Jump to content


Please note: You can easily log in to MPN using your Facebook account!

mixing midi vs audio


Chet Strzepa

Recommended Posts

I just created my first mixdown to a wave file using Cubasis. I had about 8 original midi tracks, and I recorded 2 into separate audio tracks, than recorded the other 6 into a single audio track (after mixing then using midi). So I ended up mixing down 3 audio tracks. Would there be any benefit to recording each individual midi track first to audio and then do a final mixdown of 8 tracks?

 

I'm using an AP 2496 card with 2 keyboards and an external analog mixer. Even though I raised the volume levels to get as close to clipping as possible, the mixdown on the cd is not that loud and I'm wondering why. I'm not expecting to sound pro but most of the cd's I have seem to be a lot louder (not to mention clearer and better overall). I tried normalizing in Wavelab lite, but that didn't help much.

thanks,

Chet

Link to comment
Share on other sites



  • Replies 13
  • Created
  • Last Reply
Would there be any benefit to recording each individual midi track first to audio and then do a final mixdown of 8 tracks?

 

Yes! I can't think of a reason not to at first record all midi to single audio tracks. That way, you can record your midi to audio as center for your final mix down in audio. Once everything is on that one track, you're doomed to keep it that way. Think of midi as pre-production tool - not a final mix tool. I don't use cubase but I figure most daw's are alike. I will do all pre-production in midi, transfer midi data from my instruments, do levels and pans with an idea of how they will sound as audio, but when I re-record midi to audio, I keep it center with the levels to 127.

After you're satisfied with your audio tracks - then you can bounce, let's say drums, together. You can use a limiter and some reverb on that single bounced track and save your ram/cpu for other work. But definitly if your cpu allows, do one track at a time. Your mixes will work out way better that way. ~nel

*

 

Link to comment
Share on other sites

ok I'm dumb...I'm not sure what you mean when you say you "keep it center with levels at 127". I would guess that it would be best to record each midi track with the input signal level set to max (to avoid clipping)...and then use, in my case, the Cubasis audio mixer to adjust the volume levels of each audio track before the final mixdown.

 

Or does it matter if you maximize the input recording levels for each track when recording? If you actually record a track with a higher input level, does that mean there's less inherent noise in the recording (ie better SNR)?

 

Thank you,

Chet

Link to comment
Share on other sites

I think the idea here is that (unless you have digital outputs from your synth to record from) the Digital to Analog converters on synths are generally noisier (read "of poorer quality") than those on most computer audio input devices. To make this noise as minimal as possible you would make sure the sound coming out of the synth is as "hot" as possible giving you the best signal-to-noise ratio. "Center" refers to the midi pan position.

 

-Casey

Link to comment
Share on other sites

Chet:

I don't think you're dumb at all...I have recorded digitally for years and still learn new stuff every day. Just listening to Dave and co. talk about the ins and outs of keyboards, lfo's, oscillators....I feel like a moron at times.

 

As Magpel said, to achieve the best sonic levels - I would suggest triggering your midi devices from cubase one track at a time with all midi automation turned off. Let's say you trigger the bass drum track at it's current midi automated level, panned slightly left, at 64. When that is transferred to audio if your choose later to kick the drum center and raise the level - it will always sound 'off'. However, if you record midi to audio on a track per track basis (center, full boost) - you have free reigns for the final audio mix to make any changes or additions necessary.

Yeah, it's more time consuming, but definitly worth the effort. ~nel

*

 

Link to comment
Share on other sites

Many thanks. I think I understand it now and hopefully I'll get a chance over the weekend to try recording each track separately and then do the mixing of the audio tracks as discussed.

 

It seems that I could keep asking questions. The comments about recording the synths as "hot" as possible to get the SNR down makes sense to me now. The next thing on my mind is wondering what would be the next steps in improving my sound, like going to 24 bit or getting a digital mixing for example. I will still have the limitation of the analog outputs on my synths as someone mentioned which will still be my weakest link. For now, I think I should focus on working with the gear I have since I'm still learning (and I've already overrun the music gear budget). But I am still curious as to what I would do next for sound improvements.

 

Thanks again,

Chet

Link to comment
Share on other sites

Chet:

Limitations are your ally! If you just focus on kicking out the best tunes you can with the equipment you already have, you'll master it in no time. And DAW's take years to master as it is. If you still think you're not getting the best audio input from your synths, here's some things I do:

Take off velocity controls. My early midi-audio transfers were pretty weak - the wave forms almost non-existant. Turns out it was as simple to fix as removing or at least modifying the velocity.

If you're still not happy, invest in a pre-amp. A two channel pre (I use the bluetube) costs under two hundred bucks and can not only serve for your line instruments - they can significantly boost the signal of your synths and samplers.

A big mistake would be to overuse plug in's until you're pretty well along in the mixing stage. Also, don't overcompress! Compression/Limiting works wonders on drums and bass. But all too often people will compress everything. It just sounds louder - not better.

Others may disagree here, but I would also suggest not to record too many tracks in stereo. Mono works fine for plenty of tracks, use your ears.

And don't do what I used to do which is comparing every song I've ever come up with to what I listen to. Most studio records are professionally mastered and when you (and me) get to that point of releasing our music - we can take it to mastering. Or buy Waves bundle, a Tc Electronics finalizer...whatever.

Just concentrate on your music for now and don't waste time comparing youself to anyone else. Good luck!

*

 

Link to comment
Share on other sites

Originally posted by Chet Strzepa:

the mixdown on the cd is not that loud and I'm wondering why.

The previous responses are great and cover just about everything but this.

 

The answer is compression. To get everything louder you need to compress the audio before normalizing. You may want to compress the final mix, or you may want to compress individual tracks before the mix, or some combination of the two. Compressing tracks rather than the mix gives you rather different results so you should try both and listen to the difference.

 

Some synths have "compressed" samples on them. Not sure if this is something you can dial in as an effect. If you can, using this to compress your tracks will possibly give you better results than compressing them later. At least it will keep the signal hotter - on the other hand the compressor in the synth may not be as good as your plug in.

 

BTW, if you compose your work directly in midi, you may find it useful to print out some of your tracks and play them back in. This will give them more "life". Its amazing how much better some parts can sound when played by a real musician rather than moused in, even if you have to cheat and do it at half speed!

Link to comment
Share on other sites

Originally posted by Byrdman:

The previous responses are great and cover just about everything but this.

 

The answer is compression. To get everything louder you need to compress the audio before normalizing. You may want to compress the final mix, or you may want to compress individual tracks before the mix, or some combination of the two. Compressing tracks rather than the mix gives you rather different results so you should try both and listen to the difference.

 

Some synths have "compressed" samples on them. Not sure if this is something you can dial in as an effect. If you can, using this to compress your tracks will possibly give you better results than compressing them later. At least it will keep the signal hotter - on the other hand the compressor in the synth may not be as good as your plug in.

 

BTW, if you compose your work directly in midi, you may find it useful to print out some of your tracks and play them back in. This will give them more "life". Its amazing how much better some parts can sound when played by a real musician rather than moused in, even if you have to cheat and do it at half speed!

 

I'm not really clear on what exactly a compressor does or when and where to use one. I have a VST plugin called "DMM-SC1 compressor" that I can use as an inserted effect in Cubasis VST. I don't think I can apply it to an entire mix. I haven't a clue how to use it (other than just randomly turning the virtual knobs). There's a "threshold" knob which seems to work like a gain boost, and there's a "knee" button. Wavelab Lite doesn't seem to have a compression feature.

 

I haven't done much yet with with any plugin effects for that matter. My Yamaha synth has pretty good effects so I've been using them so far. I don't know about compressed samples though...

 

Today I recorded all my midi tracks separately. I have a much stronger signal. It turned out that when I recorded the 6 midi tracks together, there was a point in the music that the volume peaked due to a crescendo, and this caused most of the rest of the track to end up being recorded too weakly.

 

Also, I'm running my synths through a Mackie 1202 mixer, and I seem to be able to get a nice strong signal. I recorded each track with the level of the inputs set so I would just avoid clipping. I haven't done the mixdown yet. I think that will be for another day.

 

I agree about recording the midi. I almost always record the midi track by playing "live" (rather than writing with the mouse). Then I might do some midi editing, then I'll play it back and record it. For this particular song, I actually reduced the tempo and recorded the drums by punching the beat out on the keyboard. This took some practice, but didn't turn out too bad. Later I went in and added some fills with the mouse.

 

Thanks,

Chet Strzepa

Link to comment
Share on other sites

Well, I'm using my compressor plugin for the drum and organ tracks. It seems to help them cut through the mix. The organ in particular was getting buried it seemed.

 

Something weird happened in Cubasis. Over the weekend I had no audio performance issues with 8 audio tracks, and then last night the computer was choking with more than 4 audio tracks. I ended up changing the disk cache scheme in Cubasis and then it seemed ok. So I haven't done the mixdown yet. Maybe tonight I hope.

Regards,

Chet

Link to comment
Share on other sites

the computer was choking with more than 4 audio tracks

That's no good. Do you use a mac? If so, learn how to create a separate ext. set for audio. Another tip is to zap your PRAM every so often. I can't help much on the PC front, but 4 tracks seems unusually low for any modern machine. Did you record everything in stereo or mono? If it was stereo it's actually 8 tracks. Also, plug ins can take up lots of RAM so be conservative. I've crashed using the sonic modulator in DP many, many times. Good luck mixing!

*

 

Link to comment
Share on other sites

First - your tracks will never be as loud as commercialy released CD's until you learn a little about compressors and Mastering tech. So don't waste time trying to compare. Make your mix as good as you can - if it will be destin to be commercialy released, it will get mastered and louder...

Second, since it seems like you have a strugling computer don't record each midi as an audio track. Use your analog mixer as we all did in the good old days. Create a great mix there then record that to cubassis.

 

 

HTH

Danny

Rotshtein Danny - Studio Engineer

Jingles show-reel

 

Visit DarlingNikkie.com To discover the sounds of "Darling Nikkie"(aka Jade 4U). . . .

New exciting project Goddess of Destruction

Link to comment
Share on other sites

I seem to have the computer problem resolved. I'm using a PC by the way and now it's working fine with 8 stereo audio tracks and a couple of plugins going. For whatever reason I have to use disk cache scheme 2 in Cubasis now. Over the weekend I was using Scheme 1 with no problems and I'm not sure what has changed. Scheme 1 doesn't work. With scheme 1 I can run 4 stereo audio tracks but when I add a fifth, it starts "tripping" and then eventually crashes. The program seems to be finicky overall though.

 

THanks,

Chet

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...