What's Happening at Speaking Access
Technology can move fast, and paradoxically, a lot of times it can move slowly. When it comes to accessibility technology in software, one thing that we have noticed is that changes happen, problems arise, exciting new developments come along, and the community has no idea what's going on.
This website was created to keep our community up to date, not just with the positive, punchy news, but with the real issues too. You'll see updates here about bugs we're fixing, features we're building, and the progress we're making. This page is also here to thank the many people who are helping us keep Speaking Access affordable by offering their time, testing, feedback, and support. We truly appreciate it - a real pat on the back to all of them.
Double Tap Interview with Christopher Mutch
Steven Scott and Shaun Preece of Double Tap recently interviewed Christopher Mutch about Speaking Access. The conversation serves as a strong introduction to the software, the thinking behind it, and why Christopher chose a blind-first design approach from the beginning.
One of the key points discussed was the difficulty of learning and remembering the many Windows keyboard shortcuts that blind users are often expected to master. Speaking Access was built to reduce that burden by letting people control their browser, WordPad, Notepad, and other parts of Windows through plain-language voice commands instead.
The interview also highlighted some of the program's main strengths: more than 500 commands, a fully voice-controlled audio menu, dictation tools, and a proofreading mode that reads punctuation and symbols aloud. Another important feature Christopher discussed was privacy. Speaking Access runs entirely on the local computer for standard commands, with no cloud dependency and no trigger word requirement.
Christopher also spoke about pricing and the broader audience for the software. Although it was developed with blind users in mind, it was also designed to help people with dyslexia, mobility barriers, and challenges such as essential tremors. He also mentioned an upcoming local dictation add-on using Whisper for longer-form offline transcription.
What Came Out Of The Show
One of the most helpful results from the Double Tap interview was that two people reached out about JAWS support. That is a major development for Speaking Access. JAWS support is now planned as one of the next major updates, and those who reached out have expressed genuine interest in helping with that work. It likely will not be a fast addition, and at this stage it may be closer to fall, possibly around October, before it is ready. Still, this was an extremely encouraging outcome and we are very grateful to the people who contacted us.
Another result of the interview was that Speaking Access was contacted by someone in Italy asking whether the software is available in Italian. At the moment, because of the current System.Speech and Microsoft-based technology behind the program, language support is limited. The shorter-term and longer-term direction now being considered is a larger transition away from System.Speech and toward VOSK, which supports many more languages. That will require major experimentation and development.
A third very practical outcome is that we discovered Azure now offers a new way to code-sign programs that may make it much easier to host Speaking Access directly on the website. Because of that, Christopher expects that within the next month or two, Speaking Access should be available for direct download from the website, along with a PayPal purchase option. That will make the software easier to access for people who would rather not go through the Microsoft Store.
Working with BlindHelp.net
We are grateful to BlindHelp.net and especially Abdullah for volunteering their time to help test Speaking Access. BlindHelp.net is a volunteer-based project that supports blind and visually impaired users by sharing accessible technology resources, tutorials, and community support. Their work is focused on helping remove barriers and making technology easier to use, and we truly appreciate the time and care they have put into testing our program.
Current Issues Found During Testing
During recent testing, two important issues were identified on different systems:
Windows 11 25H2
On one system running Windows 11 25H2, Speaking Access does not launch properly after installation. When the program is opened, it gives a single beep, but then nothing else happens. No interface appears, and there is no further response.
Windows 10 21H2
On another system running Windows 10 21H2, the program does launch and prompts for microphone training. However, after the training is completed, the software does not respond to voice commands. It appears to be listening, but no actions are carried out.
One particularly useful observation came up during testing: when NVDA is running and speaking, if the user says something to Speaking Access, NVDA gets interrupted. That suggests Speaking Access is actively hearing audio input, but something is going wrong in the stage where commands should be processed and executed.
Update on These Issues
We now believe we have identified the cause of the Windows 11 25H2 launch issue. In this case, the problem appears to have been caused by outdated drivers creating startup conflicts.
The Windows 10 21H2 issue turned out to be much simpler. It was caused by a setup step being missed on my end, not by a deeper software failure, and that has already been fixed.
Right now, we are continuing to test these changes carefully. Once the updated version of Speaking Access is ready, we will share the announcement here on this page and on social media.
Thank You
Thank you again to BlindHelp.net and Abdullah for volunteering to test Speaking Access and for sharing such helpful feedback. Their support has been genuinely appreciated and has helped improve the program.
Working with Brandon from That Blind Guy 527
We're also thankful to Brandon from That Blind Guy 527. He interviewed me on one of his podcasts and has also been helping test Speaking Access. Through that work, he helped move forward one of the next important areas of development: Outlook support.
What I am working on now is making Speaking Access handle Microsoft Outlook in a smarter way. One of the main challenges is that Outlook now exists in two different versions for many Windows users: the legacy version and the newer version. Because of that, the same action may require different shortcuts or different behavior depending on which version is open.
Rather than creating two separate sets of voice commands, I am building Speaking Access so it can recognize which version of Outlook is running and then send the correct command automatically. It takes more work, but the goal is to keep things simpler and more natural for the user.
Thank You
Thank you to Brandon and That Blind Guy 527 for the interview, the testing, and the practical feedback. That help has directly shaped the Outlook work now being built into Speaking Access.
A Final Message from the Inventor About Dictation
Speaking Access is a program the inventor is extremely proud of. It is designed to help people use their computers more easily, more naturally, and with less frustration in everyday life.
One area the inventor does not consider finished, and is working very hard on, is something extremely important: dictation. The current command and post mode is based on technology that is nearly 10 years old. For grammar and for building reliable commands from spoken phrases, it works very well. But when it comes to dictation, it falls short of the standard the inventor wants for users.
That is why work is now underway on a major improvement. We are working on a new dictation program that will maintain the same user-friendliness for audio users and empower them. It will also be affordable at under $15 US.
Two different approaches have been tested so far. One is VOSK, an open-source dictation system that can run almost instantly on Windows computers. Unfortunately, it produces text with no punctuation and no capital letters, which creates obvious difficulties for everyday writing. Experimentation with it is still ongoing using different approaches, but at this point it is not considered the long-term answer on its own.
Another option has shown much stronger results: the Whisper API from OpenAI, a dictation service that is extremely accurate. With this approach, the user can dictate naturally, say "stop," and the program can then turn that audio into readable text with punctuation. It works especially well with larger chunks of speech, often closer to full paragraphs.
This update should be coming within the next two or three months, most likely closer to three months, because part of the work is making it function properly for audio users. One important part of that customization is making sure the dictated text can be read back clearly, including punctuation.
Thank you for taking the time to have a look. Please keep checking back here, visit our social media, have a look at the Speaking Access and Friends Facebook group, and come back to see what's new.