Has anyone else noticed this? Punch in/out used to be seamless and undetectable, but ever since around (I think) ~v7 there is a short audio drop out between the time I hear the recorded track and my live performance (I use tapemachine mode). It’s pretty distracting and doesn’t exactly nurture a great take.
Was hoping this would go away with new versions of the software or when I upgraded my computer but nope. Hopefully I’m just missing something…?
Hi,
This is the same when you enable Record on the track. Unfortunately Cubase works like this now.
Really? Just “that’s the way it is now”? It worked fine before, so you’d think it would be able to work that way again. And this is not an insignificant detail, but a major workflow issue. Have you been in touch with Steinberg directly about it?
One factor that may explain this: punch in/out is basically an obsolete feature in DAWs. About the time you noticed that it stopped working well is the time DAWs started introducing track takes and quick comping. Recording a new take and comping to blend in the new take works so much better than punching that it has pretty much made punch in/out obsolete and a low priority for DAW makers.
You have got to be kidding me. Track takes and comping existed in Pro Tools for at least a decade before Cubase implemented them, so it’s not like this was a new concept a few short years ago. But even if it was a new feature, it does not in any way make punching in obsolete. If I do a perfect take except for a couple of flubs, you’re telling me that it’s easier to re-record the entire thing and then comp it? I can’t imagine how time consuming that must be. Punching in allows you to play to the performance so you have context and a smooth transition, whereas different takes often have different feels and may not blend together as well. If I’m doing a song section by section, I want to be able to play into the punch not create an entirely new track and play devoid of any context. Cubase has in fact developed this in some ways, by decoupling the punch points from the L/R locators, but even if comping did somehow replace punching in, it doesn’t make any sense to just make a feature suddenly not work as smoothly.
I’m just pointing out the timing: this Cubase issue received what is apparently low priority around the same time takes and comping were introduced into Cubase. It’s so much more efficient to set up monitoring and tracks to give whatever context you want, and then record takes and do comping, than to do punch in/out. Thats’ why hardly anybody uses punch in/out any more.
You are right that the track comping has nothing to do with the punch in/out problem, but you are wrong that ProTools had it a decade before Cubase.
I believe track comping was first introduced in ProTools around version 8 from 2008, Cubase had it since Version 4 from around 2006.
Cubase’s audio engine is by far the worst I’ve experienced to date. It’s a dinosaur in desperate need of modernisation. I was stunned to discover that I couldn’t insert/delete a plugin or create a send during playback without a break in audio, for example.
Agreed, almost anything you do will create a short gap in the audio.
I think I may have misinterpreted the reference to comping as pertaining to playlists (er, Track Versions). Pro Tools had that long before Cubase did, but if you’re referring to comping with lanes, yes I think Cubase was first.
GlennO, if you’re “punching in” by setting up lanes, you still need to engage recording during playback in order to hear the previous bit of the performance. Unless I’m missing something here? Could please you elaborate on your process a bit? Are you suggesting just doing a punch of takes and comping the best bits? I rarely do that as it’s way more time consuming, tedious, and I rarely get takes that are as good as when a performer can actually in on the parts that need work, get a good take, and we all move on with our lives. Saves hours of mind-numbing work.
I mean, to each his own at the end of the day. Do what works for you, but I can’t imagine where you got the idea that hardly anyone uses punch in/out.
I believe that GleenO is talking about doing a bunch of takes and then comp for a perfect take. To me it’s less time consuming and if you now how to do your edits properly you cannot hear it’s a combined take. Also keep in mind that today it’s pretty normal to record stuff riff by riff and not the whole song in one take. Then if you for example combining takes for the perfect chorus of a song, it’s just a matter of copy/paste for the next chorus.
You’re right of course about how to get the best performance. But think of it this way: At it’s core, punch in/out is constraining the extents of the recording to match the extents of the monitoring, right? In other words, the in/out of audio you will record will match the in/out of audio from the original that will be muted. That constraint of course made sense in the days of tape where those two couldn’t possibly overlap. That constraint was carried over into DAWs for a long time until takes and comping made it simple to relax that constraint. There is no longer a need to lock the monitoring extents to the recording extents. It’s more efficient to let them be independent, listen to whatever you need to to get into the groove, and record as much as you want without being locked to that same duration and use comping tools to choose after you record where the in/outs will occur.
As for whether punch in/out is used much these days…I think the answer lies in the observation that this problem has existed for years with hardly anyone noticing 
.
My guess is that ASIO-Guard and the later updates to it, made it worse.
Record enabling a track would change the cpu affinity to use Real cores for that track.
I don’t think this will change, well until they replace the audio engine.
Personally I don’t use punch in/out that much anymore, exactly for that reason.
It totally throws off any artist I have tried to record.
In a sense, it has improved my recordings (to add a positive). I now pretty much always record everything, and some of the happy mistakes that happen, have been worth keeping and using.
I record my guitartracks (or even bassplaying with a triple play midipickup and a bass vst) always in realtime and with the use of punching in and out, because i do not like to comp afterwards.
and i must say i do not experience any dropout or delay while recording and punching, never had since cubase 5 to 10.5
For recording&punching i use a footswitch connected to my Yamaha N12 for this purpose, which i use my recording interface.
Really never noticed any problems
Obviously not as you are recording midi, try do the same with a real guitar/bass and you will hear the audio dropout shortly.
Gee… I never had an issue with punching in/out audio tracks here. But maybe it’s the way I do it that saves me from hearing a drop out.
When I get a flubbed section that I want to fix with punch in/out, I quickly make a copy of the track. Lower it’s volume a bit, then do the punch in/out on the original track (punching in/out a bar or two on either side).
After that, it’s a few cuts (usually, but, not always, at zero crossing), add the crossfades, and it’s done. Repeat as necessary for other flubbed areas. When satisfied, I delete the duplicate track.
I like making the duplicate track so that I can listen over the flubbed section(s) while I’m doing the punch in/out.
Regards 
[/quote]
Obviously not as you are recording midi, try do the same with a real guitar/bass and you will here the audio dropout shortly.
[/quote]
As i was telling, i play real guitar(only bass is midi) and do not experience any dropout. I play new parts or bits and pieces while punching in and out using the footswitch. My preference setting is manual, not tapemachine or something else.
Maybe it has to do with the tapemachine setting and the track has to be put in and out of monitoring.
I don’t have that, using an external mixer for direct monitoring of the guitarplaying and cubase plays the already recorded parts.
You are not hearing it because you are using direct monitoring. Direct monitoring is not a solution to everyone. If for example you are recording your guitar as DI and are using ampsim to get the tone, you cannot use direct monitoring.
But that is a good workaround for when it is possible. I go back and forth between direct and software monitoring as I sometimes like to hear the track with a bit of processing. Would be nice for this very fundamental function to work properly (reminds me of when MIDI notes used to record with offset timing on Windows), but in the meantime it’s helpful to keep this in mind.
I know, that’s why I stil use my old POD XT(and a H&K Tube pre-amp) for Home-recording together with a D.I. track, so I have direct monitoring. Afterwards I choose a sound with an Ampsim (maybe together with the recorded tracks), or use Re-Aamping in my friends studio and use a real Tubeamp.
But I always prefer the direct monitoring(and no use of Cubase monitoring for that track) so you also don’t have to worry about lowlatency issues when you allready have a lot of tracks and plugins in a porject.