Need to sync two audio files

[happydog500]

I recorded a convention with talks (a person giving a talk). It was a 3 day convention so I have a lot of work.
I had my recorder (sony P530F) set with the wrong settings, so it's distant. My friend also recorded the same convention. Her's is OK. Not great, OK.

If I sync the two files at first, soon they become out of sync because they are slightly off. I would like to sync the two audios to get a MTX. I tried to do it in audacity, but it's not good for what I need.

Is there a free program that can sync two audio files that works good? I found a good one but it's very expensive.

Thank you,
Chris.

[johns0]

Audacity is good for what you need,you just have to learn what it can do such as stretch or shrink the audio so it matches the video,in the effects tab choose change speed by putting in a percentage,use a bit of math and it will match the video.

[happydog500]

If I put in a number, it doesn't sync, put in another number, it's not synced, I keep doing it, all day long and it's not synced. If I can sync the beginning, and the end, it's not synced all the way through. Even a slight off, will add echo.

To give an example of how bad audacity is for this, if I have one track on top, the other one on the bottom, when I put the line at the beginning of the file, it doesn't even go to the file below it. In audacity, you have to guess where they line up. That's ridiculous.

Take the morning of the first day. I have a file 2 hours long. I need it synced about every 3-5 minutes through the whole file. Any little part not synced, gives an echo. It's in a stadium, so it has echo already. Any little bit it's no good.

The expensive program I saw, you can go in at anytime and sync by sliding it over. Go another 5 minutes and sync all the way through. The line goes through both files so you can line it up accurately.

Chris.

[johns0]

Guess you're not good at math.

[smrpix]

Two thoughts.
1) Break the new track into chunks if 5 or so minutes and sync them up end to end. See if you lose the echo that way.
2) if the echo is slight, meaning barely out of sync, just toss the original track completely once you've got it timed out. It's no good anyway.

[happydog500]

Guess you're not good at math.

It's not just math, if you take a day and work in a couple hours, sync the beginning and the end, the stuff in the middle doesn't line up perfect.

[happydog500]

I see people do concerts that are two hours long, they go in and sync every 2 seconds all the way through the concert. I can guaranty they don't do that in audacity. It's not made to sync like that.

Not to mention it's 2 minutes each time you try.

Is there a way to grab the end of one of the files and slide it back longer or shorter? That would be a little easier.

Chris.

[Cornucopia]

Where were you vs. your friend, in relation to the distance to the speaker? At the normal speed of sound (at ground level, std. barometric pressure), 1 meter = ~129 samples (at 44.1kHz) or ~141 samples (at 48kHz).

Are you taking that shift/difference into account?

Then, sync via waveform PLUS via a recognizable common spike (at head AND at tail). It should now be synced down to the sample level. If it isn't, it's because one or both of your recorders are drifting in their sample clock. I hate to say it but a "voice recorder" such as your sony P530F is NOT a high quality device. Don't know about the other device (you didn't mention model).

At that point, I would: pick the track that had the best timebase (even if it isn't the best quality). If have video you need to sync to, you are very constrained in what you can do, but you can still do the 5-minute-worth segments syncing that smrpix mentioned. Time-compress/expand each one to match your reference for that segment, then lay them end-to-end (with minute crossfades in between). If you don't have video you need to sync to, I don't get what your problem is...?

BTW, what the heck is a MTX? Do you mean MIX? Seeing as how you didn't set this up from the outset as a multi-mike/multi-recorder project, you are fighting an uphill battle if you haven't done this (m.m. TC/Sync) before.

Scott

[johns0]

Bad spelling plus making no sense.

[Cornucopia]

I see people do concerts that are two hours long, they go in and sync every 2 seconds all the way through the concert. I can guaranty they don't do that in audacity. It's not made to sync like that.

Not to mention it's 2 minutes each time you try.

Is there a way to grab the end of one of the files and slide it back longer or shorter? That would be a little easier.

Chris.

When pros or even semi-pros do a talk/concert/recital/performance/event/etc., they would use a form of clapper at head & at tails, and they would use devices that are sample-clock stable. Then they just sync to head, check to see if tail is in sync. If it is, they're done. If not, they do a high quality (phase-linear) TC/X (shrink/stretch) of the whole program (assuming it was all recorded as one clip) to match the "reference". Then they're done.

One of those "ounce of prevention worth a pound of cure" sort of things.

Plus, no, I wouldn't use Audacity for such a multi-track sort of thing. It can be done, but would be a pain. You want a full-blown DAW or sample-accurate NLE. IOW, not free. You get what you pay for.

Scott

[happydog500]

There are places you can get live concerts. They have SBD (soundboard), AUD (audience recording), MTX (matrix). The most popular MTX is a SBD (soundboard) mixed with an AUD (audience). That way you get the best of both worlds. Typically, you get no crowd noise with a SBD. With a MTX, You get ambient noise, plus a strong sound.
With music, mixing the AUD with a SBD you have to sync every few seconds to get a good recording.

Audacity is a great program, for free. I was wondering if they make a open source program like audacity that will work like a good DAW for free.

My recorder doesn't need to be very good. I've gotten very good recordings in the past with it because it's just voice. I had it on the wrong settings this time (it was in a big, open air stadium, set to only record up to 3 feet away).
Mine is solid, just distant. Her's sound is up front, but has a echo. Combined, it makes the voice a lot stronger then the echo. It almost goes away it's so "good."

I was going to try and work something out with people who do the concerts, but thought I would try myself first. I could DL a good program, but I don't do illegal software (concerts are with bands permission).

I was hoping that since people make such a good program like audacity, maybe someone makes a good program that is easy to sync audio files. Guess not.

[smrpix]

As Cornucopia suggests, using better gear and procedures upfront is the real solution.
Fixing it in post with Audacity, Reaper, Audition, Pro Tools, et al always takes longer and usually compromises quality.

[Cornucopia]

Even the BEST audio programs ($$$) don't easily sync bad-timebase audio. You have to work through that yourself. They just make the process go a little faster. For FOSS DAW apps, you could try Ardor, Rosegarden, or Traverso, but I can't vouch for any of them.

SBD, AUD, MTX: well I guess that's what happens when non-audio engineers (fans) start trying to use audio terms! Audio engineers might recognize "sbd", and maybe "aud" as labelled recording category contractions, but NEVER, NEVER would they use Matrix/MTX for a mix of the other 2. That is NOT what matrix refers to (non-discreet surround downmix). And most audio engineers might use "Direct" or "Clean" or just "Board" for board audio. And Audience sound would be divided up into FOH/amped-mains and BOH/ambience. Oh, well. Guess it doesn't really matter unless/until a fan needs to consult an audio engineer and runs into a jargon barrier.

Your recorder doesn't need to be very good to record voice, but for MUSIC it needs to be much better. Just the fact that it has a "3-feet away" setting option tells me it is not optimal.

*******************************

...Actually, I just looked at your manual...

As per pg.51:

This unit is designed to automatically adjust the recording level
when recording meetings, etc., and is not suitable for music
recording.

It has a mono electret flushmount mike, AGC, at BEST 6800Hz bandwidth, and only 19.5kbps CELP-type compressed recording format. OK for voice, NOT GOOD FOR MUSIC.

That bit about "3 feet away" had to do with Hi vs. Lo sensitivity for the mike. Either way, it still always has AGC engaged.

On a quality scale of 1-10, I'd put it at 1.5. Sorry to burst your bubble, but now we can rest assured that you've got something with a difficult timebase. But if you consider this good, I wouldn't sweat the syncing up part. Getting it exact would be an uphill battle on a 60° slope!

Scott

[SameSelf]

This is an interesting thread. You sound like you know a thing or two about audio. I think smrpix gave some great advice in suggesting Reaper. I have not really used it because I use mostly Cubase. But the audio pros love it. It is a full featured DAW for a small pittance compared to what Cubase costs or a similar DAW. Plus IIRC, Reaper offers a free trial. So you don't even risk anything.

As for the other advice, sadly audio is one of those things where there is not much you can do if the initial recording had technical problems. It is why the pros spend so much on gear because the cost of doing it a second time overshadows everything else. But I feel your pain, I have often screwed up a recording as well (the most recent example was my portable recorder had the metronome on, doh!) So that's why I always have a back up now for everything. Audio and video.

Good luck, I hope you solve your problem.

[happydog500]

Even the BEST audio programs ($$$) don't easily sync bad-timebase audio. You have to work through that yourself. They just make the process go a little faster. For FOSS DAW apps, you could try Ardor, Rosegarden, or Traverso, but I can't vouch for any of them.

SBD, AUD, MTX: well I guess that's what happens when non-audio engineers (fans) start trying to use audio terms! Audio engineers might recognize "sbd", and maybe "aud" as labelled recording category contractions, but NEVER, NEVER would they use Matrix/MTX for a mix of the other 2. That is NOT what matrix refers to (non-discreet surround downmix). And most audio engineers might use "Direct" or "Clean" or just "Board" for board audio. And Audience sound would be divided up into FOH/amped-mains and BOH/ambience. Oh, well. Guess it doesn't really matter unless/until a fan needs to consult an audio engineer and runs into a jargon barrier.

Your recorder doesn't need to be very good to record voice, but for MUSIC it needs to be much better. Just the fact that it has a "3-feet away" setting option tells me it is not optimal.

*******************************

...Actually, I just looked at your manual...

As per pg.51:

This unit is designed to automatically adjust the recording level
when recording meetings, etc., and is not suitable for music
recording.

It has a mono electret flushmount mike, AGC, at BEST 6800Hz bandwidth, and only 19.5kbps CELP-type compressed recording format. OK for voice, NOT GOOD FOR MUSIC.

That bit about "3 feet away" had to do with Hi vs. Lo sensitivity for the mike. Either way, it still always has AGC engaged.

On a quality scale of 1-10, I'd put it at 1.5. Sorry to burst your bubble, but now we can rest assured that you've got something with a difficult timebase. But if you consider this good, I wouldn't sweat the syncing up part. Getting it exact would be an uphill battle on a 60° slope!

Scott

Not planning on doing music at all. I went to them because they sync audio. My recorder is exilent

[happydog500]

Even the BEST audio programs ($$$) don't easily sync bad-timebase audio. You have to work through that yourself. They just make the process go a little faster. For FOSS DAW apps, you could try Ardor, Rosegarden, or Traverso, but I can't vouch for any of them.

SBD, AUD, MTX: well I guess that's what happens when non-audio engineers (fans) start trying to use audio terms! Audio engineers might recognize "sbd", and maybe "aud" as labelled recording category contractions, but NEVER, NEVER would they use Matrix/MTX for a mix of the other 2. That is NOT what matrix refers to (non-discreet surround downmix). And most audio engineers might use "Direct" or "Clean" or just "Board" for board audio. And Audience sound would be divided up into FOH/amped-mains and BOH/ambience. Oh, well. Guess it doesn't really matter unless/until a fan needs to consult an audio engineer and runs into a jargon barrier.

Your recorder doesn't need to be very good to record voice, but for MUSIC it needs to be much better. Just the fact that it has a "3-feet away" setting option tells me it is not optimal.

*******************************

...Actually, I just looked at your manual...

As per pg.51:

This unit is designed to automatically adjust the recording level
when recording meetings, etc., and is not suitable for music
recording.

It has a mono electret flushmount mike, AGC, at BEST 6800Hz bandwidth, and only 19.5kbps CELP-type compressed recording format. OK for voice, NOT GOOD FOR MUSIC.

That bit about "3 feet away" had to do with Hi vs. Lo sensitivity for the mike. Either way, it still always has AGC engaged.

On a quality scale of 1-10, I'd put it at 1.5. Sorry to burst your bubble, but now we can rest assured that you've got something with a difficult timebase. But if you consider this good, I wouldn't sweat the syncing up part. Getting it exact would be an uphill battle on a 60° slope!

Scott

Never planed on recording music. I know they do mixing all the time, so I used them as an example. My recordings have come out, Excellent ++ for years, just a person talking. A grade 8-10.
I put titles, track numbers in there "001" "002" so they play in order with an mp3 player, if need be.

The sites, (with MTX as a category), have sound engineers all over the place. Some have been in the business for over 30 years. I would call a person who toured all over the world, for very big artists, doing sound, an audio expert.

I can sync the files for a few minutes, it does improve the sound a lot, so even tho it's not perfect, sounds way better. That's all I was looking for, a little improvement.

I see an "attachments" at the top of this page, can I attache a small audio clip here?

Chris.

[davexnet]

Yes attach some clips showing the problem. The file size limits are quite liberal.

[happydog500]

I'll get some samples soon. I see a lot of people use Adobe Audition. Here are some examples of MTX.

Genre: Rock
Audience Tape: A
Mikes: Schoeps mk4 > Actives > NBox >
Recorder: Sony M-10
ALD Feed Mix: 20% hard left with Bass Boost; 20% to slight right
Transfer: Adobe Audition CC (Time Alignment) > Reaper (Mixdown and EQ)
Goldwave (Track Splitting and Fades) Media Monkey (Flac Conversion)
************************************************** ***

Drew's Dub (4-Source Matrix) Mixed In Audition 3.0
Sources:
1.Bono's IEM > Icom R5 + Tascam Dr-07. Taper: ificould
2.OKMIIR > A3 > SONY DAT TCD-D100. Taper: masseygo
3.Church-Audio Mics CA-14 Cardioids > Church Audio Preamp CA-9100 v.4.4 > Hydra-Audio Cable (1m) > Roland Edirol R09HR v.1.04. Taper: trebor (better post production version)
4.SP-CMC-4 > SP-SPSB-1 Battery Box @107HZ > Olympus LS-10 (24/96) Taper:bedoc
Location: Right Center
************************************************** ************
Sources:
Audience Source 1
Lineage: Neumann KM140's > AETA PSP3 > Edirol R09 > HarBal 2.3

Audience Source 2
Lineage: SP-BMC-12 > SP-SPSB-1 battery box (bass roll off @ 107Hz, level 80%) > Sony mz-nhf800 himd Walkman (pcm mode) > 1gb himd > USB > Sonicstage 4.3 > WAV > Cool Edit Pro 2.0 (clap removal, amplify, fade in/out) > CD Wave (tracking) > tlh (FLAC-8, md5 and fingerprints)

Radio Source 1
Lineage: Classic 21 Rocks (RADIO FM Belgium) > Sony STR-DA1200ES > Edirol R-09 > Memory Card > Wave 16bit/44.1kHz > Adobe Audition 3.0 > FLAC-8

Radio Source 2
Lineage: Sony ST-S370 tuner > Tascam DA20mk2 DAT deck > Sony DG90M dat master at 48khz > fully digital transfer with coaxial cable > Marantz CDR630 (stand alone professional CD recorder) > CDRW clone > extraction on HL GSA-T40N DVDRAM drive with EAC v0.99 prebeta 4 > 44.100khz/16bit wav > Audacity 1.3.5 Beta (editing and track splitting) > 16bit flac's (level 8) > ZOMB
**************************************************

3 Source IEM/AUD Matrix

Source 1: Tascam DR-2D with internal mics 16/44 > USB > harddrive
Source 2: Sennheiser EW300 Receiver (keyboard player feed) > M Audio Microtrack II 16/44 > USB > harddrive
Source 3: ICOM RX7 (guitarist's feed) > Tascam DR-05 16/44 > USB > harddrive

Notes:
All sources synced and mixed in Adobe Audition. Vocals are a little low since they were extremely low in
each monitor feed but the mic source brought them up a little bit which helps. There are a few tiny
static spots but overall a nice clean mix.
************************************

This is why I posted. I've seen years and years of this stuff being upped. I have over 650 live concerts. My dumb little two audio sources can't be hard to do compared to what they do. One is even a 4 source.

I look through the pages and it's full of audio engineers. Some famous ones.

Chris.

[smrpix]

I look through the pages and it's full of audio engineers. Some famous ones.

Fine, but none of those sources are a...

(sony P530F) set with the wrong settings.

[happydog500]

I know, I'm not doing music. They need that, for music, I need the P530 for voice. I was referring to the syncing. Doesn't matter what source, if all I'm worried about is syncing.

Chris.

[smrpix]

Doesn't matter what source, if all I'm worried about is syncing.

That's incorrect. If you have a source recorder with a loose internal clock, syncing becomes difficult no matter what subject you're recording.

[happydog500]

OK, for starters, here is a 42 second clip of one I did two years a go. I put this recording with my Sony 530F at a 9. You put this recording at a 1.5.

Remember, all I need to do is put it on a CD, so a few people who couldn't go, can here what they missed. This is not going to be sold is stores, doesn't need to be supper good. I think for what I need it for, the Sony 530F records pretty good. At least better then the 1.5 rating.

Chris.

[Cornucopia]

Well, I will have to admit: that was much better than I was expecting! I'll have to upgrade your recording to a 3-3.5 in my book. Considering your equipment, you lucked out by being well situated on that occasion.

Just so you know I'm not bullshitting you: here is a sample of something I did a decade or more ago (mono, AV demonstration/presentation in warehouse using boom pole+SennheiserMKH60 shotgun > Fostex DAT > DVD, so this copy has been re-encoded from WAV >256kbpsAC3 >112kbpsVBRmp3). I would give this "AUD" a 7 on the same scale (have done higher recordings but none of those would match yours WRT environment/situation).

Scott

[happydog500]

Nice Recording.

Mine is a MP3 of the WAV so it's not as good here. I'm not talking about an overall rating of 9. I rated 9, based on what I had, and what I'm going to do with it. A little crap recorder, with CD given to an elderly couple.

If the best I can get is a 1.5, and I got three times higher after you heard it, thats a good score, [I]based on what I had.[I]

I only see two places to upload files, so I will post mine, the other persons, then in the next post I'll post a little MTX I did.

This is a person talking before the talk began, so it's not a very good sample, since this guy sounded different then the rest of it.

Chris.

P.S. feel free to see what you could do with these.

[happydog500]

Here is a little MTX I did. I'm not sure it lined up right, I just did a quick one just to post it.

Also, didn't mean to answer back every time I got a answer. I knew all along people where trying to help. Some of my posts came across harsher then intended.


Chris.