How we made records in the 90's

[Staticbuster]

Great read, thanks for sharing. Please continue posting, this is super interesting to hear how tings were done in the "major" studios in the pre-digital age. Basically like a Mixerman entry without the sarcasm (or falsifications to protect the not-so-innocent!)

 

[louderock]

Overdubs

We'd often start overdubbing guitars and everything else right after getting a good drum take. It would just vary depending on the band. Sometimes we'd record drums on 4 songs and then start building those tracks and sometimes we'd focus on one song at a time. There were no set rules. I know a lot of producers would approach things like Set up THE drum sound for the album and spend 10 days getting all drum tracks before moving onto bass for a week. We never worked like that. Everything was tailored to the song and drum sounds/ rooms would change up all the time depending on what served the song best. If we tracked some basics on a tune and everyone was really excited and "in the moment", we'd keep going on that song. Sometimes we'd finish a song and go ahead and mix it. It seemed to work best to bounce around so everyone was getting some work in and you don't really want to wait until the end of overdubs and have 12 songs without any vocals on them. That's a head trip for the singer. Best to work on stuff as you go. Plus, it gives everyone a chance to listen and absorb what's going on.

Building up the tracks was probably my favorite part of the process. This is where being a guitarist really came in handy. I've always said that my musician background probably helped me more than anything when it came to making records. What guitar and what amp is the right combination for this part? We had a lot of choices at our disposal and it was really fun digging in and learning what all the combinations sounded like. It's like the color palette for a painter. One very important thing I learned from Brendan is that from the very first downbeat of tracking drums or anything on a song, you should always feel like you're listening to the final record. In other words, there was never any "well later when we mix it...." with regards to the sound. You should always be hearing it out of the speakers as the final record. I feel that's a lost art to a certain extent in the current age. You have so many tracks available now in digital and people often just keep throwing information into their session. It's like data capture that instantly gets a ton of plug ins in order to try to make it into something. Didn't work like that on analog and the goal was always to record the source exactly as intended to work in the final mix. We'd have a good working rough mix on the console and if a guitar part wasn't working, we'd change the guitar, amp, the part....The goal was to have it fit in the mix right then and there. Make decisions. Don't put it off until later.

I have a few friends now that work in Protools and they can never finish anything. They'll track a clean DI signal as well as amp tones but also print the effects on a separate stereo track in case they want to change it all later. Option Anxiety is what I call that. Everything is constantly changing and there are 15 different compressor or EQ plugin choices plus XYZ just released their new plug in and the demo was cool so maybe that would make it all better..... They never finish anything.

For most records we did, everything was on a single 24 track reel. That's 23 tracks of audio and 1 for SMPTE (for console automation). Don't put anything loud or with a big transient on track 23 because it can screw up your SMPTE track 24. If we did need more tracks, I'd just stripe a clean reel on track 24 with SMPTE time code starting at the same location as the Master reel. SMPTE time code sounds like a chirpy 1k tone if you listen to it but it contains a data stream of numbers that allows different machines to sync up. We'd print 30 NDF SMPTE which means there are 30 frames per second. The "frames" actually does refer to frames of video/ film. The code is divided into Hours:Minutes:Seconds:Frames.SubFrames. If we start the time code at 0 on the master, we'd print SMPTE on the 2nd reel starting at 0. There were rack units called Lynx Timeline that you'd feed the SMPTE from each machine (track 24) into. You'd have 2 Lynx Timeline boxes and one would be designated the Master. The other machine would chase the Master and keep the time code locations exactly the same. This is called being "In Sync". Now we'd have a total of 46 tracks available to ruin our song. ; )

Sometimes a second reel was used for additional overdubs and big vocal parts and you'd wind up with a 46 track mix on the console. Other times, you may use this second reel as an overdub reel for takes and comping. Lock up 2 machines and print a stereo mix from your Master reel off the console onto tracks 1 and 2 of the second machine. Now you have a bunch of tracks to do vocal takes or background stacks or strings or ...... Once finished you may bounce a final Vocal Comp back onto your Master reel. Another thing you may do is fly a chorus background vocal stack around. Let's say you have loads of harmony tracks that you add to a chorus. It would be a ton of work to do them on every chorus so you want to fly them around as a stereo bounce instead. Record the parts on chorus 1. Now figure out the time code offset from chorus 1 to chorus 2 and enter that into the Lynx Timeline. In other words the downbeat of chorus 1 and chorus 2 are 25 seconds and 4 frames apart. This is your offset. Confused yet? Now you can fly the chorus 1 vocal into chorus 2.

Another option for flying things around was by sampling. The Eventide H-3000 would let you sample up to 23 seconds if I'm remembering correctly. The TC 2290 and AMS DMX also had samplers in them. Our Studer A80 1/2" tape machine worked as well. When using the H-3000, you'd basically hit the Record button on the sampler on beat 4 of the measure before the audio you wanted to record. You'd then be able to trim your start point a bit. Now go to the spot where you want to fly in your sample and hit Play on the H-3000 on beat 4 and see how you did. You really had to listen to see if you were in time because the tape machine didn't have Grid Mode to tell you that you were correct. This was often a 2 person job since you're using 2 machines at once. These same samplers could also be used in trigger mode which would allow you to add snare samples to your drum mix. It was kind of a pain in the butt. If there was a loop you wanted to add to the song, you'd load it into the sampler and then send a click track on beat one of each measure to trigger the sampler. Obviously this stuff is WAY easier now. We didn't do much of this on the rock records Brendan was making but it would often come up with R&B stuff. The hip hop guys would use their MPC 3000 to fly stuff around.

There were no hard drive backups because we weren't there yet. A SAFETY was made by attaching the ELCO multi pin output connector of your Master machine to the ELCO input of your second machine. You'd make a 1:1 transfer of your master to a second machine. Photo copy the track sheet.

That's basically it for tracking overdubs. Grown children in a room full of toys trying to make stuff sound cool. Each song would vary and each band would be different. It was never a matter of keep recording more and more stuff. We'd basically just work on a song until it sounded DONE. That's a matter of taste and ultimately, aside from the technicalities involved in a record, it's what we were all paid for.

I'll talk about vocals in another post.

 

[jmoose]

Flying stuff around via tape machines & H3000... geez...

My introduction to that was when an overly excited artist was dancing around the control room and brought his arm down on the MTR90 autolocator... wiped about about 10 seconds of audio across about 15 tracks. In overdub mode, horns, and all of sudden there was a whole lotta silence.

I turned and saw a bunch of lights on the deck showing over half the tracks in record so I hit stop. We all froze... instant vibe killer! haha. Said I think we should move onto another song. Yes... good idea. Time for a smoke break.

Wasn't a great day in the studio. Whole reason I was using the MTR90 in the first place was because the A800 freaked out and I had to switch machines. Was nothing major, I think it lost rewind or something but had to swap.

When the boss man came in a couple days later I explained what happened and we put up the reel with the tracks that ended up on bias beach. We realized it was a chorus that got wiped out and thankfully we cut to a click. So that started the process of flying tracks around to "fix" and replace the missing audio. Took a whole day and yes, the artist the was billed for it.

With analog tape there is no undo. Only redo.

 

[louderock]

Flying stuff around via tape machines & H3000... geez...

My introduction to that was when an overly excited artist was dancing around the control room and brought his arm down on the MTR90 autolocator... wiped about about 10 seconds of audio across about 15 tracks. In overdub mode, horns, and all of sudden there was a whole lotta silence.

I turned and saw a bunch of lights on the deck showing over half the tracks in record so I hit stop. We all froze... instant vibe killer! haha. Said I think we should move onto another song. Yes... good idea. Time for a smoke break.

Wasn't a great day in the studio. Whole reason I was using the MTR90 in the first place was because the A800 freaked out and I had to switch machines. Was nothing major, I think it lost rewind or something but had to swap.

When the boss man came in a couple days later I explained what happened and we put up the reel with the tracks that ended up on bias beach. We realized it was a chorus that got wiped out and thankfully we cut to a click. So that started the process of flying tracks around to "fix" and replace the missing audio. Took a whole day and yes, the artist the was billed for it.

With analog tape there is no undo. Only redo.

Sounds like an awesome day. Pretty sure I heard a story about Joe Walsh spilling a beer all over the 2 track reel of Rocky Mountain Way or one of his songs and they had the tape unrolled up and down the hallway to dry out.

Reminds me of a fun trick. Secretly mult the snare drum to channel at the end of the console. Flip that channel out of phase and keep it muted. Wait until the engineer's back is to you and he's punching in on a track. Un-mute the snare mult channel and he thinks he recorded over the snare because it disappears. Not that I'd ever do something like that.

 

[David MacNeill]

Thought it would be cool to share some of my experiences from making records in the 90's.

This is so fun to read. I remember the pre-Pro Tools era well. Thanks for the memories!

 

[drfrankencopter]

This thread is awesome!

I’m curious to learn more about that drum approach. Would you leave the kit an mic’s set up after doing 4 songs, or would you pack it up and move on to doing overdubs in the big room without the kit ringing away?

If you packed the kit away how different was the sound when you moved onto the next batch of songs? I was only doing demos in the 90s, so 4 songs was the whole scope of the session...we’d use the same drum setup throughout. We were also broke, so we couldn’t afford enough tape to record more than one take of each song (and this was at 15 ips...yeah, tape was $250 roll here). That helped me learn to not fear making decisions.

 

[drfrankencopter]

The talk of accidentally erasing tracks reminds me of the thrill and skill of punching in (and out) by hand. Especially with vocals...when you’re running out of tracks, and need to punch in a word or phrase it was pretty stressful. We’ve got it so easy now...auto punch and loop, and then comp.

 

[louderock]

This thread is awesome!

I’m curious to learn more about that drum approach. Would you leave the kit an mic’s set up after doing 4 songs, or would you pack it up and move on to doing overdubs in the big room without the kit ringing away?

If you packed the kit away how different was the sound when you moved onto the next batch of songs? I was only doing demos in the 90s, so 4 songs was the whole scope of the session...we’d use the same drum setup throughout. We were also broke, so we couldn’t afford enough tape to record more than one take of each song (and this was at 15 ips...yeah, tape was $250 roll here). That helped me learn to not fear making decisions.

We'd leave the kit set up in the drum room. We had plenty of mics and preamps so that nothing really had to be shared. We'd just mic up guitars in a separate area and start tracking away. Once we decided to track basics on another song, we'd just swap out any drums as needed and already have the mics in place. It got a bit more complicated if we decided to move the drums to a small tight room. We got good at it though and were quick.

Most drums at Southern Tracks were recorded here:

You can see the studio layout here:

We'd usually set up amps just outside the left of the control room for overdubs and the player would sit next to the engineer listening to speakers instead of being on phones out in the tracking room. The vocal room had a really nice tight sound and we'd often record drums in there for a less roomy 70's kinda sound.

 

[El Phaco]

Great thread ! Thanks so much for sharing !

 

[Ed DeGenaro]

Overdubs

We'd often start overdubbing guitars and everything else right after getting a good drum take. It would just vary depending on the band. Sometimes we'd record drums on 4 songs and then start building those tracks and sometimes we'd focus on one song at a time. There were no set rules. I know a lot of producers would approach things like Set up THE drum sound for the album and spend 10 days getting all drum tracks before moving onto bass for a week. We never worked like that. Everything was tailored to the song and drum sounds/ rooms would change up all the time depending on what served the song best. If we tracked some basics on a tune and everyone was really excited and "in the moment", we'd keep going on that song. Sometimes we'd finish a song and go ahead and mix it. It seemed to work best to bounce around so everyone was getting some work in and you don't really want to wait until the end of overdubs and have 12 songs without any vocals on them. That's a head trip for the singer. Best to work on stuff as you go. Plus, it gives everyone a chance to listen and absorb what's going on.

Building up the tracks was probably my favorite part of the process. This is where being a guitarist really came in handy. I've always said that my musician background probably helped me more than anything when it came to making records. What guitar and what amp is the right combination for this part? We had a lot of choices at our disposal and it was really fun digging in and learning what all the combinations sounded like. It's like the color palette for a painter. One very important thing I learned from Brendan is that from the very first downbeat of tracking drums or anything on a song, you should always feel like you're listening to the final record. In other words, there was never any "well later when we mix it...." with regards to the sound. You should always be hearing it out of the speakers as the final record. I feel that's a lost art to a certain extent in the current age. You have so many tracks available now in digital and people often just keep throwing information into their session. It's like data capture that instantly gets a ton of plug ins in order to try to make it into something. Didn't work like that on analog and the goal was always to record the source exactly as intended to work in the final mix. We'd have a good working rough mix on the console and if a guitar part wasn't working, we'd change the guitar, amp, the part....The goal was to have it fit in the mix right then and there. Make decisions. Don't put it off until later.

I have a few friends now that work in Protools and they can never finish anything. They'll track a clean DI signal as well as amp tones but also print the effects on a separate stereo track in case they want to change it all later. Option Anxiety is what I call that. Everything is constantly changing and there are 15 different compressor or EQ plugin choices plus XYZ just released their new plug in and the demo was cool so maybe that would make it all better..... They never finish anything.

For most records we did, everything was on a single 24 track reel. That's 23 tracks of audio and 1 for SMPTE (for console automation). Don't put anything loud or with a big transient on track 23 because it can screw up your SMPTE track 24. If we did need more tracks, I'd just stripe a clean reel on track 24 with SMPTE time code starting at the same location as the Master reel. SMPTE time code sounds like a chirpy 1k tone if you listen to it but it contains a data stream of numbers that allows different machines to sync up. We'd print 30 NDF SMPTE which means there are 30 frames per second. The "frames" actually does refer to frames of video/ film. The code is divided into Hours:Minutes:Seconds:Frames.SubFrames. If we start the time code at 0 on the master, we'd print SMPTE on the 2nd reel starting at 0. There were rack units called Lynx Timeline that you'd feed the SMPTE from each machine (track 24) into. You'd have 2 Lynx Timeline boxes and one would be designated the Master. The other machine would chase the Master and keep the time code locations exactly the same. This is called being "In Sync". Now we'd have a total of 46 tracks available to ruin our song. ; )

Sometimes a second reel was used for additional overdubs and big vocal parts and you'd wind up with a 46 track mix on the console. Other times, you may use this second reel as an overdub reel for takes and comping. Lock up 2 machines and print a stereo mix from your Master reel off the console onto tracks 1 and 2 of the second machine. Now you have a bunch of tracks to do vocal takes or background stacks or strings or ...... Once finished you may bounce a final Vocal Comp back onto your Master reel. Another thing you may do is fly a chorus background vocal stack around. Let's say you have loads of harmony tracks that you add to a chorus. It would be a ton of work to do them on every chorus so you want to fly them around as a stereo bounce instead. Record the parts on chorus 1. Now figure out the time code offset from chorus 1 to chorus 2 and enter that into the Lynx Timeline. In other words the downbeat of chorus 1 and chorus 2 are 25 seconds and 4 frames apart. This is your offset. Confused yet? Now you can fly the chorus 1 vocal into chorus 2.

Another option for flying things around was by sampling. The Eventide H-3000 would let you sample up to 23 seconds if I'm remembering correctly. The TC 2290 and AMS DMX also had samplers in them. Our Studer A80 1/2" tape machine worked as well. When using the H-3000, you'd basically hit the Record button on the sampler on beat 4 of the measure before the audio you wanted to record. You'd then be able to trim your start point a bit. Now go to the spot where you want to fly in your sample and hit Play on the H-3000 on beat 4 and see how you did. You really had to listen to see if you were in time because the tape machine didn't have Grid Mode to tell you that you were correct. This was often a 2 person job since you're using 2 machines at once. These same samplers could also be used in trigger mode which would allow you to add snare samples to your drum mix. It was kind of a pain in the butt. If there was a loop you wanted to add to the song, you'd load it into the sampler and then send a click track on beat one of each measure to trigger the sampler. Obviously this stuff is WAY easier now. We didn't do much of this on the rock records Brendan was making but it would often come up with R&B stuff. The hip hop guys would use their MPC 3000 to fly stuff around.

There were no hard drive backups because we weren't there yet. A SAFETY was made by attaching the ELCO multi pin output connector of your Master machine to the ELCO input of your second machine. You'd make a 1:1 transfer of your master to a second machine. Photo copy the track sheet.

That's basically it for tracking overdubs. Grown children in a room full of toys trying to make stuff sound cool. Each song would vary and each band would be different. It was never a matter of keep recording more and more stuff. We'd basically just work on a song until it sounded DONE. That's a matter of taste and ultimately, aside from the technicalities involved in a record, it's what we were all paid for.

I'll talk about vocals in another post.

We overused the H3000 as sampler way more than we should, this was in 1988.

 

[OctalSocket]

Question about the drum recording: you say that most were done in the drum booth, which by the looks of it, is live-sounding, but controlled. Did you ever end up opening that glass door up, and micing the large room for ambience?

I realize this works best if there isn’t a bunch of amps blaring away out there.

 

[louderock]

Question about the drum recording: you say that most were done in the drum booth, which by the looks of it, is live-sounding, but controlled. Did you ever end up opening that glass door up, and micing the large room for ambience?

I realize this works best if there isn’t a bunch of amps blaring away out there.

Yes we'd occasionally open the doors and try additional room mics. The larger room sounded great for drums, too. We used pretty much every space at some point or another depending on the sound we were going for. That drum booth had something special though. It was the right amount of liveliness without losing the punch. Those back panels were on big hinges and could be folded over to either have a wood side or a padded side. The drums on STP's Purple are a good example. Most of the ambience came from the room and not from reverb.

 

[louderock]

We overused the H3000 as sampler way more than we should, this was in 1988.

I'd often keep ours loaded with applause or a group of kids yelling YAAAAAYYYY that I could trigger when something good happened.

I got tweaky one day and figured out how to use ours for a bit of vocal help. I attached a keyboard to the Midi port and set up a program for basic pitch shift. I then set up the pitch bend wheel of the keyboard to control the pitch of the H-3000 + or - 20 cents. If we had a vocal line that we wanted to used that just wasn't quite there, I'd patch that across the buss and it would let us "play" the pitch adjustment as we were comping. Aside from that it seems everyone had those things parked on program 533.

 

[Ed DeGenaro]

I'd often keep ours loaded with applause or a group of kids yelling YAAAAAYYYY that I could trigger when something good happened.

I got tweaky one day and figured out how to use ours for a bit of vocal help. I attached a keyboard to the Midi port and set up a program for basic pitch shift. I then set up the pitch bend wheel of the keyboard to control the pitch of the H-3000 + or - 20 cents. If we had a vocal line that we wanted to used that just wasn't quite there, I'd patch that across the buss and it would let us "play" the pitch adjustment as we were comping. Aside from that it seems everyone had those things parked on program 533.

The doubler? Yeah, funnily enough I still prefer it's predecessors for that.

 

[strattele335]

Love this thread! Thanks for sharing Louderock!
If I may ask a question, when you were getting started, what was the thing that was hardest to learn and when did you "get it"? Like recording certain instruments, or maybe a mixing thing?

Craig

 

[louderock]

Love this thread! Thanks for sharing Louderock!
If I may ask a question, when you were getting started, what was the thing that was hardest to learn and when did you "get it"? Like recording certain instruments, or maybe a mixing thing?

Craig

What's interesting is how different it was to watch it all happening in the studio and feeling like you have a good grip on it all versus being the one in the "hot seat" as we called it and having to run a session. There was a lot of I hope I'm doing this right at first but I guess that's true with anything. It becomes second nature after a while.

Recording drums was probably the most difficult thing starting out. There were 11 mics and phasing concerns plus tuning the drums.... Lots of details.

 

[louderock]

Drum Stuff

We had a Noble & Cooley kit that sounded great and would just swap out snares and cymbals according to the song. That kit was on a lot of records. We also had an old Slingerland Radio King kit from the 40's with a 26" kick. Also a 60's Ludwig kit. Drummers would sometimes bring their stuff and we'd mix and match. Here's the typical routing for recording.

Kick - 421 in/ AKG D30 out. Both to SSL pres (really quick and punchy) bussed together through the eq side of the ADR Vocal Stressor to track 1.

Snr - 57 and AKG 451 taped together so the capsules aligned on top. 57 under. SSL Pres. All bussed together through a Pultec EQP-1 to track 2.

Hat - 451 or KM84. SSL pre. Track 3

Toms - 421. Later 421 top 57 bottom. I wired a XY cable with the polarity reversed for the 57. Both mics fed into a single pre through one cable. SSL pres. If there were more than 2 toms, they'd be bussed together so everything was recorded on tracks 4 & 5.

OH - KM184 - Neve 1073 pres - Summit DCL200 compressor to tracks 6 & 7.

Room - Coles 4038 - Neve 1073 pres - Distressors to tracks 8 & 9.

If we needed a ride mic, we'd mix it into the OH buss. Sometimes we'd use more Neve preamps but the SSL pres were used more often for the actual drum shells/ close mics. Later on as we incorporated the 3348 machine into our setup, we may split things out just a touch more. Drums were always tracked to analog 456 tape and we'd hit the kick, snare, toms channels pretty hard. 456 saturates and compresses really well. I never really liked 499.

 

[Ed DeGenaro]

Drum Stuff

We had a Noble & Cooley kit that sounded great and would just swap out snares and cymbals according to the song. That kit was on a lot of records. We also had an old Slingerland Radio King kit from the 40's with a 26" kick. Also a 60's Ludwig kit. Drummers would sometimes bring their stuff and we'd mix and match. Here's the typical routing for recording.

Kick - 421 in/ AKG D30 out. Both to SSL pres (really quick and punchy) bussed together through the eq side of the ADR Vocal Stressor to track 1.

Snr - 57 and AKG 451 taped together so the capsules aligned on top. 57 under. SSL Pres. All bussed together through a Pultec EQP-1 to track 2.

Hat - 451 or KM84. SSL pre. Track 3

Toms - 421. Later 421 top 57 bottom. I wired a XY cable with the polarity reversed for the 57. Both mics fed into a single pre through one cable. SSL pres. If there were more than 2 toms, they'd be bussed together so everything was recorded on tracks 4 & 5.

OH - KM184 - Neve 1073 pres - Summit DCL200 compressor to tracks 6 & 7.

Room - Coles 4038 - Neve 1073 pres - Distressors to tracks 8 & 9.

If we needed a ride mic, we'd mix it into the OH buss. Sometimes we'd use more Neve preamps but the SSL pres were used more often for the actual drum shells/ close mics. Later on as we incorporated the 3348 machine into our setup, we may split things out just a touch more. Drums were always tracked to analog 456 tape and we'd hit the kick, snare, toms channels pretty hard. 456 saturates and compresses really well. I never really liked 499.

Big fan of the stressor.

 

[drfrankencopter]

Big fan of the stressor.

The mention of the vocal stressor stood out to me. It seems that the only AD&R piece people talk about is the Compex...I never see any discussion around the vocal stressor.

In the drum application was the stressor being used as a compressor as well as an EQ?

 

[louderock]

The mention of the vocal stressor stood out to me. It seems that the only AD&R piece people talk about is the Compex...I never see any discussion around the vocal stressor.

In the drum application was the stressor being used as a compressor as well as an EQ?

You can patch into each side individually and we would only use the Eq side. The compressor side is very good for room mics and vocals of course.