Hi Mark,
markparr wrote: ↑Sat Dec 11, 2021 8:33 am
To follow up, when I play the track in Cubase, the CPU doesn't spike but when I render an individual track the CPU spikes during the rendering process.
I have rendered one individual track, and based on your advice Telefunkin, I have then exported that track and played it on another Media and there are no clicks which does suggest that this is a CPU issue.
I suspect something may be borderline, and maybe the rendering process takes it just over the line for some reason. I still think, though, that there should be some way to tell Cubase not to do real-time rendering (I know there definitely is for exports, but I don't have Cubase open at this point to poke around on the rendering front.)
Also, though, because the render will involve disk access, as does playing back the rendered audio track later, it is possible what you are seeing is a disk issue. If the rendered audio file itself doesn't have the clicks (i.e. when played outside of Cubase -- I think that is what you mean by "played it on another media and there are no clicks"), then the clicks in Cubase may mean it's not able to stream the audio from disk quickly enough. I had this problem at one point. Specifically, I was freezing tracks because I was at CPU limits, but, after freezing the tracks, the audio artifacts in playback got even worse, which didn't make sense to me short of having disk issues. I ran some tests on my disks, ultimately with a free product called CrystalDiskInfo, and found that the disk I was using for audio was going bad. I temporarily relocated the Cubase directory for the project to another disk, and it was fine then. I ultimately replaced the failing hard disk with an SSD of the same size, and I haven't had those sorts of problems since. (I still to hit the CPU wall at times when mixing, or when having a bunch of heavy duty virtual instruments active at the same time prior to freezing them.)
markparr wrote: ↑Sat Dec 11, 2021 8:33 am
The only reason that I wanted to render the Midi files to Audio is to future proof my tracks. Once I have completed a Project, I always export each track to Audio but I wasn't too sure if that was sufficient to safeguard against expired or missing plugins / sample libraries in the future.
Rick - with regards to freezing each track. Once the track has been frozen are you saying that the audio files stored in the Freeze directory, future proofs the track as it uses it own stored audio file?
Sort of. There are pros and cons of each method.
With Freeze, the virtual instrument gets frozen into an audio file in the Freeze subdirectory. You have the choice of freezing with or without inserts, but it is an either/or thing, so the frozen file will either include the inserts or not -- there is no having both. Also, if you unfreeze later (e.g. if you want to change the MIDI part or edit the insert settings), then the audio file gets deleted. The audio file also isn't included in Cubase's audio pool as far as I'm aware. However, if you Backup the project (be it by just backing up the project directory manually or by using Cubase's Backup command), the freeze files will be included. (I use the Backup command at the end of a project because it has an option to only back up files that are used in the project, so won't get stuff in disabled tracks, such as all the audio take components that didn't get used after comping.) That audio file is in a standard format, and can be imported into other DAWs, if needed. Of course, it will come into Cubase if you open the project from the backups later, too. BUT, if you then unfreeze, and there are problems with the plugin(s) in that were frozen, you could lose the audio file. (Of course, you could read it back in from your backup media if needed.) Also, if you're freezing with inserts included (which is a good safety against those insert plugins not being available in the future), they will be included in the audio, which could be a blessing or a curse if your intent is to remix the project, starting with just the virtual instrument sound. (I find that, by mix time, I am usually freezing instrument tracks with inserts to save CPU. In that case, the Pre section of Cubase's mixer -- basically, gain, high-cut, and low-cut -- are also included, though Cubase's channel strip is not by default, which still allows making some EQ changes in the mix.)
With Render in Place, you'd have whatever you include in that (and I think it can include inserts, but don't recall for certain -- I mainly use it when I want to break out specific parts from a multi-instrument virtual instrument, such as breaking out different drums in Superior Drummer or percussion instruments from Shimmer Shake Strike, where I specifically want to process the tracks differently without using multiple outputs in the instrument, and/or maybe to be able to tweak the arrangement at the audio level later in the process), but, if you're using audio inserts after that you'd have the option of freezing the audio track with the inserts later, and you'd have the raw audio tracks in the pool plus the frozen tracks with inserts in the Freeze files. (And, of course, you could do a render in place of the processed audio files, too, getting both the raw and processed ones in the pool.)
To date, I've mostly used Freeze for my future proofing of instrument tracks, though I also create stems at the end of every project to get submixes. Stems obviously aren't as flexible as the original tracks for remix purposes, but would be sufficient for, say, just lowering the lead vocal or the drums in an overall mix -- the flexibility mostly depends on what you include in the stems. I'm currently new enough to Cubase (started with 9.5, but really only made it my primary DAW for new projects at 10.5 -- I'm now on 11) that I generally haven't had to go back and remix projects or worry about missing plugins to date. Most of my remixing projects came from Cakewalk/SONAR, and, with a few exceptions, I've just remixed those in Cakewalk by BandLab for quicker turnaround. However, in both cases, I always export a Type 1 MIDI file (this gives me the original MIDI data, including markers and tempo maps, and any DAW can read this), plus I will have the audio files that are still in the project including frozen audio, so if the plugins aren't available for whatever reason (and this has happened quite a few times since my Cakewalk projects go back as far as 2000) I can at least get those tracks back to how they were as a starting point. Most of the time, part of my reason for remixing will include replacing older virtual instruments with newer, better ones, and often starting fresh on other audio processing (as my production quality has improved significantly over the decades). But there have been a few cases where I absolutely needed to keep those original tracks in their audio form (e.g. I'd used a Zero-G Vocaloid product for layering some backup vocals on one song, and that product didn't work on my newer computer, and I have no desire to use Vocaloids in the future), so having the original audio can definitely be a lifesaver.
I've only fairly recently started to become more familiar with the Render in Place capabilities, and have at least been considering whether I might want to shift my strategy here, though not so much for futureproofing purposes. Rather, I know many people prefer to render everything to audio before mixing, and my projects are all virtual instruments other than the vocals (and the extremely rare cases where I might use third party loops). That can provide the flexibility to send it to another DAW for mixing from scratch by someone with different plugins, or to start totally fresh in another project within your DAW (i.e. without having insert processing embedded in the freeze files in my typical use case) without having to worry about whether the virtual instrument still works. Of course, it would also dramatically increase my track count, and probably also the size of my project archive files since there would likely be double the audio at the track level (i.e. raw audio plus those same tracks frozen with inserts for performance reasons). For now, though, I'm keeping to the stems plus frozen instrument tracks (usually with inserts included), along with the occasional submix audio (e.g. background vocals, really just to save mix-time processing compared to dealing with the individual tracks, even frozen, going through a subgroup, where that subgroup processing can be pretty heavy -- I can always freeze the submix of the subgroup to free up that part of the processing, while still giving me the option to unfreeze it to tweak it).
Rick
.