You are here

Mixing Separate Multitrack and Stereo Recordings

All about the tools and techniques involved in capturing sound, in the studio or on location.

Mixing Separate Multitrack and Stereo Recordings

Postby tk76 » Sat Jul 06, 2019 8:58 pm

I have limited experience in live recording and want to multitrack a performance using 8 channels from a Sound Devices Mixpre 10T. I also have a Tascam DR100mkIII that I'd like to use to capture a stereo recording of the room. Is it possible to line up and mix the multitracks and stereo take in my DAW successfully?

Cheers :angel:
tk76
Poster
Posts: 23
Joined: Sun Sep 06, 2009 12:00 am

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Sam Spoons » Sat Jul 06, 2019 9:34 pm

Yes, try to record a transient sound like a handclap after you start both recorders and you will then have a reference to line to line the tracks up in the DAW.
User avatar
Sam Spoons
Jedi Poster
Posts: 9728
Joined: Thu Jan 23, 2003 1:00 am
Location: Manchester UK
Finally taking this recording lark seriously (and recording my Gypsy Jazz CD)........

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Mike Stranks » Sat Jul 06, 2019 10:18 pm

Although it's possible to sync-up the recordings as Sam says, unless the two recorders both have very good clocks then they'll fairly quickly drift out of sync. Been there, done that, got the T-shirt... :thumbdown:

The only truly satisfactory way to do this is to have a master/slave clocking arrangement. Sadly, I don't that the Tascam will let you slave to another clock.
Mike Stranks
Jedi Poster
Posts: 6566
Joined: Fri Jan 03, 2003 1:00 am

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Hugh Robjohns » Sat Jul 06, 2019 10:38 pm

You're talking about working with two separate free-running recorders here, and good as modern digital clocks are, there will be some level of drift between them over the length of a show. So while a handclap (or similar) will help you to align the start, there's no guarantee the two recordings will still be in sync at the end.... Or of how much drift there'll be.

In an ideal world you'd synchronise the wordclocks of the two recorders, and the mixpre10T does have the right facilities for that... But the Tascam doesn't. It can't output a clock and it can't recieve one either. It's just not designed for synchronised operations.

Soooo.... Your best bet is to sync the start as Sam suggests, and then listen carefully for phasing colouration between the stereo track and the multitrack which will indicate when the drift has exceeded an acceptable amount. When that occurs you'll need to locate a suitable transient at an earlier point and edit the stereo track to re-synchronise. Rinse and repeat as necessary...
User avatar
Hugh Robjohns
Moderator
Posts: 23748
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Eddy Deegan » Sat Jul 06, 2019 10:59 pm

Hugh Robjohns wrote:listen carefully for phasing colouration between the stereo track and the multitrack which will indicate when the drift has exceeded an acceptable amount.

Given that there is an acceptable amount, which will be present for much of the duration in varying amounts, would timestretching the shorter recording to match the longer one be an acceptable (and much easier) solution? Adding a transient at the end of the recording to aim for would help good synchronisation if so.

I had cause to do a little timestretching in Reaper recently and was surprised at how little it affected the audio, though I was only doing a little nudge and the duration of the clip was no more than a minute.
User avatar
Eddy Deegan
Frequent Poster (Level2)
Posts: 2198
Joined: Wed Sep 01, 2004 12:00 am
Location: Brighton & Hove, UK
Some of my musical works.
I had a weird time in Surrey once, but that was a drummer's fault.

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Sam Spoons » Sat Jul 06, 2019 11:08 pm

I was assuming the OP just wanted to use the stereo mix as room mics (or the Mix Pre tracks as spot mics) but Eddy makes a valid point......
User avatar
Sam Spoons
Jedi Poster
Posts: 9728
Joined: Thu Jan 23, 2003 1:00 am
Location: Manchester UK
Finally taking this recording lark seriously (and recording my Gypsy Jazz CD)........

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Hugh Robjohns » Sat Jul 06, 2019 11:17 pm

It's certainly worth a try and is a good suggestion.

However... There's also no guarantee that the speed difference between the two machines is constant. It may be... But it might just as easily vary with temperature, battery level, or just for the hell of it... So you might get the start and end in sync, but still find some audible drift at various points throughout the concert. If so, your best option will still be some internal re-sync editing.

Alternatively, maybe it would be worth hiring another digital recorder that can be sync'd with the 10T to save all those hours of re-sync work in the DAW. ;-)
User avatar
Hugh Robjohns
Moderator
Posts: 23748
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Eddy Deegan » Sat Jul 06, 2019 11:20 pm

Hugh Robjohns wrote:Alternatively, maybe it would be worth hiring another digital recorder that can be sync'd with the 10T to save all those hours of re-sync work in the DAW. ;-)

The best solution to any problem is to avoid having it in the first place :thumbup:
User avatar
Eddy Deegan
Frequent Poster (Level2)
Posts: 2198
Joined: Wed Sep 01, 2004 12:00 am
Location: Brighton & Hove, UK
Some of my musical works.
I had a weird time in Surrey once, but that was a drummer's fault.

Re: Mixing Separate Multitrack and Stereo Recordings

Postby James Perrett » Sun Jul 07, 2019 12:31 am

I've had to do this quite a few times where I'm using the venue's digital mixer to record the PA mics and then my own rig to record the room and anything that isn't being picked up by the PA. The mixer is often at the other end of the room to my rig so running a word clock cable isn't going to help. Fortunately room mics don't need to be sample accurate with the PA (at least not in the configuration that I use) so, in Reaper, I just line them up at the start, place a stretch marker at the start and then place another stretch marker at the end and drag it until the audio lines up. This seems to work pretty well but you need to remember to tell Reaper not to preserve pitch when stretching.
User avatar
James Perrett
Moderator
Posts: 8198
Joined: Mon Sep 10, 2001 12:00 am
Location: The wilds of Hampshire
JRP Music - Audio Mastering and Restoration. JRP Music Facebook Page

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Eddy Deegan » Sun Jul 07, 2019 12:43 am

On a related note, is there any preference to shrinking a longer take to fit, as opposed to stretching a shorter one or is the difference moot?

My logic was that stretching requires the interpolation of new data whereas shrinking doesn't but I may be misunderstanding the mechanics of the process.
User avatar
Eddy Deegan
Frequent Poster (Level2)
Posts: 2198
Joined: Wed Sep 01, 2004 12:00 am
Location: Brighton & Hove, UK
Some of my musical works.
I had a weird time in Surrey once, but that was a drummer's fault.

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Tim Gillett » Sun Jul 07, 2019 12:53 am

I concur with James. If if it's all your own gear it's easier. Assuming there is a timing difference - there may be none - make a note of it. It will be the same each time you record with that gear. I once needed to record a band in 10 tracks but only had an 8 track Zoom R16. I recorded the extra two tracks on my old H4 and time stretched the difference later.
Tim Gillett
Frequent Poster
Posts: 1718
Joined: Wed Jan 30, 2013 1:00 am
Location: Perth, Western Australia

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Tim Gillett » Sun Jul 07, 2019 12:57 am

Eddy Deegan wrote:On a related note, is there any preference to shrinking a longer take to fit, as opposed to stretching a shorter one or is the difference moot?

My logic was that stretching requires the interpolation of new data whereas shrinking doesn't but I may be misunderstanding the mechanics of the process.
I don't know but differences between digital recorders of the same nominal sample rate should be so very small that there should be no audible artifacts anyway.
Tim Gillett
Frequent Poster
Posts: 1718
Joined: Wed Jan 30, 2013 1:00 am
Location: Perth, Western Australia

Re: Mixing Separate Multitrack and Stereo Recordings

Postby hobbyist » Sun Jul 07, 2019 1:04 am

tk76 wrote:I have limited experience in live recording and want to multitrack a performance using 8 channels from a Sound Devices Mixpre 10T. I also have a Tascam DR100mkIII that I'd like to use to capture a stereo recording of the room. Is it possible to line up and mix the multitracks and stereo take in my DAW successfully?

Cheers :angel:

Yes.
Although some DAWs may be easier/better than others to do it.

Put them in the same DAW and compare them.
Find something unique on both to align them with.

If one is for ambience then it may need to be several milliseconds slower depending on the room size.

Trying to mix it in stereo with the main inputs may be a problem.
I would convert it to mono and add ambience that way. But what do I know, I am just a hobbyist not a professional.

Time stretch one of them to match the other. Do it both ways from the middle if that is where the sync point is.
The signal matching may not be perfect but will be plenty good enough. Especially if one is just to add ambience.

Worst case you have to resync the start of the ambience on occasion while you edit , if it it noticably different.

I had two old Lafayette analog recorders that I played simultaneously and that kept near perfect sync back in the 60s. I would not expect enough variation now to write home about.

I could be wrong as some cheap wrist watches keep perfect time.
Others drift a lot.

I bought two (cheap) 'atomic' (self setting) clocks that reset themselves daily from usa bureau of standards radio signal, but are still often 2-3 minutes apart. The builder used the reset to cheap out on the quartz oscillator not realising that those of us who buy such self syncing clocks want them to be perfect all the time not just once a day. But that might cost them a few pennies more.

OTOH if you can select the clock from a big bunch of them that are already running you can usually pick the one that is closest to the correct time and be sure it will be correct all the time as it had not drifted since it was manufactured months ago. I did that with a wrist watch I bought at WalMart and it keeps very accurate time.

I would hope that the makers of audio gear would try to put in a very accurate oscillator so their devices would not drift noticably.

Do let us know what results you had.
hobbyist
Regular
Posts: 94
Joined: Mon Jul 01, 2019 12:52 am

Re: Mixing Separate Multitrack and Stereo Recordings

Postby hobbyist » Sun Jul 07, 2019 1:07 am

Eddy Deegan wrote:On a related note, is there any preference to shrinking a longer take to fit, as opposed to stretching a shorter one or is the difference moot?

My logic was that stretching requires the interpolation of new data whereas shrinking doesn't but I may be misunderstanding the mechanics of the process.


It has been a long time since I had digital theory in grad school
(a VERY long time) but without going back to the books to check my intuition says the problem is equally difficult either way.

So I would take the ambience track and upsample it 4-8x if possible.
Then when you downsample while time shifting the impact is smaller.
hobbyist
Regular
Posts: 94
Joined: Mon Jul 01, 2019 12:52 am

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Hugh Robjohns » Sun Jul 07, 2019 1:22 am

Eddy Deegan wrote:My logic was that stretching requires the interpolation of new data whereas shrinking doesn't but I may be misunderstanding the mechanics of the process.

Interpolation is involved whichever direction you squeeze/stretch to calculate the correct waveform amplitude at the revised sample instants. Either way, there's both a speed change and a corrective pitch shift.
User avatar
Hugh Robjohns
Moderator
Posts: 23748
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Mixing Separate Multitrack and Stereo Recordings

Postby MOF » Sun Jul 07, 2019 1:22 am

I find it hard to believe that two digital recorders won’t be accurate to fractions of a second over the period of a concert to allow synchronisation in the edit. When DAT recorders came out we did a test at work where we copied a CD to it then played the two in sync’ the whole way through and they didn’t slip sync’ at all.
Even if I’m wrong you’re taking ambience to match with the clean direct signals, a few milliseconds difference would be like adding pre-delay to a reverb i.e. a desirable artefact. Also you tend not to have the audience faded up all the time, just for applause and at a reduced level during song announcements to give the sense of being in a room but not too much so that it reduces intelligibility.
MOF
Regular
Posts: 215
Joined: Thu Mar 06, 2003 1:00 am
Location: United Kingdom

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Eddy Deegan » Sun Jul 07, 2019 1:28 am

Hugh Robjohns wrote:
Eddy Deegan wrote:My logic was that stretching requires the interpolation of new data whereas shrinking doesn't but I may be misunderstanding the mechanics of the process.

Interpolation is involved whichever direction you squeeze/stretch to calculate the correct waveform amplitude at the revised sample instants. Either way, there's both a speed change and a corrective pitch shift.

Thank you Hugh. I was merely curious, but thinking about it that makes complete sense.
User avatar
Eddy Deegan
Frequent Poster (Level2)
Posts: 2198
Joined: Wed Sep 01, 2004 12:00 am
Location: Brighton & Hove, UK
Some of my musical works.
I had a weird time in Surrey once, but that was a drummer's fault.

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Dynamic Mike » Sun Jul 07, 2019 2:10 am

MOF wrote:Even if I’m wrong you’re taking ambience to match with the clean direct signals, a few milliseconds difference would be like adding pre-delay to a reverb i.e. a desirable artefact.

Wouldn't there be a natural delay between the close & room mic signals anyway caused by the placement?
Dynamic Mike
Frequent Poster (Level2)
Posts: 2937
Joined: Sun Dec 31, 2006 1:00 am
Always cut towards your mate 

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Hugh Robjohns » Sun Jul 07, 2019 2:14 am

When it comes to the synchronisation of two free-running machines, whether analogue or digital, sometimes you get lucky and sometimes you don't.

There are two factors to consider, again in both the analogue and digital worlds. One is the nominal long-term absolute speed (or sample rate in digi-speak), and the other is the short-term drift (or wow and flutter in the analogue world).

Digital recorders usually rely on a high frequency crystal as the clock source to set the sample rate, but not all crystals -- or clock generator circuitry -- are equally accurate.

The AES11 spec defines two levels of clock performance (and the IEC added a third lower grade for domestic equipment). The AES Grade 1 standard -- specified for reference master clocks -- requires a frequency accuracy of better than +/- 1 part per million (0.0001%). The Grade 2 standard allows a frequency accuracy of better than +/- 10 parts per million and is intended for portable devices.

Given that we have nearly 50,000 samples every second it doesn't take long to slip 10 samples per million... And while 10 samples is a very small difference it can build into something rather more significant over the duration of concert recording...

So, as I said, you could get lucky and have two devices with clocks that just happen to be very closely aligned for absolute sample rate... In which case life is very easy. And if they're not, stretching/squeezing in the DAW will compensate.

But then there is the second issue of Short term drift. Crystal-based clocks are normally very stable but, as I said, temperature and battery voltage can affect the clock rate in some cases. So, even if the start and end points are aligned to compensate for absolute speed variations, there could still be internal drift variations.

Hopefully not... But these are the engineering issues involved.

That different unsynchronised digital devices inherently run at slightly different speeds will have been made obvious if you've ever connected them to an interface or digital mixer and heard the clicks...

As to whether any speed variation is significant in the OP's situation will depend on how similar the room mic sound is to the close multitrack sound. If it is substantially reverberant it probably won't matter. But if it's very dry and similar it might...

H
User avatar
Hugh Robjohns
Moderator
Posts: 23748
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Mixing Separate Multitrack and Stereo Recordings

Postby Hugh Robjohns » Sun Jul 07, 2019 2:16 am

Dynamic Mike wrote:Wouldn't there be a natural delay between the close & room mic signals anyway caused by the placement?

:D Potentially not if it's edited out by syncing a hand clap! (although it depends where the clapper was located relative to the mics).
User avatar
Hugh Robjohns
Moderator
Posts: 23748
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Next

Who is online

Users browsing this forum: Dan LB, SeaOtter