Synchronization of Audio and Video from Different Sources

Shawn Redford · April 15th, 2006, 02:38 AM

I was just reading this very LONG online article and it seems like this is a possible explanation for what you're facing: http://www.kenstone.net/fcp_homepage...ion_sound.html - If you scroll down about 3/4 of the page to the "Dual System Sound" heading, you will find these paragraphs after that heading:

Quote:

For the past few years, there has been much discussion about using Mini Disc, DAT or hard disk recorders for dual system sound recording when shooting DV. In theory, it's really easy. Sure, your VX-2000 may record horrible, unusable audio with a lot of hiss. Just bring along an MD recorder, pop in a disk and life is good, right? Well, it's too bad that in reality, it's just not that simple.

First of all, from a mechanical standpoint, even though both of the devices (camcorder and MD, DAT or hard disc recorder) are both digital, they are also both completely independently synchronized to their own internal master timing device. You would think that if both devices are digital and both are using the same sample rate that it would be a simple matter of lining up the picture and the MD, DAT or hard disc recorded audio on the same timeline and let them go. The problem is that neither device is synced to each other's master "clock" so eventually, the two sources will lose sync with each other. The two may lose sync at 5 minutes; they may lose sync at 45 minutes. There are no hard and fast rules about how long each set of devices can hold sync. There is no way to determine exactly how long it will take unless you test the devices and notate how long it takes for the two to lose sync. Still sounding easy and straightforward? Let's say that your particular combination of DV camcorder and sound recording device will stay in sync for 19 minutes at a time before they lose more than one or two frames sync. You did your homework and determined the number. The next challenges become ones of routing, organization and media management. Let's go over each of these:

Dale Paterson · April 15th, 2006, 03:17 AM

Shawn - HOW DID YOU FIND THAT???

Thanks - that article describes EXACTLY ALL of the problems that I am having - and it would appear that the quick fix is to take a line out of the mixer (via a transmitter of course) to the camera (although as I read it it still does not guarantee that the different clock sources remain in synch with each other).

The most worrying thing of all mentioned in that article is the fact that the clocks of the different devices may / may not go out of synch and if they do go out of synch it may be at different points each time (if I read it and understand it correctly) i.e. they may be in synch for the first five minutes and then one of them goes out or they may stay in synch for two hours and then go out or even worse one 'slips' two or three times within an hour of video!!!

I unfortuanately (due to the nature of my work) no longer have a choice as to whether or not I use a seperate audio recording device (we provide the public address systems for the events as well and there are just too many audio sources to be recorded / sent to the PA without using a mixer) so I have to make sure that this seutp works for me.

Thanks again for finding that article - it contains some really great information in it (aside from the references to my problem).

Regards,

Dale.

Jack Smith · April 15th, 2006, 09:15 PM

A question along a totally different line.Although in your situation ,it may not be a problem,is are you listening to a preview and finding the sync out?
I have found that in some situations the preview may not sync and yet after a render or prerender it is infact in sync.Just 1 thing to eliminate prior to further diagnosing.

Shawn Redford · April 15th, 2006, 11:43 PM

Quote:

Originally Posted by Dale Paterson

Shawn - HOW DID YOU FIND THAT???

Hey Dale - I'm glad that the article was what you needed. It was just good fortune that I could recall it - I was reading the article due to another post from DVInfo and then I saw you post - I'm just happy I could remember where the source was for the webpage (reading too many articles turns into one big mush for me when I try to relocate them :)

Quote:

Originally Posted by Dale Paterson

... and it would appear that the quick fix is to take a line out of the mixer (via a transmitter of course) to the camera (although as I read it it still does not guarantee that the different clock sources remain in synch with each other).

I'm not quite sure I'm following you here. If you have all your sources (mics, instruments, etc.) going into the mixer, then Brockett is suggesting that you use a dual output from the mixer (via XLR or wireless in necessary), in which case your camcorder receives one output and your digital audio recorder receives the other output. This way, your camcorder and digital audio recorder have an identical source to record from. If you did this, the audio recorded on your camcorder would be in-sync because it's all done on your camcorder using the same internal clock for video and audio. However, your digital audio recorder will still go out of sync at some point. Brockett is suggesting this so that the two can be compared in the future, and he suggests that you "post conform" the audio but doesn't give a lot of detail IMHO on how to do this. The only thing I don't know about is your mixer - it sounds like you are using a digital mixer and that may also have an internal clock as well - it seems like this would have to be the case - yes?

Quote:

Originally Posted by Dale Paterson

The most worrying thing of all mentioned in that article is the fact that the clocks of the different devices may / may not go out of synch and if they do go out of synch it may be at different points each time (if I read it and understand it correctly) i.e. they may be in synch for the first five minutes and then one of them goes out or they may stay in synch for two hours and then go out or even worse one 'slips' two or three times within an hour of video!!!

Okay - I didn't get the same impression on that. I thought Brockett was saying that you can test the two devices and determine when they will go out of sync AND this will take place consistently each time if you're using those two devices.

As I look back over this whole thread, it seems like the advice from everyone is pretty similar to this article, but it's just in pieces (Seth is talking about non-sync'd clocks; Jeff suggests using a direct line out of the mixer). The only thing I now question is whether or not it's a good idea to stretch audio at all, since this would re-digitize the samples. As an alternative, I'm wondering if it might not be better to delete a small portion of audio at specific intervals related to the out of sync issue (e.g. 10min or 20min), but of course you would have to find a spot in the audio near those marks that was just room noise and nothing else. To those in the know, would this improve the audio quality or not make any difference verses stretching the audio?

Quote:

Originally Posted by Dale Paterson

I unfortuanately (due to the nature of my work) no longer have a choice as to whether or not I use a seperate audio recording device (we provide the public address systems for the events as well and there are just too many audio sources to be recorded / sent to the PA without using a mixer) so I have to make sure that this seutp works for me.

Well I'm no audio expert, but if you camcorder's audio recording has acceptable quality, can't you just record the mixer output to your camcorder and not have to mess with the sync issues? If you really need the higher quality audio, then use your audio recorder -- but if not it seems like you can get what you need with your camcorder. Based on what you've said, it seems like you have to record long uncut segments (like an event videographer) and as such, you face problems that fiction-'film' folks do not face since they would generally have takes short enough to keep everything in sync. If you can make a cut every 30 minutes or hour, it seems like that would help.

Honestly, all of this is good for me to learn because I have had my eye on the Edirol R4 as a future audio recorder. However, now I'm wondering if I wouldn't be better off with just a mixer, except that means that everything has to be just right when recording since there would be no option to mix the audio levels in post like there would be with the 4-track Edirol or something similar. I wonder also if there are any audio recorders that will allow some sort of manual tweak/compensation on the internal clock to allow it and the camcorder to remain in sync for a long time.

Please keep us posted Dale on your solutions since those should be of interest to anyone following in your footsteps. Thanks, Shawn

Dale Paterson · April 15th, 2006, 11:48 PM

Hi Jack and thanks for the comment.

I think I know what you are talking about.

The video preview (particularly on an external monitor) will always be slightly out of synch with the audio due to latency of the various hardware devices BUT I was trying to use the camera's audio track as a reference point to the 'quality' audio recorded via the mixer and THIS is what goes out of synch. In other words - no matter what the latency of the relevant audio device is I am comparing audio with audio.

Thanks for the input.

Regards,

Dale.

Shawn Redford · April 16th, 2006, 12:23 AM

Dale - I just looked up your mixer and I'm now realizing that you're using this to capture 6 separate channels (or more) using Vegas and that your mixer is essentially your second clock that's going out of sync. I think that Brockett is primarily talking about an analog mixer with a separate audio recorder, but if you can send a combined signal to your camcorder that still should help with sync in Vegas. I'm guessing that with 6 channels of audio it would be impossible to snip the audio like I was thinking, so stretching the audio now sounds like the best option. The only thing I would suggest is that you find out how long it takes until the Alesis and your camcorder go out of sync and then determine if you can shoot in shorter takes of audio and video just to keep everything in sync. I would appreciate hearing how you solve this. I also would like to know how you like the Alesis mixer as this could be much cheaper than the Edirol. Thanks again, Shawn

Dale Paterson · April 16th, 2006, 04:20 AM

Hi Shawn,

I just want you to know that I really do appreciate the interest shown.

Personally I still think that the article is spot on. I mean - if you used a Creative Labs Nomad, for example, to record a stereo main mix out of an analog mixer made up of multiple inputs the Nomad (according to the article) could go out of synch anyway. The only difference between that setup and mine is the fact that my recording device is a notebook and the Alesis mixer sends the audio as data via firewire.

One thing I did not do was to actually click on a button that says 'Set Clock Master' in the Alesis Control Panel prior to recording this last event. I can only assume that by not doing this the notebooks clock was being used as the master clock although I fail to see how this would make a difference.

Another thing that I did not try was to increase the audio buffers in the Alesis Control Panel but again - there were no audio dropouts in over two and a half hours of audio (believe me - I listened to each track in its entirety more than once to check) - and all that increasing the number of buffers would do would be to ENSURE that there were no audio dropouts.

I have, however, learned much from all of the input from everyone on the board and I therefore now are not entirely convinced that the clocks went out at all - this conclusion reached after much pondering on the subject. The reason that I say this is that because of the (potential) delay caused by the distance between the camera's mic, the public address system, the individual presenters, and my using the camera's audio track as a reference for synch purposes the possibility exists that although my ears may 'think' that I have aligned the audio tracks perfectly at the beginning of the first tape a slight error at this point on my part would be compounded by the end of the first tape i.e. after an hour of video. In addition to this, using the camera's audio track as a reference, once again given the above (potential) delay, the audio / video (lip) synch could still be out even if the clocks did not stray at all and the camera's audio track was in perfect synch with the mixers's audio track.

I still need to try a few things:

1) Generate my test tones immediately at the beginning of each video tape - in close proximity to a) the camera's mic and b) a spare mic assigned an additional channel on the mixer so that when I get back to the studio I can 'visually' align the audio tracks and theoretically everything else should line up perfectly (unless there is of course this clock synch problem).

2) Hook up another transmitter to another line out of the mixer and send this audio to the camera. This would certainly prove / disprove the 'drifting' clock theory but you would, of course, still be relyant on your audio perception to align these tracks correctly. Come to think of it a combination of 'beeps' and this extra transmitter might well do the trick.

3) User Acid Pro 6 instead of Vegas (although this should not make any difference I'm sure).

4) Try Steinberg CuBase LE that was supplied with the Alesis mixer as a comparison test.

5) Not use the mixer at all and just plug a mic into the notebooks mic in and then try and synch with this track.

Left any test out???

As far as the Alesis FireWire 16 Mixer is concerned:

It is a great, high quality, unit and has many features that work perfectly for me for example: you can mute the individual tracks being sent out to the public address system without affecting the FireWire output (to the notebook). The only thing that I find to be a problem (and I think this comes more from my lack of experience with the mixer than with the mixer itself) is that the gain knobs for each channel are EXTREMELY sensitive i.e. you can turn them up almost full and there is not much gain change but that last little bit really makes a huge difference with just a slight adjustment but I am sure that with a little more use I will get the hang of it. Other than that it appears to be quite robust and has all of the controls / filters / eq's etc. as per the specs. I have also checked whether or not it induces extra noise into the audio tracks and it would appear that if there is noise being induced by the mixer I either cannot hear it or see it.

The only thing that really upset me was the fact that it is supplied / bundled with Steinberg CuBase LE which only allows you to record up to four tracks simultaneously. As far as I am concerned this really is a con - nowhere on the Alesis website / packaging / manual are you informed of this and I did try to take Alesis to task on this and they were not interested in the least bit (you won't believe the answers that I got back from them). As far as I am concerned - make the customer aware BEFORE purchase so that they know that they are in for an additional cost if the software does not meet their requirements OR charge more for the mixer and include the necessary software.

The above is, of course, not a problem if you have Vegas / Acid Pro but what about the individual that does not have any other software and only after purchase finds out that he now has to go and spend quite a large sum of $$$ for additional software to be able to use the mixer's full capacity.

That aside - it works really well (for me anyway).

Regards,

Dale.

Dale Paterson · April 16th, 2006, 04:38 AM

Hello again Shawn,

Sorry - I did not even see your previous (longer) message - I think I must have been replying at the same time that you were sending it.

Your are right - my problem is that I shoot events which can sometimes be up to four hours long (with one break somewhere in the middle) so what I do is just start the mixer recording a couple of minutes before everyone arrives - after checking the levels of course - and just let it run until the event is over and then (theoretically) it should be (should have been) and easy job to just pull up the video segments (tapes) (I never stop the camera) and synch them together (the only obvious breaks which I have to cut is when I have to change camera tapes).

I can't use the camera's audio (even if I take it straight from the mixer) for the simple reason that I have to edit each individual audio track (by edit I mean doing things like normalizing a track and passages within a track) in order to compensate for different levels of different speakers or presenters. This (theoretically) should not be necessary but in practice it is impossible to adjust the level of each mic / input on the fly and get them all just right.

Due to the above problem I started another thread somewhere on this board looking for a way to either compress / limit a 'too strong' signal PRIOR to recording or enhance a 'too weak' signal PRIOR to recording. I have looked at the Acid Pro 6 Demo and this talks about 're-wiring' a device. As I understand this it means that I could take the inputs from Acid Pro 6 - send the audio to another software package - apply a compressor / limiter / expander and then send this audio back to Acid Pro 6 and record it BUT I cannot use Vegas as a re-wire device, I cannot re-wire Acid Pro 6 to itself, so I need MORE software (like Steinberg's CuBase BUT the full version) in order to accomplish this!!!

Regards,

Dale.

Seth Bloombaum · April 16th, 2006, 12:20 PM

Quote:

Originally Posted by Dale Paterson

...I still need to try a few things:

1) Generate my test tones immediately at the beginning of each video tape - in close proximity to a) the camera's mic and b) a spare mic assigned an additional channel on the mixer so that when I get back to the studio I can 'visually' align the audio tracks and theoretically everything else should line up perfectly (unless there is of course this clock synch problem)...

Dale, you should generate a test tone at the end of the tape as well. This should reveal to your satisfaction if there is a clock synch issue, and how much it is.

If (as I suspect) you are dealing with clock synch issues (or timebase errors)... such timing variations should be consistent. Meaning, if your Alesis clock runs at 100.05% speed of your camcorder's clock, it should always do so. As I wade through the many posts on this thread, there seem to be so many possible explanations explored, now you seem to not be as sure as you were that there even *is* a timebase error?

Timing with test tones at the beginning and end of a tape will reveal this for sure.

And here is an *easy* workflow if you have a test tone at the beginning and end:
Put the camera video/audio up on the timeline.
Trim the clip to the beginning of the first tone, and the end of the last test-tone, that is you'll have Tone-Program-Tone on the timeline.
Put the 2nd system audio up on the timeline.
Do the same trim.

Now you have visual references for everything you've been dealing with, which should make synch easy. Ctrl-drag the end of the audio track to match the end of the video track. If you've dropped a marker at the end of the video track, the audio will snap to it when you Ctrl-drag.

In fact, you don't need to wait until your next panel discussion to do this, just run a test whenever you can leave the camera and laptop sitting for an hour. This exercise should cut through the haze.

Then, there is the other timing "error", the speed of sound in the air, or offset. Sound reinforcement engineers deal with this frequently. Here's a chart I use:
20 feet = 17.9 milliseconds of delay
30' = 26.8ms
40' = 35.7ms
50' = 44.6ms
60' = 53.6ms
70' = 62.5ms
A millisecond is a thousandth of a second (.001). So, to apply a correction you would determine the difference in distance to the presenter of your camera's mic and the 2nd system mic.

Or, do as suggested above and record the second system output on the camcorder. Which will fix any offset errors, but NOT fix any timebase errors.

But I'll repeat. If lip-sync looks good, it is good. Lessee, a 30' offset = abt. 27ms = about 2/3rds of a PAL frame.

BTW, Quantize to frames has no effect in capture, only in editing on the timeline.

Shawn Redford · April 16th, 2006, 01:57 PM

Quote:

Originally Posted by Dale Paterson

... The only difference between that setup and mine is the fact that my recording device is a notebook and the Alesis mixer sends the audio as data via firewire.

One thing I did not do was to actually click on a button that says 'Set Clock Master' in the Alesis Control Panel prior to recording this last event. I can only assume that by not doing this the notebooks clock was being used as the master clock although I fail to see how this would make a difference.

Dale - I don't think that your PC's clock is being used. My impression of how all this works is that your PC is solely a storage device that is accepting already digitized data from your Alesis mixer. Your Alesis mixer must a have a clock since it is converting analog audio to digital and sending that out as data via firewire.

I looked up the Alesis manual online and the 'Set Clock Master' option in your Alesis mixer is for using an external clock. On p. 38 of the manual it states "Setting the master device: If multiple audio devices are connected, one must be designated as the clock master." Take a look at this post, because ultimately you might want to do something similar if your camcorder will sync with your Alesis.

The steps that you have mentioned are ones that you should do, especially to get rid of the sound-travel-delay to your camera, but the best solution IMHO is to ensure that there is no drift. Since you tape for such a long time, it would be suprising if there isn't some sort of drift due to using two different clocks, so syncing the camera and the mixer is another level of control. Thanks for the info on the mixer, Shawn

Dale Paterson · April 17th, 2006, 12:25 AM

Good Morning - and thanks for the input.

Seth:

Quote:

now you seem to not be as sure as you were that there even *is* a timebase error?

I probably should not have made this statement as I did. No - I am sure that there is a drift. All I was saying (trying to say) is that if I 'audibly perceive' that I have correctly synched the audio track from the camera to the mixers recorded audio tracks then - if I was slightly out - over time this would make a difference. However - this would only hold true for the audio / visual synch and not for the audio / audio synch i.e. even if my 'perception' of synch was indeed slightly off for the audio / audio synch at the beginning of a video tape I would also 'perceive' the audio / audio synch to be the same at the end of a video tape / hour and I can tell you that by the end of an hour I am no longer 'perceiving' the audio to be in synch with the audio - if you know what I mean.

This thread is not only making my own head spin - it's keeping me awake at night!!!

But you are right - I need to test this out - and I will do it today - and post the results.

Shawn:

A lot of other things are now beginning to fall into place for me as well.

Just for example: I recently had a problem with lip synch / external FireWire monitor preview / Vegas / Surround Sound Projects i.e. when I first started 'messing around' with mixing surround - sometimes (a lot of the time) I would play the clip (previewing on the external FireWire monitor) and the audio / video synch would start out OK (there is always a slight difference between the external FireWire monitor and the sound card) but would eventually drift way, way, out even over very short clips. I played around for days, buying different sound cards, messing around with DirectX and hardware acceleration etc. etc. and eventually found that the solution was to enable the midi ports in Vegas (any of the ports) and then set them up under the 'Synch' tab for (in my case) 25fps etc. etc. The problem - as if by magic - dissapeared there and then and has not reared its head again. Now - sometimes - the external preview will 'freeze' for a split second (while the workstation is obviously trying to do something else) and then go again but it obviously now 'catches up' to the audio at the right place and continues on in synch where as before enabling / setting up the midi ports in Vegas - the monitor never 'froze' now and then but the audio / video would just go further and further out of synch. Why this only happened on Surround Projects I do not know but there it is!

I digress - but for a reason.

I had a look at that thread and it really does (also) seem to be a possible solution i.e. the MidiStream from Kenton. The only thing that I cannot fathom out is how you would connect the MidiStream to my new notebook i.e. there does not seem to be a game / midi port anywhere on the notebook (and I did not specifically look for one either as it is the last thing that I thought that I would need). Be that as it may - thanks to your explanation - using the MIDI clock as a 'master clock' sure seems to be a good way to do things and like I said in my 'digression' above using the MIDI clock did solve another problem of mine.

Anyway - I am going to run some of these tests today - and I will post the results - as soon as they come off the press!

Thanks to both of you for the input.

Regards,

Dale.

Shawn Redford · April 17th, 2006, 01:06 AM

Quote:

Originally Posted by Dale Paterson

I had a look at that thread and it really does (also) seem to be a possible solution i.e. the MidiStream from Kenton. The only thing that I cannot fathom out is how you would connect the MidiStream to my new notebook i.e. there does not seem to be a game / midi port anywhere on the notebook (and I did not specifically look for one either as it is the last thing that I thought that I would need). Be that as it may - thanks to your explanation - using the MIDI clock as a 'master clock' sure seems to be a good way to do things and like I said in my 'digression' above using the MIDI clock did solve another problem of mine.

Dale, The MidiSream (or Midi Time Code) does NOT connect to your new notebook - it connects to your Alesis mixer (if the Alesis has that jack). Seth seems to know more about this than me, so please tell me if you think I'm on target Seth. The Alesis is a digital mixer (or at least has digital output), so it needs the sync. The firewire connection is just there to create a link to your computer for storage/download, but I can't imagine how your notebook's clock would be used unless the Alesis was polling the notebook's clock (which seems unlikely). However, even if that was the case the Midi cable would still go to the Alesis mixer because it would then sync to the camcorder's clock provided you choose the 'clock master' option for the Alesis (even though that is done using software on the notebook). Here's another thread describing five methods, one of which is related to that post: http://www.sonyhdvinfo.com/showthread.php?t=4621 - To recap: since the mixer is sampling the analog audio and outputing digital audio, it has to use some clock to do that (whether it's own or polled from the notebook, or from the camcorder). Therefore, I'm 90% sure that you need to connect the Midi cable to the Alesis Mixer. Does this make sense Dale?

P.S. I just looked through your manual, and I'm not seeing a midi port anywhere, so this may all be for nothing. If it does have MIDI then great. If not, you want to figure out how they connect mutiple mixers and then see if you can use the connection with your camcorder to sync them together. It's now sounding like this may not be possible, but there's always hope.

Dale Paterson · April 17th, 2006, 07:53 AM

Hi Shawn - thanks for that information - really appreciated.

I have not had a look at the last link you posted yet - I am busy testing this stuff.

These are the results of my first test:

Workflow:

Start notebook recording
Start camera recording
Place handheld Sony wireless mic next to and in line with camera mic
Generate a couple of very short 'beeps' (you would not believe what I am using for this)
At about 17 minutes into the recording generate a couple of very short 'beeps'
At the end of the tape (about one minute to go) generate a couple of very short 'beeps'

Upon completion - align initial series of 'beeps'

Result:

Camera video / audio (end of last 'beep') ended at 91,911,993 (absolute frames)
Mixer audio (end of last 'beep') ended at 91,907,971 (absolute frames)

Stretched the mixer audio and the everything aligned. This is good because it at least means that the clock differences are at least constant over time.

I must just say that aligning the audio tracks with these 'beeps' is a cinch - really easy and accurate particularly if you magnify the waveforms - you can be accurate to individual peaks and troughs - not bad.

Next result coming up.

Edit:

Something has dawned on me while I idly sit here waiting for another test to complete - is this not going backward? I mean - this reminds me of the old days of capturing analog video - you had to specify a master stream (be it audio or video) and then hope that eveything stayed in snych with no dropped frames etc. etc. and that you did not run out of disk space!

Regards,

Dale.

Jeff Mack · April 17th, 2006, 08:52 AM

Hey Dale,

A couple of comments. First, to mention your problem with your gain control. The reason it affects differently the higher you turn the knob is that increasing gain is not the same as increasing volume. Increasing volume is at a constant rate. Increasing gain acts like increasing a per centage - the higher the db increase, the higher the overall per centage increase. It works sort of like an exponential increase so as you twist the knob, the more sensitive it is to change. Knobs are the cheaper way to manufacture than sliders.

Second, I'm far from an expert but if you haven't tried sending a stereo line into your camera, bypassing your camera mic, I think you are creating more headache for yourself. I have an Alesis 24 track recorder and when I capture footage to my camera, I record audio in from the mixer to replace the camera mic and I NEVER have an issue with sync. I think if you try this, it will solve your problem. If you need the pa recorded but it doesn't go through the mixer, add another track to record by putting a crowd mic close to the mixer and you can mix it in post.

I have also gotten deep into Vegas surround sound and there are some issues of settings that are not really covered in the manual. Notably the center channel defaults to pan type instead of constant power. That means by default, the center channel gets sent to bothe the front right and left which was making my front right and left clip. Switching the pan type on the center channel to constant power keeps 100% of the channel to the center channel. You can tell this is happening when you solo the center channel and it plays through the front right, left AND center monitor mixer.

Good luck.

Jeff

Dale Paterson · April 17th, 2006, 10:10 AM

Hi Jeff - thanks.

1) Gain control - now I understand the difference.

2) Mixer - I (probably) have the same mixer as you. My problem is that I have to edit / mix / 'fine tune' as it were the individual tracks for my final product - I can't just take a stereo main mix out directly to the camera live - that is why all of this is happening.

3) Vegas Surround - I just had a problem with the audio / video synch when previewing on an external monitor and enabling the midi ports / synch seemed to solve the problem. I have noticed what you are talking about though.

On with the tests!!!

I have just recorded another hour of audio / video but this time not using the mixer - just the notebooks onboard sound card - with one wireless mic plugged straight into the notebooks sound card - and following the same workflow as described in my previous post.

The interesting thing about this is that the camera audio / notebook audio is out of synch almost EXACTLY (I would go so far as to say EXACTLY) the same amount as it was with the mixer after an hour. I do not believe in coincidence! Something is telling me to look elsewhere i.e. maybe not a clock issue at all???

Any thoughts???

I am now going to play around with some Vegas settings and I think that I should also try using my editing workstation to record the audio via the mixer just to eliminate the possibility of the notebook having a problem (although as a control I captured the exact same video tape used in my first test from the camera using the notebook and that was also out by the same amount i.e. if the notebook itself was at fault or inaccurate then the video capture on the notebook would have matched the audio originally captured using the notebook and mixer).

Regards,

Dale.

April 15th, 2006, 03:17 AM	#17
Dale Paterson Major Player Join Date: Jun 2005 Location: Johannesburg, South Africa Posts: 200	Shawn - HOW DID YOU FIND THAT??? Thanks - that article describes EXACTLY ALL of the problems that I am having - and it would appear that the quick fix is to take a line out of the mixer (via a transmitter of course) to the camera (although as I read it it still does not guarantee that the different clock sources remain in synch with each other). The most worrying thing of all mentioned in that article is the fact that the clocks of the different devices may / may not go out of synch and if they do go out of synch it may be at different points each time (if I read it and understand it correctly) i.e. they may be in synch for the first five minutes and then one of them goes out or they may stay in synch for two hours and then go out or even worse one 'slips' two or three times within an hour of video!!! I unfortuanately (due to the nature of my work) no longer have a choice as to whether or not I use a seperate audio recording device (we provide the public address systems for the events as well and there are just too many audio sources to be recorded / sent to the PA without using a mixer) so I have to make sure that this seutp works for me. Thanks again for finding that article - it contains some really great information in it (aside from the references to my problem). Regards, Dale.

April 15th, 2006, 09:15 PM	#18
Jack Smith Major Player Join Date: Nov 2003 Location: ontario Posts: 445	A question along a totally different line.Although in your situation ,it may not be a problem,is are you listening to a preview and finding the sync out? I have found that in some situations the preview may not sync and yet after a render or prerender it is infact in sync.Just 1 thing to eliminate prior to further diagnosing.

April 15th, 2006, 11:48 PM	#20
Dale Paterson Major Player Join Date: Jun 2005 Location: Johannesburg, South Africa Posts: 200	Hi Jack and thanks for the comment. I think I know what you are talking about. The video preview (particularly on an external monitor) will always be slightly out of synch with the audio due to latency of the various hardware devices BUT I was trying to use the camera's audio track as a reference point to the 'quality' audio recorded via the mixer and THIS is what goes out of synch. In other words - no matter what the latency of the relevant audio device is I am comparing audio with audio. Thanks for the input. Regards, Dale.

April 16th, 2006, 12:23 AM	#21
Shawn Redford Regular Crew Join Date: Jun 2005 Posts: 158	Dale - I just looked up your mixer and I'm now realizing that you're using this to capture 6 separate channels (or more) using Vegas and that your mixer is essentially your second clock that's going out of sync. I think that Brockett is primarily talking about an analog mixer with a separate audio recorder, but if you can send a combined signal to your camcorder that still should help with sync in Vegas. I'm guessing that with 6 channels of audio it would be impossible to snip the audio like I was thinking, so stretching the audio now sounds like the best option. The only thing I would suggest is that you find out how long it takes until the Alesis and your camcorder go out of sync and then determine if you can shoot in shorter takes of audio and video just to keep everything in sync. I would appreciate hearing how you solve this. I also would like to know how you like the Alesis mixer as this could be much cheaper than the Edirol. Thanks again, Shawn

April 16th, 2006, 04:20 AM	#22
Dale Paterson Major Player Join Date: Jun 2005 Location: Johannesburg, South Africa Posts: 200	Hi Shawn, I just want you to know that I really do appreciate the interest shown. Personally I still think that the article is spot on. I mean - if you used a Creative Labs Nomad, for example, to record a stereo main mix out of an analog mixer made up of multiple inputs the Nomad (according to the article) could go out of synch anyway. The only difference between that setup and mine is the fact that my recording device is a notebook and the Alesis mixer sends the audio as data via firewire. One thing I did not do was to actually click on a button that says 'Set Clock Master' in the Alesis Control Panel prior to recording this last event. I can only assume that by not doing this the notebooks clock was being used as the master clock although I fail to see how this would make a difference. Another thing that I did not try was to increase the audio buffers in the Alesis Control Panel but again - there were no audio dropouts in over two and a half hours of audio (believe me - I listened to each track in its entirety more than once to check) - and all that increasing the number of buffers would do would be to ENSURE that there were no audio dropouts. I have, however, learned much from all of the input from everyone on the board and I therefore now are not entirely convinced that the clocks went out at all - this conclusion reached after much pondering on the subject. The reason that I say this is that because of the (potential) delay caused by the distance between the camera's mic, the public address system, the individual presenters, and my using the camera's audio track as a reference for synch purposes the possibility exists that although my ears may 'think' that I have aligned the audio tracks perfectly at the beginning of the first tape a slight error at this point on my part would be compounded by the end of the first tape i.e. after an hour of video. In addition to this, using the camera's audio track as a reference, once again given the above (potential) delay, the audio / video (lip) synch could still be out even if the clocks did not stray at all and the camera's audio track was in perfect synch with the mixers's audio track. I still need to try a few things: 1) Generate my test tones immediately at the beginning of each video tape - in close proximity to a) the camera's mic and b) a spare mic assigned an additional channel on the mixer so that when I get back to the studio I can 'visually' align the audio tracks and theoretically everything else should line up perfectly (unless there is of course this clock synch problem). 2) Hook up another transmitter to another line out of the mixer and send this audio to the camera. This would certainly prove / disprove the 'drifting' clock theory but you would, of course, still be relyant on your audio perception to align these tracks correctly. Come to think of it a combination of 'beeps' and this extra transmitter might well do the trick. 3) User Acid Pro 6 instead of Vegas (although this should not make any difference I'm sure). 4) Try Steinberg CuBase LE that was supplied with the Alesis mixer as a comparison test. 5) Not use the mixer at all and just plug a mic into the notebooks mic in and then try and synch with this track. Left any test out??? As far as the Alesis FireWire 16 Mixer is concerned: It is a great, high quality, unit and has many features that work perfectly for me for example: you can mute the individual tracks being sent out to the public address system without affecting the FireWire output (to the notebook). The only thing that I find to be a problem (and I think this comes more from my lack of experience with the mixer than with the mixer itself) is that the gain knobs for each channel are EXTREMELY sensitive i.e. you can turn them up almost full and there is not much gain change but that last little bit really makes a huge difference with just a slight adjustment but I am sure that with a little more use I will get the hang of it. Other than that it appears to be quite robust and has all of the controls / filters / eq's etc. as per the specs. I have also checked whether or not it induces extra noise into the audio tracks and it would appear that if there is noise being induced by the mixer I either cannot hear it or see it. The only thing that really upset me was the fact that it is supplied / bundled with Steinberg CuBase LE which only allows you to record up to four tracks simultaneously. As far as I am concerned this really is a con - nowhere on the Alesis website / packaging / manual are you informed of this and I did try to take Alesis to task on this and they were not interested in the least bit (you won't believe the answers that I got back from them). As far as I am concerned - make the customer aware BEFORE purchase so that they know that they are in for an additional cost if the software does not meet their requirements OR charge more for the mixer and include the necessary software. The above is, of course, not a problem if you have Vegas / Acid Pro but what about the individual that does not have any other software and only after purchase finds out that he now has to go and spend quite a large sum of $$$ for additional software to be able to use the mixer's full capacity. That aside - it works really well (for me anyway). Regards, Dale.

April 16th, 2006, 04:38 AM	#23
Dale Paterson Major Player Join Date: Jun 2005 Location: Johannesburg, South Africa Posts: 200	Hello again Shawn, Sorry - I did not even see your previous (longer) message - I think I must have been replying at the same time that you were sending it. Your are right - my problem is that I shoot events which can sometimes be up to four hours long (with one break somewhere in the middle) so what I do is just start the mixer recording a couple of minutes before everyone arrives - after checking the levels of course - and just let it run until the event is over and then (theoretically) it should be (should have been) and easy job to just pull up the video segments (tapes) (I never stop the camera) and synch them together (the only obvious breaks which I have to cut is when I have to change camera tapes). I can't use the camera's audio (even if I take it straight from the mixer) for the simple reason that I have to edit each individual audio track (by edit I mean doing things like normalizing a track and passages within a track) in order to compensate for different levels of different speakers or presenters. This (theoretically) should not be necessary but in practice it is impossible to adjust the level of each mic / input on the fly and get them all just right. Due to the above problem I started another thread somewhere on this board looking for a way to either compress / limit a 'too strong' signal PRIOR to recording or enhance a 'too weak' signal PRIOR to recording. I have looked at the Acid Pro 6 Demo and this talks about 're-wiring' a device. As I understand this it means that I could take the inputs from Acid Pro 6 - send the audio to another software package - apply a compressor / limiter / expander and then send this audio back to Acid Pro 6 and record it BUT I cannot use Vegas as a re-wire device, I cannot re-wire Acid Pro 6 to itself, so I need MORE software (like Steinberg's CuBase BUT the full version) in order to accomplish this!!! Regards, Dale.

April 17th, 2006, 07:53 AM	#28
Dale Paterson Major Player Join Date: Jun 2005 Location: Johannesburg, South Africa Posts: 200	Hi Shawn - thanks for that information - really appreciated. I have not had a look at the last link you posted yet - I am busy testing this stuff. These are the results of my first test: Workflow: Start notebook recording Start camera recording Place handheld Sony wireless mic next to and in line with camera mic Generate a couple of very short 'beeps' (you would not believe what I am using for this) At about 17 minutes into the recording generate a couple of very short 'beeps' At the end of the tape (about one minute to go) generate a couple of very short 'beeps' Upon completion - align initial series of 'beeps' Result: Camera video / audio (end of last 'beep') ended at 91,911,993 (absolute frames) Mixer audio (end of last 'beep') ended at 91,907,971 (absolute frames) Stretched the mixer audio and the everything aligned. This is good because it at least means that the clock differences are at least constant over time. I must just say that aligning the audio tracks with these 'beeps' is a cinch - really easy and accurate particularly if you magnify the waveforms - you can be accurate to individual peaks and troughs - not bad. Next result coming up. Edit: Something has dawned on me while I idly sit here waiting for another test to complete - is this not going backward? I mean - this reminds me of the old days of capturing analog video - you had to specify a master stream (be it audio or video) and then hope that eveything stayed in snych with no dropped frames etc. etc. and that you did not run out of disk space! Regards, Dale.

April 17th, 2006, 08:52 AM	#29
Jeff Mack Regular Crew Join Date: Mar 2005 Location: McKinney, TX Posts: 195	Hey Dale, A couple of comments. First, to mention your problem with your gain control. The reason it affects differently the higher you turn the knob is that increasing gain is not the same as increasing volume. Increasing volume is at a constant rate. Increasing gain acts like increasing a per centage - the higher the db increase, the higher the overall per centage increase. It works sort of like an exponential increase so as you twist the knob, the more sensitive it is to change. Knobs are the cheaper way to manufacture than sliders. Second, I'm far from an expert but if you haven't tried sending a stereo line into your camera, bypassing your camera mic, I think you are creating more headache for yourself. I have an Alesis 24 track recorder and when I capture footage to my camera, I record audio in from the mixer to replace the camera mic and I NEVER have an issue with sync. I think if you try this, it will solve your problem. If you need the pa recorded but it doesn't go through the mixer, add another track to record by putting a crowd mic close to the mixer and you can mix it in post. I have also gotten deep into Vegas surround sound and there are some issues of settings that are not really covered in the manual. Notably the center channel defaults to pan type instead of constant power. That means by default, the center channel gets sent to bothe the front right and left which was making my front right and left clip. Switching the pan type on the center channel to constant power keeps 100% of the channel to the center channel. You can tell this is happening when you solo the center channel and it plays through the front right, left AND center monitor mixer. Good luck. Jeff

April 17th, 2006, 10:10 AM	#30
Dale Paterson Major Player Join Date: Jun 2005 Location: Johannesburg, South Africa Posts: 200	Hi Jeff - thanks. 1) Gain control - now I understand the difference. 2) Mixer - I (probably) have the same mixer as you. My problem is that I have to edit / mix / 'fine tune' as it were the individual tracks for my final product - I can't just take a stereo main mix out directly to the camera live - that is why all of this is happening. 3) Vegas Surround - I just had a problem with the audio / video synch when previewing on an external monitor and enabling the midi ports / synch seemed to solve the problem. I have noticed what you are talking about though. On with the tests!!! I have just recorded another hour of audio / video but this time not using the mixer - just the notebooks onboard sound card - with one wireless mic plugged straight into the notebooks sound card - and following the same workflow as described in my previous post. The interesting thing about this is that the camera audio / notebook audio is out of synch almost EXACTLY (I would go so far as to say EXACTLY) the same amount as it was with the mixer after an hour. I do not believe in coincidence! Something is telling me to look elsewhere i.e. maybe not a clock issue at all??? Any thoughts??? I am now going to play around with some Vegas settings and I think that I should also try using my editing workstation to record the audio via the mixer just to eliminate the possibility of the notebook having a problem (although as a control I captured the exact same video tape used in my first test from the camera using the notebook and that was also out by the same amount i.e. if the notebook itself was at fault or inaccurate then the video capture on the notebook would have matched the audio originally captured using the notebook and mixer). Regards, Dale.