i understand that what you want is a single button to freeze an entire mix, but remember that implementing any feature means that some other feature is not being implemented. What I don't know is whether substituting one click for ctrl-A+two clicks would be seen as such a significant benefit by the majority of the user base that it would be worth taking the time to implement, test, and debug something that essentially saves a shortcut and a click.
 
Furthermore, if you have any external inserts or use external DSP boxes (e.g., Universal Audio or Duende), you couldn't do a global freeze. So you'd have to do a real-time bounce anyway to include those tracks.
 
Another issue is that computers keep getting faster and latencies keep getting lower. A lot of people don't even feel the need to use track freeze any more; since the whole reason you want a global freeze is to overdub with the lowest possible latency, I think this feature might become irrelevant sooner rather than later anyway.

Mr. Anderton,

The tradeoff is that if you record, you're recording the processed sound - not recording the dry sound and "re-amping" afterwards, as happens with conventional plug-ins. However, this means you have to commit to a sound in the recording process, but that's not necessarily a bad thing.

 
Not so.  When you do this, if you split the input (which I actually do with a little external mixer, but if using the Toneport AS your recording soundcard happens internally), then you get the dry signal on the recorded track and the wet signal as "what you hear".  It really is the best of both worlds.

Hmmm...it was that way with the KB37. I covered that aspect in a review of it, and the review went through fact-check with Line 6 and they said it was right. Maybe it changed when the Gearbox software became a plug-in instead of capable of being run only in conjunction with the interface?
 
In any event, thanks for the heads-up if my info is outdated...I noticed I wrote the review in 2007 (!)

Blades
Not sure how relevant this is now, but back in the early 2000s I got a Line6 TonePort device, one of the first with near-zero monitoring, I think.  The nice thing at the time is that you could set your mix to something intolerable like 40ms and still play your guitar or record your vocals WITH effects via the Toneport software at something more like 2-5ms.  So, when you had a really full-of-effects project, you could easily add in your effect-laden guitar stack or several layers of vocals with reverb with no problem.
 
It was REALLY helpful.  Of course, now with more power in PCs, better driver stacks, etc, getting lower latencies later in a project is a lot more realistic than it was then.  I still occasionally like to use this method when I think of it, rather than going through the freeze process or similar, which are a creativity killer for me.  This is one of the reasons why the vDrums kit still trumps any sample librsary I've tried.  The latency is VERY noticeably better for getting groove nuances and feel when playing even if technically the samples are better in a drum library.  I just can't DEAL with any noticeable latency when tracking drums.  It just wrecks it for me.
 
I think if you have an option to be able to get closer to zero (regardless of what it actually measures at) you should do whatever you can to get there.  I'd rather record with no effects/completely dry (of course not reasonable with effected guitar sounds) than to have a little latency.  It takes a little getting used to, but its easier to do and the end result is simply much tighter recordings.
 
There's my $.02, anyway.

Blades - I fully concur with your experience.

I have a machine with very fast specs, lots of RAM and now SSDs. I have several excellent audio interfaces.

But for recording audio with a zero latency like experience, I use a Line 6 KB-37, a cousin of the device you mentioned. In side by side tests time and again I find the KB-37 consistently delivers a very satisfying zero latency like experience even when I'm listening to the audio track input monitor of the track I'm recording, even when it's loaded with effects - in songs with 75 other audio tracks. This is compared to other interfaces that can introduce some latencies that can be hard to work with. In contrast, the KB-37 lets me hear the fully effected signal like real time, it just works and records the dry signal cleanly. It's fantastic.

Here are two dated links about the KB-37, but it's a champ:

http://www.youtube.com/watch?v=zwBR1OfHRtM

http://www.youtube.com/watch?v=cRpCXWSC8TE

Anderton
i understand that what you want is a single button to freeze an entire mix, but remember that implementing any feature means that some other feature is not being implemented. What I don't know is whether substituting one click for ctrl-A+two clicks would be seen as such a significant benefit by the majority of the user base that it would be worth taking the time to implement, test, and debug something that essentially saves a shortcut and a click.
 
Furthermore, if you have any external inserts or use external DSP boxes (e.g., Universal Audio or Duende), you couldn't do a global freeze. So you'd have to do a real-time bounce anyway to include those tracks.
 
Another issue is that computers keep getting faster and latencies keep getting lower. A lot of people don't even feel the need to use track freeze any more; since the whole reason you want a global freeze is to overdub with the lowest possible latency, I think this feature might become irrelevant sooner rather than later anyway.

Computers do get faster, but the software gets fatter. As I mentioned earlier in this thread, if MOTU would have allowed plug-in use for their own low-latency CueMix, I wouldn't even be asking this, since I wouldn't be using the Echo function in Sonar. Even so, using the Echo function in Sonar is very inconsistent for me (sometimes merely stopping and restarting the Playback drastically changes the echo delay), and regardless of how low you can get the latency, there's still a serious battle between delay and audio-drops. We're not there yet, IMO.
 
True that some features which we ask for could be asking to reduce the resources to work on other aspects of Sonar features, but lately, what CW has put into Sonar hasn't been too useful for me. There are some cool features, to be sure, but nothing lately that has saved me any significant time in tracking and mixing beyond X1.

brconflict
 
Computers do get faster, but the software gets fatter. As I mentioned earlier in this thread, if MOTU would have allowed plug-in use for their own low-latency CueMix, I wouldn't even be asking this

 
The problem is as soon as you allow the use of plug-ins, then you're back to the latency issues that IIRC CueMix is designed to defeat by hooking directly into your interface and bypassing the computer.
 

Using the Echo function in Sonar is very inconsistent for me (sometimes merely stopping and restarting the Playback drastically changes the echo delay), and regardless of how low you can get the latency, there's still a serious battle between delay and audio-drops. We're not there yet, IMO.

 
Are you using plug-ins with lookahead? Understanding when/how to use the PDC (Path Delay Compensation) function in Sonar can be very helpful. For example limiters, compressors, "maximizers," and even some noise gates can have lookaheads of one to dozens of milliseconds; having a Waves Maximizer inserted in the master bus during tracking causes the latency to skyrocket to the point of being unusable. This is because Sonar is using PDC to delay all the other tracks to match the delay of the maximizer so they all sync up. When tracking, I simply disable plug-ins with lookahead and can then run at the lowest possible latency. When mixing, a couple dozen milliseconds of extra latency doesn't really matter except for fader moves requiring extreme precision.
 
I record consistently at 96 samples without dropouts, and these are projects with lots of tracks...in some cases over 80 tracks. With lots of virtual instruments, I need to bump up to 128 samples because I usually don't freeze them...128 samples is acceptable to me.
 

True that some features which we ask for could be asking to reduce the resources to work on other aspects of Sonar features, but lately, what CW has put into Sonar hasn't been too useful for me. There are some cool features, to be sure, but nothing lately that has saved me any significant time in tracking and mixing beyond X1.

Speed Comping, Quick Grouping, audio-to-MIDI conversion, and the Bi-Directional Browser and other browser enhancements (e.g., keyboard shortcut support) have saved me a lot of time. Also colorization makes it easier to parse tracks quickly, and using Melodyne means I don't lose time from V-Vocal crashing  However development also means things like VST3 support and ARA integration, which may or may not speed up tracking and mixing but support newer generations of tools like Melodyne, Padshop, etc. that some people consider important. Also ARA can be useful in more contexts than just pitch correction, and I expect will be exploited for other applications in the years ahead. But from a programming standpoint, from what I understand it was a bear to add and QC ARA and VST3 integration.

Anderton

brconflict
 
Computers do get faster, but the software gets fatter. As I mentioned earlier in this thread, if MOTU would have allowed plug-in use for their own low-latency CueMix, I wouldn't even be asking this

 
The problem is as soon as you allow the use of plug-ins, then you're back to the latency issues that IIRC CueMix is designed to defeat by hooking directly into your interface and bypassing the computer.
 

Using the Echo function in Sonar is very inconsistent for me (sometimes merely stopping and restarting the Playback drastically changes the echo delay), and regardless of how low you can get the latency, there's still a serious battle between delay and audio-drops. We're not there yet, IMO.

 
Are you using plug-ins with lookahead? Understanding when/how to use the PDC (Path Delay Compensation) function in Sonar can be very helpful. For example limiters, compressors, "maximizers," and even some noise gates can have lookaheads of one to dozens of milliseconds; having a Waves Maximizer inserted in the master bus during tracking causes the latency to skyrocket to the point of being unusable. This is because Sonar is using PDC to delay all the other tracks to match the delay of the maximizer so they all sync up. When tracking, I simply disable plug-ins with lookahead and can then run at the lowest possible latency. When mixing, a couple dozen milliseconds of extra latency doesn't really matter except for fader moves requiring extreme precision.
 
I record consistently at 96 samples without dropouts, and these are projects with lots of tracks...in some cases over 80 tracks. With lots of virtual instruments, I need to bump up to 128 samples because I usually don't freeze them...128 samples is acceptable to me.
 

True that some features which we ask for could be asking to reduce the resources to work on other aspects of Sonar features, but lately, what CW has put into Sonar hasn't been too useful for me. There are some cool features, to be sure, but nothing lately that has saved me any significant time in tracking and mixing beyond X1.

Speed Comping, Quick Grouping, audio-to-MIDI conversion, and the Bi-Directional Browser and other browser enhancements (e.g., keyboard shortcut support) have saved me a lot of time. Also colorization makes it easier to parse tracks quickly, and using Melodyne means I don't lose time from V-Vocal crashing  However development also means things like VST3 support and ARA integration, which may or may not speed up tracking and mixing but support newer generations of tools like Melodyne, Padshop, etc. that some people consider important. Also ARA can be useful in more contexts than just pitch correction, and I expect will be exploited for other applications in the years ahead. But from a programming standpoint, from what I understand it was a bear to add and QC ARA and VST3 integration.

I must not be making a useful argument, and spinning my wheels here.  I had a long response prepared to address each point, but it's really not going to help. Nothing added to Sonar recently has made anything faster for me beyond X1. It's only become more reliable and responsive, which I praise the bakers for that. I record and mix live bands in the studio. I was excited to see the new features, but in practice they've only moved the efficiency challenges rather than eliminated them.
 
I'm still wishing for Global freeze. In my shoes perhaps it might make more sense.
 
Thanks!
 

brconflict
I record and mix live bands in the studio.

 
Aha!! That was a crucial piece of data   Now I see where you're coming from. (And it's kind of comforting to know some people still record that way...)
 
However, if you are using plug-ins during the tracking process, please do consider the above comments regarding PDC. If you don't use plug-ins, they're irrelevant.
 
One possible solution is to avoid plug-ins altogether while tracking by using an interface with DSP. That way the singer can get reverb in the cans, the bass player can get compression, etc. Then when mixing, you can replace those with "real" plug-ins.
 
The people I know in your situation sometimes choose to separate the capture and editing media, e.g. Bruce Swedien prefers to record to tape then transfers to a DAW for editing. Maybe something like this is more relevant to your needs than a DAW.
 
Your application explains in a nutshell why ADAT was so popular...tracking was a piece of cake, then you could light pipe to a mixer or audio interface for editing/mixing.

Yeah, I'm going to be shopping for better A/D in the next year or so. Use the MOTU for full band, and use another DSP A/D for Vocals. Seems like the best way in hardware.
 
Yes, sorry I should have mentioned that before that I mainly do Live bands.
 
Nah, I'm just hoping to improve on the Echo function Sonar has, if possible. As the hardware speeds up, and knowing PT has a low-latency tracking function, I'm thinking the idea of freezing tracks is JUST almost there.
 
Then we can talk about PSG's, another suggestion I made (a bit confusing to the board) that could save me hours and hours of time in multi-song projects.   I just sent it on to Noel with a Screencast. I don't know if he's watched it or sent it on, or filed it. But I'm passionate about that one.

mike_mccue
 
 
As a comparison, if you are looking at the timing for a round trip through an analog monitor mix you can consider it would take an approximately 700 kilometer long audio cable to experience a 2 millisecond delay in the analog domain.
  

This is actually the exact length of all the cabeling I have in my studio.  True story...