Any advice for live MIDI recording?

[Tusker]

I play in a neighborhood community church. I'd like your advice. We have been recording our worship services directly into a DAW that the sound person controls. This person is located about 20-30 yards from the stage (I'm bad at distances). We have the usual problem with the live recordings and we hit upon the idea of recording MIDI as well, so we could do timbral substitution after the fact. All our core instruments are going mono individually into the DAW. This would enable us to go back and get stereo effects, better balancing of instruments, support for weak sounds, etc. At least we think so. It seems we are pretty well situated to get the core of the worhip band recorded to MIDI. There are two keyboardists with usual MIDI outs, the two guitarists are using Roland VG and GR guitar synths with MIDI outs and our drummer is using V-drums. If we give the bassist a Roland processor, we could record everyone to MIDI except horns, voices and percussion. (No, I don't have any Roland stock.) But it's an idea I'd like to pursue. Firstly I am wondering if timbral substitution is a false promise. Secondly I was wondering what issues we would confront if we did this. I was wondering if anyone has advice for me. I am aware of one potential issue: MIDI apparently doesn't travel more than 20 feet or so, so one question would be whether there is a MIDI amplifier box to amplify the signals for longer travel. We would prefer to keep the DAW at the back of the room. Thanks in advance, Jerry

[Salyphus]

I think as long as you continue to record the audio, recording the MIDI only gives you more options later and that's usually a good thing. And it doesn't take up much data, so why not? I've never heard of any 20' limitation for MIDI. I'm pretty sure you can get MIDI cables longer than that. I'd give it a try and see what happens.

[not Cereal]

i use midi everyday and have since about 1994. i can tell you it is a highly obsolete and tedious medium. it's a 7bit system running @1mhz the biggest problem with midi is that it can only transmit ONE peice of information at a time. since this happens at a 1mhz transmission speed, the effects can become noticable quickly. example: you play a solid chord on a keyboard while recording the midi into an un-quantized medium. look at the notes, they are all separatly recorded one after the other, a mini arpegio. modern daws try and account for this with quantization, that is when the computer makes its best "guess" as to what you are doing. using daw quantization without playing to a click track is IMPOSSIBLE. if you have 10 sources going into the DAW on one midi transmission line the timing will get shakey. since only one thing can happen at a time, the ten events (or more if playing chords)will become 10 [i]separate[/i] events and will be recieved by the DAW as such. this may be ok (or not) because a piano chord may be interupted by a bass players note. does this make sense? all the information will end up in the right tracks, but no two events can happen at once. this goes for pedals, mod wheels, patch changes and other expressive tools. i recommend you have someone help you try it first to see if it will do what you want. i love midi and will miss it when it is replaced by something that is not completely obsolete and incredibly slow. but it is getting to be 20+ years old, and we dont have any 20+ year old daws do we?

[Salyphus]

What Coaster says above is true. You can improve things somewhat by using a multiport MIDI interface with some sort of time stamping, but that helps playback latency more than recording. Still, don't let that scare you off. If you are doing really precise Kraftwerk type stuff you'll definitely notice, but for looser rock types of things you may not notice any problems at all. And you can always nudge the data later in the DAW. BTW, the Roland VG8 does not output the guitar performance as MIDI. It's a GREAT unit (I have a VG-8EX) but the MIDI out is only for Sysex data. I guess the GR would output MIDI, but don't expect that to sound too much like a guitar.

[Super 8]

Well, I can see where it might be usefull to do this. But it honestly sounds like a lot more work and worry than it's worth. You will obviously need a MIDI patchbay to handle all of the connections. You could daisy chain the whole thing, but I wouldn't recommend it because inevitably someone will end up hitting a patch-change parameter on someone elses synth. It would also make troubleshooting more of a pain. You can use the patchbay as a signal booster to help increase your proximity to the band without signal degradation. You might get maybe 35 feet of distance with this setup. Another thought would be to feed the patchbay into something other than the DAW which could stay close to the band. A stand alone sequencer or a laptop or something. Since you're only recording MIDI, there is no need to go with anything new or expensive. I have an old 286 laptop with MSDOS 5 and Cakewalk 5.0 for DOS, which works great for this kind of thing. I'm sure you could get something similar for next to nothing -mine was free. At that point the only problem would be transferring the MIDI tracks over to the DAW and syncing them up to the audio, but I doubt that would be too difficult -just another process. Timbre replacement works just fine IF you are willing to make the necessary edits to the sound patch and/or the MIDI performance data. The whole thing seems pretty do-able. But, like I said, is it going to worth the effort when you're done? How often do you see yourself making use of all of this MIDI data you're collecting? Which instruments do you really see yourself having to alter? Couldn't the edit be made to the audio in the DAW? Could the musician just do a punch in, or a retake? By the way, do you have a link to your church's website? I'd be interested in seeing and knowing more about your setup. Well, I can see where it might be useful to do this. But it honestly sounds like a lot more work and worry than it's worth. You will obviously need a MIDI patchbay to handle all of the connections. You could daisy chain the whole thing, but I wouldn't recommend it because inevitably someone will end up hitting a patch-change parameter on someone elses synth. It would also make troubleshooting more of a pain. You can use the patchbay as a signal booster to help increase your proximity to the band without signal degradation. You might get maybe 35 feet of distance with this setup. Another thought would be to feed the patchbay into something other than the DAW which could stay close to the band. A stand alone sequencer or a laptop or something. Since you're only recording MIDI, there is no need to go with anything new or expensive. I have an old 286 laptop with MSDOS 5 and Cakewalk 5.0 for DOS, which works great for this kind of thing. I'm sure you could get something similar for next to nothing -mine was free. At that point the only problem would be transferring the MIDI tracks over to the DAW and syncing them up to the audio, but I doubt that would be too difficult -just another process. Timbre replacement works just fine IF you are willing to make the necessary edits to the sound patch and/or the MIDI performance data. The whole thing seems pretty do-able. But, like I said, is it going to worth the effort when you're done? How often do you see yourself making use of all of this MIDI data you're collecting? Which instruments do you really see yourself having to alter? Couldn't the edit be made to the audio in the DAW? Could the musician just do a punch in, or a retake? By the way, do you have a link to your church's website? I'd be interested in seeing and knowing more about your setup.

[Super 8]

[quote]Originally posted by Sal Ivory: [b]I've never heard of any 20' limitation for MIDI. I'm pretty sure you can get MIDI cables longer than that. [/b][/quote]Actually, I think it's more like 25', but 20 is a safe bet. [quote]Originally posted by Bastard Coaster: [b]if you have 10 sources going into the DAW on one midi transmission line the timing will get shakey. since only one thing can happen at a time, the ten events (or more if playing chords)will become 10 [i]separate[/i] events and will be recieved by the DAW as such. this may be ok (or not) because a piano chord may be interupted by a bass players note. does this make sense? all the information will end up in the right tracks, but no two events can happen at once. this goes for pedals, mod wheels, patch changes and other expressive tools. ?[/b][/quote]Pitchbend and Mod wheels will DEFINITELY send a massive amount of data down the stream -aftertouch too. a sustain pedal shouldn't create more than 2 events (on & off), so I don't see any problem there. Everything else you said makes sense, Bastard (may I call you Bastard, Bastard? :D ). But I can't help but think about the data being sent to a synth in a standard 16 channel sequence. Yes, it's time stamped on one end, but it still ALL has to go through that MIDI cable 1 event at a time, and it comes out sounding fine. Often when the timing is noticeably bad it's because the synth can't handle the MIDI data at the speed it is coming. I can agree that MIDI is limited by the technology available in 1984-85 (somewhere in there), But that technology was designed to handle 16 channels of data flowing through it. I don't see how it is really obsolete. Inefficient maybe, in that if you want more than 16 channels you need multiple interfaces -and I'm sure there are plenty of other things it could do better or faster. But obsolete -to me- means it can't handle modern demands, and I don't know how true that is. For a 20 year old protocol, it has held up very well. Regardless, a new version of MIDI would most significantly only increase the number MIDI channel, and increase the baud rate which would probably do more for sample dumps or other large data transmissions. By comparison, the amount of data being transmitted via MIDI for 10 musicians would be pretty small. I'm not saying your wrong. I'm saying that if I can play a 16 channel sequence and have it sound just fine, then I should also be able to have 10 musicians recording a 10 channel sequence since there doesn't appear to be a fundamental difference between the two. :confused:

[not Cereal]

you DO have a hell of a point, and i agree to a degree. i transfer multi-track sequences into protools, lock the sequencers clock to pro tools clock (typically 8 tracks at once ) and a lot of the time they end up sounding fine as is. this is due to the daws clock being so much more stable than my sequencers clock. transfer without locking the clocks and the sequence drifts all over the place as the more data there is at the current moment the more it slows the sequencer down. also, locking to the daw allows quantizing which as you know attempts to lock the incoming midi data into the closest fraction of the grid you specify. this also works well, and MUCH tighter timing than raw unquantized midi data. this method is actually very good, and most of the time much of the data is "right" however any type of "feel" is gone in certain tracks. my guess is the success of this type of recording would rely highly on the midi box being used to connect all the instruments to the daw/midi recorder. a MOTU box or others will be able to do it well, i have used setups mildly similiar to what we are talking about with sucess. i think it can be done. BUT (always a but) MY experience with midi has had ONE overwhelming constant: situations like this CAN and WILL "go crazy" for no apparent reason and under non-repeatable curcumstances. i wouldnt stake anything TOO heavy on it right away, because theres gonna be SOME big problem with SOMETHING. SOMETHING wont get along with ANOTHER SOMETHING. it's inevitable. whats my point? i dont know. but i do love to talk midi, and use midi. i just dont put a lot of stake in it, because its sometimes "weird and buggy" but then so are DAWS. go for it, tusker. experiment and let us know how well it works. just FOR THE LOVE OF GOD dont try to quantize on the fly live.

[fantasticsound]

(Denzel Washington voice:) "Ok, explain it to me like I'm a 4 year old." I'm missing the point here. Or maybe you are. Why go to all this trouble? Are you interested in a live recording or a tweaked, studio recording? If it's a live recording you're after, my opinion is you're cheating the public by substituting sounds or performances after the fact. We're not talking multiple takes. We're talking about adding something that didn't exist in the live performance. If you're attempting to make the best possible recording of a specific composition, do it outside the context of the service. In the studio, like it or not, honest or not, people commit all kinds of cheating to accomplish the best possible recording of a song. It's accepted practice. But a live performance? I know this, too, is done, but I believe it's a cheesy, dishonest thing to do. And in a church?? ;)

[not Cereal]

i think he just wants it to sound better is all.

[not Cereal]

award for most stupidest double post: coaster

[Super 8]

Well religious/moral arguments notwithstanding, I would recommend you listen the Bastard about weird stuff happening for no apparent reason. If you do this, you should start by just recording one or two instruments, and see how that goes, work out any bugs, and then gradually add the other instruments in one by one. This is probably your greatest chance for success without overdosing on Tylenol or developing a drinking problem. Then later, if you want to commit musical sacrilege by hacking up and modifying your live performances -have at it. And may God have mercy on your soul... :cry:

[Tusker]

Hey thank you fellas. A lot of good information here. It will take time to digest. I think we will go for it eventually, but as some have pointed out, it's best to work to tighten up the live recording to the degree possible first. We'll work to improve the miking, vocal processing, etc. this year. Super 8, my church website is: [url=http://www.fmgrace.org]www.fmgrace.org[/url] It's just your typical neighborhood church. There's some span in the socio-economic spectrum, but there is definitely a concentration of "young parents" in their 30s and 40s. We attempted to do a christmas cd as a gift to the congregation but we lost the mover and shaker in recording (chap went off to Full Sail to study formally) so we are re-gathering our momentum. Now as to the deeply moral question of sonic truth, Neal..... :D We are not trying to be anything we are not. The ability to to MIDI based sonic enhancements may make for better listening. Or it may not. It may distract from the emotional impact which is the point of the exercise. We would not hide the fact that we are doing sonic substitutions. Nor is the focus on how "good" the band is. We have no illusions. I remember being deeply disturbed that Gabriel post-processed his double live CD with synth lines(in 1983?). I thought that was dishonest at the time. So I can appreciate your viewpoint. For me, the difference here is that the congregation are truly the performers. We (as accompanists) are not holding ourselves out as performers in the sense that a professional "act" would. Secondly, if we use midi data to replace sounds, [i]we[/i] are honestly playing the notes. We are not even talking about punching in new data. However, as we review that data that the daw is capturing, we do sometimes believe that some more appropriate sounds can provide more room, tighten the bottom, etc. We could do all of this with the typical eq and multi-band compression I suppose. But MIDI does offer some "free" options. On tactical issues, I had not thought about the time lag in MIDI. I had thought that MIDI events would roughly line up with audio events. So, (for example) we could try a different kickdrum on the V-Drums. Or add a sine wave to the bass if it needs it. So there's one additional step there to sync things up. My assumption was that we would be getting MIDI note ons from the bass and guitars. If that assumption does not hold, I am not sure how useful this will be. I like the idea of using a unitor or something to give the MIDI signal a boost. We'll probably try this out on the V-drums first. It's the drums that typically fail us. Every time we have used live drums, it's been much much tighter but a nightmare to avoid bleed-through even with a cage. It's a small stage and it's crammed with mikes. Thanks for your help. Regards, Jerry

[fantasticsound]

It seems like more trouble than it will be worth. And don't think you won't end up editing notes when given the chance, Jerry. ;) I don't have a problem with most of the "tricks of the trade." I simply think you're making things more difficult, and the value of the changes seems very questionable. If you really must do substitutions, use the live tracks to trigger the midi substitution after the fact, or re-record the instruments outside the service. Good luck! :)

[Tusker]

[quote]Originally posted by fantasticsound: [b]I simply think you're making things more difficult, and the value of the changes seems very questionable. If you really must do substitutions, use the live tracks to trigger the midi substitution after the fact, or re-record the instruments outside the service. Good luck! :) [/b][/quote]I think I understand where you are coming from better now. For me, it suggests we need to think a little harder about how we are getting sounds into the DAW in the first place. And the specific problems we need to solve there. Maybe it's time for a conversation with some V-Drum users. Thanks Neal. I appreciate your candor. Jerry :)

[Super 8]

[quote]Originally posted by Tusker: [b]Super 8, my church website is: [url=http://www.fmgrace.org]www.fmgrace.org[/url] [/b][/quote]Cool. Thanks for the link. I suppose I could provide one to my church too - [url=http://www.rivervalleyrockford.org]Click Here...[/url] It's pretty cool that you can work on these things at your church. At my church, you often have to take into consideration who else will be using the gear, whether or not they know how or why to use the gear. Some people are very workable, others are well intentioned but also convinced they know far more than they actually do. It can be a real pain, but it is also a good opportunity to learn to work better with various kinds of people -and to do so in constructive ways. [quote]Originally posted by Tusker: [b]Now as to the deeply moral question of sonic truth, Neal..... We are not trying to be anything we are not. The ability to to MIDI based sonic enhancements may make for better listening. Or it may not. I remember being deeply disturbed that Gabriel post-processed his double live CD with synth lines(in 1983?). I thought that was dishonest at the time. [/b][/quote]Nah -don't buy into that stuff! You can take that to every possible extreme. No electronic drums, no synths, no electric guitars, no amplification -on and on and on. I have one rule: when the sound I'm hearing matches the sound I have in my head -I'm done. That's all of the ethical and moral issues I'm concerned with. Anything else simply halts progress. If only a real grand piano will do, then that's what I use. If I use electronic drums, and like what I'm getting from them, then that's what I use. Same for editing after the fact. I could fix things, change them, leave the mistakes in and edit nothing. As long as it feels right to me, that's what I'll do. [quote]Originally posted by Tusker: [b] On tactical issues, I had not thought about the time lag in MIDI. I had thought that MIDI events would roughly line up with audio events. So, (for example) we could try a different kickdrum on the V-Drums.[/b][/quote]I'll be interested in how all of this turns out, and what unexpected problems arise, and how you go about solving them. [quote]Originally posted by Tusker: [b] using a unitor or something to give the MIDI signal a boost. We'll probably try this out on the V-drums first. It's the drums that typically fail us. Every time we have used live drums, it's been much much tighter but a nightmare to avoid bleed-through even with a cage. It's a small stage and it's crammed with mikes.[/b][/quote]Yeah, I use a V-drum kit at my church too. It's not the top 'O' the line model, it's a step down. I wanted the top one, but that's just me :D , and I'm very impressed with this one. The sounds and how they respond to velocity is amazing. The new nylon drum heads are great too. There is no doubt in my mind that electronic drum pads have surpassed acoustic drum heads in terms of playability. Mylar has nothing on these nylon heads! It's probably a god-send for anyone concerned about Carpal Tunnel Syndrome as well. I think the about of energy transferred into your hands must be greatly reduced. The only things I DON'T like are: The cheap-O plastic clamps Roland provides to hold the pads to the frame. Sorry, but they suck. What's wrong with good ol metal??? Also, the rubber coating on the snare drum pad is a nice touch, but the rubber breaks down after a relatively short period of time from all of the rimshots. So, I've used black duct tape in layers over the area that has worn though. It works pretty well. But I don't think Roland did extensive road testing on that stuff.