I have 2 interfaces.  My main is a Focusrite Scarlett 6i6 but I still have my old UX2 Pod Studio. 
The Scarlett requires being plugged into wall AC while the UX2 gets it's power from the computer.
 
So, I'm wondering why, on my Scarlett, I can't monitor with a lower latency buffer to record vocals with a tiny bit of reverb.  I have to use the Scarlett Mix Control low latency setting, which does not allow for any plugins to run at all so I have to do vocal tracks completely dry.  Somehow this doesn't make sense to me being this unit gets external power.
 
OTOH, the UX2, using power from the PC allows me to take the buffer down to an almost negligible level and I can monitor recording vocals with reverb, chorusing, delay or anything I want.  I'd like to use a bit of verb in recording but perhaps change to something else later in the mix.
 
What's the purpose of having to plug the Scarlett into wall AC if it does nothing to conserve processor joules?
 
John B.

The latency of an interface is a matter of its device driver interaction with the operating system and the DAW.  The power requirements, and the source of that power are not a factor the achievable latency.  Also the wattage consumed by the CPU is independent of the wattage requirements of the interface, but to some extent the lower the latency achieved by the driver, the higher the wattage consumption of the CPU because it'll be driven to a higher utilization percentage. 
 
USB 2.0 ports are limited to 2.5-watts of power, so a device that needs to support condenser microphones at a max of 10mA (about 1/2 watt) each, some A/D and D/A converters, preamp(s), a headphone amp, and some lights can quickly be pinching the power budget.  An external power source eliminates that concern.  There are also field recorders that operate on batteries that can connect to a computer with USB. 

Johnbee58
So, I'm wondering why, on my Scarlett, I can't monitor with a lower latency buffer to record vocals with a tiny bit of reverb.  I have to use the Scarlett Mix Control low latency setting, which does not allow for any plugins to run at all so I have to do vocal tracks completely dry.  Somehow this doesn't make sense to me being this unit gets external power.

 
I'm surprised you can't monitor with a low latency on the 6i6. I can run most projects with a buffer size of 32, with only some CPU hungry plugins having to use 64.
 
Your CPU is comparable to mine, so you should be able to get similar results.
 
Could maybe try using a "cheaper" reverb plugin on a bus while you're tracking, but record your vocals dry, then add a better reverb afterwards?
 
[Edit] - Meant to say... have you checked your Power Plan is set to High Performance?

msmcleod

Johnbee58
So, I'm wondering why, on my Scarlett, I can't monitor with a lower latency buffer to record vocals with a tiny bit of reverb.  I have to use the Scarlett Mix Control low latency setting, which does not allow for any plugins to run at all so I have to do vocal tracks completely dry.  Somehow this doesn't make sense to me being this unit gets external power.

 
I'm surprised you can't monitor with a low latency on the 6i6. I can run most projects with a buffer size of 32, with only some CPU hungry plugins having to use 64.
 
Your CPU is comparable to mine, so you should be able to get similar results.
 
Could maybe try using a "cheaper" reverb plugin on a bus while you're tracking, but record your vocals dry, then add a better reverb afterwards?
 
[Edit] - Meant to say... have you checked your Power Plan is set to High Performance?

I did try it yesterday and I managed to get it to work.  I moved the buffer all the way down to 1 (I don't know how many bits that is.  They only show arbitrary numbers, not real numbers in data bits). While the performance number on the bottom (CPU usage) remained low, I still heard a little crackling noise which indicates a little burp on the processor.  When I record a vocal (or anything else by microphone) I mix everything instrumentally that I need to monitor down to a single audio track.  Then I only have to run 2 audio tracks in the project to record the vocal.  After I'm done with the recording and tuning (if necessary) I'll bounce all the vocal takes down to a single track and import it into my final mix.
 
Anyway, when I applied the the reverb to what I had set up it still sounded crappyer than what I had been dealing with before so I decided to hell with it. 
 
JB

Johnbee58
 I moved the buffer all the way down to 1 (I don't know how many bits that is.  They only show arbitrary numbers, not real numbers in data bits).

 
I'm assuming you are using ASIO and not the WDM drivers?.
 
If you are definitely using ASIO, it sounds like you've got a 1st gen 6i6 where the buffer is shown is ms.  I found 1ms hit and miss, but it did work most of the time. As far as I can remember 2ms equated to 47 samples, which worked best for me when I ran that version.
 
The Windows 1809 update made my 6i6 and 18i20 not work with the released 1st gen drivers, which meant me upgrading to the latest beta drivers. This new beta driver shows everything in the standard power of 2 buffer size.
 

Johnbee58
While the performance number on the bottom (CPU usage) remained low, I still heard a little crackling noise which indicates a little burp on the processor.  

 
What plugin are you using?
 
If you're getting the symptoms of high CPU (crackles & pops), but your CPU usage is clearly low, then although it could be your buffer size, it might well be another issue with CbB and the plugin.
 
You can try gradually upping your buffer size, but if you're still getting issues with your buffer size at 6ms or above then you could try the following:
 
1. If it's a VST3, try using the VST2 version. Antares plugins for example, behave like they're maxing the CPU out using VST3, but the VST2 works like a dream.
 
2. If you've got JBridge, try enabling the "Load with jBridge wrapper" option within plugin options. I had to up my latency up to 10ms for Virtual Guitarist to avoid pops/crackles even though the CPU was clearly coping. Using the jBridge wrapper option meant I could go back down to 2ms.
 
 

There is no magic in this world:
1) check everything in your system is really tuned for audio recording: http://www.azslow.com/index.php?topic=395.0
2) use direct monitoring for voice (I mean dry signal to output which your 6i6 can do, NOT "ultra low latency" marketing tricks), till you are ready for top specially build for audio computer with top low latency interface (like Presonus Quantum)
3) think about really good audio interface... I could not understand why people suggest that even for hobby players like myself, till I have got one. You can tune your system to death, get super computer, perform "ritual computer dance" for a week or believe Zoom outperform everything for one third of the price.
---
Or you can connect bus powered Babyface Pro to 10 years old celeron computer, start REAPER, and record what you want with 3.2ms RTL under 48kHz/48 samples buffer (or EQ/Reverb on board for direct monitoring, if your can notice inconvenience even at 3ms RTL), without rendering anything. With your current computer you probably can do this in Sonar/CbB as well (probably with rendering, Sonar has no anticipative engine). Note that you still need to do (1) .

Thanks for all the suggestions.  A new interface is out of the question.  It's not that much of a problem that I need to pour that much money into.  I tried it again today and it worked better.  I took the buffer down to 2 and used the "DAW Monitoring" setting in the Control panel.  That way I was able to run not only the Sonitus Reverb, but an SSL channel strip and the red compressor too.  I have several different reverbs that I can try.  And if worse comes to worse, I can always just continue as I have been.  It's really no big crisis.  I was just curious about the AC powered interface and the lack of RAM.
 
Thanks!
JB

If you are using plug-ins for monitoring, it is always good to know which own latency they have, so called PDC "Plug-in Delay Compensation". That is not CPU nor buffer size dependent and defined by plug-in itself (can be particular settings dependent).
In Sonar/CbB you can use PDC INFO VST (from http://forum.cakewalk.com/SONAR-Cakewalk-by-BandLab-CbB-Resources-and-Utilities-m3392713.aspx).
Of just load the plug-in in REAPER, it will display that immediately.

It may be confusing at first, but you can combine hardware-based and software-based monitoring.
Let me explain...
 
Monitor your dry vocal signal via the Focusrite's onboard hardware monitoring.
Create a "Reverb Bus" for the reverb effect... and use an Aux Send to send the "Dry" vocal to the Reverb Bus.
The Reverb will have latency, but think of it as some additional Pre-Delay.
Since the Reverb is setup as Send/Return effect, you want to set the Reverb plugin to 100% wet in the Reverb Bus.
The Dry vocal signal will be near zero latency... so the latent Reverb won't be problematic.
Also, the lower the ASIO buffer size, the lower the latency on the Reverb.
Even with round-trip latency of ~6ms on the Reverb, that's but a small amount of Pre-Delay (not at all distracting).