|
Post by sirthought on Jun 12, 2019 14:42:59 GMT -6
I know there are tools like Auto Align, but I've never learned how one calculates how many samples a track might need to move back to get better phase alignment.
Using my ears and Logic's correlation meter, I can tell when flipping the phase offers better alignment. But I've heard of people using things like the free Voxengo Sound Delay to move things a set number of samples. I've messed with that but it always seems to me such a shot in the dark of how many samples and which track is best to delay.
Mostly looking at drums lately, but I know this must be useful for guitars and other sources too.
I might be overthinking it, but if any of you can point to a source where I can learn more, or have tips, I'd appreciate some good reference points.
|
|
|
Post by christopher on Jun 12, 2019 23:04:12 GMT -6
I’m not sure how it’s done so it would be nice to hear what others do. What I’ve tried quite a few times back when I was new to DAWs was start with overheads, drag the other drum clips to line up with them, zoom all the way in and verify they are in phase not nulls. However the close mics might be now sounding weird or nulling against each other, something to keep in mind. Also I think it’s common to chop up the bass line up the string strikes exactly behind each kick.
|
|
|
Post by Blackdawg on Jun 13, 2019 1:26:08 GMT -6
Fairly simple math if you think about it.
First you need to know the smaller rate. 44.1k is different than 96k in that there is literally 44100 sample in 1 second vs 96,000. Then you just divide that up. There's 1000ms in 1 second. So 96,000/1000=96 samples per 1ms@96kHz.
Now that you know that, delaying 1 sample @96khz is 1/96th a Ms or 0.0104166666 Ms or 0.0000104166666 of a second.
Now knowing how much to delay something like like a snare to your overheads. That can vary a lot and also not always mean it's going to sound better necessarily. There are tools that will help you do this though like Waves InPhase plugin.
I tend to not worry about it and try and get it sounding right with the microphones from the get go. Or if it's a mix I'll check the kick with 180° flip with the overheads and with the snare to see if ones better or not. But otherwise just go with it.
|
|
|
Post by jeremygillespie on Jun 13, 2019 6:07:49 GMT -6
Are you mixing other people’s work or mixing something that you track?
If you’re tracking it, there is no reason to get into the slipping around of tracks. It’s always going to lead to problems.
Best way to get around this is to move microphones and get the polarity and phase right first.
Moving a room mic so the initial transient hits at the exact same time as the snare drum close mic defeats the purpose of even having a room mic IMO. The slight time delay is part of the effect of having a room mic.
For guitars, If you are using more than 1 mic, place the mics. Then push one of the strings down until it hits the pickup and makes a pop noise. Record that sound and then zoom in as far as your daw will allow. That will let you see not only phase, but also alignment. If you’ve got 2 close mics, start moving the microphones independently of one another until they line up perfectly.
|
|
|
Post by swurveman on Jun 13, 2019 7:32:12 GMT -6
For mixing, I use Waves InPhase phase correction plugin. You could have bought it for $29.00 when they had their sale. Worth every penny.
|
|
|
Post by Guitar on Jun 13, 2019 8:25:10 GMT -6
Cubase 10 has an automatic aligning feature. I haven't used it yet, but I do have an application. Mostly for doubled takes like a wall of 3 distortion guitars. I want to try aligning something like that.
I agree with Jeremy about drums, and 2-mic guitar recordings. You really want to spend the extra couple of minutes to get it right before pressing record.
I have a few phase tweak plugins but I don't think I have ever really used them. I thought they would be more useful than they are. It seems more like a "fix it" thing for when you're presented with a bad problem after the recording.
|
|
|
Post by sirthought on Jun 13, 2019 18:06:38 GMT -6
Thanks for the replies. I'm both recording my own and mixing other's tracks. I usually have mics set up to work right, but now and then in the heat of the hunt you move something and forget to check. Also, I've noticed when checking a correlation meter that sometimes after EQ'ing a track, things that were in phase have drifted. It's something I just noticed a couple times, but it was weird.
Anyhow, I may tryout the Waves product, as I'm more concerned with delay between say close mics and overheads, or a mic/DI situation. I've tried zooming in and moving things, but sometimes I just don't feel confident that I've made the correct placement. Watching some demo videos of that plugin shows that sometimes you don't want to shift things as much as correct what frequencies need a tweak to align better.
Just trying to improve the sound before even getting to EQ a track. Cheers
|
|
|
Post by Blackdawg on Jun 13, 2019 18:16:21 GMT -6
Thanks for the replies. I'm both recording my own and mixing other's tracks. I usually have mics set up to work right, but now and then in the heat of the hunt you move something and forget to check. Also, I've noticed when checking a correlation meter that sometimes after EQ'ing a track, things that were in phase have drifted. It's something I just noticed a couple times, but it was weird. Anyhow, I may tryout the Waves product, as I'm more concerned with delay between say close mics and overheads, or a mic/DI situation. I've tried zooming in and moving things, but sometimes I just don't feel confident that I've made the correct placement. Watching some demo videos of that plugin shows that sometimes you don't want to shift things as much as correct what frequencies need a tweak to align better. Just trying to improve the sound before even getting to EQ a track. Cheers That's because EQ uses phase to work. When you are boosting your adding phase. I'm sure svart could explain that much better than I could. But basically EQ is all phase relationships. This why you see the correlation meter shift with EQ.
|
|
|
Post by matt@IAA on Jun 13, 2019 18:47:03 GMT -6
sirthought Remember when you were a kid and sloshed the water back and forth in your tub? Or jumped on a trampoline? EQ is just like that. Every system (mechanical or electrical or whatever) has a certain frequency - period of movement - where it will resonate. In your bathtub, that was the just right "back and forth" speed to make the water tidal waaave: The difficulty to move the tub's worth of water is like impedance. At just the right speed, it's super easy to move the water. Lower or higher speeds and it doesn't have the same effect. In an EQ, at the "right" frequency the components in the circuit are at some kind of low impedance. So the combination of R, C, and/or L cause a resonant peak. This either shorts the audio to ground at that frequency or allows it to pass to be boosted. Either way, the resonance is what makes the filter "work". Changing the R,L,C is like changing the size or amount of water in the tub. The reason there's a phase shift at the resonant point is because when any system goes through resonance or a critical speed, the phase of the system rolls over by 180 degrees. This is a fancy way of saying, if you imagine going on a frequency sweep in that same tub, as you start from a slow speed the water would be moving against you, then with you, then really with you (at resonance) then suddenly you speed up more and the water's against you. Simplistic way to think about it, but close enough. Here's a typical phase and gain with a filter. You can see that as the frequency approaches the filter, the phase begins to change; suddenly it drops off, then the angle changes and rolls out to -180 degrees and stays level. The gain drops off continuously. Every system with a resonant frequency works like this... balancing your tires at the Discount Tire, balancing a steam turbine rotor, even a guitar string.
|
|
|
Post by sirthought on Jun 13, 2019 18:52:37 GMT -6
Good stuff! Thanks
|
|
|
Post by rowmat on Jun 13, 2019 19:22:09 GMT -6
I generally align the drums in the DAW but not the room mics. I will adjust the rooms mics for the best combined tone/low end but won’t align these with the close mics.
|
|
|
Post by johneppstein on Jun 13, 2019 20:38:15 GMT -6
i generally just set up the drum mics so that such problems don't exist. See my numerous previous posts on the subject.
I really don't understand why so many people inmsist on using techniques that cause them problems whenm the soluitions are so simple.
To my detractors (you know who you are) - NO, I am not being condescending. It works. Try it.
EDIT: You don't have to correct it if you don't screw it up in the first place.
|
|
|
Post by svart on Jun 14, 2019 7:05:40 GMT -6
Thanks for the replies. I'm both recording my own and mixing other's tracks. I usually have mics set up to work right, but now and then in the heat of the hunt you move something and forget to check. Also, I've noticed when checking a correlation meter that sometimes after EQ'ing a track, things that were in phase have drifted. It's something I just noticed a couple times, but it was weird. Anyhow, I may tryout the Waves product, as I'm more concerned with delay between say close mics and overheads, or a mic/DI situation. I've tried zooming in and moving things, but sometimes I just don't feel confident that I've made the correct placement. Watching some demo videos of that plugin shows that sometimes you don't want to shift things as much as correct what frequencies need a tweak to align better. Just trying to improve the sound before even getting to EQ a track. Cheers That's because EQ uses phase to work. When you are boosting your adding phase. I'm sure svart could explain that much better than I could. But basically EQ is all phase relationships. This why you see the correlation meter shift with EQ. That's correct. Pretty much every analog EQ uses phase relationships to add or null frequency. Unfortunately that also tells us that frequency is phase sensitive, so when you "adjust phase" you're usually only doing it for a narrow range of frequencies. I've become adept at zooming in and aligning track timing by hand. It's easy enough to figure out with about the same amount of effort as learning a new plugin, and it doesn't require a plugin! For drums, I usually have the drummer do single hits all around the kit so I can get clean samples if needed. I can also use these to see the delay between mics and then can either nudge, or drag until they line up.
|
|
|
Post by mcirish on Jun 14, 2019 8:59:08 GMT -6
In Nuendo, I expand the tracks that I want to align till I can see the start of the waveform. Then, I use the range tool with the display set to samples. I simply select the range between the beginning of one to the beginning of the other. Then I use the free Voxengo Sound Delay to delay the one track. I typically use this to align the in and out kicks to themselves. I also use it to align the snare top mics and bottom mics. I use it to align the overheads so the snare is perfectly in phase in both the overhead channels. I am using a snare hit in the overheads as my zero point to base all delay times on. I can get a much punchier snare once things are aligned. The tom-tom channels are a bit tricky because they are picked up differently from each overhead. I may try to line them up with the overhead closest to them, but often I leave them as I do use a fair amount of gating or cutting of the tom-tom parts. That was always necessary when I was using MD421 on toms. The off-axis bleed of the cymbals into those was horrible sounding. Since moving to ATM230 mics on the toms, I don't have to be so heavy handed. But that's off-topic.
I also have the drummer give me a series of clean hits that I may end up using if there is some bad hit or a stick click in the song that needs to be swapped out.
|
|
|
Post by Guitar on Jun 14, 2019 10:38:57 GMT -6
Cubase 10 has a feature where you can visually superimpose two audio tracks and match them up by eye this way by using the "lower zone."
I haven't tried/learned this feature yet but it seems practical.
|
|
|
Post by christopher on Jun 14, 2019 13:05:32 GMT -6
Recently I setup of some room mics for guitar, and its been sounding great, I'm pretty floored with the sound and very happy about it. The best guitar sound I've gotten in a long time. Decided to zoom all the way and to my surprise I saw the waves are almost perfectly out of phase with the source. Of course. I played around with lining them up, the great sound was lost. I'm keeping it out phase. Its not like they are going to be equal level and null anyway.
|
|
|
Post by Guitar on Jun 14, 2019 13:46:17 GMT -6
Recently I setup of some room mics for guitar, and its been sounding great, I'm pretty floored with the sound and very happy about it. The best guitar sound I've gotten in a long time. Decided to zoom all the way and to my surprise I saw the waves are almost perfectly out of phase with the source. Of course. I played around with lining them up, the great sound was lost. I'm keeping it out phase. Its not like they are going to be equal level and null anyway. Acoustic or electric? Johan Segeborn (one of my favorite YouTube lectric players) says, "Put the headphones on and move the mic around the room till you find a spot that sounds good." That sounds a lot like your method, technicalities be darned.
|
|
|
Post by sirthought on Jun 15, 2019 4:10:44 GMT -6
Yes, christopher, I can agree that with guitars if you think the sound is better with the phase alignment off, it's still the better choice to use that sound! Your ears will decide and most audio rules aren't set in stone. Have you tried to move the mics some more and capture new audio to see if it can be fixed for the next time? I'm sure everyone has encountered this before.
|
|
|
Post by sirthought on Jun 15, 2019 4:16:05 GMT -6
Cubase 10 has a feature where you can visually superimpose two audio tracks and match them up by eye this way by using the "lower zone." I haven't tried/learned this feature yet but it seems practical. That seems like a great tool to be included. I used to use Cubase Studio (wow, maybe back at version 5?) when I had a PC workstation, but moving to Mac I just decided to try Logic and it's stuck for me. I hope they pick up something like this in an update.
|
|
|
Post by christopher on Jun 16, 2019 19:13:55 GMT -6
Recently I setup of some room mics for guitar, and its been sounding great, I'm pretty floored with the sound and very happy about it. The best guitar sound I've gotten in a long time. Decided to zoom all the way and to my surprise I saw the waves are almost perfectly out of phase with the source. Of course. I played around with lining them up, the great sound was lost. I'm keeping it out phase. Its not like they are going to be equal level and null anyway. Acoustic or electric? Johan Segeborn (one of my favorite YouTube lectric players) says, "Put the headphones on and move the mic around the room till you find a spot that sounds good." That sounds a lot like your method, technicalities be darned. I’d have to go back and look. I put the room mics up and they sounded awesome, and tracked a bunch of stuff with them. Electric sounded really nice, but I can’t remember exactly if it was that or the other stuff that was out of phase? I’m pretty sure it was electric. I didn’t really consider the distance changed (phase shifted) between close mics and rooms until you mentioned it.
|
|
|
Post by donr on Jun 16, 2019 19:43:45 GMT -6
sirthought Remember when you were a kid and sloshed the water back and forth in your tub? Or jumped on a trampoline? EQ is just like that. Every system (mechanical or electrical or whatever) has a certain frequency - period of movement - where it will resonate. In your bathtub, that was the just right "back and forth" speed to make the water tidal waaave: The difficulty to move the tub's worth of water is like impedance. At just the right speed, it's super easy to move the water. Lower or higher speeds and it doesn't have the same effect. In an EQ, at the "right" frequency the components in the circuit are at some kind of low impedance. So the combination of R, C, and/or L cause a resonant peak. This either shorts the audio to ground at that frequency or allows it to pass to be boosted. Either way, the resonance is what makes the filter "work". Changing the R,L,C is like changing the size or amount of water in the tub. The reason there's a phase shift at the resonant point is because when any system goes through resonance or a critical speed, the phase of the system rolls over by 180 degrees. This is a fancy way of saying, if you imagine going on a frequency sweep in that same tub, as you start from a slow speed the water would be moving against you, then with you, then really with you (at resonance) then suddenly you speed up more and the water's against you. Simplistic way to think about it, but close enough. Here's a typical phase and gain with a filter. You can see that as the frequency approaches the filter, the phase begins to change; suddenly it drops off, then the angle changes and rolls out to -180 degrees and stays level. The gain drops off continuously. Every system with a resonant frequency works like this... balancing your tires at the Discount Tire, balancing a steam turbine rotor, even a guitar string. dogears, Great analogy and example of wave motion and resonance. I didn't realize wave behavior was universal, nor applicable to analog electronic circuitry, and after your post I think I can grasp the concept as universal. As far as mic'ing and audio, 'does it sound good?' is the criteria that should dominate, no? Phase alignment is more important to reproduction than production, I imagine. When Shelly Yakus (engineer and mixer on "Agents Of Fortune,") put an 87 back a few feet from a guitar amp blended in with the 57 and Beyer 160 close up, he didn't intend to phase align the three mics that went to tape on one track. Not only that, he limited the room mic on the way in. I can see trying moving the time alignment of room mics around in a mix for best sound, but you don't want to align it with the close mics, right? Why would you?
|
|