Latency on new install, new laptop.

What are the possible causes of latency?
Which settings should I be aware of within Cakewalk and also with regards to any Windows background processes?
I have my CME UFS controller linked by midi cable to my Babyface and this set-up worked fine on a previous project on a Lenovo laptop. I did a test with 4 tracks of midi, 4 different instruments and I have to physically edit every note as they don't go where I played them.
Can anyone advise me or point me in the right direction to find out this info?
Thanks in advance

what's different between this project, and the previous one that had ok latency? a whole different laptop?

Many folks have latency issues with laptops, that are caused by the presence of a Wi-Fi adapter for the laptop to access the web.  For some reason, many Wi-Fi adapters cause massive DPC Latency spikes, which cause all kinds of nasty dropouts and such in applications such as Sonar.
 
So, a real easy thing to do is to get the Wi-Fi adapter turned off, through some method, just prior to launching a Sonar session, testing with Sonar, closing Sonar, and then enabling the Wi-Fi adapter again to conclude the test.
 
Some laptops have a little actual switch, by which you can simply turn off the Wi-Fi adapter.  This is the easiest method.  Another way to do it is to go into Windows Device Manager, then find your Wi-Fi adapter and change its status to 'Disable'.  (do not uninstall it, just disable it).  Then after testing things in Sonar, go back into Device Manager and simply change the Wi-Fi adapter's status to 'Enable'.
 
So, please give the above a try and then post back on what happens when you play back or record in Sonar when the Wi-Fi adapter is disabled or turned off.
 
In addition, four other likely places to look for causes of latency in projects in Sonar are: settings, project plugins, audio interface, and/or use of on-board sound.
 
1.  Settings: sometimes, settings can cause latency problems.  Setting the ASIO Buffer Size too high OR too low can cause dropouts and such.  For recording purposes, I have my ASIO Buffer Size set to 128, at 48k for a Sample Rate.  When you switch into the Mixing phase of your project, you will likely start to use more powerful plugins and they will need a bigger ASIO Buffer Size to do their 'thing' (look-ahead processing and/or more CPU intensive).  When I move to mixing a project, I bump my ASIO Buffer Size up to 1024.  For the rest of time, you will go back in and switch ASIO Buffer Size to be either low for recording, or high for mixing/mastering.
 
2.  Loaded Project Plugins.  There are types of plugins that are just not meant to be used during tracking/recording, because of the nature of what they do.  Cakewalk's Boost 11 is an example of one such plugin, as it uses 'look ahead processing' to read the data ahead of it, to know how it needs to process it.  That look-ahead processing cause a big jump in latency, and this plugin should NOT be used until you begin the mixing process.  There are others, from Cakewalk, and from 3rd-party companies.
 
You can quickly test to see if part of your latency issues are caused by one or more of the plugins that are loaded into a given project.  Sonar X3 has a shortcut key, which is the letter 'E' on your computer keyboard, and when you press 'E' Sonar will toggle off/on all loaded plugins in the project - it will cause Sonar to 'Bypass All Plugins'.  So, load up your problem project, start play, and hit 'E' to have Sonar bypass all the loaded plugins, and see if that also makes your dropouts go away.  (after testing, hit 'E' again to tell Sonar to stop the Bypass, so you can hear the effects again).
 
So, IF your dropouts go away with the plugins bypassed, then you KNOW that one or more of your loaded plugins are causing the problem.  MY advice is to temporarily swap out or remove any of the plugins you find to cause dropout issues when tracking, and then when you finish your actual recording, and move onto the mixing phase, THEN you can load up whatever plugins you wish, as long as you also remember to crank up the ASIO Buffer Size to give it enough room to DO things like look-ahead processing, etc...
 
3.  Audio Interface - if you are not using one, you will continue to have lots of trouble with running Sonar with ANY real complexity or even number of tracks in projects.  These range from about $140 - $lots, depending on features and quality of converters and such.  If you DO have an audio interface, make sure you have the latest proven version of drivers and firmware available for the audio interface, and make sure those drivers are expected to work in whatever version of Windows you are running.  Usually, the ASIO drivers are the ones you will want to use, though for some reason some folks on some interfaces use WDM drivers and report better success than with using ASIO.  I believe that percentage of folks is likely pretty small, so I would suggest trying it with the ASIO drivers first.
 
4.  On-Board Sound.  Some folks have audio interfaces, but for some reason still have Sonar pointed to using the computer's on-board sound, rather than the drivers of the audio interface.  I do not know why someone would do that - I do not suggest doing things this way.  Some folks also report that they disable their computer's on-board sound.  I do not myself have any issues with setting Windows to have the Default Audio Device set to go through my on-board sound, OR more likely to the speakers built into my HDTV that I use for a computer display (through the HDMI cable that connects my computer to the 32" HDTV).  Some folks do as I do, and some others have separate computer speakers that the connect to the back of their computers - lots of ways of doing things.  Bottom line is that I recommend having SONAR use the ASIO drivers of a dedicated audio interface.
 
If you have waded through all of the above, I suggest trying to run with settings similar to mine, which are:
 
ASIO Buffer Size: 128 for recording, and 1024 for mixing/mastering
 
Driver Mode in Sonar: ASIO
 
Sample Size (must match between Sonar and audio interface) either 44.1 k or 48 k (yes, I know some others run way higher - I just want to get this person stable first)
 
Bit Depth for Recording: 24-Bit
 
There are also a couple of buffers that changed from what they used to be in prior Sonar versions (used to be 500 ms or 512 ms), and now the default is 250 ms.  I changed my buffers to 500 ms like they used to be 
 
I hope the above helps, and that I haven't had any senior moments and forgot things.  If I did, I am sorry.  I DO need more coffee.
 
Best of luck with you, I will keep an eye out for any posts you respond with.
 
Bob Bone
 

You'll figure it out. Low latency on a laptop is like a combination lock - all the tumblers have to be in place before it will open.

Anderton
You'll figure it out. Low latency on a laptop is like a combination lock - all the tumblers have to be in place before it will open.

percepto - a word of encouragement.  I have a laptop, appear to have similar specs.  For me, all works well with next to no latency.  As Bob highlights approaches above, and Craig mentions, you'll get it figured out.  It will work.  Laptops are great for this kind of thing.

Anderton
You'll figure it out. Low latency on a laptop is like a combination lock - all the tumblers have to be in place before it will open.

You got that right!
I installed Platinum and everything went haywire!
It changed my audio and driver settings (everything was set to play from my Babyface ADAT output and it took me a while to figure out why I wasn't getting any sound!) and I had to start from scratch because the transport was running erratically and slow and I was getting glitches and audio pops. I am using no plug-ins and the sequence I am playing is very minimal, drums, bass, organ and synth.
I re-read everything here again and then started playing around with the settings without really understanding exactly what I was doing!
After about 1 hour of trying I ended up with the following settings and no glitches:

Audio Driver Bit Depth = 24
64-bit Double Precision Engine = ticked
Sampling Rate = 96000
Effective Latency at 96kHz/stereo: 2.7 msec
Asio reported latencies:
Input: 3.2 msec, 311 samples
Output: 3.3 msec, 319 samples
Total roundtrip: 6.6 msec, 630 samples
Playback I/O Buffer Size 256
Record I/O Buffer Size 256
Buffer Size (Latency) RME Babyface = 256 samples


I would just love to actually understand precisely which setting makes the difference, it is really frustrating, I thought I was reasonably intelligent, but this stuff baffles me!

If I was going too start mixing now and adding fx, is there anything I should change in the settings above?
Is there any article or post that you can point me to to explain these settings in depth?
Thanks for any input.

Boydie
Unless you have done it for a specific reason I would personally knock back the sample rate from 96 to 48

No, I didn't do it for a specific reason, as I said, trial and error. If I knock it down to 48 or 44.1, the latency doubles! Also when I try knocking the buffer size down to 128 the audio glitches and pops terribly. 
Could it be that my sound card "likes" working at 96000?

Thanks Bob, but as I already wrote, I can't make my system work with these settings!

I will definitely try the Wi-Fi adaptor fix and the tip about changing settings for mixing.
Cheers

Thanks Bob, of course I am not frustrated with anything you have written! I am always amazed and grateful when people much more knowledgeable and experienced than I, take their time to offer help and suggestions.
I will definitely try all your suggestions tomorrow and post the results.
However, is there something wrong with running Sonar at the settings I posted? As I said, it is now running smoothly, so it may be a case of "if it ain't broke…"
Cheers

percepto
Thanks Bob, of course I am not frustrated with anything you have written! I am always amazed and grateful when people much more knowledgeable and experienced than I, take their time to offer help and suggestions.
I will definitely try all your suggestions tomorrow and post the results.
However, is there something wrong with running Sonar at the settings I posted? As I said, it is now running smoothly, so it may be a case of "if it ain't broke…"
Cheers

If things are actually working smoothly, that's the important thing.  What though do you mean by smoothly?  (96 k does chew up disk space, though).
 
DO run LatencyMon though, and see where things are at.  It won't change or hurt anything.
 
Bob Bone
 

I just glanced thru the thread, so I might have missed this, have you patched your 840 EVOs?
 
http://www.computerworld.com/article/2836082/samsung-delivers-fix-for-ssd-slowdowns.html
 
It could be a contributing factor ...

Thanks so much for hanging in there with all of this - drives me NUTS when I am trying to record and something is glitchy somewhere. :)
 
If everything is working OK for you with 96 k and 256, AND you have a resulting low latency reported by Sonar for doing your recording, then there is absolutely nothing wrong with running your system that way.  Many folks choose to run at those settings anyways, so again, you are fine with that set of settings.  It WILL chew up a bunch more disk space because you are sampling twice as much data, and I would imagine processing it all would maybe take longer (not sure on that, because I don't run at 96 k), but if you have sufficient disk space, cool.
 
I remain curious as to what is making the difference when you were trying to test at 48 k and 128.  During that set of tests, the Wi-Fi was disabled?  (just want to make sure I understand the setup for the test).  And, also for that particular test, what plugins were loaded into the project?  Sometimes, specific plugins can induce a bunch of latency.
 
Lastly, if you haven't done it, and don't mind investing a couple more minutes into obliging me, I would like you to try one more test at 48 k and 128 buffer size, ASIO driver mode.  THIS time, please make sure that no effects at all are loaded/enabled.  Now, if your project HAS loaded effects, I am not suggesting you go through the hassle of removing them - I would suggest instead that you simply bypass the effects during this test.  To do that, you would hit the letter 'E' on your computer keyboard, and this toggles the Bypass All Effects off/on.  So, the test should be run with any effects bypassed, to keep it as pure as possible.  AFTER the test, hitting 'E' again will once again enable any effects to process again.
 
It's entirely up to you if you want to try the above - since you reported success running at 96k/256 you could simply run at those settings, but if things also work at 48k/128, then you could save some disk space and possibly some processing time when rendering things.
 
Also, that IS good news about LatencyMon finding your system OK for streaming audio applications, such as Sonar.  I didn't go back to see if your Wi-Fi was disabled when you ran LatencyMon, but in any case, anytime I am using Sonar on a laptop, I always make sure to disable Wi-Fi just prior to running Sonar, then enable it again when through.  I did see where you reported some other things clearing up with it disabled, so I again recommend running it that way when launching Sonar.
 
Bob Bone
 

I would like to get this sorted as I don't want to make the CPU and HDs work harder than they need to by running at 96k.
I tried it at 48k in Asio mode but I was still having problems, also with the Wi-Fi adapter disabled.
I will always run Sonar without the Wi-Fi adapter as it makes sense to eliminate a possible problem and it can't do any harm anyway.
I have no plug-ins or fx running, it is a very simple sequence of drums, bass, organ, synth (soft synths) and 1 x audio track (guitar)
The really frustrating thing is that the problem is intermittent.

I am now trying the WDM drivers. It seems to have stabilised now but I don't know really know why.
I set it at 24bit, 44.1k.
The buffers in playback queue is set at 2. I have no clue what this setting means.
I raised the buffer size slider a tiny bit towards the "safe" side as I was still getting problems on the "fast" setting. I only moved the slider a little so that the latency was still under 10. 
I have a latency of 8.7msec which seems acceptable and the distorted sound has gone.
When I did the Wave Profiler (WDM driver) it shows me that my sound card only has 44.1k available, all the others are unavailable.
There is just so much of this that I don't understand, I would like to comprehend the changes that I am making.
 Thanks for the link to the article, I will take a look tomorrow.

Hey Bob,
Yes, that is a thread from a few years ago and I just read it through and, quite honestly, I am a little shocked!
There was definitely a miscommunication somewhere, and I may have been at fault, (I honestly can't remember) but you made it sound like I was deliberately avoiding giving information.
I have consistently noted my gratitude at anyone offering their time and help, I really do appreciate it and I can't understand the animosity directed at me in some of your final posts in this thread. In all honesty I can't even remember reading those final ones as I am sure I would have responded then in the same way I am doing now. I wouldn't have simply left the issue hanging in the air like that. It is clear from your posts that you think I wasted your time, but I really was not aware of this and I apologise for any misunderstanding.

The issue then was definitely the laptop itself as I ended up using the Lenovo laptop to write and produce a song with no complications, latency or whatever. Unfortunately that situation didn't last and I had to give the laptop back and I had a bit of a hiatus of a couple of years where I didn't use Sonar. Now as you can see I have a brand new custom-built laptop (my own!) as I once again am about to write some music.

So, back to today.
As I said, I am using the WDM drivers and the system seems stable. No pops, glitches or drop-outs, but I am literally just playing the patterns in Matrix mode, no fx or plug-ins.
I don't want to start writing until I at least have a stable system, which is why I am only now installing Komplete 10 Ultimate after purchasing it over 2 weeks ago!
I will then be in a position to really test my system.

I am now reading all the info from the links you sent me and going back over the previous post as there are a few links there that I have to re-read.

These are my current (stable) settings:

Driver Mode: WDM/KS
Dithering: Triangular
Share Drivers with other programs - ticked
Use Multiprocessing Engine - ticked
Use MMCSS - ticked
Always Stream Audio Through FX - ticked

Playback Timing Master : Babyface Analog (1+2)1/2
Record Timing Master : Babyface Analog (1+2)
Audio Driver Bit Depth: 24 64-bit Double Precision Engine - ticked
Sampling Rate : 44100
Buffers in Playback Queue : 2
Buffer Size: 8.7 msec, 384 samples

FILE SYSTEM
Enable Read Caching - unticked
Enable Write Caching - unticked
Playback I/O Buffer Size (KB) 512 (I changed this from 256 on Bob's advice)
Record I/O Buffer Size (KB) 512  (I changed this from 256 on Bob's advice)

I still don't know what the "2" setting means in "buffers in playback" and whether or not that should be changed. I also don't know why no frequencies other than 44.1k are available on my RME Babyface sound card.

Do these settings look correct?