From midi to audio...

Hi, the simple answer is to solo the midi track or tracks in your sequencer, press the big red record button in Giga, it will prompt you for a location to save the wav file.

Play the song through (with Giga recording) then import the wav file onto an audio track in your sequencer.

Its a good idea to set the option in Giga that listens for midi stop and start messages too. (Under the capture to wav options after you hit the red record button)

This way, you wont have to line up your tracks manually in Sonar.

Thats a brief description, let me know if you need more detailed info.

the stop recognision; does it work with reverb/release trigger samples? or does it simply stop recording after the last stop message?

Hi Herman, I dont work much with Giga these days but from memory it will stop recording when you hit stop in your sequencer.

Forgot to mention before, for Giga to recieve these stop and start messages you may have to setup your sequencer to send them.

In Cubase, there are options under Synchronisation to send Midi Timecode and send Midi clock. You need to tick the Nemesys ports. (I cant remember which one, code or clock - that Giga responds to.)

So...you could give the remaining sound a few seconds and hit stop in your sequencer when all sound has died away. That's what I wanted to know. Thanks for the tip on cubase sync options.

Herman Witkam said:

So...you could give the remaining sound a few seconds and hit stop in your sequencer when all sound has died away.

Click to expand...

Yes definetely.

Well, I was able to record the track but not synced...I don't know how to sync it with sonar. Hopely some Sonar user will stop by and give me some tips hehehe (fingers crossed).

The other thing is multi-track recording, when I click on it, it doesn't take any effect. Does this means I have to record one track at the time? :cry:

I find this topic of very big interest as well

As Niah said, you can record but the output is not synced. Also , you might be picking up some weird computer noise that will add up at the end after 32+/- channels all mixed.

Isn't there a way to just render the thing to .Wav with lossles quality and perfect syncing using the samples and Vst's, or whatever ?

I use an old programm at the moment called Audio Compositor and it renders midi to wave using soundfonts.
Everything comes out perfect at the end and also , at points it may reach 900 notes of polyphony ( OVERKILL - has unlimited poly :D ).

Is there something similar for Giga/Kontakt ? A Magic button ?

That is one of my major fears for not moving to samplers like Kontakt and Giga .

Thanks, nice thread

Niah said:

Well, I was able to record the track but not synced...I don't know how to sync it with sonar. Hopely some Sonar user will stop by and give me some tips hehehe (fingers crossed).

Click to expand...

Perhaps check your Sonar manual. All sequencers have a sync function and the ability to send midi messages.

Niah said:

The other thing is multi-track recording, when I click on it, it doesn't take any effect. Does this means I have to record one track at the time? :cry:

Click to expand...

With Gia you have to record one track at a time unless you choose to bounce sections. For example, you might record the the Violins, Violas, Cellos and Basses, all articulations, to a stereo wav file called "strings"

To address Theo's questions, if you have a decent sound card you shouldnt be picking up at noise at all. The whole process remains digital and is not at all like the old analog bouncing.

As to rendering midi to wav with Kontakt, thats the beauty of it being a VST sampler, the only time you need to render (if your computer(s) is powerful enough) is for the final stereo mix.

VSTs are really the best of both worlds, you have the ability to edit in midi plus process audio in the normal way; EQ, Verb, compression, etc. You can still make changes to the part though, right up to the last moment.

Thanks for the info Scott ! It made me go a step closer to Kontakt

I didn't know you can use it in both real-time and in Render mode if needed. I didn't even know it renders to wave -> Which is just GREAT.

One more please : What sort of polyphony and how many channels can you have running with Kontakt and a very good computer and lots of memory ?

Thanks

Is not syncing really a problem when bouncing multiple tracks? It's just a matter of a bit of moving the files in a multitrack. Can't say I've had any trouble with that in the past.

It's more work Herman =) but i guess I have no other choice hehehe

If the instruments aren't timed to the exact same milisecond it would only add more realism i guess :wink:

Oh yeah I totally believe that. In fact, most of the times I don't edit or correct things that much, because the more you do the less human it sounds.

As far as not syncing the tracks I have to say I disagree.

Any looseness or musicality you create should happen in the playing of the track. If you start bringing in loose tracks that are not synced your more likely to have it sounding sloppy than musical.

Also, for parts like legato strings or portato horns, it can be almost impossible to line them up properly. If you are working to a deadline on a project, the last thing you want is to be messing around trying to line up parts and wasting precious time.

Remember, loose or unquantised playing thats sounds extremely musical is either a happy accident or skillful input in the first place. And a happy accident will only happen for a few notes at a time usually, not an entire track. This is why I dont believe offsetting the start time on a given track will somehow sound good. Added to that, most pro players have pretty rock solid timing, especially orchestral percussionists.

Well, you still have to see whether it's correct musically of course. When I drag the individual wave files into a sequencer I zoom in and compare the attacks. It takes me about 3-5 minutes, depending on the number of tracks, and the transientness of the attacks. I look at the waveforms. In an orchestra there's mostly at least one player who always times too late. You can't delay a whole violin section, but you can delay a single woodwind or brass player a few milliseconds. Some of the players (if would be sitting in a concert hall) like the timpani player will hear the instruments at the front a bit later than they are actually played. Thus he will time a bit later. Psychoacoustically, you have to watch out for too much of those delays though.

How do you quantize your playing? Manually or by some option on your sequencer?

I mostly do it manually, and when I overdo it the playing isn't all that smooth so...quantize yes but not that much. So I let some notes overlapping others for the sake of smoothness playing.

It depends on the attack of the samples I guess...

What do you guys think?