Thursday, February 28, 2008
AutoTune RTAS Not TDM
Tool Thursday w/ Jed Hackett.
This may already be common knowledge but just incase you haven’t heard of this, here goes. Use AutoTune in RTAS mode, not TDM. Enough said... see you next week.
O.K. I’ll explain it. AutoTune use to track really well in TDM mode, but right around the release of AutoTune 3 they did something that although sonically gave a better result, it messed with the tracking and all of a sudden AutoTune wouldn’t track worth a darn in TDM mode. I don’t know what they did but it was really frustrating. I found that it was particularly bad with vocals cut at low levels. I nearly gave up on Antares until a friend of mine discovered that AutoTune would track much better in RTAS mode. So I switched to using AutoTune in RTAS mode and it tracks like a charm. I’ve been using it that way ever since.
The only downfall with this is that even with delay compensation On, you still get a 2 to 3 millisecond delay and it will vary roughly 1 millisecond either direction depending on how much processing Autotune is having to do to correct the pitch. I don’t know for sure, but that delay may change depending on how fast your cpu is... and for those of you on LE, the delay will be different as well, so double check the position of the tuned track relative to the original track when you are done. The thing is, to me, I would rather nudge the final tuned track forward 2 milliseconds than deal with poor tracking. And for the most part, there are few singers that couldn’t use a little nudge back in the pocket. The end result with the delay, once you nudge the tuned track forward, is an unbelievably small fluctuation of 0 - 1 milliseconds on the tuned vocal track.
I know I am going to once again spark the Melodyne vs. WavesTune vs. AutoTune debate with this article but so be it, everyone has their own valid opinion. I tune with AutoTune RTAS in graphic mode and get great transparent results. If you use something else and like it then don’t change... but if you’re using AutoTune, I highly recommend you switch to RTAS mode.
‘til next time… take care,
Jed
This may already be common knowledge but just incase you haven’t heard of this, here goes. Use AutoTune in RTAS mode, not TDM. Enough said... see you next week.
O.K. I’ll explain it. AutoTune use to track really well in TDM mode, but right around the release of AutoTune 3 they did something that although sonically gave a better result, it messed with the tracking and all of a sudden AutoTune wouldn’t track worth a darn in TDM mode. I don’t know what they did but it was really frustrating. I found that it was particularly bad with vocals cut at low levels. I nearly gave up on Antares until a friend of mine discovered that AutoTune would track much better in RTAS mode. So I switched to using AutoTune in RTAS mode and it tracks like a charm. I’ve been using it that way ever since.
The only downfall with this is that even with delay compensation On, you still get a 2 to 3 millisecond delay and it will vary roughly 1 millisecond either direction depending on how much processing Autotune is having to do to correct the pitch. I don’t know for sure, but that delay may change depending on how fast your cpu is... and for those of you on LE, the delay will be different as well, so double check the position of the tuned track relative to the original track when you are done. The thing is, to me, I would rather nudge the final tuned track forward 2 milliseconds than deal with poor tracking. And for the most part, there are few singers that couldn’t use a little nudge back in the pocket. The end result with the delay, once you nudge the tuned track forward, is an unbelievably small fluctuation of 0 - 1 milliseconds on the tuned vocal track.
I know I am going to once again spark the Melodyne vs. WavesTune vs. AutoTune debate with this article but so be it, everyone has their own valid opinion. I tune with AutoTune RTAS in graphic mode and get great transparent results. If you use something else and like it then don’t change... but if you’re using AutoTune, I highly recommend you switch to RTAS mode.
‘til next time… take care,
Jed
AutoTune RTAS Not TDM
Tool Thursday w/ Jed Hackett.
This may already be common knowledge but just incase you haven’t heard of this, here goes. Use AutoTune in RTAS mode, not TDM. Enough said... see you next week.
O.K. I’ll explain it. AutoTune use to track really well in TDM mode, but right around the release of AutoTune 3 they did something that although sonically gave a better result, it messed with the tracking and all of a sudden AutoTune wouldn’t track worth a darn in TDM mode. I don’t know what they did but it was really frustrating. I found that it was particularly bad with vocals cut at low levels. I nearly gave up on Antares until a friend of mine discovered that AutoTune would track much better in RTAS mode. So I switched to using AutoTune in RTAS mode and it tracks like a charm. I’ve been using it that way ever since.
The only downfall with this is that even with delay compensation On, you still get a 2 to 3 millisecond delay and it will vary roughly 1 millisecond either direction depending on how much processing Autotune is having to do to correct the pitch. I don’t know for sure, but that delay may change depending on how fast your cpu is... and for those of you on LE, the delay will be different as well, so double check the position of the tuned track relative to the original track when you are done. The thing is, to me, I would rather nudge the final tuned track forward 2 milliseconds than deal with poor tracking. And for the most part, there are few singers that couldn’t use a little nudge back in the pocket. The end result with the delay, once you nudge the tuned track forward, is an unbelievably small fluctuation of 0 - 1 milliseconds on the tuned vocal track.
I know I am going to once again spark the Melodyne vs. WavesTune vs. AutoTune debate with this article but so be it, everyone has their own valid opinion. I tune with AutoTune RTAS in graphic mode and get great transparent results. If you use something else and like it then don’t change... but if you’re using AutoTune, I highly recommend you switch to RTAS mode.
‘til next time… take care,
Jed
This may already be common knowledge but just incase you haven’t heard of this, here goes. Use AutoTune in RTAS mode, not TDM. Enough said... see you next week.
O.K. I’ll explain it. AutoTune use to track really well in TDM mode, but right around the release of AutoTune 3 they did something that although sonically gave a better result, it messed with the tracking and all of a sudden AutoTune wouldn’t track worth a darn in TDM mode. I don’t know what they did but it was really frustrating. I found that it was particularly bad with vocals cut at low levels. I nearly gave up on Antares until a friend of mine discovered that AutoTune would track much better in RTAS mode. So I switched to using AutoTune in RTAS mode and it tracks like a charm. I’ve been using it that way ever since.
The only downfall with this is that even with delay compensation On, you still get a 2 to 3 millisecond delay and it will vary roughly 1 millisecond either direction depending on how much processing Autotune is having to do to correct the pitch. I don’t know for sure, but that delay may change depending on how fast your cpu is... and for those of you on LE, the delay will be different as well, so double check the position of the tuned track relative to the original track when you are done. The thing is, to me, I would rather nudge the final tuned track forward 2 milliseconds than deal with poor tracking. And for the most part, there are few singers that couldn’t use a little nudge back in the pocket. The end result with the delay, once you nudge the tuned track forward, is an unbelievably small fluctuation of 0 - 1 milliseconds on the tuned vocal track.
I know I am going to once again spark the Melodyne vs. WavesTune vs. AutoTune debate with this article but so be it, everyone has their own valid opinion. I tune with AutoTune RTAS in graphic mode and get great transparent results. If you use something else and like it then don’t change... but if you’re using AutoTune, I highly recommend you switch to RTAS mode.
‘til next time… take care,
Jed
Thursday, February 21, 2008
The Grid Ain’t Right!
Tool Thursday w/ Jed Hackett.
I’m often really glad that with the coming of Pro Tools we are now able to see what is going on, not just hear it. The grid is an invaluable tool both with time and pitch. I use to dread flying backgrounds from one chorus to another, either with a sampler or a 1/2” machine. Thank god for Pro Tools and the grid! But the fact that we can now see the grid has brought about it’s overuse.
I get lots of people who tell me I should give beat detective another try. That it is much improved and it could save me a lot of time lining up drums. They know I sit and edit the song manually and they just want to help save me some time. I usually just nod and politely say thanks because the advice is coming from good friends who are also great engineers and I don’t wish to offend them. The problem is, the grid ain’t right.
The classical world calls it rubato, the pop world calls it groove... and it doesn’t live on the grid. There is a reason that the drummer is laying back the high hat in between the kick and snare, or rushing a fill, or pulling a fill back, or laying the snare back within the groove. There is a reason that the drums have pulled back while the guitars push forward in certain heavy grooves. There is a reason the piano is pushing and pulling in a pop ballad. An alto sax just doesn’t sound the same when it isn’t just a bit sharp. Etc. ad nauseum!
There are plenty of times when the performance can be spruced up a bit with some editing and tuning, and that’s how I make my living. However, you have to be careful that you only move those parts of a performance that detract from the whole; that you are correcting mistakes and not vibe. When we go through a song chopping the drums at every note, then quantizing them, we very often lose the groove. Overall the track may settle a bit because it now has consistency, but you’ll notice something is missing. Usually people will just deal with that lose because they just can’t take the mistakes in the original performance. I’d like to give you a few things to think about that may help you keep the musicality in a performance while editing.
1) Before you start to edit, listen to the song as a whole and get a feel for what is going on, the vibe of the track
2) Start with the instrument that is the driving force in the groove. It may not be the drums, it could be the acoustic or the piano. Go through that track first and fix what bothers you. All the while keeping in mind the overall sense for the groove that you first got when you listened to the whole track. Pay close attention with your ears, not your eyes. Use the grid only to help you verify what your ears are hearing. If the snare is right on the grid but it feels rushed, then it is rushed so pull it back. It was an eye opening experience when a producer convinced me of that. He was insisting that the snare was early and I was insisting that it was dead on the grid. I pulled it back for him and darn if he wasn’t right! That was the day I stopped worshiping at the alter of the grid.
3) Once you have that key track tweaked so that it feels really good, then go to the next key element, and so on and so forth. All the while only using the grid to help you verify what you are hearing, not necessarily to place everything on the grid.
4) When you get to the end of things, go back to the original playlists on everything and A/B it with what you did. Make sure what you have done is an improvement. If not, figure out why and make the changes.
Here are some more specific tips that can be useful to you:
1) One really helpful hint is to find a part of the song where things feel really good and zoom in, analyze what is going on and why it feels so good. If you notice a pattern in how a particular instrument was played, mimic that feel in the rest of the song where appropriate. For example, if you find that the drummer is 3 or 6 milliseconds back on the high hat in relation to the rest of the kit and you have figured out that is the reason you like those couple of bars so much, then take that in to consideration when you are tweaking the drum track and where appropriate mimic that feel.
2) Sometimes on a shuffle feel everyone has a different idea as to how far to swing, so I usually find an instrument that has the groove nailed and measure the swing on the upbeats. If I find a consistent pattern, I will use that amount of swing when I find an instrument that just didn’t get the groove and I have to fix it.
3) If you need to tighten a track, you can get away with clamping down more on drums, bass and acoustic. Then if you let the electrics, keys and vocals float more, you’ll get the stability you want in a track yet maintain the raw energy.
4) If in doubt, always tweak less. I always err on the side of less and if at the end I find out I need to go farther, then it is really easy to do so. Once you go there, it is hard to go back. If you over tighten you’ll end up starting over from the original track. And that’s no fun.
Well I think this article is long enough and it’s too nice out today not to go running. Talk to you soon.
Take care,
Jed
I’m often really glad that with the coming of Pro Tools we are now able to see what is going on, not just hear it. The grid is an invaluable tool both with time and pitch. I use to dread flying backgrounds from one chorus to another, either with a sampler or a 1/2” machine. Thank god for Pro Tools and the grid! But the fact that we can now see the grid has brought about it’s overuse.
I get lots of people who tell me I should give beat detective another try. That it is much improved and it could save me a lot of time lining up drums. They know I sit and edit the song manually and they just want to help save me some time. I usually just nod and politely say thanks because the advice is coming from good friends who are also great engineers and I don’t wish to offend them. The problem is, the grid ain’t right.
The classical world calls it rubato, the pop world calls it groove... and it doesn’t live on the grid. There is a reason that the drummer is laying back the high hat in between the kick and snare, or rushing a fill, or pulling a fill back, or laying the snare back within the groove. There is a reason that the drums have pulled back while the guitars push forward in certain heavy grooves. There is a reason the piano is pushing and pulling in a pop ballad. An alto sax just doesn’t sound the same when it isn’t just a bit sharp. Etc. ad nauseum!
There are plenty of times when the performance can be spruced up a bit with some editing and tuning, and that’s how I make my living. However, you have to be careful that you only move those parts of a performance that detract from the whole; that you are correcting mistakes and not vibe. When we go through a song chopping the drums at every note, then quantizing them, we very often lose the groove. Overall the track may settle a bit because it now has consistency, but you’ll notice something is missing. Usually people will just deal with that lose because they just can’t take the mistakes in the original performance. I’d like to give you a few things to think about that may help you keep the musicality in a performance while editing.
1) Before you start to edit, listen to the song as a whole and get a feel for what is going on, the vibe of the track
2) Start with the instrument that is the driving force in the groove. It may not be the drums, it could be the acoustic or the piano. Go through that track first and fix what bothers you. All the while keeping in mind the overall sense for the groove that you first got when you listened to the whole track. Pay close attention with your ears, not your eyes. Use the grid only to help you verify what your ears are hearing. If the snare is right on the grid but it feels rushed, then it is rushed so pull it back. It was an eye opening experience when a producer convinced me of that. He was insisting that the snare was early and I was insisting that it was dead on the grid. I pulled it back for him and darn if he wasn’t right! That was the day I stopped worshiping at the alter of the grid.
3) Once you have that key track tweaked so that it feels really good, then go to the next key element, and so on and so forth. All the while only using the grid to help you verify what you are hearing, not necessarily to place everything on the grid.
4) When you get to the end of things, go back to the original playlists on everything and A/B it with what you did. Make sure what you have done is an improvement. If not, figure out why and make the changes.
Here are some more specific tips that can be useful to you:
1) One really helpful hint is to find a part of the song where things feel really good and zoom in, analyze what is going on and why it feels so good. If you notice a pattern in how a particular instrument was played, mimic that feel in the rest of the song where appropriate. For example, if you find that the drummer is 3 or 6 milliseconds back on the high hat in relation to the rest of the kit and you have figured out that is the reason you like those couple of bars so much, then take that in to consideration when you are tweaking the drum track and where appropriate mimic that feel.
2) Sometimes on a shuffle feel everyone has a different idea as to how far to swing, so I usually find an instrument that has the groove nailed and measure the swing on the upbeats. If I find a consistent pattern, I will use that amount of swing when I find an instrument that just didn’t get the groove and I have to fix it.
3) If you need to tighten a track, you can get away with clamping down more on drums, bass and acoustic. Then if you let the electrics, keys and vocals float more, you’ll get the stability you want in a track yet maintain the raw energy.
4) If in doubt, always tweak less. I always err on the side of less and if at the end I find out I need to go farther, then it is really easy to do so. Once you go there, it is hard to go back. If you over tighten you’ll end up starting over from the original track. And that’s no fun.
Well I think this article is long enough and it’s too nice out today not to go running. Talk to you soon.
Take care,
Jed
Thursday, February 14, 2008
Knowing When To Make A U-Turn
Tool Thursday with Jed Hackett.
I just recently edited a track for a client... drums, bass, guitars, keys, vocals, etc. As I worked, everything seemed to be falling into place quite nicely as usual. However, when I finished with the track something just wasn’t settling with me about it. Usually when I get that settled feeling I know everything is right. With this track I just couldn’t get to that point. Each part individually seemed to groove but the whole picture just seemed a bit off. When I went back to the original playlist I realized that even though it had problems, the underlying groove felt better to me. I knew I couldn’t just leave the track in it’s original state because it did have a jerky feeling to it that the producer wanted to fix, but there was something that had gotten lost in all of my tweaking.
I started comparing the two versions closely and decided to switch back to the original playlists on the electrics in the choruses, let some of the lead instruments and vocals rush a bit more and pushed some of drums fills. I ended up with a compromise between the original and the tweaked track. Once I did that I finally got that settled feeling and knew it was right.
When we dive into editing, or tuning for that matter, it is easy to lose perspective, even when we are aware that it can happen. When I finish a track I always go back to the original playlists and compare it to my tweaked version just to make sure I did the song justice. I usually find a few things that I like more about the original and incorporate them into the final product. I think the hardest thing to do as an engineer is to know when to turn around. There are plenty of times when we can use technology to enhance the music we work on, but we have to possess enough humility to know when we’re just messing things up.
See you next week,
Jed
I just recently edited a track for a client... drums, bass, guitars, keys, vocals, etc. As I worked, everything seemed to be falling into place quite nicely as usual. However, when I finished with the track something just wasn’t settling with me about it. Usually when I get that settled feeling I know everything is right. With this track I just couldn’t get to that point. Each part individually seemed to groove but the whole picture just seemed a bit off. When I went back to the original playlist I realized that even though it had problems, the underlying groove felt better to me. I knew I couldn’t just leave the track in it’s original state because it did have a jerky feeling to it that the producer wanted to fix, but there was something that had gotten lost in all of my tweaking.
I started comparing the two versions closely and decided to switch back to the original playlists on the electrics in the choruses, let some of the lead instruments and vocals rush a bit more and pushed some of drums fills. I ended up with a compromise between the original and the tweaked track. Once I did that I finally got that settled feeling and knew it was right.
When we dive into editing, or tuning for that matter, it is easy to lose perspective, even when we are aware that it can happen. When I finish a track I always go back to the original playlists and compare it to my tweaked version just to make sure I did the song justice. I usually find a few things that I like more about the original and incorporate them into the final product. I think the hardest thing to do as an engineer is to know when to turn around. There are plenty of times when we can use technology to enhance the music we work on, but we have to possess enough humility to know when we’re just messing things up.
See you next week,
Jed
Subscribe to:
Posts (Atom)