Rendering For Daw Improvements

Ideas & suggestions
1

I recently started working with DAW multitrack software and found it rerally cool to use the option of rendering each track to a separate wave file in order to import it to a multitrack and mix/edit comfortably. However, I quickly found that there are still some things to be done to allow even more comfortable work with these applications. I’m not sure if some of these ideas weren’t suggested before. If so - sorry.

  1. Usually I do some kind of mixing already in Renoise. Otherwise it would be quite hard to write proper music. The problem is that after rendering each track to a spearate waveform each of them is veeery quiet. Although I usually use 24-bit mode when doing that (so the dynamic range is not really a big problem) it’s quite difficult to work with such waveforms in a DAW, simply because in the multitrack it’s very hard to see where a weveform starts (usually no track is busy through the whole song, so most of the wavefile contains silence), because the waveforms are so small. I wonder if Renoise could have an option to automatically normalize the waveforms. Obviously I can do it in some other app, but it’s problematic to normalize i.e. 20 huge wavs manually.

  2. After loading all the waves to a multitrack I have like 20 channels that play mostly silence. This causes unnecesary HDD traffic, so the obvious solution is to cut off the moments of silence. The problem is to find them. Also, there’s a problem of synchronization. After loading all the wavs it’s good to move them so that the actualy music starts in some decent location. However, I tend to leave a small pattern in the beginning of the track, where I insert track commands preparing channel DSP etc, so the start of the waveforms is not the start of the patterns. Thus I have to move them. For now I have to move them manually by choosing some characteristic moment of the waveform and move it to a logical location.

I think both of the mentioned problems could be addressed by some pattern command that would insert markers in the channel’s waveform when rendering. This would allow me, for example, to make a marker inserted when an instrument starts playing in a channel. Or insert marker in the beginning of the song. That would rock, especially if it would allow inserting markers for each track, but if it would work for all waves at the same time - it’d still be helpful. Also, ti would help synchronizing the project to the DAW etc.

Why are these things important to me?

Renoise does have very nice mixing capabilities. but still, a multitrack give a lot better overview and guarantees easier and more precise automation etc. Also, although Renoise featured recording features in the newest release, it’s certainly not a good program to work with long waveforms, for obvious reasons. I find it fun to record the bass for my songs, but doing that in Renoise is simply hard. Other reason is that ProTools and some other multitracks feature proprietary (RTAS in PT’s case) plugins, usually very good ones.

2

a few hints about these problems :

  1. there is a commandline tool around that does normalize wav-files. you could batch-normalize all your wavs this way which still leaves you with the problem of having to set the channel-volume in your other software (because when all tracks are normalized the sound will be a mess). Problem is, the one program I know of does not work with anything else but 16-bit wav-files : https://neon1.net/prog/normalizer.html
    dunno if there is a 24/32-bit version around.
    also, I don’t know what sample-editor you use, but does your one not have any batch-capabilities ? awave is also a soft that can batch-normalize, but it’s not free.

  2. I am used to Adobe Audition 1.5, so the following tips are based on this, but I am sure every decent Sample-Editor does provide the same functionality :

first of all I usually render each pattern seperately. with each track rendered seperately you get your harddisk cluttered with lots of small files, however, the good thing is that there is a cool ghetto-way to determine if there is only silence in a wav-file :
if you use wavpack (lossless-wav-compression) with the frontend you can just pack all files with minimal compression and the wavs with silence in it will result in a size of 44 bytes. Just delete all those, depack the files again and you will end up with only the wav files that really have some kind of sound in them.

next, to put “phrases” together you simply load the first wav of a track into audition and then there is a nifty “load append”-option which loads a wav-file at the end of the current wav. this way you can just rebuild the blocks your song is made up with by append-loading the parts one after another and then saving the resulting big file. This way you are also able to change the arrangment in your other software later.

alternatively Audition offers a “Beats & Bars” scale where the timeline is not presented in milliseconds but on bars based on the bpm you put in. if you take your actual bpm divided by 8 or 16 (audition does work with bpm down to the 3rd letter after the decimal) and use the “snap to ruler”-option (thats what its called there) you can cut large chunks of your audio which are perfectly loopable … or would be if you wouldnt have the PDC-problem (which is also workaround-able by adding a very small sample at the beginning of your song and later using audition/whatever to trim the delay).

I don’t really feel like writing this all down in detail at the moment, if you are interested just reply what you want to know and I will go into the details.

PS:And about the long wav-thingie : Someone will hate me for that, but all I can say is “EnergyXT” … :) This really is the software that provides all the features that are currently missing in renoise, those two together are an evil alliance. Whenever I do something with samples longer then two bars I just use Energy, it’s so nifty to load vocals in it’s sequencer, chop them up, drag them around … and you can loop a pattern/block or press this “play from here”-button (the “RMB-Click”-like key on your keyboard) and whoops, long wav plays instantly in sync. ah, I love it.

3

First of all, I know of commandline tools that can normalize wavs, but it’d be MUCH more comfortable if Renoise could do that automatically. Also, if it’s only possible, Renoise could actually do a quick max amp test before rendering and then turn the volume of each channel up BEFORE rendering. That would totally rock, as it would provide best quality of the whole process.

Of course I know after loding it to ProTools I’d have to set the volume of each track, but… that’s what I want to do. If I render each track to a separate wave file I do that because I want to mix in a different app.

I love Adobe Audition, but as I’m on a Mac these days, it’s out of the question. Also I admire it as a wave-editior, but I really think it’s multitrack isn’t comfortable. I prefer SADiE or ProTools. Or at least Ardour.

Rendering each pattern to a spearate wave is an overkill. I also don’t find it practical at all, as not always a sound ends within the end of the pattern, so as long as I don’t make loops for Magix Music Maker in Renoise, I don’t find it useful or sensible at all.

Oh yeah, I don’t use wavpack. And I won’t use it only to find silence in tracks, itsimply doesn’t make sense at all. I prefer flac for my record archives.

EnergyXT is again out of the question. First of all - it’s not Renoise, so it’s not an answer to my question. Second - I’d have to pay for it and I don’t have the money to buy every software out there. Third - it won’t work on a Mac.

There’s also another reason to do it in ProTools instead of EnergyXT or anything else, and it’s a VERY important reason for me. I study sound engineering in the Music Academy in Warsaw and we’re doing most of our work on ProTools (sometimes also on SADiE). The more I work with it - the more I know, the better marks I get - the brighter my future. :) In addition, we can use the Academy’s studios for our own stuff (if there’s some studio-time left of course), so if I was able to bring a session to our studio I’d be able to mix my songs on an awesome TDM, fully featured version of ProTools with Control 24 and very nice JBL 3.1 monitor system and that would be TEH ultimate ultra-hypah-coolness one can imagine. Especially as in such case I’d be able to do some really good-quality recordings for my projects in a proper studio environment.

Yeah, infact I already can do that right now, but it requires me to do some stuff manually, while simple feature of Renoise inserting markers would already make my job a LOT easier.

Let alone my idea with an additonal pattern command. A simple marker on each pattern beginning would already be one of my most desired features.

Normalizing is another story. I guess I could live without it, but I beg you for the markers.

4

the fact that you are on mac surely makes my suggestions useless. and please bear in mind that those were merely suggestions or an insight on the way how I tend to solve those problems you mentioned, because I have run into them myself before. They aint pretty, but they work and I prefer some hacking-style stuff over “can’t get it to work at all”.

and as for the markers I can only repeat that you should look for a bars&beats-option in your sample-editor.

5

I know there are are lots of other ways to do stuff. I could probably sequence my music in some other application generally, but I truly think an option of adding markers would make a LOT of stuff a LOT easier you know. using some beat-detection options in other software is of course a way to figure stuff out, but it will never be 100% precise or comfortable. Not to mention it doesn’t always work, if there isn’t a strong kick on the beginning of each measure, changing time signatures and tempo
would also most certainly confuse the system.

Displaying the timeline in a beats format and manually adjusting the beginning of the first pattern “by eye” is what I do now, but it’s not really a great way to do stuff.

6

+1, because everything that gets put into the .wav by Renoise and therefore hasn’t to be redone each time you render is good.