Soft synth limitations ?

If the soft synth is multi timbral you should be able to do it by installing the synth as a First Audio Output track (if your Sonar is same as mine - I'm not sure exactly what my version is as it came with a controller keyboard) allocating different patches to different midi channels and then routing each midi track to the relevant midi channel using the track properties menu

Welcome to the forum.
As LordElpus wrote, some synths can accept input from up to 16 MIDI channels, such as Cakewalk's TTS-1. Others, like Cakewalk's Sound Center can only handle 1 MIDI track at a time.
And if you do find yourself running out of synth space, you can always Bounce the synth to audio, freeing up that synth to be used for another track.

Thank you all for the great help and precious advice you provide! :)

@JoeGBradford: "If the soft synth is multi timbral you should be able to do it by installing the synth as a First Audio Output track"

OK, I have this setting indeed, and I've tried this, but it's still the same: all tracks play the same patch, although I seem to have the ability to choose a different channel for each. I wonder how I will know for sure whether a soft synth is multitimbral or not...? I mean, will I see something specific in the settings, panels or whatever?

@LordElpus: "The TTS-1 Soft Synth may be what you need."

OK, this might be perfect, indeed. The sounds seem great (at least, enough for my all-purpose, generic use).
If I refer to: http://www.cakewalk.com/P...tor/feature.aspx/TTS-1
I understand that TTS-1 comes as a feature of "Music Creator 6"...?

@RobertB: "You might want to consider downloading MC6. MC6 itself is virtually identical to X1LE"

"MC6" = Music Creator 6 I presume? Music Creator is an enhanced version of Sonar X1 LE, right? So I could upgrade from Sonar to Music Creator 6? I'm sorry to ask, but I get a bit confused with all these software names...

@57Gregy: "And if you do find yourself running out of synth space, you can always Bounce the synth to audio, freeing up that synth to be used for another track."

I'm not sure exactly what you mean by this. Does that mean that I'd completely convert the MIDI/synth track to an audio track, so that I somehow freeze the track and get rid of all MIDI events to replace them with audio data? If yes, this is something I'd rather not do, because MIDI tracks are so convenient (lightweight, staff view, etc.). But maybe "bouncing" is a completely different process – and you can see how much I still have to learn about modern DAWs...

@RobertB: "Can you tell us exactly which synth(s) you are trying to work with? If you can provide a screen shot, that would be an immense help, as it allows us to see what you see."

I've tried with various soft synths, including Cakewalk Sound Center featured in Sonar X1 LE. I've also tested Acoustica Instruments, since it seemed to be the best GM sound bank I could get my hands on, as a VST plugin. But I couldn't figure out how to use one instance as a multitimbral synth, so I used as many instances as I could, until I reached the limit. I know this soft synth is featured as the default sound bank in a sequencer called Mixcraft, and it allows playback of GM MIDI files.
Anyway, thank you again for your help. If all it takes to solve my issue is get TTS-1 by purchasing Music Creator 6 (€40) as I understand, then I think I will follow your advice.

Use different channels for each MIDI track so they can use different patches.
In some synths, like TTS-1, you select the patch in the MIDI track; in others, like Sound Center, you choose the patch in the synth.
A multitimbral synth will have a mixer-type window with up to 16 channels in the mixer.
This is the TTS-1 GUI:
 

 
And when you Bounce a synth to audio, you don't have to get rid of the MIDI track; just change the output to... something other than it's current synth then Archive the MIDI track. That removes the MIDI track from processing but allows you to keep it in case you want to work on it again at a later date. Just unarchive it and make the needed changes, redo the output to the synth and bounce it again.

Glad we could help.
Just a few side notes. With regard to TTS-1, we have found it's best to select your bank and patch settings from the MIDI track, not from the GUI. Expand your MIDI track just a bit more than he did in the video to reveal these fields.
And I like Sample Tank, but that red default screen is a bit much. At the right edge, you have controls to change the color. I changed mine to a light blue background, and it's much easier on my eyes.

You'll have plenty more questions, I'm sure. Most of us did, and still do from time to time. Ask away. Somebody will be around to help.

Hi Deensay,
Glad we could help and that you are happy!  I am new to this forum too and I have found the people here are exceedingly helpful and supportive - quite rare these days!  Like you I was an amateur 'expert' * with midi back in my youth and I am thoroughly enjoying re-living it, but now I have a bit more money to indulge my hobby!!

 MC6 is a great piece of software and the interface is very similar to the top of the line X2 Producer, so if the bug does bite the upgrade path should be relatively painless.  I started out with MC6 on my laptop and a simple keyboard.  I have now upgraded to X2 Producer and have 'upgraded' a lot of hardware and I am in the process of building a dedicated PC just for my amateur recordings - once you start.....  
Gary

* Expert: definition: 'X' is an unknown quantity and a spurt is a drip under pressure! 

Good to see you making headway.
The administrator privileges thing is coming from W7.
In Windows Explorer, go to the Sample Tank folder and right click on the executable file(SampleTank2.5.exe)
Select Run as Administrator.
There's nothing wrong with WDM. If the drivers are well written, they can be as good or better than ASIO.
One may work better than the other, depending on system configuration.
Since WDM seems to be working well on yours, run with it.
Super low latency is good for recording audio, particularly vocal tracks, but as you've noticed, can be problematic with soft synths. Synths need a little headroom. 5-9ms is a good range. You will not hear or feel a difference, and it gives the CPU the time it needs to do its work.
Carry on, and have fun.