Time to move on from this thread. 

FWIW - in the last year or so for reasons I dont really understand - I have been getting and doing a lot of Euro / Trance Dance.  Huge numbers of VSTi's all multi-outputed lots of convo reverbs etc.....  I can assure, at 48 samples, you can get to very high cpu useage very quickly ... anyway, I digress - maybe the folowing is easier to try:-

Try this test:-

=> Set your Buffer Latency to 48 or 64 Samples  [ 128 is OK but it will take a lot more tracks as per below ]
=> open a clean / new / empty project
=> insert x 1 mono audio track
=> MUTE it and enable REC and Input Monitoring
=> slide the Vol / Vel  all to  zero
=> insert  5 different Sonitus Plugs - Rev / Mod / Delay / Mutli Comp / EQ
=> load a preset for each Sonitus plug and enable each of them
=> copy / clone that one Track  10 / 20 / 50 / 100 etc....  times until your Windows Task Manager is sitting around  %50 <-> %60
=> insert one more audio track with some audio on it - say a 4 bar drum loop - anything will do - and set it to loop
=> hit play in the looped section so that Playback does not stop / audio engine does not dis-engage
=> whislt the project is playing back, insert one instance of Rapture - does everything remain playing back without artifact or any sort of audible glitch ?
=> whislt the project is playing back, open any Rapture sound - does everything remain playing back without artifact or any sort of audible glitch ?
=> whislt the project is playing back, replace the Rapture instance with one Dimension Pro instance - does everything remain playing back without artifact or any sort of audible glitch ?
=> whislt the project is playing back, open any Dim Pro sound bank - does everything remain playing back without artifact or any sort of audible glitch ?
=> whislt the project is playing back, SAVE the project - does everything remain playing back without artifact or any sort of audible glitch ?

Again ...... for me, with my DAW, the above cannot be done with X1C Expanded.

Agsin ......  purely of audio engine comparison, the exact same above at 48 Samples @ 88.2k / 24 Bit is perfect / glitch free in S1V2 - there is absolutly no change to audio playback and stabilty - it is utterly invisible and transparent.

Again ....... all the best
Tim

 Why on earth do you need to be running as low as 48 samples? I could understand, maybe, 64 samples if you were doing live performances. But why this low when in the studio? For me even 128 samples is fine playing live.

Bit of a tangent this, but I happen to think low latency operation doesn't need (note I don't say "isn't") to be important.

I think a sensible overall system design is to get yourself a set-up where you're mostly direct monitoring, rather than monitoring through Sonar. That takes care of live overdubs. For recording soft synths, the PDC button can get you through most use cases.

It's really worth stopping and thinking what you're using low latency for, and seeing if there's another way of achieving your goals, that's more friendly to the rest of the working process.

Hang on, what are you saying here? Are you saying you're not going to spend money on making your studio work right fr paying customers on a point of principle?

I mean, I imagine this was never going to come up anyway, but thanks for helping me cross you off my list of studios I'd ever use.

Wow who would have thought this would have gotten so nasty....

Anyway...here is the deal. The OP is monitoring with a 4.3 ms latency. That is NOT necessary and if you complain about CPU useage and your buffers are set that low then you really are crying into the wind.

No human being should need to monitor with less than 8 ms latency. I understand it is all wonderful to have lowest latency but not at the sacrifice of useable CPU headroom.

This thread really should end at this point. Seriously, what more is there to say? I am not trying to be insulting but just observing the facts as I see them.

Ken

Monitoring at 48 samples: that's your problem right there, that and Euro Dance 

timboe

For goodness sake, if 48 / 64 / 128 / 256 samples is good enough for you thats GREAT - for me, I prefer 48 and my audio interface  [  FFace UC ]  handles it with ease.

Anyone here remeber when  24 Track 2"  Tape machines arrived in a big way  - " Why would anyone need more than 16 Tracks to make music " ...... " Its a waste " ....... " Proper musicians dont need that many thracks "  .............precislely the same arguments when the industry moved from  mono to 2 tack Stereo / 2 to 4 track / 4 to 8 track / 8 to 16 track etc..... 

This is where I have to dissagree with you.  Running at 48 samples is right on the bleeding edge of what any host-based DAW is capable of and you are asking for trouble at that latency - especially running multiple synths & FX.  The slightest system latency caused by a PCI device or whatever will cause a dropout at that latency.  I am running a MOTU 2408 Mk3 and 24I/O and at latency that low, even windows media player pops & clicks!
 
If you want to be able to run at that super low latency, you should be looking at something like Protools HD, where all the audio work is handled out-of-band.  I ask again - what possible benefit do you get from working at 48ms?
 
What I'm getting at here is that you are asking CW to solve a problem that isn't a problem, because who really needs to work at those latencies?  CW must focus on the issues of the masses - not a silly request to get everything smooth at 48ms.

John T


Well, bozo, let's say I'm in your studio and the issue you describe becomes a problem. I'm going to hold you liable and saying "hey, don't blame me it's this crappy software" isn't going to get you a repeat booking. 

Hasn't happened yet in 10 years in my studio.  You guys are on another planet if you think that the inability to run at 48ms is going to cost you jobs.  I live in something called "the real world" and make good money playing live and recording.

daryl1968


bill - John was responding to another post by elijahlucian

That's what I get for reading in flat-viewing mode, I guess.  I don't like the tree structure on the CW forums, but reading flat leads to these problems. 

backwoods


Monitoring at 48 samples: that's your problem right there, that and Euro Dance 

..... now that made me laugh big time !!!!!!!!
 
Bill ..... I guess we'll just have to respectfully agree to disagree.
 
You see operating at 48 samples with large loads as extreme / unnecessary / asking for trouble.
 
I see it as  " close-to-hardware "  as I can run.
 
48 Samples is not bleeding edge or extreme - RME / Echo Audio / Lynx  all have rock solid stable drivers that run at  32 samples for their PCI cards - I'm only running at 48 samples now  because I moved to a more portable RME FFUC  [ USB ]  which doesnt have a 32 sample buffer setting due to current windows architecture limitations - I  *was*  using an EchoAudio Layla 3G at 32 samples and again, like my RME it was essentially, uncrashable  - it ran all day long at 32  samples doing whatever you threw at it  [ back in the days of 8.5.3 ]
 
As an aside, your  MOTU wont do it is because their drivers are simply not what they used to be.
 
And to a larg'ish extent,  I do agree with you about your   " needs of the many outweighing the needs of the few "  argument.
 
But .... again ..... you seem to be missing my point.
 
I am just trying to highlight that Sonar's Audio Engine is falling behind and I gave  2  [ albeit very full-on ]  examples .... again Bill, doesnt it seem odd / concerning to you that S1V2 can effortlessly do what I am  " exampling "  [  woohooo a new word ]   yet  X1c Expanded cant ..... to my way of thinking, simple   "stress-testing"  like this is almost certainly indicative of other deeper problems
 
I'm  not  bagging CW  -  not at all  -  as I said before ...... I just want the Sonar engine to get better [ much better ] and my concern is that it simply is not a " Tier One " priority for CW ...... why it is not a priority just baffles me ....... and its even more baffling when new competitors [ as I have mentioned ] seem to be much further down this road than a  15 - 20 year old DAW veteran.
 
All the best Bill.
 
Tim
 
 
 
 

daryl1968


you're probably right musicroom - you're George Martin comment is absolutely correct. Necessity is the mother of invention - have you read "Here There and Everywhere" by Geoff Emerick?

I just looked it up Daryl - looks like a great read. I'll put that one on my Amazon wish list.


Thanks




I think we all see where the OP is going with this, and don't blame him. It's just not that important to me. I record seamlessly at 64 samples and feel grateful for the privilege to do so with practically no realistic boundaries. It seems a little spoiled to me to be impatient about this. For I know the audio engine in Sonar will always be on the improvement to-do list at Cake. Where they have it now is more than fine for me. But for the OP, I hope he gets what he wants. Just not sure it is a need at this point.

it ran all day long at 32  samples doing whatever you threw at it  [ back in the days of 8.5.3 ]

Now I'm confused. Are you saying they've made the audio engine worse in X1? I though it was 'untouched'?

i don't think they've "made the audio engine worse" but they've definitely added a load of extra overhead with new gui... and yes, after 20+ years i think it's time to rewrite some code...

Curiosity has got the better of me, and I've plenty of time today

timboe


FWIW - in the last year or so for reasons I dont really understand - I have been getting and doing a lot of Euro / Trance Dance.  Huge numbers of VSTi's all multi-outputed lots of convo reverbs etc.....  I can assure, at 48 samples, you can get to very high cpu useage very quickly ... anyway, I digress - maybe the folowing is easier to try:-

Try this test:-

=> Set your Buffer Latency to 48 or 64 Samples  [ 128 is OK but it will take a lot more tracks as per below ]

Left mine on my normal settings. Experience has taught me that settings tuned to a system is a very important part of what Sonar can and can't do

=> open a clean / new / empty project
=> insert x 1 mono audio track
=> MUTE it and enable REC and Input Monitoring
=> slide the Vol / Vel  all to  zero
=> insert  5 different Sonitus Plugs - Rev / Mod / Delay / Mutli Comp / EQ
=> load a preset for each Sonitus plug and enable each of them
=> copy / clone that one Track  10 / 20 / 50 / 100 etc....  times until your Windows Task Manager is sitting around  %50 <-> %60

Done all of that. Can't say I'll ever see myself needing 150 audio tracks all with input monitoring but whatever takes your fancy.

=> insert one more audio track with some audio on it - say a 4 bar drum loop - anything will do - and set it to loop
=> hit play in the looped section so that Playback does not stop / audio engine does not dis-engage
=> whislt the project is playing back, insert one instance of Rapture - does everything remain playing back without artifact or any sort of audible glitch ?

 Slight pause as I predicted

=> whislt the project is playing back, open any Rapture sound - does everything remain playing back without artifact or any sort of audible glitch ?

Yes, CPU at 75% no glitch or pause

=> whislt the project is playing back, replace the Rapture instance with one Dimension Pro instance - does everything remain playing back without artifact or any sort of audible glitch ?

Slight pause as experienced when inserting rapture

=> whislt the project is playing back, open any Dim Pro sound bank - does everything remain playing back without artifact or any sort of audible glitch ?

Yes, no pause or glitch

=> whislt the project is playing back, SAVE the project - does everything remain playing back without artifact or any sort of audible glitch ?

Some slight glitches on opening the save as dialogue and while the save takes place.

Again ...... for me, with my DAW, the above cannot be done with X1C Expanded.

Again, I'd check your system over..........

Agsin ......  purely of audio engine comparison, the exact same above at 48 Samples @ 88.2k / 24 Bit is perfect / glitch free in S1V2 - there is absolutly no change to audio playback and stabilty - it is utterly invisible and transparent.

Is working at 48 Samples really necessary especially if it stops you working? Drop it back a little.

Out of curiosity why do you need 150 plus tracks all with 5 effects on and input monitoring, got a very large live room?