SpeechExec With Dragon Speech Recognition
Of course we’re here to talk about using Dragon® with Philips® SpeechExec Pro Dictate. But, before we dive into the details, we need to cover several key requirements.
When we refer to SpeechExec Pro (SEP), we mean the latest version, SEP v8.7. For more information, see our write-up on v8.6, which is nearly identical.
In order for Dragon and Philips to work together in a supported configuration, you must have Dragon installed on the same computer as SEP, and it must be one of these versions/editions:
Philips has released SpeechExec Pro v10, which officially supports Dragon 14 and 15.
You need the right equipment. For front-end speech recognition, use a Nuance-approved microphone. At minimum, the headset that came with Dragon will do. But, in case you’re interested, Philips makes quite a nice mic themselves. Take note that only Philips’ mics have buttons that can control the SpeechExec Pro Dictate application.
For back-end / background speech recognition, you’ll need a digital audio recorder capable of producing files with one of these extensions: WAV, WMA, MP3, DSS, or DS2. Allow us to recommend, again, Philips’ own offering, the DPM8000. Be sure to check out these tips we give for setup. Also, be sure to review the “Guidelines for Using a Recorder With Dragon” for more helpful information.
You must be running the Windows operating system. SpeechExec Pro does not support the integration of speech recognition on the Mac® side. It’s actually a completely different application, with different functionality and a slightly different name.
Advantages of Pairing SpeechExec Pro With Dragon
Now that we’ve assembled the elements, let’s go over some of the advantages Dragon can bring to your dictation experience. We touched on front-end and back-end speech recognition, which are the two fundamental modes of the speech-to-text process.
- Front-end refers to the live process of translating speech to text, so that words appear on your computer screen as you speak. Philips refers to this as online speech recognition.
- Back-end refers to the asynchronous processing of your dictation files, such that you record the audio first, and then text is produced by Dragon at a later time. Philips refers to this as offline speech recognition.
Now, if you already have Dragon, Philips allows you do all of your front-and-back-end speech recognition within SpeechExec Pro. There’s no need to open up Dragon separately.
- Many of the commands you would use to dictate into and edit a text document will work in SEP.
- Your custom Dragon templates work too.
- You can create and automate a workflow that runs dictations through speech recognition and delivers them to the appropriate transcriptionist(s).
We’re going to review a few strategies for each of these setups, in several of their possible combinations. While we aren’t covering every inch of ground here, we will hit many of the highlights.
Dragon and SpeechExec: Setup Tips
First thing’s first: You should install Dragon and create a profile. In order for Philips to pair up properly for online speech recognition, your profile needs at least one source capable of front-end recording. For most people, that means one of three inputs:
- USB microphone
- PowerMic II
Be sure to choose the source that matches your microphone. The Philips SpeechMike Premium (LFH-3500) would be a USB microphone source, for instance.
If you want to incorporate offline speech recognition capabilities, you’ll need to add the digital audio recorder source to your Dragon profile.
Once it’s added, you’ll be prompted to do some training. This process can be a little confusing, but we have already covered it for you here. Look for the section called “Configuring Dragon for Your Digital Audio Recorder” and follow the step-by-step instructions.
Finally, take note of your Dragon profile name. You’re going to need to create an author profile within SEP – and if you’re planning on using offline speech recognition, you’re going to want to use the exact same name, down to the character.
SpeechExec Pro Front-End Setup Tips
As we said before, this is the process where you speak into a microphone and watch as your words appear as text in real time. In order for this to work smoothly, you’ll need to ensure a few things are in place:
- We covered Dragon setup in the previous section, so if you skipped it, go back.
- If you haven’t yet installed SpeechExec Pro, do that now.
- Close Dragon.
- Plug in your microphone.
Finished? Great. Now, open SpeechExec Pro Dictate and go to Settings > General Settings. You’ll see that the menu is broken up into several nodes and subheadings. Don’t let the number of menus overwhelm you; we’re going to point out the main areas of interest for online speech recognition:
Dictation > Dictation properties
The top box in this subheading gives you four different options on when to enter metadata for your recording. Choose whichever makes the most sense for you and your workflow. Don’t check any boxes if you don’t need this feature.
Speech Recognition > Recognition settings
This is the most important screen for your purposes. If it isn’t set up correctly, speech recognition won’t work.
- Check the “Speech recognition enabled” box.
- Select “Use local profiles only.” Unless you have specifically set up a roaming profile in Dragon, this is definitely the setting you want.
- In the “Dragon user profile settings” box, you’ll need to set “User name” to the name of your Dragon profile.
- Below that, you’ll see that “Topic” only has one option, unless you’ve added multiple vocabularies to your Dragon profile. If you have a choice, choose a vocabulary appropriate to the current subject area of your dictations.
- Next, set “Dictation source for dictations created within SpeechExec” to the source you’re using for online speech recognition. This will most likely be “USB microphone.”
- Click Apply.
Speech Recognition > Recording settings
Looking at the “Export settings” box, you’ll see a dropdown menu. If you have something specific in mind that you want to do every time you’re finished with a speech recognition session, select it. Our advice would be to just choose “Display the Finish speech recognition dialog” for now. That way, you can decide what to do after each session.
Speech Recognition > Recognition & Adaptation
There are two main sections here—but only one of which applies to online speech recognition. The lower half of the settings window displays “Adaptation settings.” These fold the corrections and speech data you generate back into your Dragon profile… and we recommend that you disable this feature.
It’s not that you don’t want your profile to improve. Rather, we have found that you’ll have a better experience using the Dragon interface to process your voice information. Your dictation is already running through Dragon anyways, and any adaptation that you would have done in Philips will be included. See here for more information about accuracy tuning in Dragon.
Initiating Speech Recognition
To start a new session, if you have a Philips SpeechMike, simply press the record button. This will open the Speech Recognition Recorder window in “Record Standby” mode. Then, press the play/pause button to begin dictating. When gathering your thoughts, use play/pause as a toggle between live and standby modes.
If you don’t have a Philips SpeechMike, click the green record button at the far left of the SpeechExec Pro toolbar (in the “New” section). Once the voice recognition window opens, click the red record button to begin.
Once you’ve completed your session, either click Finish to send your recording to the Archive folder, or Send for correction to move it to the Finished dictations folder, where it can then be routed to the appropriate transcriptionist for final editing.
If you want to save a speech recognition session and come back later, click on the clock icon at the far left of the voice recognition window. Depending upon how you have SEP configured, the file will either go into your Finished Dictations or Suspended folder.
SpeechExec Pro Back-End Setup Tips
Also known as offline speech recognition, this is the type of workflow where you create voice files on a digital recorder, upload them to your computer, and then have them run through Dragon, creating a transcribed document. This procedure can be fully automated, if you so desire.
Before you begin, there’s a bit of prep work to do. Review the Dragon Setup Tips section above and confirm you are ready to proceed.
Once you’ve done everything required in Dragon, close the program and open SEP. Turn on your digital audio recorder, plug it in, and allow the drivers to install. Navigate to Settings > General Settings from the menu, and let’s work on configuration. Note that some of these configurations will only apply to the DPM8000 recorder from Philips.
Rules > Author profiles
If you don’t see any author names here yet, you’re going to need to add one. Click the [Add] button, and in the text box for the author’s name, type in the exact name of your Dragon profile, character for character. This is vitally important to making this process work.
After you click [OK] to accept the author name, click [Apply] at the bottom of the window. If you have a DPM8000, you will be asked if you want to transfer that author name to your recorder. Click [Yes].
If you don’t have a DPM8000, you’re going to have to work a little harder to get the correct author profile configured. If your recorder came with software, you can install that and try to set the correct author.
Or, you can use SEP to change the author name of any recordings that you make (after download). You can do this by right-clicking a file and selecting “Dictation Properties.” Change the name in the “Author” field to match the name of your Dragon profile, then click [OK]. Obviously, this is an inefficient workaround; but it will get you where you need to go.
Worklist > Startup
Check these two boxes:
- “Automatically start SpeechExec”
- “Start SpeechExec in hidden mode”
This will ensure that SEP is running when you start your computer, so that you don’t have to remember to open it up. Additionally, it will minimize itself to the System Tray, keeping out of the way until you need it. When you dock your DPM8000, SEP will be running and ready to receive your dictations.
Worklist > Visibility
Check or uncheck “Show Worklist when connecting the DPM,” depending upon whether or not you want to see the SpeechExec main window appear when you dock your recorder.
As we move into this section of the settings menu, note that there are many similarities to the front end configuration. I’m only going to make a few pit stops to point out the differences. Assume that all other settings need to match the online speech recognition setup above.
Speech Recognition > Recognition Settings
For the dropdown menu listed under “Dictation source for dictations from external sources,” select “Digital recorder using sound files on disk.”
Speech Recognition > Correction Settings
Set this function to “Display the Finish correction dialog”—for the same reasons we made this setting in the Recording Settings menu.
Speech Recognition > Recognition & Adaptation
Check “Start offline recognition queue automatically…,” so that as soon as your dictations are downloaded and put into speech recognition, the process starts by itself.
Also check “Use specified Dragon profile….” This will override any author name your recording has attached to it and run it against the profile set in Speech Recognition > Recognition Settings.
DPM Configuration > Speech Recognition
Check “Send files for speech recognition after DPM download,” so that when you download your files, they automatically go into the speech recognition queue. Also, verify that the “Available speech recognition user” is set to the correct Dragon profile, and click “Synchronize.”
Occasionally, getting files to go to speech recognition after download can be troublesome. Here’s a checklist of things to try when you get hung up. Remember, any changes you make need to be sent to the device.
- Enable auto-download, if you didn’t do that as part of your initial DPM8000 setup.
- Go to DPM Configuration > DPM settings and click on “Start configuration.” In the Buttons section, make sure that “Delivery” is configured for one of your smart buttons—and that you are selecting this delivery method when recording on your DPM8000.
- Check your keywords. Return to DPM Configuration > DPM settings and click on “Start Configuration.” Under Keywords, make sure “SR” is listed in the “Delivery” column.
- Go to DPM Configuration > DPM Wizard and click “Start Wizard.” Make sure to choose the settings for speech recognition; they’re clearly labeled.
Initiating Speech Recognition
We’ve done a lot of work so far—but it’s about to pay off!
Create a short recording on your device, and then dock it or plug it in. Download your files to SpeechExec Pro (or let them auto-download, if you set that up). The files will now automatically pop into the speech recognition queue at the bottom of your screen and process, one at a time. When each one finishes, you’ll see a Microsoft Word document icon appear next to that dictation (in the main dictation window). Double-click the icon to see your file.
Philips Offline Speech Recognition Isn’t Working
There are some classic mistakes that may stunt the offline speech recognition process in SpeechExec Pro. Here are the ones we see most often:
- The author name of the file and/or recorder do not match the name of the Dragon profile.
- Dragon is open at the same time as SEP.
- There’s not a digital audio recorder source added to the Dragon profile.
- The Dragon profile is locked/missing/unavailable/corrupt.
- The status of the file is not set to “Transcription pending.”
- The audio extension of the recording is not supported.
Without a doubt, this was a lot of information. While this process is not as complex and involved as it may seem given the amount of documentation, there is a learning curve. As you’re reviewing this guide, try and focus on mastering each section before moving on to the next one. Also, be sure to click through to the various resources we’ve hyperlinked. Not only is there additional information, but in some cases, we refer to these links for steps in the Dragon/Philips setup.