Freezing vs. Bouncing

Hi Guys I think I'm pretty close to understanding things but despite all of my reading i still just need to ask and make sure i'm correct with my interpretation
 
-is there a difference between freezing a synth or freezing a track?  they seem to do the same thing?
-is there a difference between freezing and bouncing other than one being permanent and the other not?  if no difference, why would anyone ever bother to bounce when you can freeze and keep that option open?
-what am i really losing in terms of adjusting and flexibility etc when i freeze a track/synth?  i've played around and see that if i alter the midi i need to unfreeze/re-freeze to hear my changes.
 
is freezing essentially "turning off" a track or synth temporarily?  as if to say "i've made all the tweaks i need to this synth/track right now and so i can freeze it to save cpu while i work on other stuff.....but if i need to come back and adjust that's no problem at all - just unfreeze"?
 
all said i have tracks etc. that i'm not really bothering with right now....but later might want to come back and put the final touches on them.....no problem to just freeze away as unfreezing is as if nothing ever happened?
 
 
 
tremendous help from you guys the past week a huge door has been opening for me - it's amazing!
 
 

seed
 
-is there a difference between freezing a synth or freezing a track?  they seem to do the same thing?

Freezing a synth is usually taken to mean only freezing the sound generated by the synth. Freezing a track involves applying the effects while the synth is frozen.
 

seed
-is there a difference between freezing and bouncing other than one being permanent and the other not?

There was a time when the freeze function did not exist. Freezing does bounce the track in place and removes the synth (and effects if the option is selected) from memory. Bouncing does nothing to the original track(s). Before the freeze option was added to SONAR, bounce and archive was the technique used to achieve a result similar to freeze.
 

seed
if no difference, why would anyone ever bother to bounce when you can freeze and keep that option open?

Reducing resource requirements is only one result of freezing (or bounce and archive). There is nothing permanent about a bounce, all a bounce does is create a new track; the original tracks are untouched. Some find mixing easier when exclusively working with audio. Some synth patches have a random component to them. It may be easier to work with a static audio copy of a synth's random output. There are those that copy the frozen audio to another track instead of bouncing.
 

seed
-what am i really losing in terms of adjusting and flexibility etc when i freeze a track/synth?  i've played around and see that if i alter the midi i need to unfreeze/re-freeze to hear my changes.

Nothing, it is just a way of generating an audios tracks.
 
 

seed
is freezing essentially "turning off" a track or synth temporarily?  as if to say "i've made all the tweaks i need to this synth/track right now and so i can freeze it to save cpu while i work on other stuff.....but if i need to come back and adjust that's no problem at all - just unfreeze"?
 
 
all said i have tracks etc. that i'm not really bothering with right now....but later might want to come back and put the final touches on them.....no problem to just freeze away as unfreezing is as if nothing ever happened?
 

Yes. But taken to the extreme, it is not hard to imagine creating a project that would not run if tracks were thawed.
 

Actually I thought your initial post had a good grasp of freeze and bounce+archive.
 
Look at the freeze options. There is a check box which is selected by default which should freeze the effects with the synth. It should not require extra freeze step, just check the options you want.
 
From a computer resource perspective, freeze and bounce+archive do the same thing. There are other motivations for bouncing tracks but if all you are interested in doing is free up memory and CPU, freeze will do it.
 
In General, frozen tracks can always be thawed to make changes. No need to worry about glitches or corruption from freeze but it is possible to keep adding stuff to a project and use optimizing tricks like freezing until nothing else can be added to the project. At that point, the project will not be corrupted but it may not play if a track is thawed. IOW, it is always possible to create a project larger than a machine can handle.
 
DimPro can be multi-tibral but it is not as flexible as TTS-1. With 4GB of RAM there is not much difference running the 64bit version of SONAR vs. 32bit version. If you are running many instances of synths that load samples in RAM such as DimPro and Session Drummer, adding another 4GB of RAM may be a good idea.

seedtotally makes sense and i think my idea is that i would be unfreezing pieces at a time - never unfreezing everything all at once.

Luckily, there is no way to unfreeze everything at once to run into that scenario, unfreezing happens on a softsyth by softsynth basis or audio track by audio track basis.
 
 
 

seedone last question for now - when i'm slamming my cpu etc it has to do with everything playing at that moment correct?  what i mean is....if i were to selectively freeze some tracks to help out with the crashing, my best bet would be to freeze some tracks that are at a point where alot of synths/effects are working at the same time?  i do notice a certain point in my tune where it seems to fail the most often.

 
The more synths that are actively being used, the more cpu cycles are being taken.
 
 
 

seedfreezing say an intro only track that lasts the first 45 seconds wouldn't have any impact on the load my CPU is taking say 3 minutes into the tune.  correct?

Even though the softsynth isn't being used after the first 45 seconds, it is taking up some cpu resources unless it's a vst3 softsynth and you're using Sonar X3. Therefore, freezing it would give back some of your cpu cycles.

WRT managing CPU load, it might a case of rethinking the workflow. It sounds like tracks are being built up sequentially with effects added while new tracks are still being added. The is the worst case for CPU management. It precludes the ability to raise mixing latency. Raising latency reduces CPU load. A better way to work may be to track as much as possible. Then when tracking is complete, raise the latency to reduce CPU load. Then add effects. Without the ability to raise latency, you will be forced to freeze sooner and requires a better knowledge of the resource requirements of the plug-ins.
 
It is hard to definitely say what the impacts of freezing the intro would have on the rest of the project but if the intro is complete, freeze it. It can't hurt and might help.

Soft synth resource usage varies quite a lot. With my primary synth instrument (Zebra), one patch can impose a very light load while another patch might bring the CPU to its knees. The same is true - to varying degrees - of other soft synths I've used. Some, such as Z3ta and TTS-1, are relatively consistent (although the TTS-1's CPU usage is fairly high). It's entirely possible that 19 soft synths could either be too much for your CPU, or no problem at all.
 
At any rate, freezing a synth not only removes the CPU overhead, it also reclaims whatever RAM the synth had been using. With all synths frozen, you now have an all-audio project capable of literally hundreds of tracks, even on a very modest system. The downside is you have to un-freeze in order to make edits to the underlying MIDI data or synth configuration.
 
Note that 27% CPU usage is low, nowhere near the danger zone. I have projects that run at 70-80% CPU with no problems. If you're experiencing crashes (real crashes, e.g. GPFs) then something else is awry. It may be that one of those synths is failing internally, which can be an especially common occurrence with many free synths. You might find that when one particular synth is frozen your problems go away, in which case you might consider retiring that one from future use.

thanks for the advice guys
 
i freezed about 7 or so synths and now i hover around 17%-22% which is great
sad to read that bitflipper is suggesting something else is wrong but there's not much i can do about it.
thankfully it works well enough most of the time.
 
question about the frozen tracks:
-if a track has a volume envelope in it.....i'll need to adjust that envelope if i want that track to have more or less volume yes?  last night i grouped every track so that I could move ALL of my sliders up by the same amount.  i'm pretty sure that the frozen tracks with volume envelopes did not change.  i then went into the frozen track and moved the node up a bit.  does this sound about right?   am i losing this change if i were to unfreeze the track?
i guess i just had the thought that once a track is converted to audio the volume envelope changes are written in stone like something in the FX bin would be.  thinking that by moving all of the sliders up by 15%....that this envelope would be raised by the same.  i do not at this point believe that is the case though?
 
-what happens to a frozen track if i insert some blank measures in the middle of the project?
for these tracks with the volume envelopes.....they are now "broken" instead of that continuous WAV/audio file that spans the entire tune.  do i need to unfreeze everything....insert the measures....then refreeze again?
 
 
amazing weekend for my understanding
cheers to you all!

definitely helps!
 
man this is all so inspiring.....every time i get a chance to devote 3 or so hours to this I end up having a breakthrough;  either in my understanding or in learning something new.....or even musically with what i'm currently constructing.
last night i bounced a snippet of a track, reversed it....and instantly discovered what will be the bridge melody between the first and 2nd half of the tune :)
 
 
fyi moving all of the faders was an effort to just get them all up around the middle or upper half of their slots on the mixer board
 
i have two tracks that for whatever reason need to be MUCH louder than the rest and so for quite some time all of my tracks are VERY low and i have to turn the volume up on my speakers just to hear my tune.
i went into the groovesynth to make sure the level was all the way up and i also maxed the velocity and that enabled me to drop this slider and raise the sliders on everything else a bit but there still is quite a gap between these two tracks and the rest of the project.
 
i'm not sure if there is a problem per se with most of my sliders being so low....i suppose i could boost the whole track when i'm at the mixing/mastering stage?
or perhaps if the ideal is to have all sliders hovering around 1/2 or 3/4 then i'll just need to find a way to boost these two tracks or perhaps limit the rest of the tune.
i'm about 90% finished my garrison book.....which means i haven't reached the mixing/mastering chapter yet so i'm just brainstorming out loud :)
 
 
thanks guitarhacker hope yall are thawing out down there!
 
 
 

yeah knowledge surely is power!
 
the "forbidden 0db mark".....that's the slider raised all the way up in the red correct?  and will cause clipping?
if so yeah i've managed to keep everything away from there thus far.
thanks for the advice on the gain adjust - i'm hoping it will be just that simple
 
 
referencing your first paragraph again.......how to you guys sort the bazillion synths and samples we have at our fingertips?
i spent a few DAYS just going through pads and leads for this track i'm making.  i can't imagine doing that process every time so i'm thinking i'll have to devise some way of taking efficient notes so i can find the sounds i want quicker.  i know this is a common issue for people who use this type of software and that some of it just can't be avoided.  i have to believe there are ways to simplify this process tho!
 
 

If you are worried about that "destructive" thing Bitflipper (Dave) spoke about,,,,, simply clone or copy the track first. Then mute the clone and work on the original. (or vice versa)
 
Besides, the destructive "can't ever go back" only exists after you save and close the project. Until that time, you have EDIT>UNDO to bail you out of trouble.
 
0db? .... pffft... rules are for sissies 
 
Yeah, I try not to violate that by keeping it all out of the red.... but the occasional red flash isn't going to kill you.... you certainly don't want to run solid red.
 
On the tracks that you say are very low in volume... reference the pic I posted above.... are the track waves nice and full and fat looking like the wave in track 14 or are they looking more like a thin weak looking straight line for the most part? You may not have sufficient gain on the signal going in to provide you with a nice full wave.
 
Midi, for the most part doesn't have that sort of level issue.... at least not for me. Most times they are nice and full when converted. 
 
Yeah, have a look at the wave form in the track. It should be taking up a large area top to bottom ... 75% or so..... up to just shy of 100%.. you don't want "overs".  When I first started I was having 10% waves and that's a matter of getting the inputs set properly to you get better levels into the inputs. In my experience,  (see track 11 in my pic) even a thinner wave will sound good. Although it will not be as loud, there are things that can be done should you find that a given track is not loud enough. In most cases, I will bring the other tracks down rather than try to pump a weaker track up to their level. Not always, but certainly most of the time.  As long as you're not getting close to the levels where the noise becomes audible.... it's all good.
 
I normally end up with a weaker looking track when I track my acoustic instruments.....acoustic guitar and mandolin. But again, the wave picture is only part of the story..... you have to listen to it in the mix before you start trying to pump it up or apply gain to it. Some of my song tracks with mandolin in them have a pathetic looking (according to me) wave, but they sound just fine.
 
If you have a weak looking and low volume sounding wave..... then copy it or close it, and mute one copy like above with  my first paragraph...then for grins.... use the process audio> Gain> +3db on that track.... this increases the gain/volume by 3 db every time you apply it. You may need to apply it several times to get it up to where it needs to be.... you can also reduce the same way if something is too loud.   OR.... go for broke and simply apply the NORMALIZE function. This takes you right to the top with that one click. I would set the level at 95% or so... this leaves some head room in the track and going to 95% is generally more than enough. When I use that I then have to pull my faders down about half way which is fine. Realize that the down side to either of these volume increasing methods is that the noise in the track is also increased by the same amount as the music. Use these only as a last resort. The goal is to get a nice full wave that doesn't need any help with the volume. Not always possible but that is the goal...at least around here in this studio.
 
Glad to hear that you are having fun.....