By Martel Electronics · Updated November 9, 2026 · Shop all seven stenomasks

A Dragon stenomask setup succeeds when three things line up: the intended microphone reaches the computer, the recognition application uses that input, and the resulting text is checked against what was actually spoken. Connecting the USB cable is the beginning of that process. It is not the final test.

There is also an important naming distinction. Martel's Dragon Stenomask and Double Dragon 2 are microphone products. Dragon speech recognition software is a separate part of a workflow. A product name containing “Dragon” does not establish which software license, version, or reporting application is included with an order. Confirm those details on the actual listing and with the relevant supplier.

This guide provides an original setup and review process, including controlled practice passages and a way to separate audio problems from text problems. It does not report laboratory tests or promise a recognition percentage. Start with the Dragon Stenomask product page for direct USB, or review Double Dragon 2 when you also need a separate 3.5mm microphone path.

Define the intended result before changing settings

Write a one-sentence description of the working setup. For example: “My mask voice should reach my approved recognition application through the USB device, and the practice document should save in this folder.” If you also use a separate recorder, give that path its own sentence.

Then list the exact computer, operating system, recognition software version, reporting application if used, and mask model. A support conversation becomes much more productive when “Dragon” is not being used to mean both the microphone and every program on the computer.

Choose a small acceptance test before making adjustments. A short passage containing ordinary words, an invented name, a date, and a number is enough to establish a baseline. You are looking for a repeatable test, not a flattering demonstration. A passage you can read and review carefully is more informative than a long recording you never examine.

Keep the initial goal limited: clear input, expected application response, and a saved result you can reopen. Vocabulary development, professional reporting technique, and speed practice come afterward. Separating those goals prevents an unrecognized proper name from being mistaken for proof that the USB microphone is malfunctioning.

Establish the physical connection

Inspect the intended connection arrangement and identify any adapter, hub, dock, or extension between the mask and computer. Record what is necessary for your machine, then follow the device's supported setup instructions. When troubleshooting a USB microphone, Nuance's guidance recommends checking the connection and testing a direct computer USB connection rather than a hub or extension.

Do not interpret that as a reason to disrupt a working system without a purpose. It is a way to simplify a problem when the path is not yet verified. Once the basic arrangement works, evaluate any required extra component deliberately and keep a record of the result.

Position the cable so routine movement does not pull at the connection or alter how you hold the mask. Use the same practical desk arrangement for the baseline test and the later application test. Otherwise, you may change both the audio path and the physical position at once without realizing it.

For Double Dragon 2, identify the USB and separate 3.5mm destinations clearly. The analog path does not configure itself merely because the USB path works. Treat it as another connection to verify with the receiving device and its own saved recording.

Confirm Windows and application access

Check that the intended microphone is selected in the computer's input settings, then verify the source inside the recording or recognition application. A computer can have several available microphones, and a successful-looking input meter may belong to a webcam or built-in microphone instead of the mask.

If Windows access settings prevent an application from using a microphone, follow Microsoft's microphone permission instructions. Desktop applications and Microsoft Store applications are handled differently in those settings. Use the guidance for the installed system and permit only the access needed for the intended workflow.

Now make a short recording with suitable practice material in an appropriate recording application. Listen to the saved sample before judging recognition performance. You should be able to understand the intended words and identify whether the recording came from the expected path.

Write down the selected device name as it appears on your computer. Do not assume a generic label will be identical on another machine. If you later replace the computer or interface, compare the new choices with the actual connected equipment rather than selecting a familiar-sounding name from memory.

Run the recognition software's microphone check

For Dragon Professional 16, Nuance documents the microphone check under Settings, Microphone, Check Microphone on the DragonBar. Follow the wizard for the installed edition. Other products or versions may present different controls, so use their own help rather than forcing these menu names onto a different interface.

A completed microphone check is one piece of evidence. Continue with your own short passage in the application where you intend to work. Verify that the program is listening, that the intended field or document is active, and that the resulting text appears where expected.

Keep your normal working speech and physical arrangement consistent during this process. If you conduct the check in a special demonstration posture and then change the setup for everyday use, the baseline becomes less useful. The aim is a configuration you can reproduce comfortably and inspect honestly.

Do not repeatedly rerun checks to avoid investigating an obvious routing problem. If the computer is recording the wrong device, fix that selection first. If audio is clear but text is going into the wrong window, investigate the application context. Match the next action to the observed failure.

Create an original baseline passage

Use material that is safe to store and easy to compare. Here is an original example written for this guide: “Morgan Vale scheduled the equipment review for Tuesday morning. The first folder contains twelve practice files. The second folder contains the corrected draft. Please check the address before sending the final copy.”

Read the passage once using your normal working approach. Save the audio if the approved workflow supports it, save the text, and write a brief note describing the setup. Then compare the three: intended passage, captured speech, and recognized text.

The passage deliberately mixes ordinary words, a fictional name, a number, and a sequence of actions. If the name is wrong but the remaining words are clear, that is a more specific observation than “accuracy is low.” If entire phrases are missing from the audio, look earlier in the path. If text stops after the active window changes, examine the software interaction.

Keep this baseline unchanged for configuration comparisons. Use additional passages for broader practice later. Changing both the microphone setup and the test material makes the result harder to interpret because you no longer know which difference explains the outcome.

Sort problems into audio, recognition, and workflow categories

An audio problem concerns the signal captured from the mask: silence, interruptions, distortion, or speech that cannot be understood adequately. A recognition problem occurs when usable captured speech is interpreted incorrectly. A workflow problem concerns where text goes, how it is organized, or whether the needed files are saved and reopened.

These categories can overlap, but they suggest different next steps. If a saved sample is silent, adding vocabulary is not the first useful experiment. If the audio clearly contains an unfamiliar surname and the software spells it differently, replacing the cable is not the most direct first response. If the text was correct but saved under an unexpected name, the microphone may not need attention at all.

Create a small issue log with the passage, observed result, suspected stage, one change, and the retest outcome. Use factual language such as “the first three words are absent in the saved audio.” Avoid conclusions such as “this model cannot recognize my voice” until the actual evidence supports a narrower explanation.

This approach also makes support more efficient. You can provide a precise description and a suitable nonconfidential example instead of asking someone to diagnose a complicated system from a general feeling that it should work better.

Handle names, numbers, and specialized terms deliberately

Ordinary words and unfamiliar names present different review demands. A sentence can look fluent while containing the wrong spelling of a person, place, organization, or technical term. Numbers can also appear plausible even when they differ from the intended value. Treat these as items to verify, not as details that become reliable because the paragraph reads smoothly.

For practice, create a small list of invented names and quantities with a known reference spelling. Dictate them in complete sentences and review the result against that reference. Keep the exercise separate from a hardware comparison: a name the software has never encountered should not become the sole basis for ranking two microphones.

Use the vocabulary and correction features documented for your software and training method when addressing recurring recognition issues. Do not invent unsupported command sequences or apply instructions from another edition without checking them. If reporting software is involved, confirm how the recognition and reporting layers interact before changing a shared setting.

The practical goal is a controlled review process. You should be able to say which errors recur, under what conditions, and whether the attempted correction changed the result. That is more useful than chasing a single overall accuracy number with no record of how it was measured.

Build a separate test for Double Dragon 2's additional path

The reason to investigate Double Dragon 2 rather than a USB-only mask is often the separate 3.5mm microphone output. Begin by naming the destination and confirming its input requirements. Then test that path with a short recording, independent of the computer recognition result.

Label the files by purpose. For example, an invented practice session might produce a recognition draft and a separate mask-audio recording. The second file contains the mask source; it is not automatically a recording of the original room speakers. If room audio is required, establish that source separately.

During the combined acceptance test, verify that both destinations are actually operating. Afterward, open the saved recording and inspect a portion near the start and another later in the practice session. This checks more than the initial connection and helps expose an unintended stop or an incorrect saved file.

Describe the remaining shared dependencies honestly. A separate output may be useful, but the protection it provides depends on the whole arrangement. The purchasing question is whether that additional path solves your defined need, not whether the word “double” guarantees a particular level of system resilience.

Compare every Martel model before deciding

Dragon and Double Dragon 2 are not the only paths into a voice writing setup. An existing verified analog interface or recorder may make another model relevant. Compare the seven arrangements below and open the current product pages for price, contents, and availability.

For this article's direct USB workflow, start with Dragon Stenomask or Double Dragon 2. For an established analog workflow, confirm the interface and receiving requirements before choosing among MM-ST1, MM-ST2, MM-18, MM-XLR, or MM-14. Martel specifically excludes MM-14 compatibility with Andrea or USB sound cards.

Do not interpret this table as a recognition benchmark. No controlled comparative test is reported here. Its purpose is to show which hardware arrangements merit investigation for your actual computer and recording plan.

Use a practice log that leads to a decision

A useful log does not need to be elaborate. Record the date, device arrangement, application versions, sample passage, and what happened. Add one planned change and the result of repeating the same test. This creates a trail of evidence without turning everyday practice into a paperwork project.

Here is an original example entry: “Baseline passage recorded clearly. Fictional surname appeared as two common words. No missing audio. Next action: review the software's documented vocabulary method with the instructor.” Another entry might read: “Recording stops when the connection is disturbed. Next action: inspect the physical path and obtain hardware-specific support before further recognition testing.”

The difference between those entries matters. One concerns word interpretation; the other concerns capture. Treating both as an unspecified accuracy problem can lead to unnecessary purchases and repeated adjustments that never address the cause.

Retain only the practice information useful for your purpose, and handle actual assignment material according to its requirements. A troubleshooting log should help reproduce a problem without becoming an uncontrolled collection of sensitive recordings or documents.

Prepare for computer and software changes

A working configuration is valuable, so document it before making a planned change. Record the mask, input selection, application versions, and relevant supported settings. Preserve appropriate backups through the application's documented process rather than assuming a copied folder captures everything needed.

After a computer replacement, software update, or interface change, repeat the baseline test. Check the physical connection, selected source, microphone check where applicable, text destination, and saved-file location. A familiar-looking desktop does not prove the underlying audio choices stayed the same.

Avoid scheduling your first acceptance test immediately before an important session. Allow time to identify a changed input label, an application permission, or a missing component. This is practical planning rather than a prediction that an update will fail.

If a new problem appears, describe the specific change and the earliest observable failure. “After the computer replacement, the saved audio uses the built-in microphone” is a stronger starting point than “the mask stopped being accurate.” The former points toward an input-selection investigation; the latter can send troubleshooting in several unrelated directions.

Read a test result without cherry-picking

Imagine two invented practice runs of the same passage. In the first, the text contains a wrong fictional surname and a missing number. In the second, the surname is correct but an entire phrase is absent. It would be misleading to call the second run better simply because the troublesome name improved. Review all the intended content and identify the source of each difference.

Use a simple comparison sheet with four columns: intended words, audible words, recognized words, and your observation. You do not need to calculate a performance percentage to learn from it. The sheet should make clear whether a word was spoken, captured, interpreted, and placed correctly. If you cannot hear the relevant audio, mark that limitation rather than guessing what happened.

Repeat a comparison when the result is ambiguous, keeping the passage and configuration consistent. A single unusually good or bad sample should not become a claim about a microphone model. Your personal acceptance check has a narrower purpose: deciding whether this arrangement is ready for the work you intend and which issue remains unresolved.

Also separate a controlled passage from live listening practice. Reading your own prepared text tests different demands from following unfamiliar speakers. Both can be useful, but their results answer different questions. A clean reading sample establishes a baseline; it does not prove professional listening, repetition, and transcript review skills under a changing conversation.

Check the boundary between recognition and reporting software

When a workflow includes a reporting application, describe which program receives the microphone and which program organizes the resulting text. Do not assume that a successful demonstration in a plain text field verifies every integration feature. Your instructor or software provider should identify the supported arrangement for the versions in use.

Create a harmless practice task that exercises the actual boundary. Start the approved programs, select the intended input, enter a short passage through the normal method, save the work, and reopen it in the place where you will review it. Observe where corrections are made and whether the expected information remains available after reopening.

If something fails, state the boundary precisely: audio is present in the capture program but text does not reach the reporting document, or text appears but the saved project does not contain the expected material. That observation helps the appropriate provider address the issue without beginning from the assumption that the mask is defective.

Keep the baseline configuration modest. Add optional integrations only after the required path works. This produces a known starting point if a later change causes trouble and gives you a clearer explanation of what each application contributes to the finished workflow.

Rehearse normal changes in attention

A real working session involves looking at notes, checking a document, and returning to the main application. In a controlled practice exercise, observe what happens when you perform those ordinary transitions. Does the intended text field remain selected? Does the microphone state remain clear? Can you tell where the next dictated sentence will appear?

Use invented material for this exercise, and stop the test if the application begins placing text somewhere unexpected. The lesson is to recognize the state of your workflow before continuing, rather than to memorize a universal sequence of clicks. Different programs handle focus and input differently.

Record one practical recovery step from the approved application instructions and test it. A short, understood recovery method is easier to use than a page of miscellaneous tips. After returning to the intended document, make a brief sample and verify the result. This checks the complete transition instead of assuming that returning to a familiar-looking window restored every relevant state.

Make a short daily start and finish routine

At the beginning of a session, confirm the intended equipment and input, make a brief appropriate sample, and check the required recording destinations. Use a repeatable sequence that fits your approved workflow. The purpose is to catch a changed selection or missing connection while it is still easy to correct.

At the end, confirm what was saved and where. Reopen the intended files, verify the naming, and complete any required backup or transfer process. If an issue occurred, record it while the details are fresh rather than reconstructing it days later from memory.

Keep the routine short enough to use consistently. A long checklist that no one follows is less useful than a focused process that checks the actual dependencies of your system. Review it when your equipment changes, and remove steps that no longer apply rather than allowing obsolete instructions to accumulate.

For a purchase, review Dragon Stenomask for direct USB or Double Dragon 2 for USB with a separate 3.5mm path. Check current stock before planning around a delivery date, and confirm the software and accessories included in the actual order.

Additional original practice passages

Passage two: dates, numbers, and deliberate review

The fictional delivery was scheduled for November 12 at 9:30 in the morning. The request listed 16 folders, 240 pages, and three separate envelopes. A revised note changed the page count to 214 but left the number of folders unchanged. The coordinator asked the assistant to confirm the date and the corrected total before sending the package.

Numbers are useful for review because a small difference can change the meaning substantially. Compare what you intended to say, what you actually spoke, and how the application displayed it. Formatting may differ from spoken wording, so distinguish a formatting preference from a wrong value. Follow the number and punctuation procedures taught for your particular software workflow rather than inventing commands during the exercise.

Mark each date, time, and quantity in your comparison note. If the text shows 240 where the corrected sentence should contain 214, listen to that exact mask passage before deciding why. You may have repeated the earlier number, spoken unclearly, or received an unexpected interpretation. The saved sample lets you investigate the specific event without relying on memory of what you thought you said.

Passage three: fictional speaker changes

Question: When did you first review the revised schedule? Answer: I reviewed it after lunch on Tuesday. Question: Did the revised schedule include the afternoon session? Answer: It included the session, but the starting time was still marked as tentative. Question: What did you do next? Answer: I asked the coordinator to confirm the starting time in writing.

Use the speaker identification and punctuation method taught in your reporting program. The written labels in this practice passage identify its structure; they are not universal software commands. Keep the first attempt slow enough to preserve complete questions and answers. Review whether you supplied the intended speaker information consistently and whether the saved audio allows you to locate each transition.

A helpful second attempt introduces an ordinary pause before one answer. Check whether you resume speaking and recording as intended. If an application stops listening during the pause, investigate that application's operating behavior. Do not assume that the mask disconnected merely because the text stopped appearing. Check the input, application status, and saved audio in the same order used during the original setup.

Passage four: room and mask source identification

Room helper: This is the room source speaking from the participant chair. The practice folder is on the left side of the table. Voice writer: This is the mask source repeating the practice sentence. The practice folder is on the left side of the table. Room helper: The source check is complete. Voice writer: The mask source check is complete.

Use this exercise only with a suitable, authorized room recording arrangement and an informed helper. Save both intended sources and reopen them. Select the room source for playback, then the mask source. Confirm which voice and phrase each contains. Some sound from elsewhere may be audible depending on placement and isolation; the important task is identifying the intended source and confirming the route actually saved it.

Then locate the matching folder sentence in both recordings. Note the time reference or playback method you used. If the recordings are separate files, confirm alignment through the supported workflow rather than assuming identical elapsed times. This small test helps you practice toggling between room and mask review without introducing confidential material or waiting until a real assignment to learn the controls.

References and scope

Product connections are drawn from Martel's stenomask catalog. Software steps are linked to Dragon Professional 16's microphone-check documentation, Nuance's microphone troubleshooting page, and Microsoft's microphone permissions guidance. Practice passages, diagnostic categories, and log examples are original material, not customer testimonials or measured product-performance results.

30 questions and answers about Dragon stenomask setup

1. Is the Dragon Stenomask the same product as Dragon software?

No. The mask is microphone hardware; Dragon speech recognition software is a separate component of a possible workflow. Confirm which software license, version, and accessories are included in the actual order. Similar names do not establish a complete package. Keeping the hardware and software names distinct also makes setup instructions and support conversations much easier to follow.

2. What should I verify before connecting a new mask?

Identify the mask model, computer, intended application, and any required intermediate component. Write down the expected result and choose a short nonconfidential test passage. This gives you a clear baseline for the first session. Without that plan, a moving input meter or a few correctly recognized words can be mistaken for a complete test of the workflow.

3. Why test basic audio before recognition accuracy?

Recognition depends on the input it receives. A saved audio sample helps reveal silence, interruptions, or the wrong microphone source before you spend time adjusting vocabulary or interpreting text errors. If the speech is clearly captured, you can investigate the recognition layer more specifically. Testing in that order reduces guesswork and makes each later observation more useful.

4. Where is the microphone check in Dragon Professional 16?

Nuance's documented path is DragonBar Settings, then Microphone, then Check Microphone. Follow the wizard and the help for your installed edition. Other versions may differ. Afterward, make a short test in the application where you actually plan to work; completing the wizard alone does not verify text placement, project handling, or saved-file review.

5. Can I follow these menu instructions in every Dragon product?

No. The specific menu path cited here is for the documented Dragon Professional 16 interface. Use the official help for a different edition or product. The broader method still applies: verify the input, test the captured speech, check the intended application response, and inspect the saved result. Do not force an instruction onto a screen that presents different controls.

6. What if the computer sees the device but the application cannot use it?

Check the application's selected input and the applicable microphone access settings. Then test with appropriate practice material. Device detection and application access are separate observations. Describe both when seeking support, including the operating system and software versions. This is more useful than assuming detection alone proves the entire audio path has been configured correctly.

7. Why might speech appear in the wrong document?

The active application or text field may differ from the one you intended. Check where input will go before continuing a practice test. Treat this as a workflow observation rather than immediate evidence of a microphone fault. Rehearse normal transitions between notes and the main application using harmless material so you understand how your supported setup behaves.

8. Should I speak unusually slowly to improve the test result?

Use the working approach taught for your setup and keep it consistent between samples. A special demonstration style may not represent the way you will actually use the system. The purpose of a baseline is to make practical comparisons. If your instructor recommends a technique change, record that as a deliberate change and evaluate it separately from hardware adjustments.

9. What makes a good original test passage?

Use a short passage you can review accurately, combining ordinary words, an invented name, a number, and a date or sequence. Avoid confidential material. Keep the passage unchanged when comparing configurations, and save the corresponding observations. The goal is to distinguish kinds of errors, not to create a passage that makes every result look impressive.

10. What should I do about a repeatedly misspelled name?

First verify that the intended name is clearly present in the captured speech. Then investigate the documented vocabulary or correction method for your software and training approach. A recurring spelling issue with otherwise clear audio is different from a recording failure. Keep a reference spelling and retest in context so you can tell whether the attempted correction helped.

11. Does fluent-looking text mean the result is accurate?

No. A paragraph can read naturally while containing a wrong name, quantity, date, or attribution. Compare important details with the appropriate source and review process. Do not let smooth wording replace verification. In practice exercises, a known reference passage makes these differences visible and helps you distinguish plausible substitutions from missing or incorrectly captured speech.

12. Can I calculate a reliable accuracy percentage from one sample?

One informal sample is too limited to support a broad performance claim about a product or an entire professional workflow. You can still use it to identify specific problems and establish a baseline. State what was tested, what changed, and what you observed. Avoid presenting a convenient personal result as a manufacturer benchmark or a guarantee for another user.

13. When should I repeat the baseline test?

Repeat it after a relevant change, such as replacing the computer, changing the interface, updating an application, or correcting an observed problem. Keep the material and procedure consistent enough to interpret the comparison. There is little value in repeatedly testing an unchanged working arrangement without a question to answer; focus on the part whose behavior may have changed.

14. What belongs in a useful troubleshooting log?

Record the configuration, short test passage, observed result, suspected stage, one change, and retest outcome. Use specific observations rather than broad judgments about accuracy. Include versions and device models when relevant. A concise log helps you explain the issue to support and prevents you from repeating an unsuccessful adjustment because you cannot remember what was already tried.

15. Does Double Dragon 2 automatically save a backup recording?

No. It provides USB plus a separate 3.5mm microphone path, but a compatible destination must actually receive and record that second signal. Test the destination and reopen its saved file. The hardware feature makes a second path possible; your configured recording process determines whether a usable additional recording is produced during the session.

16. Does the second path contain the original courtroom sound?

It concerns the mask microphone source, meaning the reporter's voice into the mask. It should not be mistaken for a separately positioned room microphone. If original room sound is required, plan and authorize that recording arrangement separately. Label files by source so later review does not confuse a second mask recording with a recording of the participants.

17. Can an analog Martel mask be part of a recognition setup?

An analog arrangement may be appropriate when the mask, interface, computer, and software are confirmed compatible. Review the exact combination rather than assuming every analog mask works with every interface. The seven-model table helps identify connection candidates, but it is not a universal compatibility matrix or a measured recognition ranking across all possible receiving devices.

18. Is MM-14 an alternative for an Andrea USB setup?

Martel's comparison specifically says the MM-14 is not compatible with Andrea or USB sound cards. Do not override that restriction by assuming a plug adapter will make it suitable. Review a verified compatible model and interface arrangement instead. The relevant choice is the complete supported path, rather than the ability to make two connectors physically fit.

19. Should I replace the mask when recognition stops responding?

First establish where the failure occurs. Check whether the intended audio is captured, whether the application is listening, and whether the intended document is active. A clear recording with no text response suggests a different investigation from missing audio. Replacement is appropriate when it addresses a demonstrated equipment problem or unmet requirement, rather than an unexplained application symptom.

20. Can several changes make troubleshooting faster?

They can make the result harder to interpret. If you change the input, cable arrangement, microphone position, and software settings together, an improved sample does not tell you which change mattered. Preserve a baseline, make one justified change, and repeat the same test. This approach produces a more useful explanation if the problem later returns.

21. How should I test a reporting-software integration?

Use the supported configuration for the actual versions involved, then perform a small complete task with practice material. Capture input, place text through the normal method, save the project, and reopen it for review. A successful test in an unrelated text box does not establish that every part of the reporting application's workflow is functioning as intended.

22. What if audio is clear but entire words are missing from the text?

Preserve a short appropriate example and note exactly where the difference occurs. Check the software's documented troubleshooting and the approved workflow, including input state and application context. Do not assume all omissions have the same cause. The comparison between intended speech, audible speech, and text gives support a much clearer starting point than a general accuracy complaint.

23. Should I use confidential recordings for setup experiments?

Use invented or otherwise suitable practice material while establishing the system. This keeps the test focused and avoids unnecessary handling complications. When actual assignment material is involved, follow its access, storage, and transfer requirements. Do not introduce an unfamiliar cloud service into the workflow simply because it appears to offer a convenient way to test transcription.

24. What should I save before a planned software change?

Document the working configuration and use the application's supported backup or export process for relevant user data. Record versions, selected inputs, and the baseline result. The exact items depend on the software. A casual copy of an unfamiliar folder should not be assumed to preserve everything needed to restore the prior working arrangement.

25. Can a new computer change which microphone is selected?

Treat input selection as something to verify after moving the setup. Device labels, available inputs, and application choices can differ. Reconnect the approved arrangement and repeat a short capture and playback test. The mask may be unchanged, but the complete system is new enough that the previous computer's successful result does not establish the new configuration.

26. How do I compare two practice runs honestly?

Review all the intended content and identify where each difference arose. Do not select only the word that improved while ignoring a newly missing phrase. Keep the passage and setup consistent unless a specific change is being tested. Report the limits of the comparison; an informal exercise can guide your setup without becoming a general product-performance claim.

27. What should a daily startup check accomplish?

It should establish that the intended equipment is connected, the correct source is selected, and every required destination receives usable input. Use a brief appropriate sample and a repeatable sequence. Tailor the routine to your verified workflow rather than collecting unrelated checks. Its value comes from catching a real changed dependency before relying on the session.

28. What should I check when the session ends?

Confirm the intended files exist, are named clearly, and can be reopened from the expected location. Complete any required backup or transfer procedure and record unresolved issues while they are fresh. A successful recognition display during the session does not by itself prove that the material needed for later review was saved correctly.

29. Are Dragon and Double Dragon 2 currently in stock?

Availability must be checked on the live product pages when you are ready to order. Inclusion in this guide is not a stock or delivery promise. If a deadline matters, confirm the actual availability and required components before planning around the purchase. Do not infer shipment timing from a product name or an informational comparison.

30. Which product page should I open first?

Open Dragon Stenomask if your verified plan needs direct USB only. Open Double Dragon 2 if it also needs a separate compatible 3.5mm destination. If neither arrangement matches your equipment, use the full seven-model catalog to compare the other connection options before ordering.