Please check me on this Sonar ASIO panel fundamental question..

Dear forum people,

I have had this doubt, but today I spent some time testing it, and would like to be checked if my conclussion is ok.

1. I do not use any VSTi in sonar for what I play.

2. I do not use any plug-ins in sonar for the voice. I have them outside of sonar. Sometimes hardware, but lately software. I don't need to set Echo-On to hear my voice, nor to hear the instruments I am playing. I have no problem monitoring.

3. All my effects, synths, etc., are stand-alone, except AFTER recording, on playback, and those are for Audio only; the recorded MIDI tracks playback exactly the same instruments, which ARE NOT VSTi's in Sonar. So, all sounds are EXTERNAL to Sonar.

4. I use Sonar LE to record, simultaneously, one track of voice, typically 9 or so MIDI channels on two tracks, plus the total of the sounds played and the voice in one track, whose resultant I don't really use or probably need.

Now, based on what I had learned here, and so forth, I have had the Options for Audio/ASIOpanel low, like 7 or 10 milliseconds "to avoid latency", which for my case as explained, I am changing that criteria, as follows.

I am pretty certain, now, unless corrected, that in my case, it is not relevant for me to try to keep the setting as low as I had it, in the ASIO control panel, and in fact, I have increased it to 20, 40, and even 60 milliseconds, for now I believe that in my case it is even better to have more time for Sonar to do its RECORDING job. I don't now believe that it doees really AT ALL relate to my latency on what I play or sing or singing. I don't use a USB interface. I have external mixer/pre-amp for the only external sound which is vocals, and sometimes guitar or harmonica. In fact, if I keep the ASIO value of Sonar low, it probably uses more resources needlessly.

So, now I am poised to leave the ASIO settings for Sonar, on the high side, for the reason said.

The voice monitoring I do is via the card's hardware, so it has no latency.

After recording, I play the MIDI and it plays fine, without any noticeable latency beyond what the stand-alone synths or emulators I use, do have for latency, which is not noticeable. I have those at ASIO 10 milliseconds, although they seem to be fine at 7. Those instruments do have high demands on CPU and ASIO, though.

When I RECORD, the Sonar CPU indicator stays constant real low, like .03, and when I play back it jumps to 14 percent or so. I add maybe 2 audio vst effects to audio once they are recorded, for playback, or for bouncing. But that, I don't believe cares for the latency. The CPU increases like 2% for each audio VST I add to raw voice, and sometimes to audio tracks that I do render from the MIDI tracks, after recording.

So, unless I am missing something, I am jumped  in Sonar ASIO from the 7 or 10 milliseconds I had before for a long time, to 60 milliseconds or so, for I think this is even better for the way I use Sonar, unless someone has some comment that corrects my view.

In other words, Sonar is basically a recorder for my use, and so its ASIO setting is to be higher, rather than as low as possible. Does this make good sense?

Thanks for any replies.

+1 to Matts post. It covers pretty much everything. I just wanted to add some simple statements...
If your not using the effect in sonar to monitor, you do not need a low latency
The higher your latency (ASIO buffer), the less your CPU works and the less your CPU % is.
The lower your latency(ASIO buffer), the more your CPU works and the higher the CPU % is.

The higher your sample rate, the lower you can set your latency. You can achieve lower latency with a sample rate of 48kHz ,then with a sample rate of 44.1kHz

Those are just some "fun" facts...
Cj

Philip


 

I didn't know that!  Does anyone currently record at 48kHz (or higher) for vox tracks?

I'd bet on that most forumites use 48 kHz (but only a few euros).
It doesn't matter if it's vocals or tuba. If your project is set for 48 kHz then that's the only one you can use. 

I don't use a USB interface. I have external mixer/pre-amp for the only external sound which is vocals, and sometimes guitar or harmonica. In fact, if I keep the ASIO value of Sonar low, it probably uses more resources needlessly. So, now I am poised to leave the ASIO settings for Sonar, on the high side, for the reason said. The voice monitoring I do is via the card's hardware, so it has no latency.

 
 
What card are you refering to? How do you get your Midi into the computer? How do get your audio into the computer?
Somethings missing in your post.

Crg

I don't use a USB interface. I have external mixer/pre-amp for the only external sound which is vocals, and sometimes guitar or harmonica. In fact, if I keep the ASIO value of Sonar low, it probably uses more resources needlessly. So, now I am poised to leave the ASIO settings for Sonar, on the high side, for the reason said. The voice monitoring I do is via the card's hardware, so it has no latency.

 
 
What card are you refering to? How do you get your Midi into the computer? How do get your audio into the computer?
Somethings missing in your post.

Crg, "no usb interface" means no usb AUDIO interface..I get the MIDI in from two MIDI keyboards, simultaneously, one is an Axiom-61 and the other is a Roland-A33..The roland goes in via the Axiom, using the Axiom's MIDI-in port, so it results as another MIDI-ASIO input, pretty much independent and very efficient, even though the A-33 roland keyboard has no USB port (it connects to the Axiom via a classic MIDI cable that the Axiom gladly accepts and sends via its USB cable.

The 9 or so simultaneously recorded MIDI channels are the product of an arranger type software called OneManBand. I use Tyros Yamaha styles, and they follow my playing in real time. This is the software equivalent of typical Yamaha arranger type keyboards. This produces quite a bit of MIDI activity and sound generation.

At the same time, I play chords and melody using for th top manual-keyboard sounds from the B4, and Piano or other synth sounds from Proteus VX. ..By pushing buttons, I can instantly play both B4 manuals upper and lower, and also the MIDI pedals (another source of MIDI)..These pedals feed in through a Turtle-Beach MIDI-USB converter (the MIDI pedals I built myself from e-bay electromechanical swithches of old hammond home organs).

I also use a foot pedal that goes into the expression port of the Roland A-33, and that sends A LOT of MIDI data when used.

I have the following sound modules running: 1. The Proteus VX (stand alone) responds instantly as a Genera MIDI module, even though it is sample based. 2. The B4 software of Native instruments (standalone), and a third sond module that I somehow got from Japan which I now have hacked and run it as a VSTi but effectively runs as a "stand-alone" by the use of VST-Host. I use this japanese soft synth because it gives me pretty good drum kits being XG compatible, and drum-kits switch automatically via bank-patch select, instantly.

For the mic and voice, I run TWO VST effects, also independently from Cakewalk, using Savi-Host. One provides "ambience" and the other an effect for voice of my choice. They are from Voxengo and Ambience, free ones.

All of this is currently on my laptop, which has an older Audigy2 ZS card. This card has three physical outputs in stero pairs: Front Stereo, RearStereo, and a third for Bass. All of the stero pairs, though, are available via its ASIO-2 driver for Sonar at the individual level..Like Front-left, Rear-right, Center Ch, and so forth, and can be used  actually for any thing.

It has only one input for analog audio: but it works very well and can be Mic-level or Line-level, also left and right are independent.

The good thing is that this card ADDITIONALLY has many other internal hardware inputs and outputs that are also available for Sonar independently.

Another good thing it has, is that in addition to the above said analog otuputs, it has ALSO a stereo output for the earphone (or headphone)..THIS beauty of headphone output is the one that allows me to hear every thing that is playing and monitor it, WITHOUT ANY  LATENCY...The other otuputs go to my amplifiers for speakers use.

So I can monitor my voice and all the sounds being produced, while sonar can record the mic only in a separate audio track and it DOES NOT record the effects I use to guide myself when singing.

In Sonar, I record simultaneously the following tracks typically:

1. Records the Mic raw voice. Echo is OFF. (This track, though, has TWO VST AUDIO effects that I use for playback, or whatever.

2. Another track records all of the 9 or so MIDI channels of the accompaniment produced by OMB, which follows my playing.

3. Another MIDI track records only my key-strokes simultaneously for the upper and lower keyboards..I don't need to record them in a separate track (I could), but they are differentiated by being in different MIDI channels. Those are channels not used by the accomp, so every thing is independent.

4. Another track records simultaneously ALL the performance's AUDIO. This is exactly what I hear when I monitor by headphones.. I hardlly ever use this one..I just keep it for reference..But I can play it back simultaneously, along with the MIDI tracks playing back the original MIDI sounds modules, if I want to. This doubles the voice, also, and it is useable too, but I have so far prefered to play back every thing from the MIDI tracks.

All of this uses a complicated (for me) routing that uses MIDI-Yoke and MIDI-Ox. I say complicated because took me a long time to figure MIDI-Ox's mor advanced capabilities called MIDI mapping and MIDI-routing. But this allows me to record and sonar all of this, push the Stop button, and then Push the Play button, and hear and see all the devices being activated back EXACTLY as I played them, withut having to do anything to any track in Sonar. Of course then I go into Sonar to tweak the tracks, mute what I want, edit, or whatever.

So, as can be seen, the job of Sonar is basically that of a recorder of MIDI and of Audio. No sound producing whatsoever is under its control. The voice is already there and just given to Sonar for it to record, the Full audio performance is also already there for it to just record, and the MIDI data it just receives it for it to record it.

So sonar is doing rather simple tasks while I play and record all of this. These two tasks are a)Recording Audio and b)Recording MIDI...And I have verified that these two basic functions are a piece of cake (Cake-walk) for Sonar, so if the Sonar's latency is 20 or more milliseconds, it is better so than to have the latency at 7 or so miliseconds, because its job DOES NOT HAVE TO BE in real-time as I use it. The additional delay I have given to Sonar is time to do its job with less CPU demand.

Now when I play back, it's basically the same criteria...Sonar's job give the MIDI data back to the stand-alone instruments, just as when I played them, and it plays the wave of the voice, this time I do have AUDIO VST-effects for the voice, and those do not seem to burden Sonar into delays or such things..Each of those Audio effect VST's add 2 or 3 percent of CPU load...My CPU load total is typically 70%. The heaviest user is Proteus VX.

I have SonarLE set for  48KHZ (native for the card) and 24 bit depth. The card is capable of more and less, but this is reasonable for me.

I also run simultaneously with the above (when I want) the Battery-3 demo, or/AND the Adictive Drums demo..I have not bought any of those, because so far they have not convinced me that thye will work like I need them to, since the demos are limitted in ways to that effect.

I beleive that what I say relates mostly to my idea of using only Stand-Alone sources of sound, rather than have Sonar handle the VSTi's for the same things, PLUS the use of MIDI-Yoke and MIDI-Ox.

And, as said, either VST-Host or Savi-Host can be used to provide vocal mic effects ouside of Sonar.

By the way, this card also has many ambience and hardware vocal effects that I can chose to use in the same way as the software vst's that I mention above. I was using that aproach for a long time, but now have found it more convenient to do it with the software free VST effects.

I went ahead on so much detail, for it were the case for some to find any useful info from this limited setup, but can in some ways apply to more advanced users and better equipment too, possibly.

And I am open for confirmation, or correction, in the spirit of common interest.


PS. I do use external mixer/mic preamp where I can put various types of mics, be it XLR, or 1/4 jack, 48 volts, and so forth. This cards don't work wiell without those, of course, for they don't really have those type of  receptacles, nor proper mic-preamplification.

It sounds to me like you have a really good handle on how latency can affect your work-flow and for you it is not critical to have it as low as you can go and you've already found that the figures you are quoting sound in the realm of good sense for achieving stability.

I think on a forum such as this you can be forgiven for thinking that the lowest latency is the holy grail of daw recording as really low figures shows that a system can cope under extreme circumstances should it need to.

The really important thing though, which you seem to already have going on, is the ability to be able to work comfortably and reliably in order to perform the tasks you need to without having to worry or even think about what the system is doing while you are in the process of creation.

What more can you ask of your DAW?

I guess that depends on whether you want to spend your time tweaking to achieve great performance figures or whether you'd rather spend that time engaged in the process of production.

For me it sounds like you have the perfect understanding on the known trade-off between performance and stability WRT to latency and now you're ready to rock!  It's working for you so work it...

Hi brundlefly,

Due to the lengthy explanation, the above may not be too clear, so I believe those concerns you mention, wouldn't be the case here, due to the folowing:

The cables are virtual cables (MIDI-Yoke MIDI-Ox), and every thing is running in the same laptop.

What gets recorded is pretty much exactly the same quality as I hear when Sonar plays back what's recorded, both when it's wave, or when Sonar sends the data to be played by back by the same MIDI instruments that were played.

The only analog signal is the vocals from the mic, or guitar, or harmonica when I use those, and those are OK too. I have a decent AKG C-535 EB mic, and I also use a lapel type Telex that I have managed to attach to a set of AKG-240 headphones. Both give pretty good results, but I am still learning by testing audio VST effects.

One thing is that I can not bounce MIDI, but it does not matter much to me, because I can record as I play after making usual post-recording adjustments. I can bounce audio tracks easily, though, or the wave from MIDI instruments only, too.

It also helps that I have hardware sliders for each midi channel volume, when making those adjustments. I adjust the self-standing volumes directly with those, rather than each track of MIDI in Sonar, for a final recording. In making those adjustemens, it also helps me a lot that Proteus-VX shows all 16 midi channels very clearly, and the voulmes I set..So, if I want to, I can go and put those volumes in the sonar tracks too, afterwards.

I am glad you mention that you find the aproach of higher latency for Sonar in uses as I have, since being relatively new I did find this somewhat of removing a false notion that I somehow had gathered and was stuck to. So I appreciate your concurrent view.

Crg, "no usb interface" means no usb AUDIO interface..I get the MIDI in from two MIDI keyboards, simultaneously, one is an Axiom-61 and the other is a Roland-A33..The roland goes in via the Axiom, using the Axiom's MIDI-in port, so it results as another MIDI-ASIO input, pretty much independent and very efficient, even though the A-33 roland keyboard has no USB port (it connects to the Axiom via a classic MIDI cable that the Axiom gladly accepts and sends via its USB cable.

 
The Axiom sends ASIO via USB? Or just Midi? You say it results in another Midi-ASIO input. Does it go directly into Sonar or through the Audigy? I think it's probably just a Midi input. You can't have two ASIO drivers active. The Latency with your soundcard, which should be your only ASIO device may not be tweakable to the extent you need. It's going to be something you're going to have to experiment with.
Monitoring via the input card is fine but do you hear the Axiom and Roland when doing so? I'm thinking your large amount of latency is due to some routing through the soundcard.  
 
Edit: It seems to me you should be using Input Echo for the Axiom setup. 

Crg

Crg, "no usb interface" means no usb AUDIO interface..I get the MIDI in from two MIDI keyboards, simultaneously, one is an Axiom-61 and the other is a Roland-A33..The roland goes in via the Axiom, using the Axiom's MIDI-in port, so it results as another MIDI-ASIO input, pretty much independent and very efficient, even though the A-33 roland keyboard has no USB port (it connects to the Axiom via a classic MIDI cable that the Axiom gladly accepts and sends via its USB cable.

 
The Axiom sends ASIO via USB? Or just Midi? You say it results in another Midi-ASIO input. Does it go directly into Sonar or through the Audigy? I think it's probably just a Midi input. You can't have two ASIO drivers active. The Latency with your soundcard, which should be your only ASIO device may not be tweakable to the extent you need. It's going to be something you're going to have to experiment with.
Monitoring via the input card is fine but do you hear the Axiom and Roland when doing so? I'm thinking your large amount of latency is due to some routing through the soundcard.  
 
Edit: It seems to me you should be using Input Echo for the Axiom setup. 

To be crisp, ASIO is a technology for audio I/O only it does not do MIDI.   If a device has both audio and MIDI it will require two drivers in Windows, one for audio and one for MIDI.  It may "appear" like it's all one driver because they both install at the same time but they are not connected in any way.

Also, the audio I/O can have more then one driver installed and available. For example a single interface may have a MME, WDM and ASIO driver installed in Windows, all available at the same time. 

The MIDI part of the interface will only have one driver in Windows.

Due to the lengthy explanation, the above may not be too clear, so I believe those concerns you mention, wouldn't be the case here, due to the folowing: The cables are virtual cables (MIDI-Yoke MIDI-Ox), and every thing is running in the same laptop.

Okay. I get it now. But I still don't see the point. What does this buy you other than the inability to save and load a whole project as one file with all the soft synths, effects and settings preserved in one place?


It all seems pointlessly complex and prone to configuration SNAFUs.





 

brundlefly

Due to the lengthy explanation, the above may not be too clear, so I believe those concerns you mention, wouldn't be the case here, due to the folowing: The cables are virtual cables (MIDI-Yoke MIDI-Ox), and every thing is running in the same laptop.

Okay. I get it now. But I still don't see the point. What does this buy you other than the inability to save and load a whole project as one file with all the soft synths, effects and settings preserved in one place?


It all seems pointlessly complex and prone to configuration SNAFUs.



Pointless? No, not at all.

As I said, I save the whole project easily and when I open it, all MIDI, all Waves are there as a TRUE representation of everything I play and plays along with me.

Every MIDI nuance,  I play, including patch changes, from hardware controllers, and Keyboards is recorded. All waves are recorded exactly as to vocals, and any thing I typically do, without any worry of latency. Each sound source has its own ASIO panel setting, unlike having Sonar have ONE ASIO panel that relates to all sound sources.

My view is performance first, recording second. The way I have it is I can start playing and do vocals without having to load Sonar..Why load Sonar if I just want to play music with my system?

THEN, at any one moment, even without stopping play, I bring up Sonar via a template, play does not even need to stop. Then I hit the RECORD button, and instantly, recording starts.

I finish playing and hit the Sonar STOP.

Then I push the PLAY button, and Sonar plays back EVERYTHING exactly as was being played. No need to go into any track, even though everything is there, recorded on the sonar tracks, for individual tweaking as wished.

Perhaps for most, the main goal is to record, in pieces and fragments, but for me, as said, it is a little different.

Now, if I only want to record acoustic quitar and vocals via mic, I still use the SAME template, and exactly the same procedure.

It is complex because my setup is in most ways the equivalent of a software arranger, and I have it so, because a typical off-the shelf arranger is basically tied to its own built in sound module, and its hardware buttons and sliders are pretty much dedicated to its factory function.

The way I have it, it still works as an arranger, but uses various sound sources with more versatility than I get...With VST'is, it'd be even much more complex, if it even were possible.

Again, it was a lot of work and learning, but I do have Sonar basically working as a rather simple recorder of MIDI and audio, even though yes, it was complex to get it to that simplicity. And, no way I'm saying I am done yet.

Thanks for your comment and concerns, I hope some are seen in a little different way after this explanation.

Brundlefly,

Among many other that I see no point in mentioning are:

1. I use only ONE template

2. On any thing I have already saved as a project, I can change, without going into Sonar the sound characteristics, meaning I can change easily from software to hardware synths, or viceversa, without having to alter tracks, because I based everything in Genera MIDI sounds. So the buttons I push por patches will always make sense, no matter what kind of GM synth synth I use.

3. I can still, if I wanted put VST's in any MIDI track if I wanted. But that is too cumberson for me.

4. Each Yamaha style of thousands that there are and hundreds that I have from years of using them, DEFINES its own instrument patch..So imagine if I had to load a template for each? Imagine if I had to tell each of the many MIDI tracks it uses what patch to use?

So I do have external hardware MIDI XG/GM Yamaha sound module, and I can easily replace the Proteus VX softsynth if I wanted to, and it would sound OK. And the better sounding it would be, the better result, WITHOUT having to worry about a VSTi in each track of each project.. That would really be horror for me.

This way I have as close as possible as a band of musicians ready to play with me any time, and recording the total perfomance and worry about tracks and vsti's for it is not there.

Thanks , again for the observations, but they really do not apply to what I have described, so far. I do realize that this I do is not the most common, and I hope you understand that, for, even if I wanted to, I couldn't force it in.