I have often seen the word “screen reader” applied and understand now it is intended for visually impaired users. I have then been thinking that this way of playing IF may not be widespread in smaller countries with their own local language. For instance, I would like to bring IF to visually impaired in my own country Denmark, first by introducing it, and then perhaps translate some games with permission or just make the IFDB / English games more easily accessible locally and other needs that may pop up. Note that I am not myself visually impaired so my experience with this topic is very limited.
However, first I would like to hear from other visually impaired users, if there are any things I should be aware of. I might have naïve ideas about the use of screen readers, free ones especially, and how difficult/easy it is to surf the web with a screen reader etc.
Or perhaps some of you are aware of good resources on this topic?
You may be able to just try it yourself? IIRC they’re built in to most operating systems these days: MacOS/iOS VoiceOver, Windows Narrator (or on Windows you can install NVDA for free if you’d rather). There’s definitely a bit of a learning curve but if you have the time it’s worth trying out so you can see how it works and what the barriers are…
There are quite a few visually-impaired people on this forum, so I’m sure they’ll be able to give feedback if they read this post. It would be interesting to know how many of them use English as a second language and whether they would like to play IF in their native language.
I have been dealing with low vision for most of my life. The vision in my right eye waxes and wanes depending on retina issues. The left eye is mostly a place holder.
I have used Freedom Scientific’s Zoomtext and/or JAWS since 2000 or so. They are both commercial text to speech applicaitions for Windows. They are very expensive. Trial editions are available to experiment with.
I’m currently working on a project with Raspberry Pi computers that looks promising. It includes TTS and magnification along with braille display support. It looks promising for those on a meager budget.
Yes, I saw JAWS being pretty expensive and that the pro version was three times as expensive.
I wonder if NVDA is sufficient for most purposes?
(I can imagine that the more expensive screen readers may sound less “robot like” which would be preferred for atmospheric text descriptions etc)
Would be good to know for sure. However, those who manage to find the forum are probably pretty good at English already. In any case, the more feedback the better
Freedom Scientific offers a number of different languages.
I still have a license for both Zoomtext and Jaws. I have chosen to use the annual subscription option since it is easier than the four digit purchase price.
For me, NVDA is more than adequate. I have never liked Windows Narrator and their earlier versions. In addition, Zoomtext’s screen magnification has not kept up well with the internets newest graphics display formats.
Apple’s Voiceover is pretty good. I have a new Mac Mini M4 that does well. Though just as with Windows, Mac’s OS has limited the user interface options. I prefer the fredom on Linux/Gnu software.
I’ve only ever used JAWS and NVDA in a school-like setting, but from what I’ve used of them, their operation was basically indistinguishable from Orca, the dominant graphical screen reader for Linux. That said, some JAWS users swear it works better for MS Office and other software common in a workplace setting, though the price puts it out of reach for most Blind Windows users without the benefit of a school, government agency, or employer footing the bill. I’ve also heard the Windows 11 version of Narrator is consider an improvement over the Windows 7 version and older, though I still get the impression it’s mostly the screen reader you use to install NVDA or JAWS on a new Windows machine.
Aside from Orca for the Linux desktop, there are also several text-mode screen readers for the Linux Console, though the big two are probably espeakup and Fenrir.
Also, speech synthesizer is a different component of the accessibility stack from the screen reader and most screen readers can be used with a variety of speech synths and most speech synths can be invoked independently of a screen reader. And whereas there is not really any such thing as a cross-platform screen reader(even Voiceover for OSX and Voiceover for iOS have little in common beyond sharing a name), some synths are cross platform. Also, while some speech dependant blind computer users value natural sounding voices, others prioritize lack of latency or comprehensibility at very fast speeds(a typical default speed is around 150-180 wpm, around normal conversational speed… I’ve known blind folks to crank their synths up to things like 500 wpm or even 800 wpm, their speech output sounding like total gibberish or not even sounding like speech at all to my slow ears.
Personally, I’m on Linux, using Orca on the desktop for Browsing the web and espeakup in the console for everything else, using espeak-ng’s default British English voice with default speed and pitch for both, which is clearly on the artificial side without sounding like a robot from old school sci-Fi. My experience is that more natural sounding voices tend to fall into the uncanny valley(AI speech seems to be climbing the far side, but AI speech isn’t really suitable for real time application) and I have trouble keeping up at even moderate speed increases. My preferred way to play IF is to run the story file through a console interpreter(Frotz for Z machine, Glulxe for glulxe, I haven’t gotten around to figuring out other formats), and generally, espeakup reads game text automatically as it’s printed to the screen. Web-based IF is trickier since I generally need to find where the output begins to start reading it manually then find where to input my next action, and there aren’t always convenient web page elements to help me out via the navigational hotkeys most graphical screen readers provide(at least, my experience is that all of the navigational hotkeys I use regularly with Orca work basically the same in JAWS and NVDA).
Thank you for the detailed reply. There is a lot of useful information here I feel. I was most surprised that Web-based IF was tricky, as I had the idea that a full screen browser windows with pure text, especially run by the Parchment interpreter, would be well handled by screen readers.
On Linux, are most Web-based interpreters problematic in this aspect or are there some you could say are better?
For now, I am quite curious about web-based interpreters, as they are usually cross-platform and no need to install software. However, it may also be very game dependent?
If there e.g. are pictures in a game, the player may be in doubt if they missed some crucial information, or if the pictures are not really important to the game.
I use HTML Tads for TADS games on Windows and Windows Frotz for Z format games. I particularly like HTML Tads. It supports colorized text and text linking.
On linux I use frob for TADS and frotz for Z-machine games. On linux, I have never had much luck with gargoyle-free. Note: I generally use a raspberry pi as my linux machine.
On the Apple mac mini m4 I use a Mac app called Yasmin that is Z machine centric. I also use frotz and frob from the command line.
For the record, my previous comments on my own IF habits are my own opinion. Blind Linux users are as diverse and opinionated as any other group and I cannot speak for anyone else as to their preferred way of doing things. I was already accustomed to using console interpreters like Frotz or Fizmo and was very comfortable in the command line in general, before ever trying a web-based interpreter with a screen reader. If I spent more time with web interpreters, I might develop a play flow that, if not efficient, at least doesn’t feel clunky compared to just doing things the old school way, and someone who first played IF with a web interpreter could very well find using a console interpreter as clunky as I find the web interpreters I’ve tried.
I should of course be careful not to conclude too much from two persons’ feedback but I do find it interesting, that none of you are using web interpreters nor multi-format interpreters and you both use several single-format interpreters.
If anyone has additional comments, please keep it coming
