I am new to Nebula. I've been using it for less than a month. Done some research about rendering tracks to put less pressure to CPU processing. Couldn't find proper info that could help me in that regard. I am using Reaper in a mac. My machine is pretty good. Mac mini i7 quad core 16GB RAM. Anyway I am using a lot of Nebula instances and just as a start point I am putting a console channel to each track and than rendering tracks before the mix process.
My question is about render audio quality. Because after rendering I feel that the original tracks summed with the Nebula running are a tad better than the rendered tracks summed. I am doing a A/B check with all tracks playing together, soloing the set of tracks running nebula instances and soloing the set the tracks rendered.
I render all the tracks at the same time in Reaper and this seem to be fine?
Do any of you have any experience doing this kind of render?
What is the best configuration sound quality wise to render tracks with Nebula in reaper?
I changed the quality figure inside Nebula mast page to 5, but I'm not sure this is used for that.
I think quality slider is best left alone. It's supposed to jump up when you render offline, otherwise you'll run out of CPU quickly when mixing I guess .
Normally, rendering should not lessen quality in anyway. In Reaper there is a bit of a 'but'. Do you use the 'render tracks' option from 'right clicking' (no clue how that works on a mac) or 'freeze a track'? (This is called bouncing in DAW land ).
The 'caveat' with Reaper is when you start to use the batch-render function (File -> Batch item/file converter) or the Render function (to render tracks instead of a mix, File -> Render).
The issue is when there is a difference in samplerate that your soundcard is using and if you're rendering to a different samplerate.
If I'm working in Reaper like normal, you 'll see the samplerate of the soundcard in the top right. If I now load a Nebula instance on a track, it's get 'initialized' to that samplerate. Nebula now (rightly so) expects that samplerate.
If I go to File -> Render and choose a _Different_ samplerate, Reaper will happily say to Nebula "No, you're not getting the 96Khz I told you earlier, you're now suddenly gettnig 44khz". Ofcourse, results don't sound right then. Nebula expects 96khz in this eaxmple, while Reaper suddenly switches to a different samplerate.
If this is the case, I urge you to remember (because you may need to check it everytime) to set the samplerate in the _project properties_ (File -> Project Properties). Make sure that you set the samplerate there, and that it matches what your soundcard is using (or better said, it must match the samplerate that was active when Nebula got loaded). If you now go to File -> Render, Reaper understands that your whole project should be rendered at 96khz, and then converted to whatever samplerate you desire (as it should be), instead of suddenly running your project at a different samplerate.
This is the only issue I can think off ( which is kinda Reaper specific).. otherwise rendering should work just fine and (of course) have no quality loss.
Maybe a thing to check is what file format and settings are used for rendering and bouncing. Make sure it's a lossless format (like good ol' WAV), and that the samplerate matches that of the source. Also I recommend to set the output format to '32bit floating point' (or even more), even if your source is 24bit integer or 16bit integer. That way you absolutely safe anything of the track as is. In the project properties (File -> Project Properties) in the Media tab, at the bottom you can set the format used for bouncing ("Media format for Apply FX, Glue, Freeze...), and underneath it you can set the default format for rendering/mixdown ("Default media format for project/region")
I just have bunch of tracks in which I put all the Nebula stuff I want on it 'per default'. Then I select all the tracks (note, select the tracks, not the items on the tracks), I right-click the track / a track and choose 'Render/freeze tracks -> Render tracks to mono stem tracks (and mute originals)'. (or stereo if you have a stereo source of course). This renders all the tracks (using multiple CPU cores) and writes it to a file, creating new tracks with the nebula (or other VSTs) already on it. The originals are there, but muted (so they don't use CPU) and you can start mixing with your new stem-tracks.
In your original muted tracks, I now go to the plugin-list on it, and right click and choose 'toggle selected plugin offline' (note, not bypass). This unloads the Nebula instance, so that loading the project next time goes faster (Nebula is not loaded when loading the project). But if you ever want to see what you did or change it, the settings of Nebula are still there. Put the plugin back online and it gets loaded and you can change + render again if you want.
thanks a lot for your replies. I actually wasn't sure about that mast page change in the quality set. I'll return that back to default (1) and try rendering that again. I just put it to 5 because Nebula manual say something about doing that for rendering.
I also use the same render process as described by Jorismak:
"I just have bunch of tracks in which I put all the Nebula stuff I want on it 'per default'. Then I select all the tracks (note, select the tracks, not the items on the tracks), I right-click the track / a track and choose 'Render/freeze tracks -> Render tracks to mono stem tracks (and mute originals)'. (or stereo if you have a stereo source of course). This renders all the tracks (using multiple CPU cores) and writes it to a file, creating new tracks with the nebula (or other VSTs) already on it. The originals are there, but muted (so they don't use CPU) and you can start mixing with your new stem-tracks."
I'll try and render it again without changing the quality parameter in the mast guru page.
Man!! Apparently it was the quality set in the mast page, which I had put to 5 (maximum). After returning it to default (1) I now A/B the rendered tracks against the tracks playing "live" with Nebula instances and they are the same.
Thank you again for the help. That doubt was killing me