Sound Splitter

• Authors: Joseph Lee and contributors
• Download stable version
• NVDA compatibility: 2019.3 and later

This add-on, partly based on Tony's Enhancements by Tony Malykh, adds the ability to split audio from NVDA and other sounds onto separate audio channels.

Commands:

• Alt+NvDA+S: toggle sound splitter. If enabled, NVDA will be heard through the right channel.

Sound Splitter settings

You can configure add-on settings from NVDA menu/Preferences/Settings/Sound Splitter category.

• Split NVDA sound and applications' sounds into left and right channels: checkinb this checkbox will enable sound splitting feature.
• Switch left and right during sound split: by default, NVDA will be heard through the right channel if sound splitting is on. You can instead hear NVDA through the left channel by checking this checkbox.

Version 22.02

• Initial version based on Tony Malykh's Tony's Enhancements add-on.

Why not put the issue on github and discuss it there, as if you do not the feature suggestion won't exist.

As I indicated in my original post Window Eyes had the option to change
the order of name first or control name first. That doesn't mean NVDA
needs to do the same but I liked the feature in Window Eyes.

Thunderbird doesn't support browse mode when writing e-mail.  I don't know if browse mode is supported in any other e-mail programs with that command.

Gene

Hi,

In theory, yes, but again it runs into practical issues. There is a way to simulate what's termed "rearrange list" where you can change the order of a list, and then have various speech functions modified to consult this rearrange list. But remember that different speech functions are used for different purposes, ranging from just speaking a message to handling format attributes as you move through a document. Also note that we have people who rely on braille, so there is a question of making an equivalent setting for braille users. Ultimately, there are places where role is not even announced, for example in lists and Excel spreadsheet cells.

Cheers,

Joseph

Hello!

Where you compose your e-mails?

In Word and Outlook, in browse mode, w will move to next error...

Rui Fontes

Rui Fontes

I was thinking today about a new feature that might speed up work a good deal when writing documents.  I generally don't use f7 spell check these days.  I read my messages before sending them and change flagged spelling errors, thus doing two things at once that I would do separately otherwise. 

My proposal for a new feature, which I'd like comments about in terms of demand and feasibility is this:

There could be an NVDA command that moves to the next flagged misspelled word in a document.  Even if I read a document and hear a word flagged, the cursor is almost never on the word.  Its often at the beginning of the line and the misspelled word may take a lot of movement by word to get to.  If I could hear the flag, then issue the move to command, moving to the word could be sped up a lot most of the time.   As I think about it, this adds the equivalent to what a sighted person can do now.  A word is flagged, the sighted person can immediately move to or click on it, thus placing focus on it, and either manually correct it or use the context menu to do so.  Or a sighted person can do so later by seeing the flag, clicking on it, and correcting the word.

I'm surprised this hasn't been discussed before.  I did a Google search and I didn't find a Github issue, though I didn't search Github directly.

Gene

Here is a message I received off list concerning JAWS and this feature.


I don't know if the person wants his name used for attribution so I won't do so.


Before I reproduce the comment, I'll ask a question.  This may not make sense.  I don't have the technical knowledge to evaluate it.  My question is, can the way the role and other information is handled be left as it is?  Could only which output is spoken first be changed?  Would it be placed in a tiny buffer and presented in reverse order for speech or reversed in some other way?  Reversing the out;put would seem to me to be the most simple means of accomplishing what is being discussed.


At times, when I [present an idea, I wish there were some way to poll users about it.  I may present an idea that  a lot of people would like if they knew of it but there may appear to be little or no demand because the idea is largely unknown or, if known, users don't have a structured way of reporting it and so they don't.


Here is the message:


In Jaws You're looking for "Speak control type first" and "Speak control atate first" in the Settings Manager.
With both unchecked the label is spoken first (I.E., Start button).
With Control Type First you'll get button Start.
With State First it depends on whether Control Type First is checked or not.


Gene

Apologies, all. I have 2010 which isn't supported any longer, so not
being used, & that's where it was then. It is, I believe, on by
default, & it sure messes up NDVA. I don't know if the online version
has it or not.

Thanks for helping out, Brian V. It's what I get for having
ancient-of-days software.

--
Subscribe to a WordPress for Newbies Mailing List by sending a message to:
wp4newbs-request@... with 'subscribe' in the Subject field OR by
visiting the list page at http://www.freelists.org/list/wp4newbs
& check out my sites at www.brightstarsweb.com & www.mysitesbeenhacked.com

Jackie,

         It's actually located in the General Pane of Excel Options.  ALT + F, T, then tab over into the pane then either tab to the checkbox or hit ALT + L to be taken straight to it and toggle whatever its state was.

         Mine was on, which I have to believe is the default as I never recall having touched this, and when I did the above using the ALT + L navigation technique when I landed on it the Enable Live Preview option was changed to unchecked.
--

Brian - Windows 10, 64-Bit, Version 21H2, Build 19044  

The instinctive need to be the member of a closely-knit group fighting for common ideals may grow so strong that it becomes inessential what these ideals are.

       ~ Konrad Lorenz (1903-1989)

 

It's located in the 'view' menu. I'm sorry--I don't have Office 2019
to give specific instructions but hopefully you can find it there.

--
Subscribe to a WordPress for Newbies Mailing List by sending a message to:
wp4newbs-request@... with 'subscribe' in the Subject field OR by
visiting the list page at http://www.freelists.org/list/wp4newbs
& check out my sites at www.brightstarsweb.com & www.mysitesbeenhacked.com

I am using Windows 10 and Microsoft 2019, and I think the latest version of NVDA. I even tried using google sheets as an alternative but that does way too much talking and delay when you enter data in a cell and move on. By the way, what is this live preview that you mentioned and how can I change it?

TheeQuinn, there's quite a bit you don't tell us that might be
important, ie, what version of Excel you're using, your version of
Windows, etc.

Having thus said, 1 thing I do know that adversely affects Excel is
having live preview turned on, & I advise everyone to turn it off, cuz
it *really* makes NVDA laggy. Maximizing the screen can often help
also. Chances are you already know all this.

If you're using 1 of the free online Office apps, the problem may
simply be your connection. I don't know of any other solution for that
except to just ignore what NVDA first reads & then do a read current
cell when the screen has finished stabilizing. PITA, I know. Obviously
getting a desktop version might assist w/that, but it's also
expensive. They got you by the short ones. Whatchu gonna do?

--
Subscribe to a WordPress for Newbies Mailing List by sending a message to:
wp4newbs-request@... with 'subscribe' in the Subject field OR by
visiting the list page at http://www.freelists.org/list/wp4newbs
& check out my sites at www.brightstarsweb.com & www.mysitesbeenhacked.com

Good night everyone,

This is an issue that I have been experiencing for a while but it is really getting on my nerves now. This problem that I am facing is very frustrating and could affect me negatively by slowing me down and being inaccurate. I have realized that when I am navigating cells in NVDA it tends to report the cells incorrectly sometimes. for example, I may have the numbers 1, 2, 3 typed in cells A1, B1, C1 respectively. if I move from cell A1 to B1 it would still say "1 A1". I find this very misleading especially when I am trying to do working out for my school work. Does anyone know if there is a fix for this? It would be greatly appreciated and I could share it with others as well. Thanks in advance for your answers :).

Hi all,

I wrote that comment because I figured it might be possible to change this in the future; I didn't say it should be changed, but backed off by saying that perhaps it could be changed when given a chance. That was stated back when I didn't have the time to investigate the actual internals of what I described for the last two days. Now that I know about the actual inner workings, I may need to visit that issue and provide an updated comment.

If folks think that it is possible to change the internals described in this thread, then please do suggest. Although I can't speak for NV Access (again, I'm not an employee of that organization), what I can advise is:

When discussing NVDA changes, for anyone familiar with NVDA screen reader source code, please do not just simply say, "oh, it is possible to change things." We (the community) want to see your work, and that work is understanding the internals around the thing you wish to change and offering a solution that is easily understood by audience members (in this case, users). If change is impossible, provide a justification as to why. I do believe that third-party pull requests are essential, but it is also important to think about multiple variables and impact in many areas.

Personally, I tend to gravitate toward small pull requests with good reasoning and structure. After all, you are essentially writing a small essay, so it is best to provide useful structure and reasons as to why something should change. When in doubt, write GitHub issues and pull request descriptions as though you are presenting a case to judges or a business proposal to a senior manager; I won't go into the discussion of formal versus informal in detail, but personally I might look at pull request descriptions that are structured and provides reasons (just saying "I like this approach better" without telling NV Access, third-party developers, and users as to why is unlikely to get traction). In some ways, even if we say we are volunteering, you are ultimately interacting with an organization and a community that demands high standards and professionalism from code contributors (that's why we have a code of conduct document). Come to think of it, writing pull requests will teach you to speak at least five languages or styles of communication:

1. Python: since NVDA's front-end is written in this language.
2. C++: because if you need to work on some back-end details such as dislay model, knowing C++ is a must.
3. Formal writing: you are effectively proposing something that can have an impact on lives of thousands, so treat it as a professional level engagement.
4. Explanations to users: ultimately, your change can (and will) be noticed by users.
5. People outside of NVDA community: sometimes you need to talk about your changes with folks who have no clue as to how screen readers work. This group is distinct from the fourth item - members of the users group are familiar with the screen reader, whereas the latter do not.

Although I begin to sound as though I am some kind of a salesperson, if you want to see the style I personally employ when writing GitHub issues and pull requests, take a look at the following example:

Publish NVDA desktop app to Microsoft Store: https://github.com/nvaccess/nvda/issues/13058

I don't expect everyone to use this formal writing style, but when it comes to suggesting changes, please do so while providing practical (or sometimes theoretical) ways to do so whenever possible.

Cheers,

Joseph

On Sun, Jan 30, 2022 at 02:31 PM, Rui Fontes wrote:

Someone said that was not possible.

-
Because it is not, at present.

When someone asks, about a piece of software, "Can you make it do this thing?" and "this thing" is not presently supported, the answer is normally, "No."

The original question was posed by Don H in message  #92388, and reads:  "Is there a way to change this from within NVDA to say Button Start instead of what is now saying Start Button?"

The appropriate answer to that is, "No."  These answers always carry the non-stated because it's not necessary to state it, "as things stand at the moment."


 
--

Brian - Windows 10, 64-Bit, Version 21H2, Build 19044  

The instinctive need to be the member of a closely-knit group fighting for common ideals may grow so strong that it becomes inessential what these ideals are.

       ~ Konrad Lorenz (1903-1989)

 

The problem is:

Someone said that was not possible.

My point is:

It is possible.

Now, if it is a very requested feature, if it is reasonable to implement it, if NV Access wants to do it, bla, bla, bla... is another question very different...

Regarding user requests:

Feature request toggle order of what is read out re controls
#10752

In this issue:

@josephsl
Add your reactionShow options
Collaborator
This comment was left via email reply.josephslcommented on 3 Feb 2020
Hi, surprise! You are not the first one to request this, and perhaps one day, it will be possible to configure this. Thanks.

and:

@Brian1Gaff
Add your reactionShow options
This comment was left via email reply.Brian1Gaffcommented on 4 Feb 2020
Yes indeed, I seem to recall there are issues of lag introduced when a work
around to do this was tried last time several years ago. It would be nice
though to have this sort of thing under a verbosity tab.
 Brian

and another issue:

Allow users to change the order in which object properties are reported in speech and braillecomponent/braillecomponent/speechfeature#7232 opened on 31 May 2017 by leonardder

Rui Fontes

Hi,

Ah, the difference between two dimensions... Let me preface the below explanation by saying that I'm still trying to process what turns out to be a really complex function that ultimately generates a list of strings. I hope this can also serve as a backing to Brian's statement sent a few minutes ago as to why it cannot be done.

The difference in behavior between when you press Tab or K or up/down arrows is because these commands traverse two different yet related dimensions of a control. Specifically, whereas Tab and K commands attempt to retrieve system focus, up/down arrows attempt to traverse document text. In fact, these commands are found ion separate modules and operate differently.

When you press Tab or K from a browse mode document, NVDA will see if you are navigating between groups of focusable controls. If pressing these keys results in system focus changes, then system focus will indeed be moved to the new control as far as NVDA is concerned, resulting in NVDA saying the control label followed by role (this is because the same sequence of events I described yesterday (see an earlier message in this thread regarding Start button) is executed). On the other hand, when you press up or down arrow keys, NVDA will traverse through the next or previous line, gathering attributes (internally called format and control fields). The reason for NVDA saying "link label" is so you can distinguish attribute changes throughout the text. Just like system focus movement, this sequence of events does generate a speech sequence, but unlike focus movement, a generator is used (a generator is code that can yield execution to the calling code while performing its task, and then resume from where it left off next time it is invoked; I know, it's a bit hard to explain).

In summary, the idea of speech sequence reordering is not really generalizable across all situations because we are dealing with two fundamentally different ways to looking at a control: system focus versus text. We can argue that they are one and the same because we are dealing with a control label and role in both situations. However, think about what's the most significant piece of information required at this moment after performing a task:

• Focus movement: where am I?
• Text navigation: what am I reading?

Tab and K keys fall under the first question, whereas arrow keys fall under the second question provided that they are involved in reading text. There is one prominent situation where arrow keys fall under the first question, and that's when pressing arrow keys results in system focus changes. Why? Because:

• Pressing arrow keys from say, File Explorer, changes from one control to the next.
• Pressing arrow keys from say, Word 365 document, results in caret movement while focused on the same document.

Therefore, changing the speech sequence when you press arrow keys (to let it say label followed by role) cannot be implemented unless you are willing to argue that everything should be treated the same. To that argument, I offer a line from Zen of Python:

Although practicality beats purity.

Hope this helps.

Cheers,

Joseph