Taperssection.com
Gear / Technical Help => Post-Processing, Computer / Streaming / Internet Devices & Related Activity => Topic started by: BradleyJY15 on September 14, 2021, 12:15:30 AM
-
Didn't see any recent topics on this...
When you tape AUD & SBD sources, how do you account for the delay?
I walked distances off and went with 1,125 ft/sec @ ~68 F.
Do you normally go by distance/equation or line up wave forms?
-
I am an avid golfer but use GPS not a range/pin finder. Shooting the distance to a stack seems best. Is this going to change my golf game?
-
If already time aligned on the same deck, then just find a nice clean high frequency transient (think hi-hat) and align on that.
-
I am not great at this, but for my 12 show SBD/AUD adventure in 2018, I found doing it by ear/eye worked better than the actual math (Audacity). Although, I estimated, no golfing GPS for me. (no golfing period, but that's a different topic-OT)
-
I typically just line it up later by sight and by ear using spoken stage dialog between songs and/or drum kit transients. But I still like to note a distance estimate in my recording notes by walking off the linear distance from PA or stage to the recording position, at which point I can approximate the needed delay using the 1ms = 1.1' rule of thumb. That gets within putting range.
The latest recorders I've been using have the ability to apply a delay to each channel at the input of the machine, and I've thought about potentially using using that capability to create a live matrix mix for a patcher. But more often than not I'm using all channels of one recorder for mics and just record the SBD if available to a separate machine and deal with it later.
Of course you need to time it right to sink one past the miniature windmill hazard.
-
I don’t bother with delaying while recording. It’s way easier, and probably more accurate, to just move the file by site/ear in a daw
-
I applied the calculation and it sounded great. I think I'll first apply the calculation then check by site/ear in Audition. The distance I walked off to the stacks isn't as precise as I would like - I walked up to the barricade in the pit, so probably short a little. I am going to eyeball the final distance to stacks in the future. My pacing is very good (I am a civil engineer and had to learn good pacing as part of a surveying class 20+ years ago - and as said above as a golfer, my paces are damn near precise).
And a BIG THANKS to everyone here for all the help you have provided to get my setup to where I am now. Couldn't have done it without this website/forum. So much good information for such a small niche of a hobby.
-
I start with 1ms/foot and go from there.
-
Unaligned to some degree helps define the size of the room, sometimes that's useful and sounds more realistic than fully aligned, as if there is no room predelay. I frequently make a large room smaller, but not all the way to 'reverb emanating directly from band'. You can also measure song tempos and set the offset to a musical interval so the slap is in time.
-
I line up recordings by eye/ear using the signal peaks during quiet spots if I can.
Like when the stage is totally quiet, but someone clicks a guitar cable in, that's just about perfect. Or a count-in of rim clicks from the drummer...
Once I align them, THEN I do the math to double check, and see if it makes sense.
If the distance is some impossibly large number, I recheck the peak alignments and the math.
Note that audio system alignment can be done with software like Meyer SIM or Rational Acoustics SMAART to get the most precise results, but they are not necessary for post production work.
-
At larger shows I ask the FOH for the distance since a lot of digital boards will have that info already. I align in post production using the math (around 1ms per foot), the phase scope in Wavelab and using ears. The comb filtering from misaligned tracks is most noticeable to me in the low frequencies.
I have often used the delay on recording decks if sending a mix to an output either for another taper or a stream.
-
As EmRR mentions, sometimes the best alignment is a slight misalignment. I like to start by getting it as close as possible, then shift it around by ~millisecond increments while listening, going with whatever sounds best.
The when and why? There is a number of things going on-
It will depend somewhat on the relationship of the AUD and SBD in terms of how much direct-arrival content is shared between them and how similar it is (that is, direct-arrival of band sound, not direct-arrival of audience and ambient sound present in the AUD but absent from the SBD). The SBD provides dry close-mic'd direct-arrival sound, maybe with some 'verb on it, while the AUD generally contains some combination of direct-arrival and reverberant sound. The AUD provides the reverberation and reverberant tail, while both AUD and SBD typically provide direct-arrival content, and that relationship is key. An AUD pair of omnis is less likely to require precise alignment (or rather, may allow for more misalignment, intentional or not) with the SBD than a pair of PAS hypercards from the same AUD location.
When there is a lot of clear direct-arrival band sound in the AUD, misalignment with the SBD acts somewhat like slap-back echo. When the AUD is dominated by reverberant sound, misalignment of the two acts more like the predelay control on a reverb unit. That control over the predelay-like aspect is where EmRR's comment- "I frequently make a large room smaller, but not all the way to 'reverb emanating directly from band'." comes into play. The closer the alignment, the less pre-delay before the reverberant tail occurs, emulating the acoustic behavior of a smaller room.
In the impulse over time graph below, this relates to changing the distance between the direct impulse and the early-reflections/late-reverberation part. Apologies for the huge image.
(http://info.audiokinetic.com/hubfs/impulseresponse.jpg)
Slapback echo comes from misaligned direct impulses. The room size / pre-delay aspect relates to the gap between the direct impulse and the reflections and reverberation portion that follows.
More on some esoteric aspects later..
-
Somewhat more esoteric-
For those familiar with the multichannel OMT (oddball microphone technique) recording approaches that I and some other tapers use, some arrangements of which incorporate rear-facing microphones, one reason for using more-directional microphones in the rear-facing positions is to limit the pickup of direct-arrival content from the band and everything else in front into those channels to the greatest degree practical. With less front-arriving direct content in them they can be manipulated more freely without incurring audible problems, which might include delaying them slightly in some instances. A slight delay can sometimes be helpful in uncluttering the sound by allowing the front-direct sound to breathe a bit more, and/or make the reverberant room sound more lush without actually increasing its level. It is an example of intentionally misaligning the bleed from the front into the rear facing ambient channels.
When an OMT recording using rear-facing channels is used for surround playback rather than a 2ch stereo mix, a slight delay can sometimes help with keeping the band sound locked to the front as the level to the surround channels is increased to a desirable level, allowing for greater surround immersion before sounds from in front start to get pulled too strongly into the sides and rear. Again, this is an example of intentional time misalignment of the direct sound arrival from the front with its bleed into the ambient/surround speakers.
-
Even more esoteric, yet potentially more applicable for tapers using traditional AUD microphone setups as well as OMT microphone arrangements-
There is a technique for mixing orchestral spot microphones with the main microphone channels that I believe might also be applicable to the SBD part of a matrix. After all, a SBD feed is essentially a collective stem of close spot mics. This technique treats the AUD as the foundation of the recording - Like a classical main microphone pair, aiming to preserve it's primary role and imaging qualities. It essentially treats the spot (SBD) signal in such a way as to emulate and reinforce the early-reflections part of the graph I posted above, letting the main pair provide the initial direct-impulse and reverberation parts.
It time-delays the spot mic (SBD in our case) so that it's initial impulse is aligned with the main (AUD) pair, but routes that through a multi-tap stereo delay/reverb unit set in such a was as to emulate random early reflections. The unit is is set to 100% wet, eliminating the direct impulse, and set to provide no reverb or tail, leaving both of those parts to the main (AUD) pair alone. It produces only a series of short dry delays, each randomly panned and with random decreasing levels within the appropriate time range (note the downward trend and range of amplitudes in the early ref. portion of the graph above), mixed in with a level lower than that of the direct signal from the main (AUD) pair but higher than the reverberant tail.
Ironically, this actually delays the SBD to be slightly behind the AUD, but only by a few milliseconds. This is what preserves the stereo imaging cues provided by the AUD pair.
This technique would apply only to an already great AUD that might be made even better with some clarity sweetening. It might also work as a way of effectively stereoizing a mono SBD. I have a paper on this somewhere and will search for it. A brief online search didn't pull it up.
-
Even more esoteric, yet potentially more applicable for tapers using traditional AUD microphone setups as well as OMT microphone arrangements-
There is a technique for mixing orchestral spot microphones with the main microphone channels that I believe might also be applicable to the SBD part of a matrix. After all, a SBD feed is essentially a collective stem of close spot mics. This technique treats the AUD as the foundation of the recording - Like a classical main microphone pair, aiming to preserve it's primary role and imaging qualities. It essentially treats the spot (SBD) signal in such a way as to emulate and reinforce the early-reflections part of the graph I posted above, letting the main pair provide the initial direct-impulse and reverberation parts.
It time-delays the spot mic (SBD in our case) so that it's initial impulse is aligned with the main (AUD) pair, but routes that through a multi-tap stereo delay/reverb unit set in such a was as to emulate random early reflections. The unit is is set to 100% wet, eliminating the direct impulse, and set to provide no reverb or tail, leaving both of those parts to the main (AUD) pair alone. It produces only a series of short dry delays, each randomly panned and with random decreasing levels within the appropriate time range (note the downward trend and range of amplitudes in the early ref. portion of the graph above), mixed in with a level lower than that of the direct signal from the main (AUD) pair but higher than the reverberant tail.
Ironically, this actually delays the SBD to be slightly behind the AUD, but only by a few milliseconds. This is what preserves the stereo imaging cues provided by the AUD pair.
This technique would apply only to an already great AUD that might be made even better with some clarity sweetening. It might also work as a way of effectively stereoizing a mono SBD. I have a paper on this somewhere and will search for it. A brief online search didn't pull it up.
This is genius!
I've always made sure that my SBD feed happens before the audience mics, but the early reflection technique is a way to get around that, and preserve the natural stereo image.
Just to clarify, in this technique, there would not be ANY direct soundboard feed used, only the 100% wet early-reflected convolution of the dry feed?
-
As I understand it, it's basically a series of rapid delays. Each identical to the dry SBD signal, but with various levels with an an overall attenuating trend, and with various pannings.
So a delay more than a 'verb thing.. convolution could do it of course, but so could a basic digital stereo multi-tap delay. The 100% wet thing is just a way of initially lining up the direct-arrival impulse of both AUD and SBD like we typically do, but then only letting the SBD contribution occur during the early reflection time zone without having to do a subsequent time shift.
-
Even more esoteric, yet potentially more applicable for tapers using traditional AUD microphone setups as well as OMT microphone arrangements-
There is a technique for mixing orchestral spot microphones with the main microphone channels that I believe might also be applicable to the SBD part of a matrix. After all, a SBD feed is essentially a collective stem of close spot mics. This technique treats the AUD as the foundation of the recording - Like a classical main microphone pair, aiming to preserve it's primary role and imaging qualities. It essentially treats the spot (SBD) signal in such a way as to emulate and reinforce the early-reflections part of the graph I posted above, letting the main pair provide the initial direct-impulse and reverberation parts.
It time-delays the spot mic (SBD in our case) so that it's initial impulse is aligned with the main (AUD) pair, but routes that through a multi-tap stereo delay/reverb unit set in such a was as to emulate random early reflections. The unit is is set to 100% wet, eliminating the direct impulse, and set to provide no reverb or tail, leaving both of those parts to the main (AUD) pair alone. It produces only a series of short dry delays, each randomly panned and with random decreasing levels within the appropriate time range (note the downward trend and range of amplitudes in the early ref. portion of the graph above), mixed in with a level lower than that of the direct signal from the main (AUD) pair but higher than the reverberant tail.
Ironically, this actually delays the SBD to be slightly behind the AUD, but only by a few milliseconds. This is what preserves the stereo imaging cues provided by the AUD pair.
This technique would apply only to an already great AUD that might be made even better with some clarity sweetening. It might also work as a way of effectively stereoizing a mono SBD. I have a paper on this somewhere and will search for it. A brief online search didn't pull it up.
This is genius!
I've always made sure that my SBD feed happens before the audience mics, but the early reflection technique is a way to get around that, and preserve the natural stereo image.
Just to clarify, in this technique, there would not be ANY direct soundboard feed used, only the 100% wet early-reflected convolution of the dry feed?
Not sure about the classical use case, but you can do it both ways. Plenty of verb plugins include "early reflections" and using a 100% wet early reflections verb bus is generally great way to create space on DI'd instruments, so I'd assume it works the same way here.
-
Found it. Günther Thiele's Room Related Balancing Technique-
Link to the original 1990 paper (scanned) available at the Hauptmikrofon site:
https://hauptmikrofon.de/theile/1990-10_Room-related-Balancing%20_88.AES-Preprint2886.pdf
I came across it in a later 2001 paper of his on multichannel recording. This one is well formatted, easier to read and has better illustrations of the idea. Skip to page 36 for the relevant section:
https://hauptmikrofon.de/theile/2001-2_Multich.-Recording_Nordic-Symp.-2001.pdf
-
If already time aligned on the same deck, then just find a nice clean high frequency transient (think hi-hat) and align on that.
+1
I trust my eyes to a wave form, more than mathematics, or my ears. Then again, Mr. Hose is my time alignment guru.
-
I reread the papers I posted above, along with a couple others, and my main take away remains: Align by eye, confirm and tweak by ear (bumping it back and forth a bit by a few ms to confirm the sweet spot) as being a good practical approach.
Not sure how applicable the Room Related Balancing thing is for TS tapers. It's mostly about preserving depth cues present from the main pair (AUD in our case) to present a more convincing illusion of source depth upon playback. That's not especially important or desirable for most TS tapers recording PA amplified performances where a primary intent of mixing in some SBD is increasing the sense of proximity / decreasing the sense of distance from the source in addition to reinforcing clarity and detail.
Thinking back about listening to my recordings and those of others, I can't recall hearing any real sense of depth within the source at PA amplified concerts. There is depth in audience and hall sound which can be good, and the overall sense of distance to the source which is generally best minimized as much as possible, but little if any depth of sources withing the band itself. Of the times I can recall such a perception of depth it was problematic - electric jazz acts in classical halls, not fully PA'd where some performers are clear but others are lost in distant mush, comes to mind. Were I do hear and value a great sense of depth is with non PA amplified acoustic music, particularly classical music in a good hall. Getting a clear sense of front/back depth without loosing too much clarity and detail of the more distant sounding stuff in back is one of my favorite things to get right with an orchestra. Everything on stage sounding equally close and clear just doesn't sound correct or natural for that, it looses a delicious 3-dimensional sense of reality, unlike PA amplified stuff where that is more or less the expected goal of a good AUD/SBD mix and any sense of depth within and between the various sources is derived from effects and reverb.
Regardless, one's listening preference comes first so the align by eye, tweak by ear approach applies either way.
In that light, I take the Room Related Balancing technique to be most useful as a way of understanding what's going on, more than an approach that applies necessarily to the kind of taping most are doing around here.
It might apply in circumstances where upon dialing in the best AUD/SBD balance everything sounds appropriately clear yet otherwise too flat and soundboardy, as a way of retaining that SBD clarity yet gaining some depth. Sort of the opposite of the way in which I generally use SBD as clarity seasoning, yet the way many tapers mix things using more SBD and less AUD.