Topic: is a Matrix w/ 2 JB3's possible?

[T.J.]

hey everyone,

i have a question regarding a matrix using 2 JB3 sources: is it possible using wavelab 5.0?  i have recorded one one DAUD source 483>V3>JB3#1 and another SBD>RCA>JB3#2. Taking these two sources and dumping them in wavelab using the montage option, is it possible to sync the two sources even though they don't run off the same word clock? i have already tracked one source out using cd wave. the other i left untouched. it seems impossible get them lined up. has anyone ever successfully done this? any tips?

thanks in advance...

[halleyscomet8]

this should do it. i've use this for every matrix i have done.

http://www.sloppyart.com/wavelab_matrix/

[Chanher]

This method can sound just as good as any other matrix, it is just more difficult and time consuming. I'm done with this method so I bought a busman R4.

Search or check the archives, there's a guide to doing this in Wavlab.

[Church-Audio]

hey everyone,

i have a question regarding a matrix using 2 JB3 sources: is it possible using wavelab 5.0?  i have recorded one one DAUD source 483>V3>JB3#1 and another SBD>RCA>JB3#2. Taking these two sources and dumping them in wavelab using the montage option, is it possible to sync the two sources even though they don't run off the same word clock? i have already tracked one source out using cd wave. the other i left untouched. it seems impossible get them lined up. has anyone ever successfully done this? any tips?

thanks in advance...

The only issue I could see is sample rates not being dead on between the two units. One of the methods I use for two recorders is using a 1k Burst about 1 second, and feeding it into both machines at the same time then I use the wave form viewer in wavlab to sync them up works like a charm if the sample rates are not the same then you have time issues but they should be solved if you sync the recording with a burst tone. In order to do this you must disconnect the inputs and hook up your burst tone and leave them in record and connect the inputs again so that you do not stop the record on the machines. This will give you a time frame. Because you can not possibly hit record at the same time on two machines its the only way to sync them up and be 100% accurate.

Chris Church


Chris Church

[T.J.]

this should do it. i've use this for every matrix i have done.

http://www.sloppyart.com/wavelab_matrix/

thanks guys. that is the link i used to initially educate myself on how to do a matrix and it is a huge help. however, i can't seem to sync the two sources correctly. i zoom in to line up the wav files and repeatidly(sp?) listen to hear them lined up, but can't get to sound perfect.

i am really leaning towards getting an R4, i just don't matrix that often and have to justify the need for four channels.

[Church-Audio]

this should do it. i've use this for every matrix i have done.

http://www.sloppyart.com/wavelab_matrix/

thanks guys. that is the link i used to initially educate myself on how to do a matrix and it is a huge help. however, i can't seem to sync the two sources correctly. i zoom in to line up the wav files and repeatidly(sp?) listen to hear them lined up, but can't get to sound perfect.

i am really leaning towards getting an R4, i just don't matrix that often and have to justify the need for four channels.

Read what I said above there is a way to do it. If you dont have a tone burst unit very cheap. find a tone burst file on the internet and use a simple cd player with a Y cable going into both recorders at the same time then you can sync them up.

Chris Church

[T.J.]

this should do it. i've use this for every matrix i have done.

http://www.sloppyart.com/wavelab_matrix/

thanks guys. that is the link i used to initially educate myself on how to do a matrix and it is a huge help. however, i can't seem to sync the two sources correctly. i zoom in to line up the wav files and repeatidly(sp?) listen to hear them lined up, but can't get to sound perfect.

i am really leaning towards getting an R4, i just don't matrix that often and have to justify the need for four channels.

Read what I said above there is a way to do it. If you dont have a tone burst unit very cheap. find a tone burst file on the internet and use a simple cd player with a Y cable going into both recorders at the same time then you can sync them up.

Chris Church

i hear what you're saying chris, but just fail to see how this can help me with the project i have already started. +T for the help

[easy jim]

The problem you are having is drift.  Drift occurs between the two sources as a result of the difference between the two clocks.  While it should be very close coming from two JB3s, it would not be unusual to have as much as 50-100 milisecs of drift over a long show.  To correct drift, you need to time-scale one of the sources to the other, which will serve as your master clock.  (Time-scaling can also be referred to as changing 'pitch'.)

I find chopping up one source and pasting it in aligned to the beginning of each track on the other to be unduly cumbersome.  And, it is not an accurate way to make a matrix because you will still have drift between the sources which may be noticeable by the end of each track, especially if it is a long one.  So, you may end up with things perfectly aligned at the beginning of a track, but it will sound echo-ey or chorus-ey by the end.  It is much better to put the two sources next to each other in your edit window and then time-scale one to the other.

To do so, find a loud transient sound (snare hit, mic pop, etc.) as close to the beginning as possible and line them up.  Then, go to the end and find a similar loud transient as close to the end as possible and see how far off the two sources are on your time ruler.  Calculate the amount of drift and use whatever time-scaling function is available in Wavelab to either stretch the shorter source or shrink the longer source to get them to line up over the entire show.  In some software, you grab then end of one source and drag it using the time ruler in the edit window to time-scale by the proper amount (this is how AudioDesk & Digital Performer do it); in other software, you calculate the percentage of the difference between the source in original form and the properly time-scaled outcome you desire:  i.e., shrinking a source by 99.999975 % or something like that to make the longer source match with the shorter one.  Generally, your editing software will create a new file/track for the time-scaled version and it will take some processing time for it to complete before you may then get your mix together to go from 4 -> 2 tracks.

I do not know Wavelab and whether or not it has a time-scale function, but I would be surprised if it did not.  Once you mix two sources from different clocks a few times, correcting drift gets a lot easier and should become something you may consider a relatively easy step to make a matrix from two separate sources.  Now, I tend to agonize more over setting the levels of the two sources and whether or not to spot mute any artifacts out of the mix if they are bad and only present in one source.

Maybe someone who knows about time-scaling in Wavelab will chime in.  Feel free to PM me if you get stuck.

By the way, you should use the V3 source as the 'master clock' assuming you patched in the JB3 digitally.  The V3's clock will be far superior and more accurate than the internal clock in the JB3.

[Church-Audio]

The problem you are having is drift.  Drift occurs between the two sources as a result of the difference between the two clocks.  While it should be very close coming from two JB3s, it would not be unusual to have as much as 50-100 milisecs of drift over a long show.  To correct drift, you need to time-scale one of the sources to the other, which will serve as your master clock.  (Time-scaling can also be referred to as changing 'pitch'.)

I find chopping up one source and pasting it in aligned to the beginning of each track on the other to be unduly cumbersome.  And, it is not an accurate way to make a matrix because you will still have drift between the sources which may be noticeable by the end of each track, especially if it is a long one.  So, you may end up with things perfectly aligned at the beginning of a track, but it will sound echo-ey or chorus-ey by the end.  It is much better to put the two sources next to each other in your edit window and then time-scale one to the other.

To do so, find a loud transient sound (snare hit, mic pop, etc.) as close to the beginning as possible and line them up.  Then, go to the end and find a similar loud transient as close to the end as possible and see how far off the two sources are on your time ruler.  Calculate the amount of drift and use whatever time-scaling function is available in Wavelab to either stretch the shorter source or shrink the longer source to get them to line up over the entire show.  In some software, you grab then end of one source and drag it using the time ruler in the edit window to time-scale by the proper amount (this is how AudioDesk & Digital Performer do it); in other software, you calculate the percentage of the difference between the source in original form and the properly time-scaled outcome you desire:  i.e., shrinking a source by 99.999975 % or something like that to make the longer source match with the shorter one.  Generally, your editing software will create a new file/track for the time-scaled version and it will take some processing time for it to complete before you may then get your mix together to go from 4 -> 2 tracks.

I do not know Wavelab and whether or not it has a time-scale function, but I would be surprised if it did not.  Once you mix two sources from different clocks a few times, correcting drift gets a lot easier and should become something you may consider a relatively easy step to make a matrix from two separate sources.  Now, I tend to agonize more over setting the levels of the two sources and whether or not to spot mute any artifacts out of the mix if they are bad and only present in one source.

Maybe someone who knows about time-scaling in Wavelab will chime in.  Feel free to PM me if you get stuck.

By the way, you should use the V3 source as the 'master clock' assuming you patched in the JB3 digitally.  The V3's clock will be far superior and more accurate than the internal clock in the JB3.

The best way to deal with already recorded tracks Is to use a program like Nuendo or Qbase. And put two sets of stereo tracks into two stereo interleave tracks and then use your mouse to grab the wave and line it up with the other stereo track. and open your view until you get a really good lineup. Then once it looks good it will be close. Then hit play and listen, then slide one of the waves left then listen to the improvement if its worse go to your right and then you can figure out if its late or early thats all you need to do. Then send me $50.00 US in an envelope for a "thank you gift"

Chris Church

[poorlyconditioned]

The problem you are having is drift.  Drift occurs between the two sources as a result of the difference between the two clocks.  While it should be very close coming from two JB3s, it would not be unusual to have as much as 50-100 milisecs of drift over a long show.  To correct drift, you need to time-scale one of the sources to the other, which will serve as your master clock.  (Time-scaling can also be referred to as changing 'pitch'.)

I find chopping up one source and pasting it in aligned to the beginning of each track on the other to be unduly cumbersome.  And, it is not an accurate way to make a matrix because you will still have drift between the sources which may be noticeable by the end of each track, especially if it is a long one.  So, you may end up with things perfectly aligned at the beginning of a track, but it will sound echo-ey or chorus-ey by the end.  It is much better to put the two sources next to each other in your edit window and then time-scale one to the other.

To do so, find a loud transient sound (snare hit, mic pop, etc.) as close to the beginning as possible and line them up.  Then, go to the end and find a similar loud transient as close to the end as possible and see how far off the two sources are on your time ruler.  Calculate the amount of drift and use whatever time-scaling function is available in Wavelab to either stretch the shorter source or shrink the longer source to get them to line up over the entire show.  In some software, you grab then end of one source and drag it using the time ruler in the edit window to time-scale by the proper amount (this is how AudioDesk & Digital Performer do it); in other software, you calculate the percentage of the difference between the source in original form and the properly time-scaled outcome you desire:  i.e., shrinking a source by 99.999975 % or something like that to make the longer source match with the shorter one.  Generally, your editing software will create a new file/track for the time-scaled version and it will take some processing time for it to complete before you may then get your mix together to go from 4 -> 2 tracks.

I do not know Wavelab and whether or not it has a time-scale function, but I would be surprised if it did not.  Once you mix two sources from different clocks a few times, correcting drift gets a lot easier and should become something you may consider a relatively easy step to make a matrix from two separate sources.  Now, I tend to agonize more over setting the levels of the two sources and whether or not to spot mute any artifacts out of the mix if they are bad and only present in one source.

Maybe someone who knows about time-scaling in Wavelab will chime in.  Feel free to PM me if you get stuck.

By the way, you should use the V3 source as the 'master clock' assuming you patched in the JB3 digitally.  The V3's clock will be far superior and more accurate than the internal clock in the JB3.

The best way to deal with already recorded tracks Is to use a program like Nuendo or Qbase. And put two sets of stereo tracks into two stereo interleave tracks and then use your mouse to grab the wave and line it up with the other stereo track. and open your view until you get a really good lineup. Then once it looks good it will be close. Then hit play and listen, then slide one of the waves left then listen to the improvement if its worse go to your right and then you can figure out if its late or early thats all you need to do. Then send me $50.00 US in an envelope for a "thank you gift"

Chris Church

(+T is all I can afford right now.)  I've tried this (Cubase).  The problem is the clocks move a little bit with time.  The human ear is very good at hearing *changes* in alignment.  You can be out 10 or 15ms as long as the clocks are matched.  But if there are two different clocks, it can be difficult to get great sound throughout.  Hence the need to split a show into several pieces and move each one.

  Richard

[Church-Audio]

The problem you are having is drift.  Drift occurs between the two sources as a result of the difference between the two clocks.  While it should be very close coming from two JB3s, it would not be unusual to have as much as 50-100 milisecs of drift over a long show.  To correct drift, you need to time-scale one of the sources to the other, which will serve as your master clock.  (Time-scaling can also be referred to as changing 'pitch'.)

I find chopping up one source and pasting it in aligned to the beginning of each track on the other to be unduly cumbersome.  And, it is not an accurate way to make a matrix because you will still have drift between the sources which may be noticeable by the end of each track, especially if it is a long one.  So, you may end up with things perfectly aligned at the beginning of a track, but it will sound echo-ey or chorus-ey by the end.  It is much better to put the two sources next to each other in your edit window and then time-scale one to the other.

To do so, find a loud transient sound (snare hit, mic pop, etc.) as close to the beginning as possible and line them up.  Then, go to the end and find a similar loud transient as close to the end as possible and see how far off the two sources are on your time ruler.  Calculate the amount of drift and use whatever time-scaling function is available in Wavelab to either stretch the shorter source or shrink the longer source to get them to line up over the entire show.  In some software, you grab then end of one source and drag it using the time ruler in the edit window to time-scale by the proper amount (this is how AudioDesk & Digital Performer do it); in other software, you calculate the percentage of the difference between the source in original form and the properly time-scaled outcome you desire:  i.e., shrinking a source by 99.999975 % or something like that to make the longer source match with the shorter one.  Generally, your editing software will create a new file/track for the time-scaled version and it will take some processing time for it to complete before you may then get your mix together to go from 4 -> 2 tracks.

I do not know Wavelab and whether or not it has a time-scale function, but I would be surprised if it did not.  Once you mix two sources from different clocks a few times, correcting drift gets a lot easier and should become something you may consider a relatively easy step to make a matrix from two separate sources.  Now, I tend to agonize more over setting the levels of the two sources and whether or not to spot mute any artifacts out of the mix if they are bad and only present in one source.

Maybe someone who knows about time-scaling in Wavelab will chime in.  Feel free to PM me if you get stuck.

By the way, you should use the V3 source as the 'master clock' assuming you patched in the JB3 digitally.  The V3's clock will be far superior and more accurate than the internal clock in the JB3.

The best way to deal with already recorded tracks Is to use a program like Nuendo or Qbase. And put two sets of stereo tracks into two stereo interleave tracks and then use your mouse to grab the wave and line it up with the other stereo track. and open your view until you get a really good lineup. Then once it looks good it will be close. Then hit play and listen, then slide one of the waves left then listen to the improvement if its worse go to your right and then you can figure out if its late or early thats all you need to do. Then send me $50.00 US in an envelope for a "thank you gift"

Chris Church

(+T is all I can afford right now.)  I've tried this (Cubase).  The problem is the clocks move a little bit with time.  The human ear is very good at hearing *changes* in alignment.  You can be out 10 or 15ms as long as the clocks are matched.  But if there are two different clocks, it can be difficult to get great sound throughout.  Hence the need to split a show into several pieces and move each one.

  Richard

Richard the clocks do not jump around its sold for each recorder. If they are jumping around the pitch would be changing since its just flanging between tracks its the difference between two recorders clocks that is the problem one could be 44.100 the other could be 44.155 this difference creates a problem called flanging between the two recorders trust me your clock is not moving on the the recorders if it was there would be wow and flutter something digital has virtually eliminated. The clocks must be solid on a digital recorder! if not your pitch will change and the recorder would be a peace of shit.

Chris Church

[T.J.]

good advise by all. a round of T's is all i can afford right now too 
i'm going to play around with with the pitch in wavelab, amongst other things. there is no "deadline" in which this needs to be done.

By the way, you should use the V3 source as the 'master clock' assuming you patched in the JB3 digitally.  The V3's clock will be far superior and more accurate than the internal clock in the JB3.

yes i did use the opti out on the V3 to record. i'm assuming when you refer to the "master" clock you mean i should use the V3 source as the "base" source. i should try to strech the other source to match the V3 source, correct?

[Gordon]

I have done a number of matrix's like this.

mics > v3 > jb3

& sbd > rca > jb3

break the sbd into tracks and leave the aud as one long wave file.  then open the aud in the monatge.  add sbd track one, do it track by track.  I have never had any problems with drift.  it takes some time and practice to line them up but it's not that hard.

[T.J.]

I have done a number of matrix's like this.

mics > v3 > jb3

& sbd > rca > jb3

break the sbd into tracks and leave the aud as one long wave file.  then open the aud in the monatge.  add sbd track one, do it track by track.  I have never had any problems with drift.  it takes some time and practice to line them up but it's not that hard.

exactly what i have and how i tried. to me, it's a pain in the a$$. i must not have the patience for lining the sources up in post.

[easy jim]

yes i did use the opti out on the V3 to record. i'm assuming when you refer to the "master" clock you mean i should use the V3 source as the "base" source. i should try to strech the other source to match the V3 source, correct?

Yep, that is how I would do it.

break the sbd into tracks and leave the aud as one long wave file.  then open the aud in the monatge.  add sbd track one, do it track by track.  I have never had any problems with drift.  it takes some time and practice to line them up but it's not that hard.

exactly what i have and how i tried. to me, it's a pain in the a$$. i must not have the patience for lining the sources up in post.

I have found that drift is still a problem when you chop up one source and align every track.  I think you will find, as well, that with this approach you will have brief gaps of a few milisecs +/- between each segment of the chopped up source when it is lined up with the continuous source.  I can often hear the added delay from drift by the end of a 10 -15 min track if the clocks differ enough.  It is a major pain to line them up so many times taking the chop-up-one-source approach, so I try and only do it once at the very beginning and then time-scale one source to the other.  If the time-scaling is done accurately, your two sources would then be as if recorded both slaved to the same clock.

The problem is the clocks move a little bit with time.  The human ear is very good at hearing *changes* in alignment.  You can be out 10 or 15ms as long as the clocks are matched.  But if there are two different clocks, it can be difficult to get great sound throughout.  Hence the need to split a show into several pieces and move each one.

Richard the clocks do not jump around its sold for each recorder. If they are jumping around the pitch would be changing since its just flanging between tracks its the difference between two recorders clocks that is the problem one could be 44.100 the other could be 44.155 this difference creates a problem called flanging between the two recorders trust me your clock is not moving on the the recorders if it was there would be wow and flutter something digital has virtually eliminated. The clocks must be solid on a digital recorder! if not your pitch will change and the recorder would be a peace of shit.

I think we're referring to a few different things here.  As for the clocks, which I've learned a lot more about as I have been multitracking, and also often grabbing additional sources from different recorders:  digital clocks will be very internally consistent in their measurement of time in general whether they are really high end (apogee, for instance) or lower end.  However, two different clocks, even two of the exact same clocks by the exact same manufacturer, will not maintain a consistency in reference to each other as they register the passage of time.  This is why drift occurs, and it is why master clocking devices like the Apogee Big Ben, etc. are necessary when using multiple digital front ends for a multitrack.  Master clocks like the Big Ben allow you to sync/slave all the digital sources to whichever device is assigned as the 'master.'  Otherwise, you get jitter (clocking errors) due to the slight inconsistency between each devices' internal clocks' measurement of the passge of time.  The longer the recording, the more apparent the drift between two sources from different clocks.

[BayTaynt3d]

I do this all the time, and I wouldn't bother with chopping it up first, what a hassle. It's already been said, but the general idea is this. First you need a decent editor (not the montage crap). Then sync the beginning up so it looks/sounds right. Then go to the end of the timeline and STRETCH one source (or smoosh it) until the ending looks/sounds sync'ed. At that point the whole show should be synced. Basically, you get the beginning synced, then you fix the drift by uniformly stretching/compressing (not compression, but making shorter) one of the sources. It's a little time-consuming at first, and you have to know your editor, but with practice, it's easy.

[Roving Sign]

Then go to the end of the timeline and STRETCH one source (or smoosh it) until the ending looks/sounds sync'ed. At that point the whole show should be synced.

That's cool from a geometric point of view...(and maybe that's all that's really important.) But smooshing and stretching - - how do those processes affect the music? Does anyone know what - technically - those processes are achieving?

And to be clear - Word Clocks control Word Length (I presume thats the same as bit depth) and Sampling Rate? Or is sampling rate set by another gadget?

[Roving Sign]

Is there a way to determine the precise sampling rate of a wav? Like out to a few decimals...? EXACTLY...

not just 44.1?

[Church-Audio]

yes i did use the opti out on the V3 to record. i'm assuming when you refer to the "master" clock you mean i should use the V3 source as the "base" source. i should try to strech the other source to match the V3 source, correct?

Yep, that is how I would do it.

break the sbd into tracks and leave the aud as one long wave file.  then open the aud in the monatge.  add sbd track one, do it track by track.  I have never had any problems with drift.  it takes some time and practice to line them up but it's not that hard.

exactly what i have and how i tried. to me, it's a pain in the a$$. i must not have the patience for lining the sources up in post.

I have found that drift is still a problem when you chop up one source and align every track.  I think you will find, as well, that with this approach you will have brief gaps of a few milisecs +/- between each segment of the chopped up source when it is lined up with the continuous source.  I can often hear the added delay from drift by the end of a 10 -15 min track if the clocks differ enough.  It is a major pain to line them up so many times taking the chop-up-one-source approach, so I try and only do it once at the very beginning and then time-scale one source to the other.  If the time-scaling is done accurately, your two sources would then be as if recorded both slaved to the same clock.

The problem is the clocks move a little bit with time.  The human ear is very good at hearing *changes* in alignment.  You can be out 10 or 15ms as long as the clocks are matched.  But if there are two different clocks, it can be difficult to get great sound throughout.  Hence the need to split a show into several pieces and move each one.

Richard the clocks do not jump around its sold for each recorder. If they are jumping around the pitch would be changing since its just flanging between tracks its the difference between two recorders clocks that is the problem one could be 44.100 the other could be 44.155 this difference creates a problem called flanging between the two recorders trust me your clock is not moving on the the recorders if it was there would be wow and flutter something digital has virtually eliminated. The clocks must be solid on a digital recorder! if not your pitch will change and the recorder would be a peace of shit.

I think we're referring to a few different things here.  As for the clocks, which I've learned a lot more about as I have been multitracking, and also often grabbing additional sources from different recorders:  digital clocks will be very internally consistent in their measurement of time in general whether they are really high end (apogee, for instance) or lower end.  However, two different clocks, even two of the exact same clocks by the exact same manufacturer, will not maintain a consistency in reference to each other as they register the passage of time.  This is why drift occurs, and it is why master clocking devices like the Apogee Big Ben, etc. are necessary when using multiple digital front ends for a multitrack.  Master clocks like the Big Ben allow you to sync/slave all the digital sources to whichever device is assigned as the 'master.'  Otherwise, you get jitter (clocking errors) due to the slight inconsistency between each devices' internal clocks' measurement of the passge of time.  The longer the recording, the more apparent the drift between two sources from different clocks.

Unless your using a Fred flintstone A/D most clocks get generated via a crystal but they are then fed into a microprocessor and then they are time corrected and they are balls on accurate to it self THEY HAVE TO BE if not you change the pitch of the track This would not be acceptable This is why when you do see wow and flutter specs they are listed as two low to measure.

The days of clocks that are not consistent in a device are over. This does not mean you don't have problems with two different devices clocks not being synced. If you have a problem in two different tracks from the same machine its because you have done different processing to the tracks in question and increased the latency of the track but even this can be corrected by increasing the other tracks latency to compensate.

So if you were to take a set of stereo tracks and not sync them to each other first and then process them together in the same time domain you could introduce latency but that is not the clock on the recorder that is jumping. There has been much confusion about this subject. When your doing multitracking your always increasing latency of tracks because of processing that is done to each individual track most software will compensate for this even with out you being aware of it. This is how things stay in pitch with each other if it did not work this way a guitar tuned to 440hz could end up being 440.33 instead of 440.00hz

[BayTaynt3d]

Then go to the end of the timeline and STRETCH one source (or smoosh it) until the ending looks/sounds sync'ed. At that point the whole show should be synced.

That's cool from a geometric point of view...(and maybe that's all that's really important.) But smooshing and stretching - - how do those processes affect the music? Does anyone know what - technically - those processes are achieving?

And to be clear - Word Clocks control Word Length (I presume thats the same as bit depth) and Sampling Rate? Or is sampling rate set by another gadget?

Sure they can affect the music, but we're talking about spreading a very small amount of time across hours of music, and most NLEs pitch-correct while they do it -- Unless your drift is like minutes or something (which it won't be), it's inaudible. More importantly, that doesn't really matter, because if you want them to be in sync you have NO OTHER CHOICE. The exception to that statement might be splitting individually and sync'ing each track individually -- the difference there is you are just substituting getting a minor delay by the end of each song. They each have there downsides. That said, I've NEVER noticed a problem with this approach, and I've done it a lot in the past year (many times taking 2 channels from a JB3 or H120 and syncing with four channels from my R4), but technically, yeah, you're messing with it. But I've never had a source more than a second or so off, so divide that into two hours of a set, and the stretching is absurdley minimal. Compare the workflow of syncing the whole thing in one go with syncing each one individually, and then combine that with perfect sync everywhere vs. drift creating a delay in longer songs (subtle, but it might be audible, and I dare you to find this method to be audible). Anyway, just my two cents...

[Church-Audio]

Then go to the end of the timeline and STRETCH one source (or smoosh it) until the ending looks/sounds sync'ed. At that point the whole show should be synced.

That's cool from a geometric point of view...(and maybe that's all that's really important.) But smooshing and stretching - - how do those processes affect the music? Does anyone know what - technically - those processes are achieving?

And to be clear - Word Clocks control Word Length (I presume thats the same as bit depth) and Sampling Rate? Or is sampling rate set by another gadget?

Sure they can affect the music, but we're talking about spreading a very small amount of time across hours of music, and most NLEs pitch-correct while they do it -- Unless your drift is like minutes or something (which it won't be), it's inaudible. More importantly, that doesn't really matter, because if you want them to be in sync you have NO OTHER CHOICE. The exception to that statement might be splitting individually and sync'ing each track individually -- the difference there is you are just substituting getting a minor delay by the end of each song. They each have there downsides. That said, I've NEVER noticed a problem with this approach, and I've done it a lot in the past year (many times taking 2 channels from a JB3 or H120 and syncing with four channels from my R4), but technically, yeah, you're messing with it. But I've never had a source more than a second or so off, so divide that into two hours of a set, and the stretching is absurdley minimal. Compare the workflow of syncing the whole thing in one go with syncing each one individually, and then combine that with perfect sync everywhere vs. drift creating a delay in longer songs (subtle, but it might be audible, and I dare you to find this method to be audible). Anyway, just my two cents...

Thank you Anyone that says they can hear sample shift in a digital recorder has bat ears. It's a non existent problem now a days. Between two different machines it's possible this is why we have clock master out and clock inputs on professional gear. When doing a matrix or bringing in audio from two different machines its not the individual machines them selves that are the problem. Its the fact the clocks on the two machines are different.

[easy jim]

Sure they can affect the music, but we're talking about spreading a very small amount of time across hours of music, and most NLEs pitch-correct while they do it -- Unless your drift is like minutes or something (which it won't be), it's inaudible. More importantly, that doesn't really matter, because if you want them to be in sync you have NO OTHER CHOICE. The exception to that statement might be splitting individually and sync'ing each track individually -- the difference there is you are just substituting getting a minor delay by the end of each song. They each have there downsides. That said, I've NEVER noticed a problem with this approach, and I've done it a lot in the past year (many times taking 2 channels from a JB3 or H120 and syncing with four channels from my R4), but technically, yeah, you're messing with it. But I've never had a source more than a second or so off, so divide that into two hours of a set, and the stretching is absurdley minimal. Compare the workflow of syncing the whole thing in one go with syncing each one individually, and then combine that with perfect sync everywhere vs. drift creating a delay in longer songs (subtle, but it might be audible, and I dare you to find this method to be audible). Anyway, just my two cents...

I second this completely.

Thank you Anyone that says they can hear sample shift in a digital recorder has bat ears. It's a non existent problem now a days. Between two different machines it's possible this is why we have clock master out and clock inputs on professional gear. When doing a matrix or bringing in audio from two different machines its not the individual machines them selves that are the problem. Its the fact the clocks on the two machines are different.

And am in complete agreement here too.

[Javier Cinakowski]

I have enjoyed this discussion.  Thanks!

[SparkE!]

Most common crystals are spec'd at 100 ppm accuracy.  The result is that if you get two clocks, both at the worst case error, each in a different direction, you can get as much as 720 ms (that's almost 3/4 second) of skew between them in an hour.  In practice, you'll never find two crystals that are that far off from each other.  Typically, you'll see less than 20 ppm error between two of the same model of crystal.  You can buy 10 ppm crystals, but it's not common for those to be used in audio equipment.  Most high spec crystals are used in radio communication equipment.

Anyway, I'm lucky to have two UA-5's that produce timing skews, relative to each other, of less than about 2 or 3 ms per hour, so I can just drop one waveform on top of each other.  Back when I used my JB3's alone, I would get more like 15 ms per hour between them, which meant that I had to align them track by track.

Now here's where I was really trying to go with this:  When you line up two sources and haven't used a tone generator like Chris was suggesting, look for drums by themselves between tracks in the middle of the show.  It's pretty easy to line those up.  If you have to resort to lining up track by track instead of a whole show, then you might want to consider lining the tracks up in the middle of the track.  That way, your tracks end up with less average skew than if you lined them up at the first of the track and let them drift apart.  When you line them up in the middle of the track, they drift together, then drift back apart.

[poorlyconditioned]

Most common crystals are spec'd at 100 ppm accuracy.  The result is that if you get two clocks, both at the worst case error, each in a different direction, you can get as much as 720 ms (that's almost 3/4 second) of skew between them in an hour.  In practice, you'll never find two crystals that are that far off from each other.  Typically, you'll see less than 20 ppm error between two of the same model of crystal.  You can buy 10 ppm crystals, but it's not common for those to be used in audio equipment.  Most high spec crystals are used in radio communication equipment.

Anyway, I'm lucky to have two UA-5's that produce timing skews, relative to each other, of less than about 2 or 3 ms per hour, so I can just drop one waveform on top of each other.  Back when I used my JB3's alone, I would get more like 15 ms per hour between them, which meant that I had to align them track by track.

Now here's where I was really trying to go with this:  When you line up two sources and haven't used a tone generator like Chris was suggesting, look for drums by themselves between tracks in the middle of the show.  It's pretty easy to line those up.  If you have to resort to lining up track by track instead of a whole show, then you might want to consider lining the tracks up in the middle of the track.  That way, your tracks end up with less average skew than if you lined them up at the first of the track and let them drift apart.  When you line them up in the middle of the track, they drift together, then drift back apart.

This is exactly my experience.  Song by song is *usually* fine.  But I've noticed that if there is a small amount of drift, then I will hear slight changes (in particular, volume changes of vocals) from one part of the song to another.  Remember, a typcial voice may be between 1 and 2kHz (is this right?).  Well, that is under 1ms.  And it is very easy for sources to be out this much.  When soundboard and mics are out by even 0.5ms they can interfere contructively or destructively to change the vocal volume by a fair amount.

I have not tried time scaling yet.  But I've got an R4 now, so this is no longer a problem for me .

  Richard

[Church-Audio]

Most common crystals are spec'd at 100 ppm accuracy.  The result is that if you get two clocks, both at the worst case error, each in a different direction, you can get as much as 720 ms (that's almost 3/4 second) of skew between them in an hour.  In practice, you'll never find two crystals that are that far off from each other.  Typically, you'll see less than 20 ppm error between two of the same model of crystal.  You can buy 10 ppm crystals, but it's not common for those to be used in audio equipment.  Most high spec crystals are used in radio communication equipment.

Anyway, I'm lucky to have two UA-5's that produce timing skews, relative to each other, of less than about 2 or 3 ms per hour, so I can just drop one waveform on top of each other.  Back when I used my JB3's alone, I would get more like 15 ms per hour between them, which meant that I had to align them track by track.

Now here's where I was really trying to go with this:  When you line up two sources and haven't used a tone generator like Chris was suggesting, look for drums by themselves between tracks in the middle of the show.  It's pretty easy to line those up.  If you have to resort to lining up track by track instead of a whole show, then you might want to consider lining the tracks up in the middle of the track.  That way, your tracks end up with less average skew than if you lined them up at the first of the track and let them drift apart.  When you line them up in the middle of the track, they drift together, then drift back apart.

This is exactly my experience.  Song by song is *usually* fine.  But I've noticed that if there is a small amount of drift, then I will hear slight changes (in particular, volume changes of vocals) from one part of the song to another.  Remember, a typcial voice may be between 1 and 2kHz (is this right?).  Well, that is under 1ms.  And it is very easy for sources to be out this much.  When soundboard and mics are out by even 0.5ms they can interfere contructively or destructively to change the vocal volume by a fair amount.

I have not tried time scaling yet.  But I've got an R4 now, so this is no longer a problem for me .

  Richard

Richard I disagree how do you know this is because of drift? I have spent 1000's of hours on editing tracks on a computer and never heard a recorder drift by it self and how do you know these changes in vocals are not from "DYNAMICS" of the performance???

[BobW]

Richard I disagree how do you know this is because of drift? I have spent 1000's of hours on editing tracks on a computer and never heard a recorder drift by it self and how do you know these changes in vocals are not from "DYNAMICS" of the performance???

I'm unsure if at audio level the crystal clocks' drift are significant, but in telecom and radio we calibrate and use master clocks religiously. Heat affects the crystal frequency significantly.
We used to put the crystals in "ovens" called OCXOs, aka small thermal-controlled enviroments inside of radios and calibrators.

Also, isn't phase a "subset" of pitch. That is, can't phase be viewed as a fractional pitch change?
So, then, is sync more important than just for timing and pitch??

[F.O.Bean]

FWIW, you can 'time-stretch' in WL's montage but i am unsure exactly how to do it

I usually chop up the wavs at a few places on the EXACT same spot and do this about every 30-40 minutes to avoid too big of a drift

works out well for the most part, but is TIME CONSUMING and EASILY take a day or two for me to get just the right mix of the two down

[Church-Audio]

Richard I disagree how do you know this is because of drift? I have spent 1000's of hours on editing tracks on a computer and never heard a recorder drift by it self and how do you know these changes in vocals are not from "DYNAMICS" of the performance???

I'm unsure if at audio level the crystal clocks' drift are significant, but in telecom and radio we calibrate and use master clocks religiously. Heat affects the crystal frequency significantly.
We used to put the crystals in "ovens" called OCXOs, aka small thermal-controlled enviroments inside of radios and calibrators.

Also, isn't phase a "subset" of pitch. That is, can't phase be viewed as a fractional pitch change?
So, then, is sync more important than just for timing and pitch??

My whole argument is, yes we still use crystals for sampling rate clocks. Just like we use them for alarm clocks, but the major difference is after the crystal generates the frequency its feed into a microprocessor and time corrected. Even a very simple recorder works this way. So you might get drift but it’s not over a period of hours it would be days! Think of how long it takes your alarm clock to start to lose time, or your crystal controlled wrist watch and you will get my point. This rule applies to 99% of the digital recorders made after 1995. There is still an issue between recorders because of the lack of standard pitch calibration at the factory. This is because of cheap manufacturing, and a lack of checking the crystals for exact frequency operation the variance between parts can be as much as 10-15% thus the need for the microprocessor in the first place. So that is where your variance comes from. Now temperature does effect the operation of a crystal. This is why we have temperature controlled crystals. They are put into miniature ovens on the circuit board to control temperature, on higher end digital gear like a very high end console.

So the solution around the oven was the microprocessor. This MP sample’s the frequency of the crystal and then corrects it in real time and then spits it out to the A-D converter. I have never used gear that drifted by it self. If I had a recorder that did that I would trash it, because how can you capture a performance when the PITCH OF THE SHOW IS CHANGING??? How would that be a good thing? Level is not affected by sampling rates, it’s affected by amplitude. This is a function of an analog circuit you can change the sampling rate all you want your level does not change. Your bandwidth does. meaning at 44.1 you can expect 20hz to 20k, at 48khz you can expect 20hz to 30k, at 96khz you can expect 15hz to 47k, and 192 you can expect 5hz to 96khz. But the level of frequencies is not changed by discrepancies in the frequency of the sampling clock. Now when you combine two different recorders into the mix you can get comb filtering effects due to the change in pitch, and the frequency amplitude due to comb filtering. This is what happens is two frequency’s are just slightly off in arrival time but at the same amplitude and frequency * like when you do a matrix* this can cause smearing of the time arrival of the frequency’s thus may cause a hump or dip in the frequency response. But only when two or more stereo tracks are combined from two different recorders with unlike clocks or different arrival points that remain uncorrected.

Chris Church

[cyfan]

Many of my recordings are post-matrixes of the following sources:

Soundboard > JB3
Mics > Edirol UA-5 > JB3

I use Cool Edit Pro's multi-track mode to synch them.
Line up the start of the recordings so they are synched at the beginning.
Find a loud snare pop or something near the end and find out the time difference, then divide by the # of minutes. (I've found in my case that the soundboard recording is always .004 second per minute longer than the mic recording, so I don't even check anymore).

If through the mixing process I find the soundboard will be the dominate source in the mix, I go through and splice a .004 second silent portion in at every minute interval. (this takes just a few minutes using copy/paste functions.

If I determine the mic recording is the dominant source in the matrix, I go through the sbd track and remove .004 seconds every minute (which is a bit more time consuming).

In the past, I tried to compress or expand one source to match but found it introduced some weird digital noises in the recording.

In any case, once I have both files transferred to the PC, I'm able to synch a 2 hour show in less than 30 minutes.

One thing to keep in mind, if the mic source is boomy for whatever reason, and you couple it with a very crisp soundboard mix, the end result may still sound like it has some drift even if it doesn't.

[T.J.]

all great advise guys. thanks a bunch!

i have one recording from last week which is only about 45 minutes (about 5 songs with some noticable spikes from the bass drum). i think this is a good opportunity to play around and find out what works best for me. i'll let you know how i make out.

+T's

[Gordon]

I guess I have been lucky with my matrix's.  I've never had to do any time stretch etc and have never noticed any drift etc.

[Church-Audio]

all great advise guys. thanks a bunch!

i have one recording from last week which is only about 45 minutes (about 5 songs with some noticable spikes from the bass drum). i think this is a good opportunity to play around and find out what works best for me. i'll let you know how i make out.

+T's

Your better off with a snare hit then a bass drum ( a bass drum has a much longer sustain then a snare shot. A snare shot is a very quick event with a sharp spike very easy to identify and line up. Also it should be mentioned one of the best ways is to zoom right into the peek as much as you can so that its easy to look at.

Chris Church

[T.J.]

that makes sense. there is also some tuning in between songs which should help.

[cyfan]

Your better off with a snare hit then a bass drum ( a bass drum has a much longer sustain then a snare shot. A snare shot is a very quick event with a sharp spike very easy to identify and line up.

I've found the absolute best peak to find at the end of the set is the loud pop when some guitar player pulls his cable out before the sound guy has muted the channel 

tim

[Church-Audio]

Your better off with a snare hit then a bass drum ( a bass drum has a much longer sustain then a snare shot. A snare shot is a very quick event with a sharp spike very easy to identify and line up.

I've found the absolute best peak to find at the end of the set is the loud pop when some guitar player pulls his cable out before the sound guy has muted the channel 

tim

Nope thats no good, because the scream of the sound engineer or system tech has to long of a decay