#Studio16 3.0 track drift
34 messages in this thread
Has anyone here experienced this problem and/or have a solution?
I recently bought the Studio16 3.0 upgrade from Sunrize and I have been using
the program for a few weeks now. First of all it is terrific, even the manual.
I am making the sound track for a 50 minute video, using only two tracks. I am
using LTC SMPTE to trigger the CueList. I am experiencing a drift where the
sound comes in increasingly earlier, and by half way through, the sound is too
early by over 2 seconds. The amount of the error would seem to be related to
the difference between dropframe and non-drop frame timecode. I have checked
out everything many times, restriped my tape to make sure I was using
non-dropframe code, and had discussions with tech support at Sunrize, and so
far no solutions. I have been working on resolving the problem for several
days now, and I am getting frustrated. Manually altering the CueList to make
allowances for this is tedious, because it is an exponentially multiplying
error and must be done from the beginning through to the end in gradual
increments. Fortunately I am not doing matchframe editing, but still it is a
serious problem.
Any help or advice would be much appreciated.
Please put my address in your address book and let me know the resolution, if
any.
I would like to get a sunrize and use it exactly as you are, re-edit the entire
soundtrak of a completed (video-wise) production. If your problem can't be
solved, it
is not the solution I am looking for. Brian
Brian,
The few people I know who use the Sunrize and use it this way, aren't
experiencing this problem. The tech people at Sunrize tell me they have had
three cases of it, including someone in New Zealand. Just about everything has
been tried, including putting the board into another computer, but they have no
solution yet. I am going to try one more thing, restriping my tape with 30
dropframe TC and change my settings in Studio16, and see if that makes any
differences. I am sure there is a solution. I just haven't found it yet.
Apart from this, albeit major problem, this new version 3.0 of Studio16
is a wonderful program to use, very easy, to put together quite complex sound
tracks mixing up to 8 tracks of different sound, and you can see in a glance
exactly what is happening, or supposed to happen, at any given frame, including
all of the dynamics of cross-fades, etc. And it is very easy to make
adjustments.
I will certainly let you know if I find the solution.
Glenn.
Glenn, how are you generating the LC time code? Is whatever your source synced
with the video signal, or is it freerunning independent? Just a thought. Brian
Not your favorite topic, but I am using the TC Gen program that came with
MediaEditor. I have tried non-dropframe and dropframe. However, I don't think
that is the problem. The TC is reading properly.
Tonight I discovered an anomaly in the Sunrize program, between the software
and the hardware. Preferences in the Sunrize program allows you to choose
between Internal or External TC. I have found with my CueList, by checking it
while the tape is running and the CueList performing, that Preferences
sometimes resets itself to Internal. This may be because I have tried so many
solutions, and re positioning my cue points, that I have the program thoroughly
confused. Nevertheless, I cannot get it to stick with reading TC from the
tape.
I have sent two faxes to my contact at Sunrize Tech support tonight, and I hope
to have a solution tomorrow. Here's hoping. This has cost me four days, which
I can't afford right now. With most settings on Preferences, when you save
your setup, they are saved, but I have definitely discovered, that the source
of your SMPTE code is not. I always reverts to the default which is Internal.
Is there another forum I should post this to? Where do all of the users of
Sunrize check into? MultiMedia has a digital video section.
I'll keep you posted.
Glenn
Glenn,
We've had some problems witht the Studio 16 3.0 software ourselves. I do have
a suggestion, reduce the buffering to about 64KB. This is assuming that your
SCSI interface and hard drive can handle it. Sometimes if you decrease the
buffering too much, you may get audio "drop outs".
Regards,
John Chang
Technical Support Manager
Scala, Inc.
I reduced the buffering to 256, but not as far down as 64. I have had a number
of phone and fax exchanges with Sunrize since the last message, and we got
their programmers onto the problem. As my last test, I made a TC window dub
tape and used that to create a new CueList. At 27 minutes into the CueList, it
was triggering the sound 2 seconds early.
They think that the TC on my tape is drifting which is causing a cumulative
error. The TC would be alright when I was laying in my cue points in the list,
but on playback, (and I certainly had to use the Ignore SMPTE errors function),
this slippage was causing the sound to come in early. The program I have been
using to stripe my tapes may be at fault.
As a result I have ordered their SMPTE output module which uses the Toaster to
genlock the signal to the video (or so I understand). At one point they
offered to send me a beta version of a 3.1 upgrade they are working with which
includes, among other things, a TC type auto-sensing feature. I can certainly
try lowering the buffer and see if it works with my system. Thanks for the
suggestion. I will also post the results when I get their TC output program
and see if that solves it.
Glenn Masterpieces in Video
Glenn,
During your conversation on the new Studio 16 Ver. 3.1, did they ever mention a
new feature like "export cuelist"? This is the feature we have been waiting
for. It will allow our Studio 16 EX module to work with multiple tracks. We
have an in-house alpha version that only does mono/stereo playback.
Regards,
John Chang
Technical Support Manager
Scala, Inc.
I really didn't quiz them on the features of the new version, although maybe I
should have. That is the only one I know about. I did suggest adding a
"governor" to the mixer so that you can set maximums and minimums for each
track, so that when you are setting mixer levels on the fly, you can be sure
you are staying within your desired range. They thought that sounded "cool"
and have added it to their list.
Why don't you call Doug at their tech support. He is some kind of head honcho
in that department. From what they told me, he only comes in to help on Tech
Support Mon thru Wed.
Glenn
Glenn,
Sounds like a good idea. I'll grill him for the information. Good luck on
working out your sync problems.
Regards,
John Chang
Techncial Support Manager
Scala, Inc.
I tried your suggestion of lowering my Channel Buffer to 64K. My system
handled it with no problems, except the drift in the sound tracks was just as
bad.
Glenn,
Too bad. If I think of anything else, I'll let you know.
Regards,
John Chang
Technical Support Manager
Scala, Inc.
Glenn, are you syncing your Amiga to the video while running TC GEN now? You
should be able to do this by firing up the toaster and running the video into
input 1. This shoud keep your Amiga in video sync while the TC Gen program is
running striping you time code. If you are using MPhile to get your cue list
number or to do your TC burn ins, that could also be a problem, as I believe
thier update rate is slow and wavers. Brian
By testing, MediaPhile's TC Gen won't output TC with the Toaster running.
However, I have loaded the Toaster and then exited, which leaves its Genloack
function still operation. This did not solve the problem.
One of the techies mentioned that I milght have an Amiga timing problem. I
hope not, because I have no idea of how to correct that.
At the moment I am waiting for the arrival of their SMPTE output module which
will work with the Toaster running (this is what I understand, anyway) I have
enough memory and an acceleration board that should handle it.
I will let you know. I have the audio and video parts of a project comleted
now but without being able to accurately cue in the Audio, I have a problem.
Glenn
Glen,
Interesting. I have to run "ignore SMPTE Errors" to avoid lots of dropout
and intermittent recording problems. I run VITC from the tape through a
Horita VLT-50 to the SunRize card. None of the video editing hardware
notices these problems so I always thought it was in the VLT-50 or
SunRize. Naturally each company blamed the other…
Actually, I just thought of a cross check. I can run the LTC out of the
VLT-50 to the audio-2 of then deck generating the VITC and see if they are
in synch or if there are any errors… Hmmmm
Keep us posted on your progress.
wmc – via Autopilot!
I see what you are getting at. You could also create a TC window dub for each
of the TCs and compare. I finally did that to show Sunrize how much drift
there actually was – about 2 secs every 30 minutes. In a 50 min video, this
plays hell with the production. It is practically impossible to simple correct
the CueList to allow for this, since when you stop the tape in the middle and
start it again, it starts on cue and then gradually drifts again. I'll let you
know what happens when I get the SMPTE module from them, hopefully Monday. I
have a 50 minute video with the video and audio completed, but I can't put them
together to master it. In the meantime, on to the next project…
Glenn,
Most of my assembling of audio with Studio-16 is done is 10 – 30 second
spurts (animation stuff). So I haven't witnessed drastic dirfting. Is it
possible to break up your audio into shorter "scenes" when you lay down
the track in order to avoid significant drift?
You've probably checked this, but you do have the Studio_16 set for
non-drop frame, right?
wmc – via Autopilot!
My longest audio clip is just over a minute, so that isn't the problem, and I
am only using two tracks. I have tried non-dropframe, dropframe (even stripped
a copy of the master with dropframe LTC and tried that but it didn't make any
difference. It is only in a final assembly or test, when I want to run the
CueList all the way through, that the problem shows up significantly. If I
start the cuelist anywhere in the middle, it starts off OK and then gradually
the error accumulates.
I have had one new idea from Gary Kleiner: the control track, particularly if
it wasn't laid in continuously, may be interferring with the LTC, so I am going
to try that. Make a blacked tape and lay in the SMPTE continuously. Then put
in the video in video insert. This would preferably be when mastering the
video. It's worth a try. I am going to waint for the SMPTE module, though. I
have spent so much time testing out new ideas to overcome this problem already,
that my production schedule is seriously behind. I will let you know.
Thanks for the suggestions.
Glenn
Glenn,
I think the control track thing is grasping at straws, so you ar probably
right not to waste time testing it out. The reason I say that is, if the
problem you are experiencing is generic to Studio 16, there is no
indication of control track funnies in my set up, yet I have Studio 16s
erratic reading and report of time code errors all the time – errors my
decks don't see and which are not in the VITC. Also, my decks will abort
an edit if a control track error occurs. And I've only had that happen a
few times (alway with Ampex tape, so I quit using it and not use Fuji an
3M). So, I dont think that is it.
wmc – via Autopilot!
Last night and this morning, I tried a few more experiments. With a fair
amount of effort, I removed BigBlocks from my Audio partition, so I am back to
512 blocks — no difference.
I had also noticed before, that when I played back the narration track, it
seemed to run faster than how it was recording. I simply seemed to be speaking
considerably faster. I have a SETCPU Cache Fastrom statement in my
startup-sequence, so I commented that out and rebooted, and ran my window dub
test tape again. This is very curious. That did slow down the narration to
the speed at which it was recorded, BUT I still had exactly the same timing
problem — the TC discrepancy over 30 minutes was virtually the same to the
frame. What do you make of that?
Still haven't received the Sunrize SMPTE module. That is getting to be my last
hope.
Glenn
Glen,
The problem you're having (slipping audio tracks vs. SMPTE) I have
never seen. The other symptom (changing from external to internal) I have seen
and it may be related to your slipping problem. I have found that the Studio 16
software seems to "auto sense" where the SMPTE time code is. That is, if you
have it set to use internal time code but you have time code coming in the time
code input on the back, it switches to external time code. Likewise, if you
have it set to external time code, but none is coming in (or it thinks none is
coming in) it switches to internal time code. I have observed this more
switching from external to internal, but that is due to the way I tend to use
it in my setup.
Also, make sure that in preferences you have "Lock On" selected. This
is sort of a chase sync mode where it constantly adjust the timing of long
samples to match the incoming time code.
I have never seen this slipping problem and I use this all the time.
Somebody else suggested to make sure that you're time code was synched
correctly to the source you're trying to sync to (VCR?Multi Track Rec?) and I
would reiterate that suggestion. It may be that you're other source is the one
slipping. In my case, I record my time code on the eighth track of my
multi-track recorder and use the output of that track as my source there-after.
Because it's recorded on the tape, it has to be in sync with all the other
tracks on the tape, and the Sunrize always stays in sync with it.
By the way, how long are you talking about? I always work in time
frames from 1 Minute to about 6 Minutes as my main interests are music videos
with a few commercials on the side. I've never attempted any synchronizations
longer than about 6 minutes. Incidentally, my video is all digitized and played
back through a PAR/TBC IV combo synchronized through the SMPTE on the Sunrize.
Hope This Helps You,
Terry Tabor
Terry,
Many thanks for taking the time to offer your experience. All of the
things you have mentioned, I have been aware of, and have experimented with.
One thing I did notice was, although I had set Preferences to external TC, when
I checked on Preferences while the CueList was playing through and slipping, it
had changed back to Internal. The folks at Sunrize decided that there was two
different TC sources at work, and so far I agree. I always keep Lock ON, but
have experimented with it off too, to no success. I have checked my DMA Mask
settings; reformatted the partition to get rid of BigBlocks (no improvement in
slippage); prepared a new master tape with new LTC (no improvement), etc. etc.
At this point, the finger of guilt seems to point to my TC generating
program. It will not work if I try to use the Toaster to Genlock the audio
signal to the video — with any configuration of the Toaster's genlock
functions. The Sunrize SMPTE output module works with the Toaster to achieve
this. I am still waiting for that to arrive — perhaps tomorrow?
My source is Hi8 video tape with the LTC on the left PCM audio track. I
agree with you in that no many people have experienced this problem: ergo there
must be something wrong with the way I am doing things or the way my system is
set up. My video, by the way, is played back through a TBC IV. All of my sound
clips are a minute or less, on two tracks. I am using the 1012 board. I am
working on the problem with a guy in Colorado who has similar equipment to mine
(same GVP board, etc) and he has both the 1012 and 516 boards. He is also a
beta tester for Sunrize. I hope I will have good news on this soon. GLenn.
Glenn,
I am coming to this late, but…. Is it possible that your problem has
to do with hard drive access during playback? Are you working with 50 minute
samples that are continuous? If so, you might try editing some of those samples
into smaller bits of about 3 minutes long and, for the purposes of testing, lay
the shorter bits into the timeline at various points along the total length of
your program. In the edit screen, you can select a short section of sample, use
"COPY AS" and the orginal TC should be retained. If the short samples TRIGGER at
the correct time as you roll your video, but still seem to DRIFT, I would
suggest your problem is not related to TC at all, but may be a playback speed
problem. You mentioned a test you did where you detected a difference in pitch
of a voice sample. So it seems likely that a playback problem related to your
SCSI controller could be the culprit.
Just a thought.
Regards,
Geoff Adams
My samples are already mostly less that a minute in length. The problem is
that they trigger early, at least according to a TC window dub tape I made for
testing purposes. After 30 minutes, they are triggering about 2 seconds before
the TC where the sample is entered in the CueList.
However, the results of my experiment in slowing my system down, and having the
sound return to the speed at which it was recorded, makes me believe there is
something to what you say about the SCSCI controller. There is a difference
between what GVP says the DMA Mask setting MUST be set at, and what Sunrize
recommends, although I have actually tried that with both settings and it
doesn't make any difference. Even when I slowed the system down, I had the
same timing error.
The discouraging news is that I have been e-mailing back and forth with one of
SunRize's beta testers who has been having the same problem. He has similar
equipment to mine (i.e. same computer, GVP card, etc.) but he has both the 1012
and 516 cards. I have only the 1012. I am going to send my most recent
findings to Tech support at Sunrize and see what they make of it. I am waiting
for their SMPTE output module, hoping that would solve the problem, but this
beta tester is using the module, and he is still having the same problem. The
really odd thing is that most users of the boards and software aren't
experiencing this difficulty. If I could only lay my finger on the area in the
setup that is causing the problem, I could probably find a fix to it. I may go
back to ver 2.0 to test it there, and complete this project … but after using
3.0, it is a real PITA to go back.
Glenn,
It sounds confounding and frustrating. Good luck, let's hope somebody
can figure this thing out.
Regards,
Geoff Adams
Glenn, Sorry none of my suggestions helped you. I hope you work your problems
out though. The only problems I have had with my entire setup is getting enough
tracks to play at once. I max out at about six tracks simultaneously. That's
only for long tracks. I can get eight tracks if 3 or 4 of them are only short
sound F/X every once and a while. But synchronization has been a piece of cake
since I figured out that it gets confused if you expect it to run off internal
time code when time code is coming into the external time code spigot.
Incidentally, the internal/external time code box in preferences on mine
actually changes automatically to external if TC is present at the TC input,
and back to internal if TC is not present at the TC input. Kind of begs the
question of why they have the selection box there at all, doesn't it? Why not
just have a internal/external "status" display.
Incidentally, my time code source which I use to stripe my multi-track
machine and/or my S-VHS machine was a compact disc full of test tones
(including 10 minutes of all 4 SMPTE formats) that I bought at the Camelot
store in the mall for $9.95! How's that for synchronization on a budget!
Good Lock (sic!),
Terry Tabor
You should send the suggestion of an Internal/External message to Doug at
Sunrize. Its a good one. Read my message I posted tonight. Sorry about the
formatting. I did it on my wordprocessor and got the margins wrong. I have a
couple of people on e-mail who wanted to know the outcome. Thanks for your
input.
Glenn
Glenn, Thinking about you're problem a little further, I came up with a few
more thoughts. The fact that you're seeing the external TC switch back to
internal TC in preferences indicates to me (at least from the way the software
works in my setup), that for whatever reason, the Sunrize board isn't
recognizing the external TC as valid time code, and (as mine does under the
same conditions) is therefore switching automatically to internal time code
which is essentially free wheeling compared to any external time code.
Then in your comments, you mention that you're recording you're time
code on the PCM track of a Hi-8 tape. I've never tried doing that, and although
the specs would indicate that Hi-8 PCM tracks should be able to faithfully
playback recorded time code, the problem might be that the Sunrize board
doesn't like its time code PCM'ed. Or it might be as simple as an incorrect
(too hot, too little) level setting into the Sunrize board. I would suggest
some experimentation with the levels first. Then maybe try using the linear
tracks for the time code instead of the PCM. My guess is this might not be
practical real world as you can't post stripe using the linear tracks, but it
might give you an indication whether the Sunrize is choking on the PCM'ed time
code. If that's the problem, one solution would be to buy one of those time
code reconstructors (probably not the right word) that accepts somewhat corrupt
time code and outputs a reconstructed version of the time code. They're not
real expensive; I bought mine for about $200 bucks from a company called
"MIDIMAN". The box is called the SYNCMAN Plus. I seldom need it in my setup
anymore, but it's a versatile box well worth the money. It generates time code
for striping to tape which I use on shoots for striping right to the linear
tracks of my Hi-8 cameras (I don't have PCM capability). It does this
regeneration of time code trick. And it generates Midi time code synchronized
to either the time code coming in to it or the time code it's generating.
In any case, my main point is that I would be kind of suspicious of the
PCM tracks on Hi-8 capability to record time code and then play it back for the
Sunrize board. And even if other things work all right off the recorded PCM TC,
the Sunrize might be more finicky. And there are ways around that problem as
well (Linear tracks, TC regenerators). Keep me updated. I hate to hear when
people have invested serious money into a setup and can't get it to fly right.
Terry Tabor
I have now read your second message. I experimented with recording levels and
Doug and Sunrize told me what the tolerances were for the Sunrize board. Those
experiments did not resolve the problem. I still think that my most recent
test with ver 2.08, which I posted, puts the blame directly on the 3.0 program.
I have never had problems with my PCM TC in an editing situation, at least none
that I can't write off to other factors.
But what is strange to me is that, are most of the people who have been reading
this thread using STudio16 3.0 only for animation in short clips of only a
couple of minutes? In my experience, after 10 minutes, the miscueing is
noticeable, and after 20 mins, drastic action has to be taken.
I will keep everyone posted on what Sunrize comes up with. They are currently
working on 3.1 close to a beta version.
Glenn
Glenn,
I am most curious in your question to the thread about how people are using the
Sunrize audio boards. My original interest in the board was much the same as
your application…
I wanted to do all my video edits and carry forward all of the sound to my
edited master. Then dump the entirety of the audio into the sunrize with the
Emaster's TC. Edit my soundtrack using the sunrizes four channels and then dump
back on to the Edited master in sync. This would be for programs up to 1 hour
in length. I am also suprised as you are that no one else seems to be doing
this type of thing.
BTW, glad to hear that TCGen doesn't seem to be the problem. Regarding that, I
think that once the toaster has been fired up and then exited, it still syncs
your Amiga with the incoming video on input 1. I'm not 100% sure of this, but
it seems like I remember reading that a long time ago somewhere.
Brian
You're right about the Toaster. Loading the Toaster sets up its genlock
function which stays operational after you quit Toaster. That's why you get
that overlay of Workbench on the Preview monitor until you shut the entire
system down. That is how you make a TC window dub tape which is the only real
way to test how your editing software is performing. Start the Toaster and
then shut it down. Then go into Sunrize and start the SMPTE Monitor module.
Click on the menu choice that turns off the header leaving only the TC window.
You can now lay this window over incoming video. Record a copy of this and you
have a tape that shows exactly what frame you are at anywhere in the video.
It's a good idea to have one of these around, because they are a good test for
any program that is trying to sync to video. You probably already know all of
this, but there might be some readers who don't.
I'll let you know if Sunrize finger the bug. I hope there is some easy fix.
Going back to the old version of the program is a PITA. Once you've seen
Paris, …
Glenn, Indeed it sounds from your experience that the problem is due to version
3.01 (3.??). I can't speak for anyone else on this thread, but as I mentioned
earlier, I do indeed work almost ex- clusively with shorter clips as most of my
work is for commercials (30s to 1m) and for music videos (4m to 6m). In fact
the only things that I have done longer than that have been "artsy" videos
similar to "Chronos" and "Koyanisqatsi" and in those cases, I composed and
recorded the music first, and edited the video to "sync" to the music. I put
the "sync" in quotes because there were no synchronized sound F/X involved,
just getting the video cuts close to the section/tempo changes in the music;
not very sync critical. The shorter cuts have had critical sync needs with
synchronized sound F/X, but again, max length has been about 6 minutes.
Hope Sunrize works out this problem though, as I am sure sooner or
later I will need perfect sync on longer video.
Keep us posted,
Terry Tabor
I think you probably picked up my long reply from Sunrize when you were posting
this one. I was using version 3.0. Version 3.01 is not available yet and is
just coming to a beta version. Sunrize said they would be offering this to
their 3.0 owners free of charge, BUT this is not official word in any way. I
am not doing match-frame editing either, but the error becomes so great, that
is not acceptable in any way, after about 10 minutes.
Glenn
Glen, Hi, I have also just had a similar problem with my 516 3.0, but I still
have a feeling that it was something that I screwed up. I would like to stay in
touch tough to find out how you resolve your problem, and just to have another
Studio 16 user to correspond with. By the way, I do use the SMPTE output
module. I don't get to this forum too often, please E-mail me any fun facts and
stuff. Thanks, Gary Kleiner, Mobile Pro Video, Fort Collins, Colorado.