Re: MacSpeech Dictate review
- From: Mark Conrad <noneof@xxxxxxxxxxxxxxxxxx>
- Date: Fri, 14 Mar 2008 20:50:49 -0700
In article <michelle-8B9695.13582214032008@xxxxxxxxxxxxxxxxx>,
Michelle Steiner <michelle@xxxxxxxxxxxx> wrote:
Yeah, magic. If true, I would worry that to miniaturize it that
much, they would have to cut corners in the circuitry, making the
microphone unsuitable for speech recognition.
"miniaturization" does not mean "cutting corners" or inferior product.
Otherwise, we'd all be using mainframe computers today.
True enough, however I would feel better about buying a
headset for speech recognition if the headset manufacturer
had the balls to advertise it for that purpose.
I do not see Logitech's headsets on the MacSpeech or
Dragon recommended headset lists.
All that said, remember I am the guy who stated to
actually _try_ a headset with MacSpeech/Dragon - - -
_before_ making any judgement about it being unsuitable
for speech recognition purposes, as regards accuracy.
My personal complaint about using 'recommended' headsets
is that the majority of them are downright uncomfortable.
....but they _are_ accurate, as advertised. No speech app'
vendor is going to recommend a headset that makes his software
look bad.
Logitech's headsets appear to look more comfortable to wear,
but I would personally would have to test their accuracy and
stray noise rejection before I would used them.
Accuracy is a Big Thing with me, because I spend as much time
correcting text, as I do "speaking-it-in".
Rule of thumb assuming 99% raw accuracy:
One hour dictating at 120wpm, one hour correcting.
OVERALL speed therefore drops to only 60wpm, not to mention
that correction is a tedious pain in the ass.
....and I am talking about high 99% accuracy of the raw text,
which MacSpeech has yet to achieve.
Drop down to MacSpeech 96%, then you are going to be tied up
correcting mistakes for _four_ hours, on the same
7,200 words with their 288 mistakes.
Your overall speed for your dictation drops to 24wpm,
hardly an advertisement for using speech recognition!
THAT is why raw accuracy is so darn important, and why guys
like me grab at any and all minor improvements to raw
accuracy, _AND_ minor improvements to anything that
speeds up the correction process.
I am not knocking MacSpeech, just stating facts gathered
from my own testing of their brand new MacSpeech Dictate 1.0
Your milage may vary.
MacSpeech has stated that they will offer free updates to
fix most of the rough edges of their software, before the
end of the year.
Remember how lousy Mac OS 10.0.0 was when it first came out?
I for one am very happy with the MacSpeech Dictate initial
offering, they have made all the right choices thus far.
Particularly happy with their decision to include a separate
translator with their headset, thereby turning that headset
into a "USB headset", with all the attendant advantages.
Advantages such as being able to switch to automatic tracking
array microphones, that sit in the middle of a conference table,
and automatically track the voice of separate speakers at the
table, by electronically "pointing" the array microphone at them.
Mark Conrad:
Michelle:Here are some of the phony answers
that I got from other Logitech flunkies:
1) MacBooks do not have an internal sound card.
Depends on what is meant by sound card.
I prefer to use the term 'digitizer', because that is the
main task that a Mac sound card does.
digitize - - -
verb [ trans. ] [usu. as adj. ] ( digitized)
convert (pictures or sound) into a digital form
that can be processed by a computer.
But the MacBook does have an
audio-in jack.
Yep.
However, that is a line-in level circuit.
Yep, not digitized, therefore can not be processed by the
computer until and unless it is "digitized", i.e. converted
to those famous binary digits 'one' and 'zero'.
I don't know whether microphones output is
line-in level or a lower level.
A simple microphones output is ANALOG as chuck and I have
repeatedly told you.
Egad you are dense, sometimes I suspect that you are
a Mark Conrad in drag.
2) Microphone itself generates digital output.
Theirs does.
Not their MICROPHONE, which is what was being discussed
Smarten up, I hate dealing with dense people.
3) USB jack on MacBook goes to input of
the Mac's internal sound card.
Essentially, it does.
Wrong! Internal sound card is totally bypassed when the USB
jack is in use. You could rip out the internal sound card,
and the "USB headset" would still work!
That is one of the main reasons a "USB headset" is used in the
first place, to bypass the potentially 'electrically noisy'
internal Mac sound card.
A "USB headset" has its own EXTERNAL sound card, physically
removed from all the potentially harmful 'electrical noise'
bouncing around inside a cramped MacBook.
4) All you gotta do is select the headset model
in Mac Preferences and everything will work
without any stinkin' digitizer being involved.
That's right; because when you select the headset in Mac Preferences,
all you're doing is choosing among built-in, audio-in, and USB.
Dead wrong.
You clipped the crucial part where we were discussing USB input.
A external digitizer (sound card) is 100% necessary when using USB.
No amount of fiddling with Preferences will make the headset work
if the external digitizer is missing.
Mark-
.
- References:
- MacSpeech Dictate review
- From: Mark Conrad
- Re: MacSpeech Dictate review
- From: Mark Conrad
- Re: MacSpeech Dictate review
- From: Mark Conrad
- Re: MacSpeech Dictate review
- From: Jolly Roger
- Re: MacSpeech Dictate review
- From: Mark Conrad
- MacSpeech Dictate review
- Prev by Date: Re: Apple design process
- Next by Date: Re: MacSpeech Dictate review
- Previous by thread: Re: MacSpeech Dictate review
- Next by thread: Re: MacSpeech Dictate review
- Index(es):
Relevant Pages
|