Uniwire
-
- KVRist
- 55 posts since 12 Apr, 2005
I've written the guys at Muse asking when Uniwire would be available for all of us. Last May, they said late June and now, they're not replying to any of my emails I've sent for the last two weeks.
Anyone have any info?
Thanks!
rw
Anyone have any info?
Thanks!
rw
-
- MUSEician
- 682 posts since 20 Aug, 2004 from California
Sorry, we are late. Unfortunately, I cannot be more specific than very, very soon. I hope in a week or two we will be able to be more specific.
Dan Timis
Software Developer
Muse Research, Inc.
Software Developer
Muse Research, Inc.
-
- KVRAF
- 3971 posts since 19 Apr, 2005 from Brissie
I'm just teasing, Dan! Still, it IS quite funny saying "in a week or two we will be able to be more specific" - you guys are getting quite good at bureaucratic-speak! 
I've joined Lurkers Anonymous.
-
- KVRist
- Topic Starter
- 55 posts since 12 Apr, 2005
3 weeks and counting. 
-
- MUSEician
- 682 posts since 20 Aug, 2004 from California
I was hoping that we would ship by AES NY (October 7-10). I thought that there was a very good chance that in those two weeks we will know if we can ship at AES and I was also hoping that I can give you this news.
Unfortunately, it does not look likely that we will be ready to ship by AES. We will be showing the beta version in NY, though. What we showed at NAMM and Messe were early prototypes; now we will show the beta version with all the features in place. This is probably of little consolation to you, unless you happen to attend the AES show. When you see the demo you will understand that this is a very complex and powerful feature.
I'm really sorry for not being able to give you a firm date. It is the proverbial last 10% of the task that takes 90% of the time to finish.
Unfortunately, it does not look likely that we will be ready to ship by AES. We will be showing the beta version in NY, though. What we showed at NAMM and Messe were early prototypes; now we will show the beta version with all the features in place. This is probably of little consolation to you, unless you happen to attend the AES show. When you see the demo you will understand that this is a very complex and powerful feature.
I'm really sorry for not being able to give you a firm date. It is the proverbial last 10% of the task that takes 90% of the time to finish.
Dan Timis
Software Developer
Muse Research, Inc.
Software Developer
Muse Research, Inc.
-
- KVRist
- Topic Starter
- 55 posts since 12 Apr, 2005
Thanks for the info Dan. I was under pressure to expand a scoring for TV system. I'll just make a few adjustments and await uniwire. I do hope it works well when it is released as I'd love to run a big stack of Receptors.
rw
rw
-
- MUSEician
- 682 posts since 20 Aug, 2004 from California
What platform and what host are you using? Can you tell me more about what you are trying to do? Like how many Receptors, how many channels per Receptor, how many channels total, what king of Ethernet router or switch would you be using, would your Mac or PC have gigabit Ethernet, would your router be a gigabit one or 100 baseT?
This would help us with testing since there are many configurations to cover.
This would help us with testing since there are many configurations to cover.
Dan Timis
Software Developer
Muse Research, Inc.
Software Developer
Muse Research, Inc.
-
- KVRist
- Topic Starter
- 55 posts since 12 Apr, 2005
Hi Dan,
Mac G5 Dual 2.0 - OSX Tiger - Pro Tools HD3 - Airlink 101 ethernet switch - The G5 is 100baseT.
I was thinking 3 receptors (perhaps this will not work?) running 8-12 channels each. This way no single unit would be bogged down.
Right now I'm mainly using a Coyote R&D Two-Can for Gigasampler which works great, but I'd like to also add in Receptors dedicated to soft synths. I do have nearly every (great) soft synth for Mac running in TDM/RTAS or VST shell mode (Fxpansion) but running more than 5-6 weighs down the CPU substantially. Especially plugs like the CS-80V and Moog Modular V.
So my dream setup is this. Gigasampler (which now uses ethernet for midi only and shows up as 8 banks of 16 channels in Pro Tools. Adat for audio) would be switched over to regular midi (I'll just use that unit to old fashioned way) and then using ethernet for the receptors ESPECIALLY because the audio would come up as a PT channel.
Since my projects are so diverse on almost a daily basis, band, commercials and all sorts of TV projects, I'm sure you can imagine how many different styles and thus templates for those style I need to pretty much instantly recall.
Let me know if you need any further info and thank you for asking!
Ron Wasserman
Mac G5 Dual 2.0 - OSX Tiger - Pro Tools HD3 - Airlink 101 ethernet switch - The G5 is 100baseT.
I was thinking 3 receptors (perhaps this will not work?) running 8-12 channels each. This way no single unit would be bogged down.
Right now I'm mainly using a Coyote R&D Two-Can for Gigasampler which works great, but I'd like to also add in Receptors dedicated to soft synths. I do have nearly every (great) soft synth for Mac running in TDM/RTAS or VST shell mode (Fxpansion) but running more than 5-6 weighs down the CPU substantially. Especially plugs like the CS-80V and Moog Modular V.
So my dream setup is this. Gigasampler (which now uses ethernet for midi only and shows up as 8 banks of 16 channels in Pro Tools. Adat for audio) would be switched over to regular midi (I'll just use that unit to old fashioned way) and then using ethernet for the receptors ESPECIALLY because the audio would come up as a PT channel.
Since my projects are so diverse on almost a daily basis, band, commercials and all sorts of TV projects, I'm sure you can imagine how many different styles and thus templates for those style I need to pretty much instantly recall.
Let me know if you need any further info and thank you for asking!
Ron Wasserman
-
- KVRist
- Topic Starter
- 55 posts since 12 Apr, 2005
PS - I TOTALLY understand how complicated an undertaking this must be for your code writers.
-
- KVRist
- 401 posts since 24 Nov, 2003 from Waiheke, Auckland, New Zealand
Thats why they invented cofee,,,fishermusic wrote:PS - I TOTALLY understand how complicated an undertaking this must be for your code writers.
-
- KVRian
- 1116 posts since 22 Apr, 2005 from Nashville, TN USA
I'm curious, will Uniwire make Receptor look just like a VST to my host (Sonar)? If so, will it offer multichannel audio via the the Ethernet? Lastly, how can it do this if the Ethernet is only 100mbit?? I'm currently using FXTeleport and two P4 Dells, both with GigaBit Ethernet, and I can get 4-8 channels of audio at low-latency. When I throttle the connect back to 100mbit, I can barely get one (24bit/44.1k)...So I'm curious if this Uniwire thing is really so feasible. Or perhaps it DOESN'T offer multiple channels of audio....Sorry if these are ignorant/uninformed questions! 
-
- MUSEician
- 682 posts since 20 Aug, 2004 from California
I cannot talk in detail about UniWire performance until we actually ship the feature. I also don't know what are the performance limitations of FX Teleport.
But, let's talk theoretically. If we use 32-bit floating point samples, which is the format used by VST, 100 mega bits per second / 32 bits per sample = 3.125 mega samples per second, or 3,125,000 samples per second. At a sample rate of 44100 samples per second, this gives us:
3,125,000 / 44100 = 70.86 channels.
This assumes that we can use the entire bandwidth of the network. Of course, there is a lot of overhead, but even if we could use only half of the theoretical bandwidth, we would still get 35 channels. If instead of 32-bit samples we used 24-bit samples we would get 46-47 channels.
Of course, this is theoretical. In practice we may be able to get a little less, but what you are describing is so far outside the ballpark. I am really surprized that you can barely get one 24-bit/44.1 audio channel with FX Teleport. Is this similar to what other FX Teleport users report?
But, let's talk theoretically. If we use 32-bit floating point samples, which is the format used by VST, 100 mega bits per second / 32 bits per sample = 3.125 mega samples per second, or 3,125,000 samples per second. At a sample rate of 44100 samples per second, this gives us:
3,125,000 / 44100 = 70.86 channels.
This assumes that we can use the entire bandwidth of the network. Of course, there is a lot of overhead, but even if we could use only half of the theoretical bandwidth, we would still get 35 channels. If instead of 32-bit samples we used 24-bit samples we would get 46-47 channels.
Of course, this is theoretical. In practice we may be able to get a little less, but what you are describing is so far outside the ballpark. I am really surprized that you can barely get one 24-bit/44.1 audio channel with FX Teleport. Is this similar to what other FX Teleport users report?
Dan Timis
Software Developer
Muse Research, Inc.
Software Developer
Muse Research, Inc.
-
- KVRian
- 1116 posts since 22 Apr, 2005 from Nashville, TN USA
Dan, thanks for the quick and informed response. I should have thought a little more about the other variables before I stated a number of channels...I just ran another "stress test", and in reality, it's the CPU of the remote host that causes the limitations.
I set up 6 instances of something that is fairly easy on the CPU (miniMogueVA-Luxus, which just recently got "receptorized"). Played a simple part with about 5 note polyphony and cloned it 5 times.
Upon playback, it was the slave CPU getting "stressed" that was causing the glitching. I could mute one or two and it got better, or I could up the latency a little and that helped. Then I changed my network card on one machine to only allow 100mbit, and guess what? Not much difference at all. The VST's instantiated a good bit slower, but the playback was about the same.
Sounds like you have done your math. Uniwire is the main thing making me want a Receptor (and the ability to use all my softSynths live). I don't know about other users experiences with FX Teleport, and I've only been using it a couple of months myself. It's nice, but Receptor is sexier (And I need to return my 2nd Dell to work from where I "borrowed" it
I just read the "sneak preview" document on your site, and that actually answers the questions I bothered you with:
1) Receptor with uniwire appears as VST plug in within host
2) 32 channels of audio
3) 16 channels of MIDI
This is gonna ROCK!
I set up 6 instances of something that is fairly easy on the CPU (miniMogueVA-Luxus, which just recently got "receptorized"). Played a simple part with about 5 note polyphony and cloned it 5 times.
Upon playback, it was the slave CPU getting "stressed" that was causing the glitching. I could mute one or two and it got better, or I could up the latency a little and that helped. Then I changed my network card on one machine to only allow 100mbit, and guess what? Not much difference at all. The VST's instantiated a good bit slower, but the playback was about the same.
Sounds like you have done your math. Uniwire is the main thing making me want a Receptor (and the ability to use all my softSynths live). I don't know about other users experiences with FX Teleport, and I've only been using it a couple of months myself. It's nice, but Receptor is sexier (And I need to return my 2nd Dell to work from where I "borrowed" it
I just read the "sneak preview" document on your site, and that actually answers the questions I bothered you with:
1) Receptor with uniwire appears as VST plug in within host
2) 32 channels of audio
3) 16 channels of MIDI
This is gonna ROCK!
-
- KVRian
- 1116 posts since 22 Apr, 2005 from Nashville, TN USA
YIKES! The question then is WHEN?!?!?!?Hybernation wrote:This is gonna ROCK!
Just read all the annoucements from AES. VERY happy that IK ST and SS2 are being supported.
And VERY DISAPPOINTED that V1.4 is available and DOES NOT INCLUDE UNIWIRE!
WHEN WHEN WHEN...
