feature requests for future fersions

[synthany]

timestamp on the UI texter
when people comment.
on the chatter you know, put a timestamp
when people comment.

i like to match it up to the recordings.

o dream...

[trece8]

Hi everybody.

A very cool thing would be mixing the client's volumes by the server, so that the server sends only one track to each client [not (#users-1) tracks, as it is now].
That way, there would be much less upload bandwidth needed and we could use cheaper internet connections for server stuff.
At least, having that as an option would be VERY useful.

The only thing would be that the sliders on the clients would send to the server the information of their state when they change, and the server would use more CPU and RAM.

P.S.: I seem to remember that I had posted this idea before... my memory is not that good, it seems.

[AndyMc]

That could be a lot of extra load on the machine to mix live audio into 1 stream.

The Time Stamp is a good idea but which time would be used, local pc, server or other?
This could maybe be a update on the reaninjam.dll and use local then maybe option to log chat too.

[Brennan]

That was a Google Summer of Code assigment, to mix on the server (optionally.) Unfortunately the dude did not come thru with the doing. But it would still be nice to have, I agree.

Timestamp on the chat lines would be nice too.

[trece8]

Andy... it seems to me that it wouldn't be so much load on the machine. I mean: every client machine mixes the tracks and it doesn't seem a lot of work, doesn't it? I know that can be easily done using JACK (in Linux), but not so sure in a stable performing way.

The problem would be resources regarding RAM or CPU cycles?

P.S.: I'm SO sorry for my lack of coding abilities: I'd really like to help... I know what a PC can do, but don't really code... unless some very basic applications for JACK.

[AndyMc]

Sorry I wasn't clear on my previous post, how I meant by a lot of work, if it was automated it would need an auto limit of some kind, possibly auto eq which then timed by 6 to 8 people with possible mixing of 2 to 3 tracks a person into a stereo with these fx could be a bit much work for the machine with all the rest the machine has to handle like networking and codec stuff.

Sorry I wasn't too clear before.

If this wasn't done as a minimum then how crowded it can get atm it would be a lot more crowded sounding squashed to 1 stereo pair with no eq and limiting.

If the client took on some of the work it would be achievable keeping server load low, but with some machines just adding 2 fx can cause pops and clicks and then becomes a whole new set of issues to fix in the NINJAM community.

But the idea is good, would save on bandwidth and allow more to run servers on home connections.

Time stamp could probably be added to the reaninjam client easily enough, just a timed pulled from the local system time on ever received message and sent message then saved to an open file as a log.
Would be nice to have avatar icons eventually too, which could maybe be pulled from this forum board, this would encourage people to sign up to the forum to set an avatar.

:O its almost 2am, I'm off to bed. :)

[trece8]

Another perhaps-related-thing could be the allowing of "DJ" users.
"DJ" users could be allowed to manage the mix that everybody gets (So it can be a first step to test server-side-mixing, avoiding the need of automated mixing by the server) by vote, for example.
What they do is receive all the different audio channels and mix them as they wish.
For example, all the different player tracks can be sent to the DJ and the DJ send his mix to the server, to broadcast it to the players** (that way, the DJ doesn't need high upload bandwidth, just download)
OR he could just send his mix levels to the server, for the server to apply them.
The guy could even apply effects (would be awesome), so somebody could "direct" the jams, switching the attention to the different players, so somebody could direct the alternating of the otherwise-eternal solo players, for example.


**BTW, if your computer keeps your own audio, you can just send the same audio stream to all clients, and then just make the musician hear all the other players (avoiding him to listen his delayed audio).
An example of aplicability is: if two guys play and you send the sum of that two tracks to both (0 dB each, supposing the mix doesn't clip), each guy can have the audio of the OTHER guy by mixing the incoming audio with his inverted audio of one loop before. (If you want me to explain it better, I could draw some images)

I don't know if there can be some streaming improvement with that, but by having that DJ thing, you could just send the same stream to everybody (so the server mixes all the audio only one time).

[AndyMc]

Most of this can be done already in REAPER or even NINJAM standalone but atm last ability to add fx per channel or channel pair, we've talked about this before and what you mention is an example of how it would be needed and used.

When mentioning before we discussed the idea of allowing channels to be routed out on multiple ASIO strips.
This would allow for use of hardware sends/receives also which for anyone with the gear it could be a lot better produced in realtime.

I think the route table on ReaNINJAM FX allows for more outputs to be set but don't know if they all become the same or route individually.

[trece8]

Sorry, Andy... perhaps is because I don't use REAPER, or because I have explained myself poorly, but I can't see any connection between your post and mine. (Except for the FX, but it was a tiny detail in my post)