Topic: matrix delay

[DaryanLenz]

So, I have this idea up my arse, that says it wants to start making matrix tapes with a new mixer I just picked up.  Now I know delay issues exist in order to time align the mic and sbd sources.  So, theoretically, if I were to run mics say, at the sbd, how much delay are we talking here?  How about closer up?  Second, what kind of unit does one need to do this...preferrably not too exspensive...ie, 100 bucks?  Thanks for the help, I really appreciate it!
Thanks

Daryan

[Tim]

if the mics are within 100 feet of the board you should be okay without a digital delay unit.

[MattD]

Tim, I don't follow your reasoning. Perhaps a further explanation would shed some light on things. Why does distance from the board matter? Isn't the distance from the sound source also important? Assuming instantaneous transmission (no cable delay), if the mics are 100 feet away from the PA, then the mixer will receive the sound from the PA about 0.09 seconds later than the board feed. This is an audible difference.

-Matt

[Tim]

it was late, I meant within 100 feet of the PA, thanks!

rethinking my idiocy in the light of day and after a vigorous workout at the gym the rule of thumb is 1ms for every foot. I guess 100 feet might be a bit much  

[JeffK]

You won't notice a huge delay if you're within about 30 ft of the source.  So... most small clubs, you'll be fine running sbd/aud mixes, but outdoors or larger theaters, you'll wanna have a digital delay unit.  Even in smaller clubs, you'll notice that your sbd/aud mix tapes will sound MUCH cleaner with the delay unit in place.  The only one that I'm familiar with is the TC Electronics D-Two, which is a rackmount unit that you'll need AC for, sells new for a bit over $400, IIRC.

Tim is right though, with a guestimate of about 1ms/ft. of delay.  The actual delay formula has to do with temperature as well though, so it will actually change throughout the show.  Either way though, you're still a lot closer to being exact than not using a delay unit altogether.

[Todd R]

I think you'll need to be no further away from the stacks than 30 to maybe 40 feet.  I did a SBD+Aud matrix of SCI a few years back without a delay for the board feed.  Jon-O asked what I was doing, and I told him I was trying a matrix.  He sortof laughed, but told me to go ahead and try it.  I was recording by the board about 50 feet back.  Jon-O said it would work when you were 30 feet or so back, but had his doubts it would work 50 feet back.  Turns out he was right--it's not completely unlistenable, but has quite a bit of echo/reverb to it.  Recorded outdoors, but sounds like it was recorded from the back of a big hockey arena.  I'd never try it without a delay as far back as 100 feet--that'd be about 100ms delay difference and I bet would sound like ass.

I never found much in the way of a cheap, portable, DC-powered delay.  I did pick up, but never got around to using, an Alesis nano-verb.  You can pick one up for $50 or less.  About the size of a D8, it is an 18-bit digital stereo effects processor for guitar.  Lots of effects, including a digital delay.  No way of setting your exact millisecond delay, but you should be able to get a feel for it and do it by ear.  Only other issue is that it is AC powered so you'd need to use a power inverter to run it.

I'm curious if other folks have found a reasonably priced, portable, DC powered digital delay that would work for this task.

-Todd

[Jason B]

I bought a Behringer UB1832FX-Pro mixer especially for multi-mic and matrix mixes. It has a built in, 24 bit effects processor with 99 effects. I think there are 6 or 7 delay effects, but they are not individually setable. They are all preset delays.

I have made some board/aud matrixes with it, but have yet to do so in a room where my mics weren't within 30 feet, thus I have not yet had a chance to use the delay.

-JB

[John R]

jeff's right not only temp, but barometric pressure and wind of course.  jon-o uses the yamaha digital board for peters tapes and the stuff i get at blues and brews.  he IS the man.

jr

[Tim]

John are you actually in Telluride? God I love that place...

[John R]

moved down to seaside fla in dec after 12 years in tride.  going back for bluegrass, four corners, and hopefully blues and brews again if i can put it together.  you going to be there this summer?  have we met there?

jr

[Marc Nutter]

Hi All,

While you may not hear a discrete delay (echo), assuming equal levels of both sources) until about 50mS (about 55 feet) or more, the common sources being combined but out of time will cause significant comb filtering (cancellations in the frequency response).  Even a time arrival difference of 1mS causes combing to begin at 500 Hz and have intervals every 1k on up.

A good way to observe the impact of misalignment is to take two recordings of the same show, align them perfectly and then offset one of them by .5mS, then 1mS, and more to see how it works.  This can be achieved with a good software package that allows working with more than two audio channels.  The offset can be performed by dragging one file or through setting a delay on an insert or effects send.  

It won't be until about 40mS or more that you can actually detect a discrete echo.

While the initial tendency is to hear the combined board/audience ("matrix") recording as fuller and richer, due to the level increase of two common sources, you will gradually start noticing the changing frequency response.  I often find that it is one way to take two good recordings and make one mediocre one, unless great care is given to proper time alignment, including the changes that occur as the show goes on (heat and humidity), and changing the time offset as needed to maintain alignment.

As far as those who are interested in finding a delay unit, they are somewhat hard to come by.  Those used for sound system driver and cabinet alignment often include numerous functions that will not be of use to most home users, unless you also want a great crossover and EQ.  Then, expect to pay over $1000.00

Less expensive units like those from Rane and Behringer, available for $200-400.00, are known for their functionality and price point but not necessarily for optimal sound quality.

Still once you are in that price range, assuming you have a computer, you are well on your way to a recording package that would perform the exact functions and have more beneficial qualities to offer as well.

Happy Recording Everybody,

Marc

[jpschust]

Marc,

A question since you have so much knowledge on the subject.  which would be prefferable to the ear:

1.  A matrix made with a very high end digital delay in a 2 mic/sbd mix (nothing terribly fancy)

or

2. A "matrix" made with an aud recording and a patch off the board post mixed through software?  Would doing this negate the fq issue faced without the digidelay unit?

Jonny

[wboswell]

Marc-
Another ?  When you say "good software package", which would you consider a good enough package to matrix in the post without the dealing with such cancellations?  I realize that my question isn't necessarily fair, so here is another go:  Most people on the hobby level use programs such as Sound Forge 6.0, Wave Lab, and Cool Edit Pro.  Will these commonly used programs produce desirable results?  Thanks for your insight!

-William

[JAH]

Marc-
Another ?  When you say "good software package", which would you consider a good enough package to matrix in the post without the dealing with such cancellations?

-William

Samplitude or ProTools are the shit.  A product that you get what you pay for.

jah

[jpschust]

I'd probably use protools.  Soundforge and Wavelab arent meant for multitrack applications so much.  Yes they can do the combinations, however thats not really what they are meant for.  I'd want to be working in software designed for multitracking purposes.

[nickgregory]

I have used Cool Edit Pro with very good results.

Nick

[Marc Nutter]

Marc,

A question since you have so much knowledge on the subject.  which would be prefferable to the ear:

1.  A matrix made with a very high end digital delay in a 2 mic/sbd mix (nothing terribly fancy)

or

2. A "matrix" made with an aud recording and a patch off the board post mixed through software?  Would doing this negate the fq issue faced without the digidelay unit?

Jonny

Hi Jonny and All,

I prefer to be able to keep the recordings separate (ie. at least in archived form).  Then, when you combine them, you can always undo your steps.

With an inline delay making a single two channel recording, you only get one shot and without a audio measurement system (SmaartLive, SIM II, others...), you will not be able to set the delay with great accuracy nor make changes as the show goes on.

By having two discrete recordings and mixing them with software, you can stop, re-align, and continue as much as necessary.  Since the speed of sound in air changes with temperature and humidity, delay times do need to be changed (usually only about 1-2mS per hour of audio).  

While there are varying methods for signal alignment, in software, you can make a cut in the space between a couple songs (on the board recording since its speed is constant) as compared to the changing ambient recording and moving that segment until it aligns nicely with the other recording.  This is more visual or graphic in nature.  The other way is to add delay through an effects send or insert in the software mixer.  The latter does not provide the visual benefit of the first.

With software, you can align visually by looking for impulses like a kick or snare drum which causes a dynamic peak above the majority of the recording.  Everything will seem perfect for three of four songs, and after a while, kick drums will start sounding mushy and impulses will no longer be visibly aligned.  This is where it becomes necessary to make further offsets.

Happy Recording,

Marc

[nickgregory]

Not sure how "precise" this method is, but what I do is take my AUD .wav as my master, cut up the SBD feed into tracks, and align them in Cool Edit as seperate tracks, finally mixing them all together.  In my experience this helps to get rid of the delay issues that would occur in just mixing 2 long wav's together.

Nick

[Marc Nutter]

Marc-
Another ?  When you say "good software package", which would you consider a good enough package to matrix in the post without the dealing with such cancellations?  I realize that my question isn't necessarily fair, so here is another go:  Most people on the hobby level use programs such as Sound Forge 6.0, Wave Lab, and Cool Edit Pro.  Will these commonly used programs produce desirable results?  Thanks for your insight!

-William

Hi William (and All),

I really like Wavelab.  While it is not designed to track more than two channels simultaneously, I believe the Audio Montage section could be utilized to do exactly what we are discussing.  Since I have Nuendo (WARNING: Expensive), I typically use it for multi-channel projects.  However, in looking at the Audio Montage in Wavelab, it seems that you can move tracks around like any other multitrack software and at does have an effect called "echo" which can be used as a delay if you like this approach better.

Otherwise, Samplitude, as already mentioned is great.  ProTools works but is once again, painfully expensive.  I think Sonic Foundry Vegas 4.0 will support similar editing as well.  It sounds like Cool Edit will too.

Any others???

Happy Recording Everybody,

Marc

[wboswell]

I have been using Sound Forge 6.0 exclusively.  Maybe its time for me to demo a new program.  However, I have had success using the 'mix' feature by copying one track and pasting it to the other as Nick mentioned he did with CE Pro.  I haven't had any luck matching long wave files, which I have learned is due to varying word clocks in different A>D's.  I have enjoyed the results, though it truely is a labor of love.  Many hours are required to match the files.

What is your preferred 'mix'?  I try to run the sbd source somewhere between 12 and 15 dbs louder than the aud source.  This, of course, is subject to the genre of music.  A Tim O'brien show that I did needed more much more sbd than aud because of the crowd response to each song.  I have since edited the crowd reactions by decreasing the response volume before the matrix is applied.

Cool topic.  I would love to find a program that was a little more user friendly.  Maybe not even user friendly, just less time consuming!

Have fun!
-William  

[Marc Nutter]

Hi William and All,

Just like so many other aspects of recording, I've never found that there is a one specific way that works best all the time.  I like to take the audience recording (as I generally like them better than SBD recordings) and EQ it to taste (usually reducing bottom end that I find excessive).  Then, I take the SBD recording and mix it in until I find the overall clarity/presence/detail that I'm looking for.  

Another approach may be to start with the SBD recording (especially if you like how it sounds) and add enough of the audience recording to give it the mix the richness and feel for the room that you are looking for.

Remember, the faders may be deceptive, since levels may be different on the recordings.  Instead of seeking a percentage or ratio, I really like to go by ear.
You may even find it fun to add a touch of compression on one or both recordings, EQ them separately and then as a mix, or even add some reverb to the SBD or the mix.  

Since we are deviating from the leave it as it is approach, go deep, make mistakes and figure out what you like best.  Sounds familiar, huh?

Happy Recording,

Marc

[wboswell]

I just got through doing that for a buddy's band.  Thank goodness it was only an hour long set!  Not exactly my favorite flavor of music, but its done now so no worries.

I went based on my ear for the mix.  The SBD source was pretty miserable.  The vocals were to high in the mix, drowning out the rest of the instruments.  However, the AUD source was a good representation of what it sounded like.  

I ran cards further back than I normally do to pick more of the room, which I think helped out the in the mix process, but not necessarily on the AUD recording.

Given the problems with the SBD mix, I had to leave the AUD source much louder than I normally do.  Instead of leaving the SBD source at 100% and decreasing the AUD source by ~12dbs like I normally do, I left the SBD source at 100% and decreased the AUD source by only 2dbs.  There is a little more chatter than I would have liked, but the end result is a very full sounding recording.

If I was really motivated, I might have played with the EQ settings a little, but I have already put more time into than I needed...  afterall, this is charity work for a friend.  I even paid my way into the show!

Later, William

[F.O.Bean]

charity work........i do WAY TOO much of that.......sometimes, it just gets under your skin tho, to see how good you can make a buddys band sound.... you know what i'm talking about......

bean

[wboswell]

charity work........i do WAY TOO much of that.......sometimes, it just gets under your skin tho, to see how good you can make a buddys band sound.... you know what i'm talking about......

bean

Yeah.  No matter how hard you try to make them sound better, they still sound like your buddy's band.  Like I have been told before, you can polish a turd, but its not going to make it any better!