| Vom Thema belegte Seiten: [1 2] > | Microphones Initiator des Themas: Oliver Pekelharing
|
|---|
I'm trying DPI 15 and am impressed with its speech recognition (more than 95% accurate straight out of the box), but not with command recognition. What I did notice is that it makes no difference whether I use a table mic (Yeti Blue; the best table mic according to Dragon themselves) or the built-in mic in my laptop. I find this odd, as everywhere you look everyone always goes on about how important the mic is. Will a headset mic make any difference?
Thanks,
Olly
| | | | Tom in London
Vereinigtes Königreich
Local time: 14:02
Mitglied (2008)
Italienisch > Englisch
Olly Pekelharing wrote:
I'm trying DPI 15 and am impressed with its speech recognition (more than 95% accurate straight out of the box), but not with command recognition. What I did notice is that it makes no difference whether I use a table mic (Yeti Blue; the best table mic according to Dragon themselves) or the built-in mic in my laptop. I find this odd, as everywhere you look everyone always goes on about how important the mic is. Will a headset mic make any difference?
Thanks,
Olly
I tried about 3 or 4 until I decided to spend a bit more and got this one. It works perfectly. I wear it by adjusting the angle and clipping it over one ear only. I've been using it since April 2015.
Plantronics Blackwire C435
[Edited at 2017-01-24 16:23 GMT]
| | | | Tony M
Frankreich
Local time: 15:02
Französisch > Englisch
+ ...
SITE LOCALIZER | No expert on S.R., but a sound engineer | Jan 24, 2017 |
Olly Pekelharing wrote:
Will a headset mic make any difference?
Olly
First of all, as you say, that DOES seem curious; are you absolutely certain that you are actually switching over between the 2 mics? I'm half suspecting that maybe your computer is 'stuck' on the default internal microphone, and isn't actually trying to use your table mic at all?
Regarding your final question: all else being equal, a headest mic should always be better, simply because it is closer to your mouth (hence less background noise) AND is in a fixed position with respect to your mouth. These two considerations alone really ought (for this purpose!) to override any other quality issues, unless the headset mic were of truly abysmal quality — far worse than any I have ever personally encountered!
[Modifié le 2017-01-24 16:26 GMT]
| | | | | Internal mic | Jan 24, 2017 |
(to Tony) No, I checked under recording devices, and the Yeti mic was definately enabled and being used (unless my computer was lying to me). So it is odd, and all the more impressive that Dragon is scoring so highly with my laptop mic; or maybe I just speak really clearly
I'm not keen on hanging a mic on my ear, but if it will make speech recognition even better and improve command recognition then I may have to give... See more (to Tony) No, I checked under recording devices, and the Yeti mic was definately enabled and being used (unless my computer was lying to me). So it is odd, and all the more impressive that Dragon is scoring so highly with my laptop mic; or maybe I just speak really clearly
I'm not keen on hanging a mic on my ear, but if it will make speech recognition even better and improve command recognition then I may have to give it a go.
Thanks,
Olly ▲ Collapse
| | |
|
|
|
Tom in London
Vereinigtes Königreich
Local time: 14:02
Mitglied (2008)
Italienisch > Englisch
Olly Pekelharing wrote:
(to Tony) No, I checked under recording devices, and the Yeti mic was definately enabled and being used (unless my computer was lying to me). So it is odd, and all the more impressive that Dragon is scoring so highly with my laptop mic; or maybe I just speak really clearly
I'm not keen on hanging a mic on my ear, but if it will make speech recognition even better and improve command recognition then I may have to give it a go.
Thanks,
Olly
I had similar misgivings but once you get used to it, you're not even aware of it and can even use the phone without taking it off. It doesn't "hang" -it kind of "projects" so that the microphone is close to your mouth which, as Tony says, is the most important thing.
| | | | Michael Beijer 
Vereinigtes Königreich
Local time: 14:02
Mitglied
Niederländisch > Englisch
+ ...
| tips/questions | Jan 24, 2017 |
Olly Pekelharing wrote:
(to Tony) No, I checked under recording devices, and the Yeti mic was definately enabled and being used (unless my computer was lying to me). So it is odd, and all the more impressive that Dragon is scoring so highly with my laptop mic; or maybe I just speak really clearly
I'm not keen on hanging a mic on my ear, but if it will make speech recognition even better and improve command recognition then I may have to give it a go.
Thanks,
Olly
Exactly which commands are you having trouble with? Dragon commands? KnowBrainer commands?
Immediately after you say one of them, say: "Show recognition history". A list will pop up showing you exactly what Dragon recognised, which might be useful in pinpointing the problem.
I've never tried one of those ear-hanging ones, or any kind of headset, and am very happy with the accuracy I get with my desktop mic (a SpeechWare 3-in-1 TableMike, bought via KnowBrainer). Hell, even my laptop's built in microphone works almost perfectly.
Dragon Professional Individual 15 has a brand-new engine based on AI/neural net stuff (yes, that thing that everyone is talking about at the moment), and I am getting amazing results with DPI15.
Michael
PS: all of these Déjà Vu X3 commands (in KnowBrainer 2017) work flawlessly:
[Edited at 2017-01-24 17:10 GMT]
| | | | Tony M
Frankreich
Local time: 15:02
Französisch > Englisch
+ ...
SITE LOCALIZER
Olly Pekelharing wrote:
No, I checked under recording devices, and the Yeti mic was definately enabled and being used (unless my computer was lying to me).
Well, since these settings are all done in software, it is not unheard of for the PC to be lying to you! Mine very often sweras blind it is doing one thing, but in fact does the opposite!
The easiest way to check would probably be to simply make a short test recording using the Windows audio recorder, and you should easily be able to hear a difference between the 2 mics when you play it back; this is the only way to be CERTAIN!
... if it will make speech recognition even better and improve command recognition then I may have to give it a go.
For the small cost, I think it's well worth the effort to try; you can buy a dirt cheap headset with mic for just about €5–10, which will be good enough to try the idea out, before investing in a better model as Tom suggests. Inevitably, the mic will be probably 3 or 4 times nearer your mouth, which will make an enormous difference; just look how many stage perofmers use a hand-held radio mic close to their mouth, or headband mic, compared with standing in front of a fixed mic.
| | | | Nora Diaz
Mexiko
Local time: 07:02
Mitglied (2002)
Englisch > Spanisch
+ ...
| Tablemike vs Headset | Jan 24, 2017 |
I have both a table mike and a headset, both are high quality, both purchased from KnowBrainer. My 2-in-1 TableMike is several years old and it still works perfectly. It's also my preferred microphone, I only use the headset when traveling. Regardless of which microphone I use, I try to follow a piece of advice I got years ago from Lunis Orcutt from KnowBrainer: I always use a USB sound card. I have two of these adapters, one for each of my microphones, and I believe it does make a significant d... See more I have both a table mike and a headset, both are high quality, both purchased from KnowBrainer. My 2-in-1 TableMike is several years old and it still works perfectly. It's also my preferred microphone, I only use the headset when traveling. Regardless of which microphone I use, I try to follow a piece of advice I got years ago from Lunis Orcutt from KnowBrainer: I always use a USB sound card. I have two of these adapters, one for each of my microphones, and I believe it does make a significant difference.
Having said that, I'm a bit intrigued by the fact that you're having command recognition issues, but dictation works fine. If it were a microphone issue, I would think both commands and dictation accuracy would be affected. ▲ Collapse
| | |
|
|
|
Michael Beijer
Vereinigtes Königreich
Local time: 14:02
Mitglied
Niederländisch > Englisch
+ ...
| I just did a quick test | Jan 24, 2017 |
Tony M wrote:
Olly Pekelharing wrote:
No, I checked under recording devices, and the Yeti mic was definately enabled and being used (unless my computer was lying to me).
Well, since these settings are all done in software, it is not unheard of for the PC to be lying to you! Mine very often sweras blind it is doing one thing, but in fact does the opposite!
The easiest way to check would probably be to simply make a short test recording using the Windows audio recorder, and you should easily be able to hear a difference between the 2 mics when you play it back; this is the only way to be CERTAIN!
... if it will make speech recognition even better and improve command recognition then I may have to give it a go.
For the small cost, I think it's well worth the effort to try; you can buy a dirt cheap headset with mic for just about €5–10, which will be good enough to try the idea out, before investing in a better model as Tom suggests. Inevitably, the mic will be probably 3 or 4 times nearer your mouth, which will make an enormous difference; just look how many stage perofmers use a hand-held radio mic close to their mouth, or headband mic, compared with standing in front of a fixed mic.
I opened a Word document and dictated part of the above text,
(1) with my desktop mic held right in front of my mouth, and
(2) with my desktop mic on my desktop, approximately 50 cm away from my face, off to the left of my screen.
Apart from one tiny difference ("is" instead of "as"), which was most likely due to different pronunciation, they were exactly the same: both of them were 99.5% correct. Dragon initially got the word "mic" wrong (writing it as "Mike"), which I fixed by training it.
Since I hate wearing anything on my head, I think I'll stick with a desktop mic. DPI 15 combined with my desktop mic are actually so good that I can walk around in front of my (raised) electrically adjustable desk while dictating, with no loss of quality. The future is here!
[Edited at 2017-01-24 17:03 GMT]
[Edited at 2017-01-24 17:04 GMT]
| | | |
Well I've only been at this for a day, but I'm sending the table mic back and I've ordered a bluetooth headset to try out (Sennheiser Presence; 2nd best in a comparison on knowbrainer), although it seems the mic is no longer a big issue with DNS 15, so if the improvement isn't significant the headset's going back too. I now realise the commands weren't working when I was trying them inside Studio, where you need to use the dictation box for 'full command' (this didn't use to be the case, as far ... See more Well I've only been at this for a day, but I'm sending the table mic back and I've ordered a bluetooth headset to try out (Sennheiser Presence; 2nd best in a comparison on knowbrainer), although it seems the mic is no longer a big issue with DNS 15, so if the improvement isn't significant the headset's going back too. I now realise the commands weren't working when I was trying them inside Studio, where you need to use the dictation box for 'full command' (this didn't use to be the case, as far as I can remember).
One other wee thing is I find the delay before something actually happens after I speak a command quite long; like a few of seconds. Is this usual behaviour?
Olly ▲ Collapse
| | | | Michael Beijer
Vereinigtes Königreich
Local time: 14:02
Mitglied
Niederländisch > Englisch
+ ...
| the delay will depend on yr hardware | Jan 24, 2017 |
Olly Pekelharing wrote:
Well I've only been at this for a day, but I'm sending the table mic back and I've ordered a bluetooth headset to try out (Sennheiser Presence; 2nd best in a comparison on knowbrainer), although it seems the mic is no longer a big issue with DNS 15, so if the improvement isn't significant the headset's going back too. I now realise the commands weren't working when I was trying them inside Studio, where you need to use the dictation box for 'full command' (this didn't use to be the case, as far as I can remember).
One other wee thing is I find the delay before something actually happens after I speak a command quite long; like a few of seconds. Is this usual behaviour?
Olly
I've tried to time how long it takes before something I say hits the page, and I estimate the delay at between 1-2 secs. Never any more. It's basically there the second I say it, or at least that's how it feels.
| | | |
Am now using a bluetooth headset, but don't notice any difference in recognition accuracy. If anything, it's worse! But I'll give it more time. In any case, it's a light little gadget which I stopped noticing straight away, so no problems there. However, I just started working in a MemoQ project and to my surprise it supports a lot more commands than Studio (without the need for the dictation box). Is this deliberate? Or are there some settings I need to change for working in Studio?
| | |
|
|
|
Tony M
Frankreich
Local time: 15:02
Französisch > Englisch
+ ...
SITE LOCALIZER
Olly Pekelharing wrote:
Am now using a bluetooth headset, but don't notice any difference in recognition accuracy. If anything, it's worse!
I'm not entirely susprised; I would have thought Bluetooth wad ideal for this application; for one thing, the microphone is not ideally placed close to your mouth; and for another, the VOX / gating involved in most technology of this kind can only hinder rather than help the VR function.
I would definitely not recommend Bluetooth for this kind of application.
| | | | | | Nora Diaz
Mexiko
Local time: 07:02
Mitglied (2002)
Englisch > Spanisch
+ ...
| Studio and Dragon | Jan 25, 2017 |
Olly Pekelharing wrote:
Am now using a bluetooth headset, but don't notice any difference in recognition accuracy. If anything, it's worse! But I'll give it more time. In any case, it's a light little gadget which I stopped noticing straight away, so no problems there. However, I just started working in a MemoQ project and to my surprise it supports a lot more commands than Studio (without the need for the dictation box). Is this deliberate? Or are there some settings I need to change for working in Studio?
Hi Olly,
While Studio is not a full-text-control program, there are some things you can do to better control it. Here's a link to a webinar from August explaining how Studio works with Dragon and some tips for commands:
http://www.translationzone.com/video/beyond-dictation-how-to-control-studio-by-voice-with-dragon-naturally-speaking/111200/
| | | | | Vom Thema belegte Seiten: [1 2] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Microphones | Anycount & Translation Office 3000 | Translation Office 3000
Translation Office 3000 is an advanced accounting tool for freelance translators and small agencies. TO3000 easily and seamlessly integrates with the business life of professional freelance translators.
More info » |
| | Wordfast Pro | Translation Memory Software for Any Platform
Exclusive discount for ProZ.com users!
Save over 13% when purchasing Wordfast Pro through ProZ.com. Wordfast is the world's #1 provider of platform-independent Translation Memory software. Consistently ranked the most user-friendly and highest value
Buy now! » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
