cause the original stereo image to break down. Taking it to an extreme, if you totally remove the Side channel, you are left with a mono version of the original stereo audio. Conversely if you totally remove the Mid channel you are left with a very odd sounding (polarity inverted) version of the original stereo where the left side is simply an inverted version of the right. This is just like the old issue of wiring your speakers with polarity inverted on one channel. While it sounds like things are 'outside of the speakers', there isn't any of the original panning remaining since each channel is a summed version of both the Left AND the right channel. This Side signal on it's own will completely null if you make the mix mono, which means you have zero mono compatibility! Literally, the more you increase the Side/decrease the Mid signal the less "stereo" the signal becomes. For example, an instrument panned hard left will gradually appear less and less "hard left" as you increase the Side signal relative to the Mid.
The bottom line is the illusion of the signal being wider thus only works for a while as you increase Side/decrease Mid.
Why does this happen? I believe there is a misunderstanding of the Mid and Side signal. The Mid signal is actually a mono signal containing both sides, while the Side signal is also a mono signal which 'contains' both sides. The misconception is that the Mid signal is literally ONLY what is in the center, and the Side signal is ONLY what is in either side. That would be fantastic if it were true, but is not. The side signal is NOT a stereo signal with the middle missing any more than the Mid signal is just the center panned elements in the mix.
What IS true is that any processing that you do to one or the other will only affect the middle or the sides so long as they are both equally present to 'decode' back to the original stereo. This may come across as a paradox of sorts, because as soon as you change something in one or the other you are no longer getting the exact stereo image of the original. The two signals need the other to be present in order to keep the stereo image intact (see the math above). For small changes like cutting some low frequency energy from the sides, this would not cause a noticeable collapse of the stereo image in part because lower frequencies are not as directional sounding as higher frequencies. For bigger changes, especially non-linear effects like compression, the stereo image may change dynamically - for example, hard compression/ducking of the sides would cause the stereo image to go mono when compressing.
The "S" in mid side stands for SIDE - not SIDES. It is a mono signal, derived from subtracting one channel from the other. A common misconception is that the Side signal is a stereo signal missing the center panned information. This is only conceptional true, and only when the processing the Side signal and THEN decoding the result. As an encoded signal the side is mono.
If you think about it as a math problem it MAY become more clear (I know this helped me understand the concept).
The Deep End
The Mid signal is really just a mono signal, which (ironically?) also contains the sides by definition. It is defined as L+R, which literally means +L summed with +R.
M = L + R
The Side signal is L-R, which literally means +L summed with -R. With me so far?
S = L – R
The 'magic' comes when you decode the side signal, splitting it into two signals, inverting one and panning them hard L/R. This can be expressed as L = L -R, and R = R -L. Or literally L = +L summed with -R, and R = +R summed with -L. While at this stage it appears the signal is stereo, it is not - meaning anything panned hard L or R is actually in BOTH channels, UNTIL you add the Mid signal back into the equation.
And so when you combine those two signals Side signals with the Mid signal panned center, an interesting thing happens. The Mid signal, being in the center, appears in both the final left and right channel output and is therefor combined with each of the two Side signals described above.
Taking each channel one at a time, this means the Left channel has a +L and a -R signal (Side) added to a +L and +R (Mid) which is easy to see the +L summed with +L gives you the left channel on it's own and the -R summed with the +R nulls leaving ONLY the original left channel signal on it's own in the left channel - perfect. Same for the right, summing the L+R (Mid) signal with the R-L causes the L to null, leaving only the original right channel. Signals panned center are unaffected and as such remain in the center.
Thus the exact original stereo sound field is perfectly reproduced!
Now for the important part - changing ANYTHING with either signal will cause the original stereo image to break down. Taking it to an extreme, if you totally remove the Side channel, you are left with a mono version of the original stereo audio. Conversely if you totally remove the Mid channel you are left with a very odd sounding (polarity inverted) version of the original stereo where the left side is simply an inverted version of the right. This is just like the old issue of wiring your speakers with polarity inverted on one channel. While it sounds like things are 'outside of the speakers', there isn't any of the original panning remaining since each channel is a summed version of both the Left AND the right channel. This Side signal on it's own will completely null if you make the mix mono, which means you have zero mono compatibility! Literally, the more you increase the Side/decrease the Mid signal the less "stereo" the signal becomes. For example, an instrument panned hard left will gradually appear less and less "hard left" as you increase the Side signal relative to the Mid.
The bottom line is the illusion of the signal being wider thus only works for a while as you increase Side/decrease Mid.
Why does this happen? I believe there is a misunderstanding of the Mid and Side signal. The Mid signal is actually a mono signal containing both sides, while the Side signal is also a mono signal which 'contains' both sides. The misconception is that the Mid signal is literally ONLY what is in the center, and the Side signal is ONLY what is in either side. That would be fantastic if it were true, but is not. The side signal is NOT a stereo signal with the middle missing any more than the Mid signal is just the center panned elements in the mix.
What IS true is that any processing that you do to one or the other will only affect the middle or the sides so long as they are both equally present to 'decode' back to the original stereo. This may come across as a paradox of sorts, because as soon as you change something in one or the other you are no longer getting the exact stereo image of the original. The two signals need the other to be present in order to keep the stereo image intact (see the math above). For small changes like cutting some low frequency energy from the sides, this would not cause a noticeable collapse of the stereo image in part because lower frequencies are not as directional sounding as higher frequencies. For bigger changes, especially non-linear effects like compression, the stereo image may change dynamically - for example, hard compression/ducking of the sides would cause the stereo image to go mono when compressing.
Selig Audio, LLC
man, i really love detailed answers like this, especially when they come from gilesselig wrote: ↑18 Oct 2022The Shallow End
The "S" in mid side stands for SIDE - not SIDES. It is a mono signal, derived from subtracting one channel from the other. A common misconception is that the Side signal is a stereo signal missing the center panned information. This is only conceptional true, and only when the processing the Side signal and THEN decoding the result. As an encoded signal the side is mono.
[...]
br,
daniel
- 12 - Hobbyist
minimal techno - deep minimal dubstep - drum 'n' bass/neurofunk - brostep/deathstep - band recording
New Release: https://open.spotify.com/track/5mQ1XEQtZcVeFVfZvcS5kw
minimal techno - deep minimal dubstep - drum 'n' bass/neurofunk - brostep/deathstep - band recording
New Release: https://open.spotify.com/track/5mQ1XEQtZcVeFVfZvcS5kw
- crimsonwarlock
- Posts: 2327
- Joined: 06 Nov 2021
- Location: Close to the Edge
Thanks for taking the time and give this elaborate explanation. My gut feeling was already that there was some kind of encoding-decoding going on, and I suspected that some phase inversion was used here.
This also solved my problem with my setup. As the Stereo Splitter RE has a stereo-out for the side signal and a stereo out for the mid-signal, I thought I had to use those. It also has mono sends/returns for mid and side, and I (now obviously) had to use those to have the bypass working as it should. Problem solved AND learned quite a bit about mid/side processing from you. Double whammy
-------
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
- integerpoet
- Posts: 832
- Joined: 30 Dec 2020
- Location: East Bay, California
- Contact:
Ah. When you said "template" I thought of a document which might contain multiple combinators.crimsonwarlock wrote: ↑18 Oct 2022That one simply set ALL embedded REs to bypass. Not useful, as I have a whole stack of other stuff in the mastering stack (which is actually a combinator)…integerpoet wrote: ↑18 Oct 2022I only have a few seconds here right now, but it occurs to me a Combinator has an effects bypass button.
Such a thing seems like a cute idea now that I have some experience with what I like doing with Ozone 9. Maybe I'll build a template purely of REs instead of upgrading to Ozone 10.
Because of course having mastered an entire track to my personal satisfaction I am now a mastering authority capable of building tools.
- Shocker: I have a SoundCloud!
- crimsonwarlock
- Posts: 2327
- Joined: 06 Nov 2021
- Location: Close to the Edge
I started out with just a combinator that holds the mastering stack of REs. The idea was to add that combinator to a project when I was finished with mixing. Then I decided it would be better to master in a separate project file, because of processing power preservation and because it is advised everywhere to do this Now I have a project with the mastering combi and some stuff outside it like some metering tools, reference track setup, so I can quickly drop in reference tracks, and some listening environment correction (like SonarWorks). I'm already thinking about removing the combinator around the mastering stack, so I can add specific combinators inside the stack (if needed) as you presumed.integerpoet wrote: ↑18 Oct 2022Ah. When you said "template" I thought of a document which might contain multiple combinators.
What I like with this setup is that you can have things in there, that are normally not in the available mastering suites (like Ozone). I like to be able to choose from different types of compressors, EQs, saturation, etc.integerpoet wrote: ↑18 Oct 2022Such a thing seems like a cute idea now that I have some experience with what I like doing with Ozone 9. Maybe I'll build a template purely of REs instead of upgrading to Ozone 10.
To be honest, I suck at mastering (so far). Building this thing and trying out different REs to use, and try all kinds of mastering techniques to see what works (for me) is a great learning journey. I started about a week ago on this, and I have already learned loads of stuff... like today (thanks again Giles).integerpoet wrote: ↑18 Oct 2022Because of course having mastered an entire track to my personal satisfaction I am now a mastering authority capable of building tools.
-------
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
Analog tape ⇒ ESQ1 sequencer board ⇒ Atari/Steinberg Pro24 ⇒ Atari/Cubase ⇒ Cakewalk Sonar ⇒ Orion Pro/Platinum ⇒ Reaper ⇒ Reason DAW.
-
- Information
-
Who is online
Users browsing this forum: No registered users and 9 guests