Tracktion Software - Freezing Rendering Exporting Video Lecture | Tracktion Tutorial: Master Music Editing - Video & Sound Editing

freezing rendering and exporting when
working with large complex edits which
is easy to run out of processor cycle
each filter you add to an edit will
increase the CPU load and when the CPU
meter starts to approach the redzone he
risks dropouts in your audio as your
system struggles to keep up you may be
able to deal with more tracks by
increasing the cache size in the file
settings tab of the settings page if
your system is still struggling however
you may need to freeze some tracks
select a track or several trucks and
press the freeze tracks button in the
properties panel traction will now
render all those selected tracks to a
single 32 bit freeze file which is
stored in the temp directory and then
automatically mute those trucks and
shrink them to a thin strip to indicate
their frozen statism now any filters on
those trucks will use no processor cycle
and any audio on those trucks will not
need to be read from the hard drive but
you will still be able to hear them
because traction will automatically sync
the playback of the freeze file behind
the scene you may wish to switch to the
file settings tab of the settings page
and check that the temp directory is
located on a partition within a free
space you will not be able to make any
changes to the frozen tracks unless you
select them again and unfreeze if
several tracks are frozen and you
unfreeze just one of them you will need
to wait while traction recalculates the
freeze file actually what traction is
doing is rerender in all the frozen
tracks but this time without the one you
just unfrozen
if you cancel this process all the
trucks will be unfrozen immediately
leaving you free to make a new selection
of trucks to refreeze unfreezing all
your frozen trucks works immediately
because there is no new freeze file to
calculate trucks that are routed to
other tracks cannot be frozen directly
but it is possible to freeze the entire
subgroup by freezing just the group
track
if you freeze the entire drum group for
example you can then work in detail on
other elements of the mix without your
process of straining to calculate the
drum mix at the same time freezing a
group will only shrink the group track
however so you should be careful not to
make any changes to the individual
tracks as you will not hear the results
until you unfreeze the subgroup you can
also freeze an entire folder by freezing
its overview truck
I often choose to freeze all tracks in
an edit before tracking important
elements like vocals as this ensures
that all my process of power and hard
drive capacity are devoted to capturing
the performances and it greatly reduces
the risk of glitches or dropouts
spoiling a good take to do this simply
select one truck then press ctrl + a to
select all tracks then hit freeze
traction will pop up a warning if any of
your trucks are routed to subgroups but
it will freeze the group trunks anyway
so you can ignore this warning and let
traction sort it all out for you when it
has finished the Edit should sound
exactly as it did before freezing but
your CPU meter will now be reading
practically nothing and your hard drive
will only need to stream a single stereo
track freezing can also be used in
conjunction with all send but you need
to know a little trick freezing a track
that sends to a reverb will render that
track without the reverb return so it
will lose the effect if you select the
return track as well as the send track
and freeze them together then it won't
lose its effect
but the reverb is now frozen so any
unfrozen tracks can no longer use it the
trick is to select the returned truck
and use the standard copy and paste
functions to create a duplicate now make
sure to select one of those returned
tracks along with any others before you
hit freeze if you unfreeze all tracks
the duplicates should be muted otherwise
the effect levels will be boosted so
make sure the duplicate return track is
always either muted or frozen if you
need to free up some processor cycles
but wish to keep the ability to mix the
tracks try rendering those tracks
instead the render track button lives
underneath the freeze button and
contains options to render to a specific
file which will allow you to specify the
file name and directory of the resulting
file or render into the project
directory which will name it
automatically and save it into the
current project folder both options
allow you to specify the resolution of
the rendered file and to choose whether
the new track will replace the original
or whether it will be added to the edit
alongside the original choosing to add
the track and then muting the original
will allow you to change your mind later
if you need to rendering a MIDI track
will also allow you to edit the
resulting audio but unlike the freeze
function rendering several MIDI tracks
will increase the load on your hard
drive
while reducing the load on your CPU
which may be a problem with slower
laptop drives or high sample rates
selecting several trucks and rendering
them together will create a single file
with a mix of all those trucks
you
it is also possible to render subgroups
or trucks which return audio from rewire
devices if you press the render button
for a track with no clips traction will
check that this is really what you want
to do
once you are happy with your mix you
will probably want to export it as a
single audio file to be mustered or
burnt onto CD the export menu contains a
create an audio file option which will
open a render dialog here you can
specify the name and directory of the
resulting file choose from various file
formats specify mono or stereo and set
the resolution normally you should set
the sample rate the same as the current
driver settings as this will give the
best quality results but you can choose
a different one if you need to and
traction will sample-rate convert the
result for you the best bit depth to
choose depends on what you plan to do
with the resulting file if the song is
completely finished and ready to burn to
CD you should choose 16-bit and you
should also enable the differ option
unless you have used a third party
dither plug-in the master section if you
plan to process the resulting file in
any way afterwards however even just
with a simple volume change or fade-out
you should choose the 24 or 32 bit
options instead to avoid the div annoys
and truncation distortion caused by
converting the audio to 16-bit 32-bit
audio does not really provide any extra
quality over a 24-bit version but it
does provide extra Headroom the easiest
way to hear the difference is to
deliberately distort the mix bus really
badly by turning up the volume you can
add some extra volume filters to the
master section to make this as obvious
as possible
now export three versions at 16 24 and
32-bit resolutions using whatever sample
rate your drivers are set to if you
import these files into a fresh edit
they should all sound equally distorted
however if you turn the volume down for
the 16 and 24 bit files the result is
just a quiet and distorted signal but if
you do the same for the 32-bit file the
clipped Peaks will be restored and the
signal will clean up a bit you can also
opt to automatically trim the start and
end points of the resulting file or
render just the region between the
markers and if there are any clips or
tracks selected you can opt to render
just those clips or trucks choosing to
render each truck to a separate file
will render each track individually
which will allow the project to be mixed
in a different application if needed
while file exports from traction also
have a broadcast wave timestamp so
applications that support this will be
able to automatically place the audio at
the correct point on the timeline
likewise traction recognizes timestamp
information in broadcast wire files
created by other applications and audio
clips can be moved back to their
original position on the timeline via
the right-click menu the render process
is faster than real-time but the time it
takes to export will depend on the
complexity of the edit and the speed of
your system as well as the length of the
piece a small minority of plugs or
rewire devices may not work properly
faster than real time however
subtraction provides a render at one
times play speed option if you choose to
normalize the export traction will scan
the file find the loudest peak and then
scale the volume level of the whole file
to make that peak hit full scale 0 dbfs
this will guarantee that the file is as
loud as possible we
clipping but it does not guarantee that
it will sound the same volume as other
normalized files as our perception of
volume is based on average levels rather
than peak levels the adjust level based
on RMS option will do a much better job
of matching volume levels between
different exports but unless you set a
very low target value or render a 32-bit
file you will risk clipping the peaks
and introducing Distortion and the
general rule I would advise that you
export your mixes at 24-bit resolution
was plenty of headroom as you can then
adjust their relative levels in context
before converting them to 16-bit files
suitable for a CD to create an mp3 or
OGG file option works in much the same
way but instead of high quality audio
suitable for a CD or DVD it will produce
data compressed audio suitable for
internet distribution traction uses the
lame codec which needs to be downloaded
separately for licensing reasons and
traction needs to be directed to the
lame executable file in the file
settings tab of the settings page you
may need to relaunch traction after
changing this setting the dialog is very
similar to the audio export function but
with a choice of mp3 or OGG files
instead of woth or AIFF OGG Vorbis files
are generally considered to sound better
than mp3 but although it is freely
available from hobus com many people do
not have the codec installed on their
computers and most mp3 players will not
be able to decode them mp3 files on the
other hand will play on almost anything
including many modern mobile phones if
you wish to create an mp3 file that
plays on the widest possible range of
target devices you should choose the
constant bitrate option rather than the
variable or average alternatives you can
then choose a bitrate from the list on
the right higher values here translate
to better quality sound at the expense
of larger file sizes 128 kilobits per
second is the most commonly used
setting up files on the other hand
should normally be created using the
variable bitrate option which then
provides a quality setting from 1 to 10
again higher values will sound better
but will result in larger files a
setting of 4 or 5 will produce a file
roughly the same size as a 128 kilobits
per second constant bitrate export OGG
in mp3 exports both start by rendering a
temporary 16-bit uncompressed version
before encoding so the dealer buttons
should be on unless you have
specifically used a third-party dither
plug in the Edit you can't argue the
files by selecting add id3 or vobis info
and pressing the edit button to enter
the data finally I want to say a word on
the subject of divin dither is a clever
mathematical trick which greatly reduces
the distortion and loss of detail caused
by reducing the bit depth of your audio
signal as tractions audio engine uses a
minimum of 32-bit floating-point
resolution dealer should be used
whenever you export a 16-bit file even
if the original audio files in the edit
are also 16-bit for example here is a
very low level reverb tail which I
exported from traction at 16-bit with no
dither and then normalized in an audio
editor so he can be listened to with
normal monitor gain settings
notice how the reverb itself seems to
become more distorted as it decays and
the sudden stop when it drops below the
level of the lowest bit now listen to
this version which was exported at 16
bit with tractions dither enabled the
dinner has added precisely calculated
random noise to the signal and now the
reverb tail seems to continue for longer
and to decay smoothly into the
background noise this is not magic just
clever statistics notice that the noise
does not start until the start of the
audio and turns itself off after the
reverb has decayed to silence this is
called an auto black and it ensures that
totally silent sections remain totally
silent the last example was exported at
16-bit again but this time with
tractions dealer turned off and the MDA
did the plug used in the master section
instead
this time the noise starts at the very
beginning of the file and continues
right to the end as there is no auto
black this is sometimes preferable as
background noise may be less noticeable
if it is continuous than if it stops and
starts again if you use a third-party
develop lug or a brick wall limiter that
adds its own dither noise it is
important to make sure that the dither
is the very last plug in the chain and
also to ensure that the master fader is
at unity because the volume level of the
added noise is critical this also means
you can't use a dedicated must-have fade
in and out options in the master section
but you can add an extra volume filter
to the master section before the dither
and use automation instead if you stick
to tractions dither you can of course
use the master fade features and there
is no need to set the master fader to
unity whatever you do you should make
sure that you don't add dither twice so
if you're using plugs that are capable
of adding do the noise you should double
check they are set correctly before
exporting