recently I noticed that if one opens up managed dictation source in 15.6, there is a click box to automatically adjust the microphone level as I speak.

 

At times my microphone seems to not recognize my voice, and at times it seems much more "comfortable". I'm experimenting with unchecking automatic adjustment, which increases or decreases the gain automatically, in favor of by setting the gain by using "Check microphone"

 

I'm wondering if there are any expert experiences with this/recommendations

This has become an interesting dilemma, especially among physicians. The primary advantage of auto-gain is to let Dragon automatically adjust volume input levels. If you back away from the microphone, move-in, speak louder or softer, Dragon will regulate the volume input to an ideal volume level. This function was introduced as a non-default setting in Dragon 13 and referred to as Far Field algorithms. Nuance never advertised this feature because they probably knew that some resellers and end-users would figure it out and give it a whirl. We further suspect Nuance obtained some of their beta testing results right here. We were pushing this option in Dragon 13 and pictorially detailed this process in our Dragon 13 training guide. 

The Nuance advertising team doesn't always come up with snazzy naming conventions. This is why we coined the 70s phrase “auto-gain”. Nuance also introduced an impressive technology that they accurately named “open ended vocabulary command technology” but it didn't have the sizzle of “artificial intelligence”, which we went with here Now the disadvantages…

Auto-gain control can occasionally fail to produce the 1st word of a phrase which will kill your entire phrase. Another potential problem, for old geezers such as myself, who occasionally create unrecognizable old man sounds such as snorting, throat clearing, nailbiting, wheezing etc... is temporary freezing. Dragon is pretty good at discerning geezer sounds from actual dictation but the resampling process can take 5 to 10 seconds. While we sometimes refer to this as freezing Dragon is simply trying to sort out noise from dictation. This isn't to be confused with the Dragon 15.3 20 second+ freeze which was dramatically improved in Ver. 15.6.

The jury is still out as to whether auto-gain works well with Medical dictation but that's why the ability to disable auto-gain control (which necessitates running the Microphone Check) was added to DMPE 4 and the Group versions of Dragon. Nuance didn't find out about this potential problem until after DPI 15 was released and the complaints started rolling in. This is why you now have a toggle option. A hack was required in DPI 15.3.

Hopefully we have covered the pros and cons but only you can decide which works better for you. Our comparison results were more or less equal but we also use a SpeechWare 9-in-1 microphone which includes built-in “less extreme” auto-gain. SpeechWare incorporated auto-gain in their USB microphones long before Nuance even thought about adding it to Dragon 13.

-------------------------

Forum Mission Statement
Trial Downloads
Dragon/Sales@KnowBrainer.com 
(615) 884-4558 ext 1

I tried with and without in a busy Pathology Lab.....with the auto-gain on things slowed and recognition was reduced. Not recommended.

Kernman, Im a internist/hospitalist, use dragon for all my work, either from the cloud or DPI15.6 in the office. No question, I'm seeing a difference with auto gain off. First words don't drop, I have to repeat myself less, the dragon seems happier without my having to shout

docinfniti....I was getting "please say that again" repeatedly with auto-gain on.