Audio mistake that's been on my mind for years: An ultimate-nerd challenge.

24 track

Telefied
Ad Free Member
Joined
Nov 6, 2014
Posts
20,576
Location
kamloops bc
did you create phasing issues? or bad proximity effects?
dead batteries? forgot to turn on the phantom?
ab05ee58684e69efaeef8bc42f26b6ca.png
 
Last edited:

KokoTele

Doctor of Teleocity
Vendor Member
Joined
Mar 17, 2003
Posts
14,940
Age
47
Location
albany, ny [not chicago]
All of this is way beyond me, but I know the room and the artist Ben is talking about and I’m super curious to know the answer, even if I don’t understand it.
 

Peegoo

Doctor of Teleocity
Joined
Oct 11, 2019
Posts
14,778
Location
Beast of Bourbon
I'll play.

You had the mic set to MS mode with only a single cable on the red connector. That rejects left and right address of the capsule.
 

KeithDavies 100

Tele-Holic
Joined
May 19, 2021
Posts
759
Age
60
Location
Cambridge, UK
An audio detective story - cool!

All of my suggestions have come and gone above. All I had was about the mic's pickup pattern not being optimal for a group who move in and out like that, as others have said. Other things that might come to mind - only one cartridge powered; phantom power not on etc - would have been noticeable in the control room.

I wait in suspense...!
 

Mouth

Friend of Leo's
Joined
Feb 13, 2018
Posts
2,124
Location
U.S.
I've nerded to the point of reading the manual.

I'm in need of the solution to the mystery.

Screenshot_20220806-213734_Samsung Internet.jpg
 

bottlenecker

Poster Extraordinaire
Joined
Dec 6, 2015
Posts
6,666
Location
Wisconsin
I’m wondering what happens from a listener point of view, when a single central M-S mic is providing the stereo image, and performers step up and back from their semicircle around the mic. Does it cause a volume change and a perceived left-right sideslip, because of the M-S capsule arrangement, for the performers who are not right about inline with the M or the S capsule?

The only experience I have with the “bluegrass self-mixing semicircle” involved a cardioid LDC or an X-Y coincident pair of SDCs. Both seemed to work, but I liked the mono, volume-change-only effect of the single LDC best.

I think mid/side is perfect for a bluegrass "single mic" situation, as long as the mid capsule is cardioid. The overall pickup pattern is like a very wide cardioid pattern with a null at the back. If it gets summed to mono the fig8 capsule's signal disappears, but if the performers have been treating it like a single mic, the cardioid capsule by itself should still have a complete presentation.

I'm very curious about the answer to the riddle now.
 

Ben Harmless

Poster Extraordinaire
Joined
Mar 10, 2003
Posts
5,251
Location
Salem, Mass
Ugh. Sorry for my delay. Work has been pushing a little hard lately. Okay. Ready?

I know the VP88 is stereo, but is it mid/side? I'm trying to think of how you could sum the fig8 pattern and cancel it, but you were broadcasting stereo. Was the control room radio mono?
You're getting there. Significantly. Multiple audio nerd points.

I've nerded to the point of reading the manual.
That's really, really good, and definitely a piece of the puzzle. Points!

I think mid/side is perfect for a bluegrass "single mic" situation, as long as the mid capsule is cardioid.
I agree, except while I haven't played with it, I always worry that the way that stereo is generated with M/S actually could create some weird imaging and filtering depending on seating position and the environment in a live setting.

So y'all are right on it. Here we go:

The VP-88 is indeed a M/S mic by default. I never really knew that. I think I assumed when we were using it that it had X/Y capsules in there or something. Nope. The internal matrix does produce a two channel output, so it's not immediately obvious if you don't go looking. Now, Bottlenecker was right up there - we were broadcasting in stereo, but as Mouth also pointed out (via the manual) broadcasting and listening are two different things, and this is where the fact that it's radio, and especially public radio comes in.

In the US, the public radio audience leans educated, white, wealthy, and older. This is absolutely not universal, but them's the stats. While many of these listeners may have very nice radios in the fancy hifi rigs in their living rooms, many others - especially the older crowd, might instead be listening the way my grandparents used to listen to A Prairie Home Companion - that is, sitting at their kitchen table with a little portable radio......with one speaker.

Now, just as a quick refresh for anyone unfamiliar, M/S or Mid-Side uses magic/science to create a stereo image. I won't try to cover the whole thing, but look it up - it's great for learning about several good audio concepts. The problem here is that both channels contain information from the sides, but it's really the same stuff, just polarity flipped. As such, any radio which sums to mono will completely lose the stuff off to the side, regardless of how things were mic'd or broadcast. What's more, one or two of my station's repeaters was broadcasting AM, and almost certainly in mono.

For extra nerding, look into how stereo FM signals are broadcast, and appreciate how that system also uses a sum-and-difference matrix. Still results in summing though, so whether you're mono because your radio only picks up the main carrier, or because it's summing the signals, the problem persists.

This brings us to a major pet peeve of mine. Many argue that M/S recording is fundamentally mono-compatible, and it's celebrated as a go-to technique with phase advantages over other options. I've never felt this way. While it doesn't create major phasing issues, I just can't wrap my head around how a technique that loses much of it's audio information when summed is considered "mono-compatible." I assume the major texts in the field don't see it this way, but I can't help it. In any other stereo technique, the performers standing to the extreme side of the stereo field wouldn't be cancelled, but with M/S, they are - mostly. This article argues that M/S is the "safest" technique for broadcast, but I have to imagine they're generally picturing a field recording or a performance recorded using more distant mic'ing than the gather-'round-the-mic approach.

This is why I now default to X/Y as my go-to stereo technique. It's just as mono compatible, and while people may be concerned that it's not as "wide" a stereo image, I'd argue that for closer-mic'ing situations, that kind of width is a distortion of the actual environmental stereo field, and would be better achieved by other techniques.

So, in sum (pun intended) there's a right and less-right time for all of these things, but in my specific use case, I consider it a mistake. All the recordings were great, but the broadcast was flawed. I remain a little sad, but it's still probably not my biggest mistake in the field, so here we are.

Thanks for playing! I'll try this again if I can think of other problems I've caused!

Oh, and if you read this far, then nerd points for you. Nerd points all around!
 

bottlenecker

Poster Extraordinaire
Joined
Dec 6, 2015
Posts
6,666
Location
Wisconsin
Ugh. Sorry for my delay. Work has been pushing a little hard lately. Okay. Ready?


You're getting there. Significantly. Multiple audio nerd points.


That's really, really good, and definitely a piece of the puzzle. Points!


I agree, except while I haven't played with it, I always worry that the way that stereo is generated with M/S actually could create some weird imaging and filtering depending on seating position and the environment in a live setting.

So y'all are right on it. Here we go:

The VP-88 is indeed a M/S mic by default. I never really knew that. I think I assumed when we were using it that it had X/Y capsules in there or something. Nope. The internal matrix does produce a two channel output, so it's not immediately obvious if you don't go looking. Now, Bottlenecker was right up there - we were broadcasting in stereo, but as Mouth also pointed out (via the manual) broadcasting and listening are two different things, and this is where the fact that it's radio, and especially public radio comes in.

In the US, the public radio audience leans educated, white, wealthy, and older. This is absolutely not universal, but them's the stats. While many of these listeners may have very nice radios in the fancy hifi rigs in their living rooms, many others - especially the older crowd, might instead be listening the way my grandparents used to listen to A Prairie Home Companion - that is, sitting at their kitchen table with a little portable radio......with one speaker.

Now, just as a quick refresh for anyone unfamiliar, M/S or Mid-Side uses magic/science to create a stereo image. I won't try to cover the whole thing, but look it up - it's great for learning about several good audio concepts. The problem here is that both channels contain information from the sides, but it's really the same stuff, just polarity flipped. As such, any radio which sums to mono will completely lose the stuff off to the side, regardless of how things were mic'd or broadcast. What's more, one or two of my station's repeaters was broadcasting AM, and almost certainly in mono.

For extra nerding, look into how stereo FM signals are broadcast, and appreciate how that system also uses a sum-and-difference matrix. Still results in summing though, so whether you're mono because your radio only picks up the main carrier, or because it's summing the signals, the problem persists.

This brings us to a major pet peeve of mine. Many argue that M/S recording is fundamentally mono-compatible, and it's celebrated as a go-to technique with phase advantages over other options. I've never felt this way. While it doesn't create major phasing issues, I just can't wrap my head around how a technique that loses much of it's audio information when summed is considered "mono-compatible." I assume the major texts in the field don't see it this way, but I can't help it. In any other stereo technique, the performers standing to the extreme side of the stereo field wouldn't be cancelled, but with M/S, they are - mostly. This article argues that M/S is the "safest" technique for broadcast, but I have to imagine they're generally picturing a field recording or a performance recorded using more distant mic'ing than the gather-'round-the-mic approach.

This is why I now default to X/Y as my go-to stereo technique. It's just as mono compatible, and while people may be concerned that it's not as "wide" a stereo image, I'd argue that for closer-mic'ing situations, that kind of width is a distortion of the actual environmental stereo field, and would be better achieved by other techniques.

So, in sum (pun intended) there's a right and less-right time for all of these things, but in my specific use case, I consider it a mistake. All the recordings were great, but the broadcast was flawed. I remain a little sad, but it's still probably not my biggest mistake in the field, so here we are.

Thanks for playing! I'll try this again if I can think of other problems I've caused!

Oh, and if you read this far, then nerd points for you. Nerd points all around!

So, what was the end result? Were people off to the sides, outside of the center cardioid pattern, and their sounds were lost to summing?
I know nothing about how FM stereo works, even though I used to broadcast over it. I'm going to have to look that up.

You may be right about a bluegrass band, if they're used to omni mics. If a group is positioned all around a mic (like it's an omni in a studio), a mid/side pair would be problematic. I love how well it picks up people out to the sides, but in mono as you know, they disappear.
When I think of a bluegrass band on a single mic, I think of them playing into a cardioid live, which may be a careless assumption on my part. But any group or source that fits in the cardioid pattern will still be there if the side signals are discarded due to summing.

Larry Crane said he hates mid/side because he doesn't want his mix to change when summed to mono. This is why I'm careful to keep the level of my side signals fairly low (this is done by the stereo width control on the vp-88), so that the m/s source doesn't change volume too much in the mix, and I still like my mix in mono. I most often use m/s as part of my drum sound. If I didn't record drums in stereo, I would do everything the same way, and just not set up the figure 8 mic.
It's crucial to check mixes in mono even if not recording in mid/side, so for me there's no drawback as long as I like both results.
 

SRHmusic

Tele-Afflicted
Joined
Oct 19, 2020
Posts
1,888
Location
North Carolina, USA
So, if I read your post correctly, the mic was set to M/S, giving mid+side and mid-side (mid minus side) on the L/R feeds. Then when broadcast as L+R and L-R (FM broadcast standard), mono listeners get only middle and nothing from the sides. (Note, it's the transmit side, not the receiver, that creates the L+R,L-R, for backward compatibility with mono receivers. Edit- L+R is in the original mono band near the carrier; L-R is on a sub-band outside the audio bandwidth that only the "newer" FM stereo receivers would process.)

In my "guess" post above I was thinking M/S perhaps with a polarity inverted, which would give only the sides and no middle.

Yes, it certainly seems xy mode would be better overall.

Thanks for the thought puzzle.
 
Last edited:

thankyouguitar

Tele-Meister
Joined
Feb 24, 2020
Posts
478
Age
41
Location
Oak Park, MI
I'd originally put this in one of my replies to Mjea80's acoustic guitar recording thread, but I felt like it was a little too off-topic. This is related to a realization I had a few years ago about my old job. Since then, I've been embarrassed but haven't told anyone. To my knowledge, we never got any complaints, but it bugs me because it compromised the output for some of the audience. I'm making this a TDPRI challenge with a reward of maximum audio nerd points.

I used to work a monthly show on a public radio station with a performing arts/recording studio. The hosts were a well-regarded couple who had guests playing (broadly) American folk and old-time music. Most of the performance was done with a small group around a single central mic - in our case, a Shure VP88. They'd gather around and move in and out when they were being featured or needed to drop back in the mix. It was fun. We had extremely talented performers, and they did most of the work themselves. I didn't appreciate it enough at the time. There used to be an online archive of all the shows, but it's sadly gone now.

Now, for the challenge: Where was the screwup? There are clues throughout. Get nerdy. There may be more than one answer, in which case that's additional things for me to be embarrassed about. Thanks for that.

Takers?
VP88 is a stereo mic. Would maybe create phase issues for folks receiving a mono signal?

Edit: I read the answer above and that was what I was pointing to in my guess. My analog receiver will default to FM mono if the signal is not strong enough. I do also have a mono amplification system so if it wasn't summed at the receiver it definitely is at the amp....

OP thank you for posting. This was fun!
 
Last edited:

bottlenecker

Poster Extraordinaire
Joined
Dec 6, 2015
Posts
6,666
Location
Wisconsin
Yes, it certainly seems xy mode would be better overall.

XY has it's own problem. Sometimes mono is derived from stereo by only presenting one channel, like only the left channel, instead of the better way of summing both channels. When this happens, a source recorded XY has no mic pointed directly at it, and can have a weak off axis sound.
 

SRHmusic

Tele-Afflicted
Joined
Oct 19, 2020
Posts
1,888
Location
North Carolina, USA
XY has it's own problem. Sometimes mono is derived from stereo by only presenting one channel, like only the left channel, instead of the better way of summing both channels. When this happens, a source recorded XY has no mic pointed directly at it, and can have a weak off axis sound.
Yes, of course, if that's the case then xy would have issues. Sounds like a bad idea sending only one side as the mono channel. You'd want to sum or have a matrix (or even a mixer) to sum a mid or room mic into the channnels. The long standing FM standard is to send L+R and L-R, so not sure where that would happen in broadcast, but I guess anything is possible and bottom-line- the engineer needs to know their signal chain! :)
 

bottlenecker

Poster Extraordinaire
Joined
Dec 6, 2015
Posts
6,666
Location
Wisconsin
Yes, of course, if that's the case then xy would have issues. Sounds like a bad idea sending only one side as the mono channel. You'd want to sum or have a matrix (or even a mixer) to sum a mid or room mic into the channnels. The long standing FM standard is to send L+R and L-R, so not sure where that would happen in broadcast, but I guess anything is possible and bottom-line- the engineer needs to know their signal chain! :)

Yes, it's a bad way to do it. I don't think it's ever an engineer's choice on the production side. It's something that would happen in other media, like tv, or at a device level. I have no idea how common it is anymore, but there are a lot of cheap no name bluetooth speakers out there.
Personally, I don't care about mixing for crappy speakers or tv, I just really like the sounds and realism I get with mid/side.
 

Ben Harmless

Poster Extraordinaire
Joined
Mar 10, 2003
Posts
5,251
Location
Salem, Mass
You are all providing good data, and these are the types of discussions where I wound up learning a lot about this stuff.
The only experience I have with the “bluegrass self-mixing semicircle” involved a cardioid LDC or an X-Y coincident pair of SDCs. Both seemed to work, but I liked the mono, volume-change-only effect of the single LDC best.
Honestly, a single mic in mono would be my first choice for the classic bluegrass arrangement. It's folk music. High lonesome is emotion as much as it is performance - maybe more, and that's saying something, 'cause some of the performances are jaw-dropping. You don't get this stuff from exciters and stereo-wideners. If I could get away with it, I'd probably run an omni dynamic on top of the LDC and see if it got enough of the vibe.

I'm not sure if you've directly stated what the broadcast audience heard?
Totally fair. I wasn't directly explicit.

To clarify, whatever portion of the audience who was listening in mono would have heard the musicians standing directly in front of the mic, but much less of the people standing off to the side. All the stuff that made it stereo actually cancels itself out because we were laughing in the face of physics by trying to capture three mics worth of recording with only two mics.

M/S is kind of like a deal with the Devil. You achieve greatness, but then the Devil sums you to mono and laughs.

I have a friend who had a solution to that problem (cassette, ca. 1989):
I love this. At times, I've considered making a recording that has things that intentionally disappear from the mix if it's summed to mono - maybe even things that were masking other things. 60 degrees out is a brutal option, but I admire it. I do think that there's a lot of possibility for the future of doing some cool art with multichannel audio if/when fancier systems become more available.

XY has it's own problem. Sometimes mono is derived from stereo by only presenting one channel, like only the left channel, instead of the better way of summing both channels. When this happens, a source recorded XY has no mic pointed directly at it, and can have a weak off axis sound.
Totally true. Particularly for closer mic'ing. That, plus you're gambling with being able to capture the sweet spot on a given source if you want both channels to sound nice. M/S can still do that, but everything's a compromise. That's why we still have a claim to being art!

All of this is way beyond me, but I know the room and the artist Ben is talking about and I’m super curious to know the answer, even if I don’t understand it.
You definitely know where and who, and I believe you've played on the same stage if I'm not mistaken. I'm far enough out of that job where I can savor the positive memories, but forget the rest. I got to meet Arlo Guthrie there though, and he told me stories about dropping out of the little college in Billings, MT where my grandfather taught so that he could east to pursue a career in music and become Arlo Guthrie. That alone was worth it. I'm so far from that now that it feels like another life.
 
Last edited:

David Barnett

Doctor of Teleocity
Ad Free Member
Joined
Mar 17, 2003
Posts
18,447
Age
66
Location
The Far-Flung Isles of Langerhans
Totally fair. I wasn't directly explicit.

To clarify, the listener on a mono system in this case would get the center channel (the mid) but no part of the sides because they're generated by a single figure-8 mic which has had its polarity flipped to make two channels out of it. Those interact with the mid channel to create a fully accurate stereo image when you're listening on a stereo system, but if you sum that stereo signal, the stuff that is still there in reverse polarity will cancel and leave nothing but the mid channel. This isn't a big deal if you're using M/S to provide stereo vibe on say, a soloist, as you'd always be pointing the mid mic at an instrument/voice/whatever, and losing the sides just loses a little of the room sound. If you WANT what's on the sides but you get summed, then you're reliant on whatever off-axis pickup the mid mic might have. All the sources to the edges of that pattern (90 degrees from axis, say) are going to be a lot softer and lose their natural tone.

M/S is kind of like a deal with the Devil. You get greatness, but then the Devil sums you to mono and laughs.

So, like, did the mandolin and dobro disappear every time the violin stepped up to solo?
 

David Barnett

Doctor of Teleocity
Ad Free Member
Joined
Mar 17, 2003
Posts
18,447
Age
66
Location
The Far-Flung Isles of Langerhans
Also - does the Shure internally matrix the L+R and L-R channels into L and R outputs, or does it require an external matrix when using it in M-S mode?

And isn't it possible to fake the M-S matrix on a pro mixing console just by feeding the L+R mic into one channel, panned straight up, then Y the L-R into two channels, pan them hard L and R, with the polarity flipped on the R one?
 

loudboy

Tele-Afflicted
Joined
May 21, 2003
Posts
1,292
Location
Sedona, Arizona
And isn't it possible to fake the M-S matrix on a pro mixing console just by feeding the L+R mic into one channel, panned straight up, then Y the L-R into two channels, pan them hard L and R, with the polarity flipped on the R one?
It's easier to do it in your DAW - just duplicate the track, flip polarity on it, pan them out and group the faders. Bring them up to where you want the stereo field to be.
 

bottlenecker

Poster Extraordinaire
Joined
Dec 6, 2015
Posts
6,666
Location
Wisconsin
This isn't a big deal if you're using M/S to provide stereo vibe on say, a soloist, as you'd always be pointing the mid mic at an instrument/voice/whatever, and losing the sides just loses a little of the room sound. If you WANT what's on the sides but you get summed, then you're reliant on whatever off-axis pickup the mid mic might have. All the sources to the edges of that pattern (90 degrees from axis, say) are going to be a lot softer and lose their natural tone.

This is true, but some cardioid patterns are wide enough to park a bus in. If the mid mic does a good job with your source(s) by itself, m/s is a good choice. If not, m/s is a bad choice. I would not put instruments hard to the side of an m/s pair anyway just because it sounds too extreme to me, like panning something in a mix hard left or right.


download (4).jpeg
 




New Posts

Top