1-Latency issues for singers.
2-Trying to use direct mix with a VS-700 console and I/O.
3-Think I am following directions but it is not working.
4-If I turn ECHO on the track (Main Vocal 1), I get latency.
5-If I turn ECHO off, I hear nothing.
6-Settings are as in this photo.
7-Mic is plugged into input #8 of the VS-700 R.
8-Track input is Right VS-700 IN 7-8 (which I always use for a single vocal mic)
 
UPDATE: If track is armed and I record, signal is recorded, but I don't hear it going down.
 
UPDATE 2: I did not have mix directed to Main. Working with it to get levels sounding halfway normal.
 
https://www.facebook.com/photo.php?fbid=10154433271091486&set=a.10150103203311486.279563.549556485&type=3&theater

I get around the latency by plugging the mic into a mixer with good preamps and a direct channel out to the daw. I feed back the stereo mix to  a stereo pair on that mixer. This allows for talent controlled level of both the back track and themselves with zero latency. This permits me to run 1024 asio buffers while the talent hears no latency at all.  I even use a separate VSTi synth laptop running at low latency and it's own monitor sending also it's midi to the daw that is running latencies that would be otherwise unuseable.

First, I don't know why this is in the hardware forum when it is a Sonar software issue with latency; but that aside...
 
Using Direct Mix was a disaster.  It did remove latency, but suddenly every recorded track was drowning in reverb and things were coming out of different buses than those to which they were assigned.
 
Of course, NOT using Direct Mix means the latency is so bad the singer can't work.
 
Total mess.
 
I appreciate bitman's suggestion, but I may have to make a drawing to ensure I understand the routing.  This should not be this hard.  As loyal as I have been to Sonar, if other DAWs do not have such bad latency, I might finally decide to change software.

I can't see your posted picture ... and having used VS700R direct monitoring for several years, I don't really understand why it's such a mess in your case.
 
If you need to engage "input echo" on the vocal track in Sonar, you will be monitoring via Sonar FX (i.e. including all plugins that are in the project, even if they are not on that particular vocal track). The lowest ASIO buffer possible you can run your project on without crackles determines your actual latency.  So your latency depends ...

1. first of all on the plugs you use (and some are really bad for latency despite sounding great, e.g. WAVES Abbey Road Plates, some IR reverbs, ... some are absolute no-nos when tracking like linear phase plugs)
2. next important is your audio driver. VS700R is OK but not top, even with a strong DAW you will not be able to go below 8ms RTL, even in light weightprojects ...
3. Sonar is only third in line (try disabling all FX with the global FX button, and what will remain is only the latency from your audio interface) ...

Direct monitoring just means using the VS700 IO application (which you can also trigger from inside Sonar, I believe with VS700C buttons COMMAND+IO EDITOR), bring up the fader of your channel, unmute AND select the proper output at "DIRECT MIX OUTPUT" ... now your mic signal will go through the preamp, the DSP compressor (if you use it) and then directly out to your MAIN, SUB or DIGITAL1 output. There is no way that this can screw up your internal Sonar routing as this has nothing to do with it.
 
What you will get with direct monitoring, however, is only the try signal, which the vocalist may or may not be comfortable with, but I have also used hybrid monitoring where I had the direct signal from VS700R direct mix plus a reverb tail from a Sonar reverb bus (where latency becomes rather irrelevant since it's just the reverb tail and not the entire vocal signal)
 
 

I was able to get the dry vocal.  Regarding latency, are you saying to disable plug-ins on other tracks than the vocal track?  I have Lexicon Pantheon on three reverb buses (vocals, drums, and instruments) so I can send various tracks to them.  I don't believe I had any other plug-ins active.  All the virtual instruments were frozen.  I lowered the ASIO buffer to 3 on the slider--would have to look to tell you how many samples that was.
 
There may be no way this could screw up routing, but it did.  I was getting horrendous reverb on everything, including the kick drum, until I disabled all the reverb plug-ins on the three buses so we had to listen to all instruments and audio dry while cutting vocals. One of the string parts that was clearly being sent to the string bus was playing when the string bus was at zero.
 
At the lowest possible ASIO setting, the vocal latency gives a classic slapback sound, which is very distracting to the singer.

Plug-ins anywhere in a project can affect latency. In order to keep everything in sync all buses/tracks are delayed by the bus/track generating the most latency. Simply turning off plug-ins does not affect PDC. To eliminate PDC, the effects bin(s) must be bypassed. This may be done individually using the context menu in the FX bins or globally with the FX button in the Mix Module. PDC button may be used to disable PDC on the live track. Note: the links are to X2 help pages but the info applies to X1. The X1 help is not available on the Cakewalk web site.
 
At this time it may be easier to do a

• temporary bounce of the all the effects and instruments,
• archive all the tracks,
• bypass all the FX bins on the buses
• and use the temporary mix when tracking the vocals.

This should eliminate any surprises from the backing tracks.
 
The slapback echo on the vocal track is the result of mixing the direct signal and the signal through the DAW. Not owning a VS series interface, I cannot give you specific guidance with you Direct Mix issue. In general, ASIO buffers do not affect direct monitoring because the signal being recorded does not go through the driver. The input is directly routed to the output and mixed with the signal from the DAW.

Thanks for all the information.  In regard to the ASIO buffer, here's what I meant:
 
- A high ASIO buffer causes a lot of latency but better quality audio
- A lower buffer can cause some pops or clicks but reduces the latency.  This works fine for me, but wasn't enough for the singer. 
 
I do understand that the ASIO buffer doesn't have anything to do with the direct mix results.
 
** THE PDC ADVICE IS VERY HELPFUL.  That is first on my list to try.  Good stuff. **
 
To be clear, hitting the OFF button of the plug in the audio FX bin is not enough?  That's the only way I know to turn a plug in off (other than the keyboard shortcut e which essentially hits all the OFF buttons at once).
 
Thanks again for taking all this time to help.

konradh
To be clear, hitting the OFF button of the plug in the audio FX bin is not enough?  That's the only way I know to turn a plug in off (other than the keyboard shortcut e which essentially hits all the OFF buttons at once).

Turning off plug-ins does not affect PDC. Shortcut "E" does the same thing as the FX button in the Mix module, it bypasses ALL FX bins. It is possible to toggle FX bin bypass individually with the FX bin context menus. I don't have X1 installed at this time but X2 is similar. Here is an image of the FX bin context menu from X2.

Archiving tracks also bypasses the FX bins in the archived tracks.