First look beta voice feedback.

I've been eager to test the voice feature, more in a factfinding sense than a 'must need to use' sense, to get a good understanding of how it works in SL. I'd tried it on the beta grid, but... somehow it's different on the main grid.

Perhaps not in a technical sense, but in a 'I'm doing something other than testing at the same time and it really changes the vibe of what's going on' sense. Even so, pretty important stuff.


1. RE: The oversized communications window for voice/text.

Serious, serious, serious problem. This sucker needs to dock, needs to slam itself thin and out of the way, or at very least needs to be functional and transparent at the same time.

The effect is about as bad as driving into the sun in heavy traffic, for someone who gets a lot of messages like I do. God forbid you have to chat, open a notecard and look at something all at once. Which I have to do all the time. You can't do it, really, without a realtime symphony of refocusing, moving, sizing, and zooming while you type. And I've got a decent widescreen here.

Solution: slam it into the same size as before, with a SMALL fussy little controlset off to one side for voice ops. With a tiny button that clearly says HELP for the uninitiated, and mouseover tooltips. We can kinda overlook a large edit panel as we always have, because building can take as long as it needs to take. Not chat. It's realtime.


2. Voice itself.

All in all, it's not too bad. Plug it in, push to talk, it works, get fancy as you learn. Great. It's a bit more difficult to assign gain factors to various people (the system seems to default to 'too quiet') but once that's figured out, it's... acceptable. Very, very reminiscent of the there.com chat, and all its spatialised sound issues - there's always *some* reason for voice to be too quiet... or too quiet due to this other thing... or too quiet due to that thing... yeah.

I'm not sure but it seems that crossing sim borders requires one to reset gain channels... not fun; this only happened to me twice so I may be off base here. But if I'm right, consider this: I'm showing someone around my sims with a horse and carriage for business purposes. Every sim crossing, suddenly my voice drops out. As does theirs. And again, and again.


3. Cultural issues.

So far, we are in the very early days. It's tough to find someone using the voice beta. I went a few days and didn't see anybody at all using it in my sims; a voice mentor friend of mine had been coming by and said the same thing. "Word on the street" is very negative against even the concept of the feature. We suspected that ours was likely the first voice chat in my sim.

I know this won't last, from my there.com days. There are too many people out there who type only 7 words per minute, and find a headset a godsend or simply... 'the way it's done'. The 'adequacy as a typist' filter on Second Life users will be gone, and the next million users after implementation will include a LOT of talkers. If I had to guess, I'd predict the cultural sea-change to come in say, 4 Q 2008 or 1 Q 2009 if voice rolls out for all by this fall.

What would REALLY make a difference, I think... and I don't know if this could be done, but if say, even a mediocre voice-to-text feature was available or even third-party recommended (with instructions) by the Company... this would go far.

I've had some practise at it for years elsewhere, but even so it's not easy to talk in voice and ensure that texting people around you can follow the whole conversation. Voice-to-text might at least minimise the damage - just another tool in the kit, just another part of the overall solution.


4. Pricing.

Last I'd heard, pricing was 'undecided'. If this is true, this is a *huge* step in the right direction.

Previously, as recently as winter there was a plan to tack on another 100 USD/mo for all the old private sims (predominantly class 4's) to get voice, as documented in a zdnet article. That includes just about everyone who got a private sim in the last couple years before winter.

Bad enough to drive a model T on a grid full of class 5's now, but to force another 100 USD/mo onto that model T so it doesn't end up being a horse and buggy by comparison to capable, voice-enabled class 5's... that would turn even the most hardcore fanboy into a dissident overnight, especially fanboys that helped bring the millions to the grid by building a compelling world in the first place.

I suggest we applaud all good moves loudly and clearly, and even a simple 'undecided' is a huge step in the right direction.


Suggestion:

What I'd really like to see personally, is the voice feature tied to premium membership, or at least payment verified. Why? It assists the 'pay as you go' model - the world is still compelling for the free users, but now there is yet another incentive for them to spend money as opposed to squeezing more out of the already paying population.

In this way we won't see roving campbots with 24/7 audio streams (gods forbid!) and voicechat griefing may be kept in check somewhat (one could at least tie the verbal abuse to a paying customer in an AR). Ideally, landowners won't get stuck constantly footing the bill for people voicechatting the day away with their sweetheart in Moldavia. Let 'em use Skype or Googletalk for that, not shake us down, because most of us (me included) can't magnanimously afford to be a free voice calling service to the world any more than the Company can.


Feature feedback often has to be negative to be useful, and I want to stress that I'm still very excited and look forward to the future. I'm hoping my words here will be used or at least taken constructively.

I saw the latest Pixar movie "Ratatouille" today, and I'm looking forward to a time when the metaverse will look and sound as rich as a professional CGI movie. I'm just as eager to get there as anyone, but *how* we get there will be critical to making it happen.

Together, let's ensure voice is done right.

Unless I missed something, LL never did revise December 2007 as the cut-off date for grandfathered $195/month sims.

It's a bit pointless to rebel against a $100/month price increase for Fall to get voice, only to be forced to pay it anyway come January.

If anything I'd start nagging LL as to whether they're standing by what they said in November, or if they changed their mind and will keep grandfathered pricing indefinitely.

It's not a rebellion, it's simply a consumer decision.

Consider if you signed a car lease and then three months later, were told it would suddenly, arbitrarily jump up by 100 USD a month in a year or so. Would you keep the car?

I honestly don't think anyone at the Company would make early adopters with mostly old class 4's - people who sweated out hundreds of hours of poor grid performance and stuck through it all - start paying 295/mo.

Even with added features, even if they gave class 5 sims along with it - one can't foist BMW lease pricing on someone who may be straining to afford a car lease at all and expect them to pay.

Besides, what would they do with class 4 sims at 295/mo that people abandoned? Would anyone buy class 4 hardware brand new for 295/mo, especially when we'll be on the verge of yet another hardware upgrade past class 5?

I don't think so, and that's another reason why I really don't think they will go ahead with this idea.

* * * * *

Anyway! Back to the feature feedback.

I'm keeping an eye on the voice feature, collecting as much feedback as I can in my sims and elsewhere, but I'm getting a very real sense that us 'traditional' grid users - the ones used to typing - won't adapt very easily.

It looks like we are going to hit 8 million user signups before the beta rollout is complete. This first 8 million, or more accurately, the almost two million who seem to have actually signed in within the last 60 days, are probably going to view the grid as a text environment.

But by the time we hit ten million, the folks coming in won't know that legacy. Will most think "I can speak and hear, but how come people can't hear me, and are doing this weird typing motion in the air in front of me?"

Some of us have seen how prevalent and dominant voice became at there.com - if you didn't have it, or couldnt' hear it you were quickly out of the conversation. Sad, painful, agonising for many.

If anyone has any suggestions for say, a reasonably priced 3rd party voice-to-text app that could be used in conjunction with Second Life, I'd say that would be good medicine for the almost certain pain to come as the culture changes.

Perhaps something like this?

http://reviews.zdnet.co.uk/software/productivity/0,1000001108,10002689,00.htm

It seems to be about 30 USD... not something every voice user would be willing to do. In my wildest dreams, such a thing could be used as a sort of 'universal translator' if fed into babelfish after sentences &c &c on different channels.

Say outputting on /100 for espanol text, /101 for output in hangul, /102 for output in greek, with backtranslation from the same preset channels.

Okay, that's perhaps getting ahead of current version feedback, but I'd really like to know if anyone tried 3rd party voice to text software. I might pick it up myself...

A 'voice to text' feature has been available in Windows for some years. It comes with Microsoft office, assuming you have a couple of days to train the programme. A few basic 'voices' are supplied 'Microsoft Sam' etc.. a company recently had a stand at the ITE 2007 that offered a SL equivilant version, but still only used the 'Microsoft Sam' voice, with links to a couple of other websites that offered a few more 'natural' ones.

I tried it myself and it was pretty good. You point the programme to your chat logs (you have to have chat logs turned on for it to work), it basically read each chat log as they are written, albiet a bit of a time lapse (second or two minimum, longer when there is a lot of lag). The real benefit was it didn't take much training for it to work. SL Speech Tools I think it was called.

I have had voice enabled in FirstLook since it became live on the main grid, but can never seem to get it to work, as there are no visual signs that it is, and I haven't seen anyone else with any either. The SIM I am on is enabled and I have set all the parcel rights throughout.

Personally, I don't see how LL can justify the cost for voice in SL. There.com never charged me extra (it was included in my membership package, one payment for lifetime membership), and seeing as there are a prolifercation of free alternatives like Skype etc that work well, they will have a hard time getting people to cough up yet more money over and above the high prices they charge already to play an online 'game'.

That's "text to voice" not "voice to text."

"voice to text" is when I speak into the microphone and it spits out text into my chat bar.

LOL.. and you never used the Microsoft Text to Speech programme, whereby it works both ways? I can 'type a letter' using my voice and issue OS commands etc...

QUOTE: A 'voice to text' feature has been available in Windows for some years. It comes with Microsoft office, assuming you have a couple of days to train the programme. A few basic 'voices' are supplied 'Microsoft Sam' etc.. UNQUOTE

You don't 'Train' a programme that reads text, hence my post regarding the Windows version. The programme will work both ways and 'read' your text back.... it is this very engine that the In Game programme I spoke about further in my post, uses.

I don't happen to have a copy of Microsoft Office loaded on the machine I'm on now, and the copy I use on my RL biz accounting machine is ancient (2001? older?) so I was wondering - any 'quick tips' to get started on this?

The name of the MS app or how to find it in Office?

I know, I should just read the docs but I think I stripped off everything but Word and the spreadsheet on my other machine - it dates back to the days of the tiny hard drive.

Buying MS Office isn't really an option but I'm hoping at least there will be both Windows and Mac support... not that any of that will be effective for the Linux user, sadly.


RE: SL Speech Tools, I found this video:
http://www.youtube.com/watch?v=a209xEeJjL0

...and it looks like it is for sale:
http://www.slexchange.com/modules.php?name=Marketplace&file=item&ItemID=196663

(I'm not affiliated whatsoever with these guys and have never tried it)

While I'm constantly impressed by the creativity of residents, it looks like this version is Windows only and many of my friends in SL are ardent Linux and Mac users. Ouch.

* * * * *

Honestly I think this is the future of SL, or better put, the future metaverse in general.

Voice to text for those who have voice, text to voice for those who have text, and local speech generation to save on bandwidth like the SL Speech Tools guy seems to have done it. But without cross-platform support it's going to be a tough road ahead.

It's never pointless to save $100 bucks! After the thread about this Class 4 pay for voice while Class 5 and mainland get it for free - There was silence from Linden Lab on the subject of paying for voice and still they won't discuss it any further. If you don't comment on their 'suggestions' in the blog then they initiate the change and tell you they blogged it months ago (example: No more notices to Owner of groups when people drop the group).

When Linde Lab makes my Class 4 Server a Class 5 Server I will pay the extra $100 Bucks.

Macs have an incredibly simple text-to-voice system: highlight some text, right-click, select "Speech / Start Speaking". Or go to a Terminal window and type "say <whatever you want>" - with the "say" command one can also output to an AIFF file, or select a different voice, or read out text from a text document.

The voice isn't at all realistic, though, and there is a basic problem with making it much better, in that human beings are incredibly good at picking up the tiniest details of tone, timbre, volume and so on in voice; an immense evolutionary advantage for a social animal, and we have whole chunks of our brains doing this subconsciously. Computer-generated voice will be in the "uncanny valley" at best for the foreseeable future.

In contrast to this, though, voice-to-text products are actually quite good these days. With a bit of training a relatively inexpensive software package can result in text with fewer typos than the average SL conversation. (I would love to see LL licence one of these packages for general use but I very much doubt they will.)

While the voice may not be perfect, it would give those who arent using voice a ability to communicate, which is the main concern , isn't it? Of course it will just fuel the paranoids further, but that's OK.

My guess is that the main concern is voice-to-text. People using voice can still read text, although it does remain a mild bit of effort to constantly use both senses.

Basis:

A long long time ago, I was a member of a community called there.com. I still am, and still have enormous respect for what was done there.

There.com experienced the voice/text crisis a number of years back - those of us who were there saw every agonising bit of it.

Try being an unhearing texter in a group with 2 or 3 voicechatters... even with a friendly 'translator' it just... doesn't work. It's simply not possible to keep up.

Try as voicechatters might to be polite... politeness wears off after a bit and only the very most ardent friends will continue to drop voice for text once most people can speak and hear. I fell into a 'listen and type' sort of person myself, which is likely what I'll become in Second Life now also.

What happened there was unforgettable, and a stark warning... though now their community has largely acculturated.

It's very common to see young people talking. Type? What's that? Why? It still goes on; the older userbase still types, and new folk certainly can. But the end result is nothing like the culture in SL today.

What happened was this: a hard demographic shift toward the young... the myspace generation (the shift not induced by voice, but just by mainstreaming of the product). Even when there.com's original target audience was closer in demographic to their current audience's mom. True, Second Life isn't supposed to be for anyone less than 18, unlike there.com. But we all know how hard that is to enforce.

As Second Life mainstreams... the world as a whole wins, but by the same token we won't all be artists and tech experts and whizbang early adopter people any more.

The culture and economics of the grid will reflect that slowly at first, then with continued user churn and free access it will faster and faster mainstream itself. Expect to hear a lot of young people talking to other young people.


This is why I'm so interested in a communications toolset that allows everyone to communicate to everyone, but most critically voice to text.

Anything less will actually probably *degrade* communication between average individuals, regardless of which side of the voice/text divide they are on.

I have said it before, but all of the young people I know type more than I do in practice. They are constantly IMing and making Myspace posts and so on, and when they are not near a computer they are sending each other text messages, rather than actually using telephones to speak on (it is, after all, rather hard to speak on a phone whilst using it to play rubbish tinny music to annoy everyone else on the bus, which is the #2 use).

Part of the difference between SL and There is also that communication is actually more significant in SL, because of the business element. It is not just important to be able to speak to people in the here and now, as banter - it is also important to be able to manage multiple conversations at once without much pause, to be able to review what you said previously, to have specific records of when X promised to give you Y for L$Z. Only shysters in RL conduct their business entirely over the phone, precisely because it is transitory and can be used to confuse people.

The voice-text-voice programme is an add on to Microsoft Office, not sure about the basic module, I have the professional versions. When you install, use Custom setup and select it from the list.

XP also comes with Microsoft Narrator (Start, Programmes, Accessories, Accessibillity...) It's a text to voice reader, never used it myself, but looks potentially usefull, assuming you can point it to the chat logs in the SL folder.

With the main client I crashed ~25 times yesterday. Yet each time I used the firstlook voice client not a single problem at all... except that I totally hate it.

I too have a decent wide screen like the OP, and the IM/history/everything communicate window blocks so much that I'm no longer in a 3d world. That's right, I'm in a 2d world with an extreemly tiny portion of my screen having access to the 3d (which then takes constant juggling to see or do anything... Hence the many times retrying to log using the main client, even after noting that the FLVC didn't crash at all.

Does anyone else here detect these behind the scene attemps at subtle persuasions(use voice and SL works otherwise 'have fun')

very much like your choice is between voice OT1H and this humungus visual blockade on the other

I am a huge LL supporter but I cant deny that I too am quite dissapointed with LL's decision to integrate voice into the client without giving due consideration to that portion of their customer base who do not, for whatever reason, want voice in their SL.

And too say, "well if you don't want it just don't use" is rather short sighted because as we all know that if you are in a community in which you feel compelled to refuse voice chat with even one of your friends (and we know there will be more), a huge strain will be placed over that long nurtured friendship. There is no doubt that the integration of voice chat will put a tremendous strain on many established friendships and communities. And it is an RP killer for the many beautiful and long established RP communities who have supported SL for a long time.

There are so many ways that LL could integrate voice into SL without having to, in the process, alienate and divide its customer base. Ideas such as charging a fee to those who want voice, allowing only premium members to have voice, integrating a voice changer or a text-voice/voice-text functionality, etc., are just a few ideas that could be implemented that would lessen the pain and be more considerate and inclusive of its customer base.

SL is currently the leader in virtual worlds without voice. SL should not have to fully revert to a system used by a lesser populated world such as There.com in an effort to be competitive and successful. Yes, we can have voice. But SL should be innovative and revolutionary and put thought into implementing a communication system that is truly inclusive and considerate of all, rather than exclusive and considerate of only the few.

on the other hand, LL did opensorce the client so really we have only our poor programming skilz to blame for any disatisfaction we have with the client.

Although it's certainly not easy jumping into the middle of many programs ( SL is many programs ) and tickle out their rhymn and reason.

I tried the "New and improved" Voice First Look last night, and was once again underwhelmed. I uninstalled and trashed it within 15 minutes of installing and testing it.

They still don't get it that NO ONE wants the text chat and text IM features welded inseperably with the Voice UI controls. The chat history window is consumed for 2/3 of its surface area by a "near me" pane and an "avatar profile" pane that are of virtually NO USE if you are not doing voice chat. People have complained repeatedly that they do NOT want the chat history and text IM to be lumped into a single interface that eats over half of an 800 x 600 window, but LL blithely blunders on, smugly looking at their 20" cinema displays and not having a clue why anyone would think the UI is too big.

I've ALREADY had one friendship in SL heavily strained by my refusal to use Voice. One of my intimate friends said they wanted to at least hear my voice once. When I told them that was NEVER going to happen, it was like someone dumped a bucket of ice water on the conversation. We didn't text chat again for close to a week, and there has been a palpable distance between us ever since.

And with any friend that I have now who admits to enjoying Voice Chat, I already feel the barriers growing between us. I know they will be spending a large part of their time and attention participating in conversations that I can't possibly ever be part of. I feel like I am losing some of these friends, and the blasted Voice Chat still hasn't even become more than a First Look option. It's like watching your friends join a "Members Only" club or church that you wouldn't be caught dead setting foot inside, or that would never accept someone of your race or religion, and knowing they will be spending a lot of their free time there, instead of in the social venues that you used to enjoy sharing with them.

Yea, Ceera has a point there- I dont like the whole 1 window thing. Eats up some Viewer space if you ask me, and I dont have a 40" monitor for all those windows .
I dont like moving windows around to see what the heck is going on.

What I see others asking for also is a way for Non-Voicers to hear at least whats being said even if they cant participate vocally.

Tested First Look quite a bit over the weekend...hours in fact.

Voice works great in First Look. But the client crawls compared
to the current client. Although I have modest hardware, I maintain
around 14fps with the current client, but with First Look it consistently
drops to just about 1fps.

I've also been testing Voice via the beta client for some time which
provides something for comparison.

* * *

My other comment about First Look is the Communication tab which
now combines Chat/IM/Groups et al...it seems very crowded.

The thing I dislike the most is the Chat History window. In First Look
there is a scrolling list of entries above it...takes up too much space.

If I could have just an uncluttered Chat History window I could accept
the new Communication center a little easier.

But like most changes I've endured with the client; I tend to get
used to them...like them or not...reluctantly.

- Infrared

Tested Voice and many of had a ball, but there was the us & thems starting to happen already, and the firstloock screwed up normal chat so babd by not allowing seperat IM and local chat windows, that I went back to the old viewer.

How can you stay stable enought to run the client more then 50 mins if that?

Easy just login and play, I have no problem with any of the viewers. Believe it or not the other 35,000 people online aren't on the forums complaining, because they aren't having your problems.

Your SL has never worked stable for 9 months I've been here and yet other peoples has improved, there's got to be something different about your computer or ISP, because unless you change something at your end, I predict in 20 updates and 2 years time you will still have problems running SL.

Well as i said first look seems to be the problem viewer here and not the 1.17.3......I don`t understand why you talking for 35,000 others but then again ..........

Well for the last 9 months for me after i undated my computer i had no problems what so ever.... But when the viewer crashes after x amount of time ( first look ) then there is a problem. the 1.17.3 viewer and the one before i had no problems what so ever. As for my connection I highly doubt that because I am on a fiber optical 100mbps up and down connect. nothing. But then again as i said before 1.17.3 and before runs best . First Look crashes randomly...........Hence its the viewer and not the connect or modem.........

Usagi

The communicate window is evil.

Please vote on Jira:
https://jira.secondlife.com/browse/VWR-1076
and a similar:
https://jira.secondlife.com/browse/VWR-1550

you can vote for as many issues as you like when you are logged in.

However, the voice firstlook source code isn't yet available to the opensource community - only the non-voicelook code is available. The opensource community haven't had a chance to tackle the communicator window yet (for instance)

Matthew