Just a bit curious here as to how often do you guys fill all 8 insert slots and sends…seriously after 12 years of everyday professional use of cubase for music i must have maxed out my inserts only a handful of times…and i never thought im only using a few plugins…then again i use the eq and strip sometimes for non essential track so that may be it…
Sound design, mostly… but I can easily do over 16 inserts on a vocal, especially for a dance/electronic track with plugins that are only there to do some wild automation effect every now and again. And, by the way, when you want to automate more than 8 plugins, the fact that only 6 are pre-fader and the rest have to go on a bus is soooo painful. 
1 Like
This is a very understandable concern, however I don’t think it’s helpful to belittle people who like to use lots of insert plugins just because it’s outside of your realm of possibility/practicality/taste.
I’ve had an idea for potentially a happy solution to the 8 inserts debate, please vote in the poll IF you’re desperately unhappy with this limitation: [POLL] FAO: people annoyed about 8 inserts limitation - #2 by Hattrixx - Cubase - Steinberg Forums
i guess nobody likes to have a “broken” mixer and how more inserts would compromise stability is beyond me. i personaly dont care, if somebody cant open my cubase 9 project in cubase 5. why should i care? the filters dont work the same way, retrologue2 isnt there and halion se and groove agent se have a different plugin than before. so there are already differences. still cubase 5 could just say, cant open plugins number 6 and 7 and thats it.
i for myself cant open some of the projects my clients did on cubase 5 as halion se and groove agent se changed from that time, so that leaves me with an empty intrument tracks. i dont see this any different than when somebody cant open all plugins because he doesnt have the slots. so is not only 100 % compatible backwards it already is not forwards!
regarding to the insert slots, its so easy to fill them up with 6 plugins, i can by no reason understand why you keep on defending this.
- vms
- autotune
- virtual tape machine
- 1176
- la2a
you see i have one insert left, now i can decide rather i want a neve or pultect where i like to use both for vocals… god may be with me that the vocals dont have any problems, that would need a surgical eq as well. dont even think about special effects…
you see there is nothing really special going on. if you have some aqua plugins like lime you would need 3 slots only for the pre, comp and eq.
now you can tell me again, use a chainer. many people here have already brought up reasons why they dont like to do that, also i dont like to use the post fader inserts as well!!
anyway im sure you bring up some reasons and defend steinberg. maybe you say render in place and then you have 6 more slots???
Steinberg should just let the insert slots be fully custimizable by the user.
The design can be the excact same as now until you decide to add the 9th plugin yourself.
I understand there have to be a set structure in daws on certain things, but to limit users
to 6 slots is to me beyond belief. Everybody can win here I’m sure, why does it have to
be either or…
Would it be hard to add more inserts? 10-12 would be my sweet spot
The thought occurred to me that if reports of processing power being compromised if plugins are not spread over multiple tracks, maybe Steinberg is limiting insert slots to eight for this reason.
Don’t be so close-minded. I still cannot understand why some of you are blindly defending a limitation, on a software you use and paid for. Don’t tell us to use a chainer like we didn’t know it existed, that is an idiotic solution to a problem that should not exist. I could make a huge list of DAWs that doesn’t impose this limitation. It’s insane you have to use a third-party chainer to have more than eight inserts on each track on a $500 professional-grade DAW. I’m sure Steinberg’s engineers are capable of doing that without a “broken” mixer. Do you see any other DAW crashing and burning when someone uses ten inserts on the mixer? The limitation is hardware-bound, not software.
Very good point. It’s possible. I’ve done extensive testing in four DAWs about this very subject (back in 2013), but was not able to confirm it for Cubase, only Sonar (at the time).
This was an issue in Sonar and was confirmed with Cakewalk. It’s all about how they did “audio packet scheduling” with multicore CPU assignments. In a nutshell, Sonar could spike if a “serial signal path,” when a VSTi and lots of inserts on a single channel are used, would use more resources than any free single virtual CPU core could handle (usually there are 2 virtual cores per 1 physical core). Basically, the audio engine would have to assign the whole serial stack to a single core, so if it wouldn’t fit on one core, spikes ensued. This was why dropouts would happen even when the overall CPU (when considering all the cores) would not be maxed out. It’s also why Sonar was often working better for computers with less, but faster cores, than more cores, but slower, configurations.
I say was, because they have recently enhanced the audio engine in how it does multi-core, but I haven’t tested yet.
Back to Cubase… Cubase was also better than Sonar in this specific use-case in all my apples-to-apples testing, even without ASIO Guard. And even while Sonar seemed a bit better, overall, in how many real-time effects and VSTs could be run if this little gotcha was avoided (by spreading them out over multiple tracks/channels and avoiding a tall serial stack).
So… from a technical standpoint, this could indeed be a very real factor. Or it could not be a factor if Cubase has no such serial signal path bottleneck. I was never able to confirm this in my testing, but there were glimpses that it might also happen.
Reaper definitely had no such limitation and would balance audio packets, regardless of serial or parallel topology, and would squeeze every last drop out of available cores.
I may do another head-to-head test in 2017, but need to finish some new music first!
Yep, that’s my pet theory too, I think Cubase creates a thread for a track, it only takes one thread to run out of CPU and you’re into snap, crackle and popsville.
That one only works with Waves plug-ins, not with third-party plug-ins…
So, would you think a chainer would raise the same issues as having more than eight slots, or is it somehow different?
Thanks -
Not true. You can only use 8 sends simultaneously from one track. But you can make as much sends as you want to.
You can even use busses as sends. Or sidechains.
Alternatively you can copy the track you need more sends for, don’t select an output for it and make 8 new sends from within that track. Possibilities are endless.
I think it would make matters worse unless they did something like treating the chainer as if it were an external fx and use latency compensation accordingly, you could then run the whole slot on a new thread.
On the poll in another thread it seems a clear majority are happy with the current insert slot number.
I do find it a bit strange that this “stupid limitation”, as someone called it, is causing such grief when simple workarounds exist. But these workarounds are then criticized as convoluted.
I read that what may be needed on one track (for instance) is maybe an EQ, then a compressor, and another compressor, then a gain staging VU, then a delay, then a desk emulation, then a limiter, then a reverb, then insert another compressor so you can A/B with the compressor you already used… etc etc.
To my mind, the need for a channel strip and a workflow like that is convoluted. Just my opinion of course, but these requirements sometimes look like more to do with mucking around with plugins than making music.
I would have no problem with unlimited inserts for those that want them, if it did not cause yet more unnecessary bloat to the program, though I suspect it would. (Perhaps someone from Steinberg could comment on how efficient or not it would be to implement this.)
I’m all in favour of change, progress and innovation, but most of all I want I want Cubase to remain a stable and professional platform for making and recording music, and not turn into an over-complicated toy.
I don’t know much about coding DAW software but…
Whatever way look at it, whether it’s 16 inserts on one track, or one track routed to a group, or one track with a plugin chainer you still have a signal path of 16 plugins that has to be processed in real time which as far as I can see, has to be processed as one thread.
Create a single track in an empty project, loaded with CPU intensive plugins. Route it to a group loaded with plugins. Now put more plugins on your master bus. Now look at your cores in Windows Resource Monitor and you will see 1 core heavily loaded and all the others idling. It all gets handled as one thread.
And after you make that music, what do you do? You mix it. That’s where the workflow you deem “convoluted” comes into play - the mixing aspect.
Ever see a patch bay while a real engineer is working on a large mix?
Kinda “convoluted”, no? Yeah, it is. And that’s part of making music.
Cheers.
Why the one thread? We are talking about creating a stream out of buffers, if you tagged each buffer with an numeric sequence number and timestamp you know how to reassemble them for the next stage regardless of whether they appear in order or not, When you have all the buffers in a set in place compare the current timestamp with the oldest initial timestamp and that’s your latency. I’ve written similar code to piece together steams of info from multiple RADARs - it’s not difficult and the sort of maths involved is very fast and efficient. No plugin is in real time, there’s always latency - the question is whether the latency is acceptable or not.
Real engineers use TT/Bantam patch bays, not the 1/4" TRS ones in your photo. Is that YOUR patch bay, RiffWrath?
:HiHi: