Syntrillium Backup Forums :: View topic - Random Skipping/Jumping in CEP 2.0 Multitrack

theoria · Location: USA

(FYI, I've changed the topic of this thread to comply more with the
last post I made)

Hello,

I just received CEP 2.0 and am extremely happy with some of
the neat features that it offers. I'm annoyed by the lack of
VST support, but with all the DX effects thrown in and wrapper
possibilities, I think I've got enough to work with to keep
me busy. Smile

I'm also excited to start using the Red Rover.

Anyhow, I just started experimenting with CEP last night and haven't
had a lot of time to optimize my machine (AMD 1GHz Athalon proc.,
128MB RAM, SB Ensoniq sound card & 40GB HD). But, for the record,
I have been working on a few very involved projects using
<A HREF=http://www.fasoft.com>N-Tracks Studio</A> for the past
few years, sometimes mixing more than 25 tracks w/various
compression, eq & reverb plug-ins per track. Although I'll have
occasional burps while listening in multitrack view,
I have *never* had any glitches in my mixdowns using N-Tracks.
Everything has always sounded exactly the same in my mixes as
they do in multitrack view, when everything is functioning properly.

Thinking that CEP would offer me a more professional recording
environment (N-Tracks is a shareware program for ~$40.00), I wouldn't
have thought that the simple task of mixing down would introduce
strange glitches (not audible during multitrack playback) into the
final wave file. Unfortunately, this problem was repeatable and
I haven't gotten a clean mixdown yet -- although the glitches/sound
burps appear in different places from one mix to another.

I should also mention that random glitches appeared occasionally when
listening to my 2 track test in multitrack mode (pre-mixdown). Also,
one track was using reverb w/stock settings, I had experimented with
routing to a bus with a mix & I also experimented with the stock
limiter plug-in.

Although there are times when I get glitches while listening to
N-Tracks projects w/+25 tracks in multitrack mode, I've *never* had
random glitches appear in my final mixdown wave file -- even using
the default optimization settings. Is there a solution for this? I
would think that mixdowns would be a trivial task for a $250 program,
especially when a $40 program written mostly by one person has no problems using the same computing equip

ment.

Maybe some of the optimization tips will help (I'll try tonight), but
I'm a little disappointed since I wouldn't think optimization
settings would matter when doing mixdowns!

Any comments?

Thank you for your patience as I get acquainted with CEP! I'm sure
I'm missing something obvious...

+Steve

ozpeter · Location: Australia

Well, it shouldn't happen... maybe report back after you've tried optimisation (and seach the forum for 'optimise' and 'glitch' etc if you haven't already done so) - you may find others here won't jump in until you are sure you've tried all reasonable tweaks & settings yourself.

- Ozpeter

VoodooRadio · Location: USA

The old acronym YMMV (your mileage may vary) has certainly seemed to hold true in people's ability to get it up and running. On one hand, there are those that are elated and report no difficulties. Then, there are the ones that disclose.. they had problems, but after "optimizing", all's well. And then, there's a handful (myself included) that found it very confounding! Your system description certainly seems "more than capable", although I would plan on increasing the amount of RAM. I didn't see in your post, where you mention which version of Window's your running on, and that seems to make some difference for some folks. I am running Ver1.2a and sometimes run sessions with upwards of 30 tracks (using Win98fe) without a hitch what-so-ever. I guess that is what has me both spoiled and paranoid about making a change to "the new" version. Shy

theoria · Location: USA

OK,

First off...Sorry I forgot to mention this...my system is running
XP Professional and I agree, my system could stand a
boost of another 128MB. What's troubling is that I've been able
to push 25+ tracks (most with at least 2 VST/DirectX plug-in effects
each as well) w/N-Tracks and a few effects on the master channel
without any glitches whatsoever, using the same system.

Anyhow, I've now had some time to get a real feel for CEP 2 and have
done quite a bit of optimization -- I think I've tested at least 100
different configurations! I've even migrated most (10 tracks) of
a song I did in N-Tracks to CEP, which I must say was an interesting
test. Pardon my longwinded-ness below :)

After defragging my hard drive, I saw the original problem with
the mixdown glitches go away (I think). So, this must have had
something to do with the fact of my drive being a mess. BUT, I
still had plenty of glitches on multitrack playback when I tried
pushing more than 3 tracks with 2 plug-ins on each track (reverb &
compression). So, I tried various buffer settings, recommended XP
performance tuning, etc. and still heard glitches when playing back
in multitrack view. I even tried switching to 16-bit mixing, but
still heard skips, etc.

So, I decided to do a side by side comparison with N-Tracks
(http://www.fasoft.com), loading the same wave files in CEP2 as in
N-Tracks and playing in multitrack while checking system
performance. Right off the bat, I saw that just by opeing CEP2,
the memory footprint was about 8x (@ ~40MB) that of N-tracks,
which was about 5MB. So, CEP2 is definitely more resource
demanding. Then, while playing back songs with the same wave files,
I noticed that N-Tracks appeared to handle plug-ins much more
efficiently. This is where my glitches appear to be coming from
on the CEP2 side. Each additional plugin in CEP2 appears to drive
the CPU load up by 10-15% whereas in N-Tracks it's something closer
to 2-5%. Even when using the same plugins in each program this
appears to be the case. Everytime the CPU load hits %80+ a glitch
occurs. In N-Tracks, even with 20 tracks with 2 plug-ins on each,
loads of this high are *never* reached. Lastly, CEP2 appeared to be
making greater use of the HD than N-Tracks when playing back, which
probably also creates better conditions for glitches -- could this
have to do with my buffer settings?? I've tried countless settings...

From this it appears that CEP2's higher demand on system resources
is what is causing most of my problems, when compared to N-Tracks.
I'm a little concerned that the load is so much more higher with CEP2
when using the same # of tracks with the same plug-ins. I wonder if
this is because CEP is doing "something" in a "higher quality fashion"
or if it's just because there's a lot of sloppy code under the hood?
Regardless, I really hope a patch is coming out soon to perhaps
address some of these issues. N-Tracks is $40 by one person who
happens to release new versions practically every month!

With all this said, I must say the sound quality and many of the
features of CEP2 blows N-Tracks out of the water. I'm excited to
start making use of this. I also noticed that when I
locked effects on tracks in CEP2 (which is a neat feature) that
it DRAMATICALLY reduced the load/memory consumption, but it was
annoying that I'd have to wait for these to get process everytime
I reopened an existing session. Is there a way to save these "locks"
so that one doesn't need to wait for them to get created everytime
they bring up an existing session???

Until, I get these glitches worked out I won't be able to make good
use of this application, unfortunately. But I do hope to get some
solutions!

Maybe I'd be better off trying 1.2a?? Could a CEP2 registered user
use the older version? Seems kinda strange, but I can't stand
glitches and it appears that many who claim no problems with 1.2a
are hesitant to move to 2.0. Hmm..

Thanks for reading!
+Steve

jeff12345 · Posted - Mon Jan 13, 2003 12:11 pm

I agree with Pavell. I've got an 80G hard drive and 512MB of RAM. No glitches here.

theoria · Location: USA

Thanks for all the fabulous information!!

I'll try a RAM upgrade as soon as I can afford it -- my budget
is very small! I received CEP2 as a Christmas gift...

One question I have over and over again is why I don't have
any problems w/skipping, etc. using N-Tracks (http://www.fasoft.com)?
I believe N-Tracks tries to do as much as possible with RAM
(not HD), thus enhancing the playback performance with less
disk reads. After fully importing all the tracks from an N-Tracks
song to CEP2 with a very similar effects setup (having a few more
in N-Tracks), CEP2 appears to put a much heavier load on the system
(even after locking all the tracks with dedicated effects). I'd
be surprised that a $40 program written by one person is doing things
more efficiently/effectively than a $250 program? I'm not sure if
N-Tracks does it's mixing at 16-bit, but even after trying 16-bit
mixing in CEP2 I still get occasional skips. Is anyone familiar
with N-Tracks who could make a good comparison of the inner-workings
of it versus CEP2? I'd be interested to know what the key differences
are.

One thing I noticed last night was that if I hit Ctrl-Alt-Delete
in CEP2 while doing a multitrack playback to bring up the CPU/Memory
load meter, the whole application would freeze for a second and
the sound would studder immensely. When I tried doing the same thing
in N-Tracks (w/20 tracks & ~15 effects running real-time (no locking
capabilities)), the load meter came up no problem w/o any studdering
whatsoever. I found this to be an interesting test and perhaps
telling with regard to how efficiently the code for each application
was written. What are your thoughts?

Thanks again for your help! I'll try to scrape up some funds for
RAM and report back. I really do like a lot of the features in CEP2.
I just can't stand the skipping!

+Steve

P.S - What are the best buffer settings for the Ensoniq card?
Any recommendations? I haven't run across any suggestions for
this card yet and have tried countless settings that all appear
to have similar effects. Also, where can I obtain the latest
drivers for Ensoniq? Creative Labs appears to have dropped
support and I couldn't find anything through google. I'm using
what Microsoft threw at me when I upgraded to XP -- I believe
their dated & signed 7/2001.

MusicConductor · Location: USA

VoodooRadio · Location: USA

theoria · Location: USA

Thanks for the info! Could you let us in on what the specs of
your system? I'd be curious to know.

If anything, as I believe I mentioned previously, CoolEdit does
seem to have a more "quality" sound - especially the reverb
plugins. I'm very excited to get going with this package. I
just want to make sure I work out all the "bugs" before investing
a lot of time in a serious project and then learn of something
that stops me in my tracks. But again, thanks for sharing this
with us. I am in the process of scraping up funds for more RAM
-- I know it's cheap, I just a lot of other expenses at the moment
that must take priority :)

Last night I spent about a half hour doing some more tweaking.
After locking all the tracks w/direct real-time effects, changing
the buffers to 12 seconds x 4 and checking on all the DMA settings/
virtual disk settings, I was able to playback 12 tracks and do
a Ctrl-Alt-Delete to get the taskmgr up WITHOUT any skipping
in the audio (yay!). And then, after turning off Live Update & Auto
Scrolling, I was able to record an additional stereo track without
any clipping as well. I watched the physical RAM usage / CPU load
during all of this and saw that the whole song played back with an
average of 5-10MB free RAM and a system load of 4-40% throughout.
I also noticed that the song appeared to playback better after having
made it through once previously.

I think things are starting to come together. I know more RAM
couldn't hurt and I'd be interested to see how it affects
performance.

Re: N-Tracks -- I'm not sure what's going on inside, but it appears
to be pretty stable. Regardless, CoolEdit has a lot of really neat
features that N-Tracks doesn't. BTW, does anyone care that CoolEdit
doesn't allow for Master Channel effects to be applied? Or is it
just thought that since this isn't a good idea for mastering
purposes anyhow that it doesn't matter? Just curious...

Thanks for everyone's comments!

+Steve

andyeb · Posted - Thu Feb 20, 2003 6:26 am

I too am thinking about going over to CEP2 from n-track, but am finding CEP2.0 seemingly less efficient with the same line up of plugins on a single 44.1k, 24 bit stereo track.

System details:

Athlon 1.2GHz
1GB PC133 SDRAM
Abit KT7A M/B
M-Audio Audiophile 2496 sound card
Windows XP (thoroughly stripped down for audio).

Basically, n-track will take the same line up of 10 DX plugins with 30-60% of CPU usage, whereas CEP breaks up before I have even finished loading on the plugins.

I too am surprised at the performance difference, given the relative cost in the two programs.

Andy

otis · Location: USA

I have N-Track and have just recently purchased CEP2. I have run into somewhat different issues in CEP2 that I did not have in N-Track. In N-Track, I can record more than 2 channels simultaneously w/out a problem, but in CEP2, the program crashed when I tried to record more than 2 channels. I played around w/ the buffer settings, etc. in CEP2 but nothing helped until I tried newer drivers for my sound card (which is a Roland RPC-1 card that interfaces w/ a Roland VM3100 Pro mixer). This fixed the problem in CEP2. I was stumped though as I could not explain why N-track could record more than 2 channels at a time w/ the older driver but CEP2 could not. I switched from N-track mainly because of stability issues although I have to say it was one of the easiest to use audio apps I have ever recorded with.

andyeb · Posted - Thu Feb 27, 2003 2:08 am