Opcode Vision/Galaxy Stability
Jon Drukman
: I am interested in hearing about crashing/inconsistencies and general
: problems from users of Vision 2.0.3 and later.
: I'm still tearing my hair out nearly a year later!
I use Vision 2.0.8 extensively, for hours at a time, and it never even
burps. What problems are you having?
/jon
C J Silverio
| I am interested in hearing about crashing/inconsistencies and general
| problems from users of Vision 2.0.3 and later.
Try upgrading to 2.0.8. It seems more stable to me (not that I
had lots of problems to begin with). Galaxy still wigs out
at seemingly random intervals, though. I'm wondering when
Opcode will overhaul that dusty bit of software.
---
C J Silverio ceej@[ netcom.com | well.sf.ca.us | genmagic.com ]
Kelp Studios ke...@meer.net
<a href="http://www.spies.com/ceej/ceej.html">cj's home page</a>
Russell Pickett
>
>Try upgrading to 2.0.8. It seems more stable to me (not that I
>had lots of problems to begin with). Galaxy still wigs out
>at seemingly random intervals, though. I'm wondering when
>Opcode will overhaul that dusty bit of software.
Galaxy 2.0 will be shipping for Mac and Windows in the second quarter.
It'll support OMS 2.0, and contains various big fixes.
--
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
R Pickett <--- watch this space for exciting changes 06/95!!!
eme...@sfsu.edu Not responsible blahblahblah my own words blahblah,
eme...@opcode.com blahblah reflect the opinions blahblah get it blah?
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Russell Pickett
>
>In article <kelp-01039...@kelp.meer.net>, ke...@meer.net says...
>>
>>Try upgrading to 2.0.8. It seems more stable to me (not that I
>>had lots of problems to begin with). Galaxy still wigs out
>>at seemingly random intervals, though. I'm wondering when
>>Opcode will overhaul that dusty bit of software.
>
>Galaxy 2.0 will be shipping for Mac and Windows in the second quarter.
>It'll support OMS 2.0, and contains various big fixes.
Now, THERE'S your Freudian slip -- that was supposed to be "bug" fixes...
but I suppose some of them are big, too.
Tom Ritchford
Let me chime in with a litany of complaints!
Well, I've been using Vision since 1.0 and have had very many problems with
it, including version 2.0.8.
The last time I was in the studio, the program missed a punch-out, destroying
program material (this is on a tape deck through MMC, so there's no UNDO.)
We have two people working and a method where one checks the other before
recording... plus we could see that the punch-out point, mode, etc. were
all correctly set in the wake of the punchout.
This isn't the first time that this has happened.
I was able to make it Vision 2.0.3 (our previous version) crash 100% of the
time starting from a blank document in 6 operations! Plus it wasn't at all
reliable with MMC.
Plus there are many other reproducable bugs. Using instruments between two
systems with different OMS setups (but the same instrument names) just doesn't
work reliably (we're trying to work on two systems.) Drawing lines in the
tempo window can lead to unpredictable results, sometimes even crashes.
Starting the middle of a sequence with long sustained notes often will not
play the sustained notes (which has caused me no end of grief!)
Historically, we had the problem in 1.4 that if you created a file using the
New command and saved it, there was nothing there when you retrieved it.
(Opcode denied that this bug could even happen... they basically weren't
interested.) We lost work this way at least a half a dozen times.
Not to mention the functionality that they deleted claiming it wasn't used!
(The . key used to move to the next sequence in the current queue... I would
use that function all the time! Bye-bye since 1.4/)
I might add that I run with few extensions or control panels (basically
just QuicKeys and ATM...) and I don't have this problem with any other
program that I use, including Galaxy and Max.
--
/t
Tom Ritchford t...@mvision.com (212) 306-0414
Market Vision, 40 Rector Street, NY, NY 10006
Cerebrate April Fools Day!
CD release party for Verge's "Little Idiot" 4/1/95
atomic city
>Galaxy 2.0 will be shipping for Mac and Windows in the second quarter.
>It'll support OMS 2.0, and contains various big fixes.
Will it finally support the Roland D-70?
[to save bandwidth, insert several followup messages here from various
people inside Opcode and elsewhere, all basically saying, "Galaxy 1.2.5
DOES support the D-70 already...." to which I shake my head solemnly...]
I'm a big fan of Opcode's stuff, but if the behavior of Galaxy with a
D-70 is called "support", it's time to look at alternatives. I've
spent over an hour on the phone with various Opcode folks about this,
and after the second call, I had a costly hardware/software upgrade
done to my D-70, which still didn't make Galaxy work right with it.
The computer: A Macintosh Performa 460 or Powerbook 140 with standard
Opcode MIDI interface or Opcode Studio 3 or Opcode Studio 5LX, makes
no difference.
The synth: a Roland D-70 with ROM rev 1.11 or 1.18, makes no difference.
The software: Opcode Galaxy 1.2.2 or 1.2.5, makes no difference.
Using a bulk dump program and extended product mode on a hidden D-70
page, I can reload or dump the entire contents of the D-70's memory
in under a minute, bit-accurate and perfect every time.
Using Galaxy, a request to offload memory into a bundle appears to
work perfectly; you get a large file set on the Mac, and you can open
windows and look at the patch and tone names, etc. Try to put it back
into the D-70, and it wedges on you. The data transfer is at least
four times slower than that of the bulk dump, yet the D-70 acts as
if it's being fed too much data at once. The symptoms appear to be
that the Mac is feeding data to the D-70 in packets with long pauses
between them, but (a) not as long as the pauses between packets when
the D-70 is doing the sending and (b) too fast for the D-70 to handle,
despite the fact that at this pace it takes nearly five minutes for
a full dump.
Even Nick Rothwell, who's managed to kludge it into sort of working,
says that Rhythm maps are corrupted on transit every time. I find this
kind of frustrating; the sysex dump system works, but it requires
rebooting both the D-70 and the Mac every time. It seems that
something is mis-set in Galaxy, and I'm hoping that 2.0 addresses the
problem effectively somehow. I should note in all fairness that Galaxy
handles my other synths, including the notoriously-cranky Xpander,
with ease. It's a joy to use and organize my data with; which is why
I so desperately want it to handle my D-70, which is in far more need
than most synths of a sensible librarian.
--
metlay * atomic city * box 81175 pgh pa 15217-0675 * atomi...@netcom.com
---------------------------------------------------------------------------
Please buy my CDs. You would make me very happy, and wouldn't that be nice?
Nick Rothwell
> Even Nick Rothwell, who's managed to kludge it into sort of working,
> says that Rhythm maps are corrupted on transit every time.
I've not used Galaxy on my U-50 for a while, but as I recall the
problem is that the inter-packet gap used by Galaxy on transmit is
slightly too small, risking overruns on the U-50. The timing is
obviously marginal: it doesn't work for you at all; for me, it's
fine for everything except rhythm data. My editor/librarian (Anodyne)
works reliably for everything, but only after I increased the gap
timing.
How do I load up rhythm data? Answer: do a load via MIDI Manager
into a sequencer, then playback the sequence at a slower tempo into
the U-50. Time dilation does have its uses. But if I ever get back
to this, I'll just do the whole lot in MAX, which, time and again,
is proving to be the only reliable way of doing ANYTHING.
Aron Nelson
spent over an hour on the phone with various Opcode folks about this,
and after the second call, I had a costly hardware/software upgrade
done to my D-70, which still didn't make Galaxy work right with it.
Mike,
I've spent weeks on this with Roland.
You probably know that I wrote the D-70 editor/librarian. We didn't
charge for this editor because I told Opcode that there were many problems
with the unit.
You may not believe me but it does have problems. Ask Roland Tech support
who gave me the permission to include all of the disclaimers in the online
help.
Read it and weep.
<I had a costly hardware/software upgrade
done to my D-70, which still didn't make Galaxy work right with it.
Nowhere does it say to go and get an upgrade. In fact, the online helps
TELLS you EXACTLY what rom version we used in our testing. Roland USA got
a copy of editor and they used it as well. They were and are not going to
upgrade the ROMS so that's where we stand.
> The data transfer is at least
four times slower than that of the bulk dump, yet the D-70 acts as
if it's being fed too much data at once.
It probably is being fed too much data at once.
The librarian is probably not sending data back in the same format as the
bulk dump. There are many ways to send Roland data - handshaking,
non-handshaking as well as packetized in different ways.
> The symptoms appear to be
that the Mac is feeding data to the D-70 in packets with long pauses
between them, but (a) not as long as the pauses between packets when
the D-70 is doing the sending and (b) too fast for the D-70 to handle,
despite the fact that at this pace it takes nearly five minutes for
a full dump.
I can accept this. When I wrote the librarian I did use a slower
computer. It is quite possible that the send routines are too fast now.
This I can fix.
>Even Nick Rothwell, who's managed to kludge it into sort of working,
says that Rhythm maps are corrupted on transit every time.
Send me a bulk dump of a good Rhythm map and I will A/B against what I am
sending. If it's bad, I will fix it.
>It seems that
something is mis-set in Galaxy, and I'm hoping that 2.0 addresses the
problem effectively somehow.
2.0 will not address this if it is a bug in the librarian.
My offer stands. Mike, you know my email address (see below). If you
can't email me a rhythm dump - mail me your address and I will send a
self-address stamped envelope with a disk.
Aron Nelson
76004...@compuserve.com
--
"Cyboman did it" J. Irvin
"The more notes you play, the more you're getting ripped off!" D. Scoggin.
James Torck
In a previous posting, Tom Ritchford (t...@mvision.com) writes:
> In article <3itav6$h...@nntp.Stanford.EDU> wesley@casa (Jon Drukman) writes:
>
> Well, I've been using Vision since 1.0 and have had very many problems with
> it, including version 2.0.8.
(...more complaints)
I have been using Vision since 1.0 ..something, and am now using Studio
Vision Pro. The program rarely crashes, and I have never had the problems
that are listed in these various posts. It's one of the most stable
programs I have. When I was having timing errors in Studio Vision, I phoned
tech support, and I was sent a free update which fixed the problem the
following week.
Cheers,
James
Russell Pickett
>Let me chime in with a litany of complaints!
>
>Well, I've been using Vision since 1.0 and have had very many problems
>with it, including version 2.0.8.
>
>The last time I was in the studio, the program missed a punch-out,
>destroying program material (this is on a tape deck through MMC, so
>there's no UNDO.)
>
>We have two people working and a method where one checks the other
>before recording... plus we could see that the punch-out point, mode,
>etc. were all correctly set in the wake of the punchout.
>
>This isn't the first time that this has happened.
Did you ever take notes or try to reproduce the alleged bug? We get many
reports of errors here at Opcode, but without detailed analysis of the
alleged problem, we can't fix it. "It don't work right always" won't
point us to any point in the code to fix a bug...
Please, by all means, if you manage to make this repeatable, we'd be
happy to hear about it -- we're always listening to the users, as you ARE
the ones that decide the success of the products.
>I was able to make it Vision 2.0.3 (our previous version) crash 100% of
>the time starting from a blank document in 6 operations! Plus it wasn't
>at all reliable with MMC.
Does this still do this in 2.0.8? If not, you should be congratulating
us for fixing bugs you used to run into, not still worrying about it.
>Plus there are many other reproducable bugs. Using instruments between
>two systems with different OMS setups (but the same instrument names)
>just doesn't work reliably (we're trying to work on two systems.)
How so? What does or doesn't it do? Please send more detailed
information; otherwise, you can't reasonably expect us to find these
errors.
>Drawing lines in the tempo window can lead to unpredictable results,
>sometimes even crashes.
Again, what results? What kind of line? Using which Strip Chart tool?
Details, my man!
>Starting the middle of a sequence with long sustained notes often will
>not play the sustained notes (which has caused me no end of grief!)
Under what conditions does this happen?
>Historically, we had the problem in 1.4 that if you created a file using
>the New command and saved it, there was nothing there when you retrieved
>it. (Opcode denied that this bug could even happen... they basically
>weren't interested.) We lost work this way at least a half a dozen
>times.
Which is it -- Opcode wasn't interested, or the bug is fixed? Can't be
both....
>Not to mention the functionality that they deleted claiming it wasn't
>used! (The . key used to move to the next sequence in the current
>queue... I would use that function all the time! Bye-bye since 1.4/)
Sorry about that. If R&D decided that a feature was under- or non-used
that you found essential, that's not a happy thing. I would hope,
however, that the added functionality you've seen since 1.3 has made up
for it. Since you seem to be still using the product, I can only presume
that's true.
>I might add that I run with few extensions or control panels (basically
>just QuicKeys and ATM...) and I don't have this problem with any other
>program that I use, including Galaxy and Max.
Neither Galaxy nor Max (well, sort of Max) is a real-time clock-driven
program. Making a computer interpret and edit a realtime stream of
arbitrary data on the fly is not a first-year Computer Science lab
project.
I'm sorry you've had so much trouble with Vision; I'm glad you're still
a Vision user in spite of it. It's good to see that our products are
still preferred, even by those who are frustrated with them.
In all seriousness, if you find repeatable bugs, and can recipe them out
step-by-step, please mail me at eme...@opcode.com with all pertinant
information. I'm in Product Testing, so I can place bugs directly in the
examination loop.
Thanks in advance for your feedback.
Tr...@telalink.net
Dear Russell,
Do you know when the update to SV will be out (the one that gives
us the features we've been waiting for since the introduction of Logic
Audio?
And will there be an upgrade fee- and will it work reliably without a
bunch of bug fix versions (unlike 2.0 :)?
I need to know whether to bail on Opcode or wait and check out 'STUDIO
VISION '95!
Sincerely,
Troy
Aron Nelson
problem is that the inter-packet gap used by Galaxy on transmit is
slightly too small, risking overruns on the U-50. The timing is
obviously marginal: it doesn't work for you at all; for me,
Thanks Nick.
I think I'll simply put in a timing routine that samples the speed of the
CPU at runtime. Then the delays should be relatively close between
Macintosh models.
Aron
Tom Ritchford
>In article <TOM.95Ma...@tom.mvision.com>, t...@mvision.com says...
>>Let me chime in with a litany of complaints!
>[Russell Pickett is very polite....]
>I'm sorry you've had so much trouble with Vision; I'm glad you're still
>a Vision user in spite of it. It's good to see that our products are
>still preferred, even by those who are frustrated with them.
>
>In all seriousness, if you find repeatable bugs, and can recipe them out
>step-by-step, please mail me at eme...@opcode.com with all pertinant
>information. I'm in Product Testing, so I can place bugs directly in the
>examination loop.
>
>Thanks in advance for your feedback.
Thanks for the very nice letter. Yes, I do still use Vision for dozens
of hours a week... I was just grumpy since I had lost work the previous
night.
I'll get the stuff from my lab book and send to you... (my internet access
is at work ahd the music stuff is at home, so there may be a delay...)
--
/t
Tom Ritchford t...@mvision.com (212) 306-0414
Market Vision, 40 Rector Street, NY, NY 10006
Verge's "Little Idiot" -- coming soon to a player near you!
atomic city
>
>Mike,
>
>I've spent weeks on this with Roland.
>
>My offer stands. Mike, you know my email address (see below). If you
>can't email me a rhythm dump - mail me your address and I will send a
>self-address stamped envelope with a disk.
It's very good of you to offer, and typical of the sort of attention to
detail that I've come to expect from Opcode. I'm sorry that it got as
far as a post to rmms, but I am glad that you're willing to help me with
the problem, and will be as quick to post a notice here and on the D-70
group about any fixes you can manage as I was about the problems. I will
be in contact you via private email at my earliest opportunity, with
detailed symptoms and so forth. Many thanks.
mike
Russell Pickett
Tr...@telalink.net says...
>
>This message also posted to rec.music.makers.synth
>
>Dear Russell,
> Do you know when the update to SV will be out (the one that gives
>us the features we've been waiting for since the introduction of Logic
>Audio?
First, apologies to Troy and anyone else that might have been waiting on
the answer to this question -- I've been rather busy with the MS-OMS
deal, so I haven't had time to address this as promptly as I'd have
liked.
The simple answer is "no." Firstly, I don't know which features you're
looking for (I'm tragically uneducated as to the competition's feature
set), and secondly, I'm in the Windows end of things mostly, and so am
not in good touch with Studio Vision.
That said, I can talk a bit about Studio Vision 3.0, which we showed at
NAMM preliminarily.
SV 3.0 is going to feature a new, more modular architecture, that allows
for easier support of new hardwarem, in addition to a plug-in DSP
ability. At NAMM, we were showing our time compression, pitch shifting,
Audio -> MIDI, and MIDI -> Audio (!) algorithms being used. There will
also be a new "consoles" feature that replaces Vision's rather dated
Faders, with up to 64 completely customizable consoles available.
Back to the DSP. The Audio -> MIDI -> Audio bits are very exciting.
Imagine being able to take your singer's vocal track, the one where she
rushes the end and goes flat. Convert it to MIDI, and voila! You have
all of the data from that performance converted to notes, controllers,
and pitch bend. All of the nuances in pitch and brightness and such are
there in the MIDI.
Now, you want to correct it? Just extend the time of the last few notes
to smooth out the rushing tempo, and delete the extraneous pitch bend of
the flat notes. THEN, select MIDI -> Audio, and the corrected
information is re-applied to the audio, seamlessly. Basically, we are
offering a way to use MIDI, the protocol we know how to alter and
blenderize in a million ways, as a control construct for digital audio,
which is much harder to work with directly.
I'm virtually certain that it's also going to support the new OMS 2.0.
So, that's SV 3.0... more might show up later, as the feature freeze has
not yet happened.
>And will there be an upgrade fee- and will it work reliably without a
>bunch of bug fix versions (unlike 2.0 :)?
Probably an upgrade fee, as it will be a major number release. Also,
though, there will likely be interim bug and maintenance (and feature? I
don't know...) releases of SV 2.0, which will be under standard update
policy.
As far as it working on the first run... well, we ALL want that, now,
don't we? SV 2.0 did open our eyes to some of our QA problems, so I bet
you'll find future releases to be more solid, if a little later than
expected.
>I need to know whether to bail on Opcode or wait and check out 'STUDIO
>VISION '95!
Why would anyone bail on Opcode? (*mock horror)
You should definately check out the new run of products we have coming
out... many new features... Also remember, OMS is rapidly becoming the
de facto standard on Mac, and will be the explicit standard on Windows
after we finish our work with MS to integrate it into the operating
system...
Opcode is the future... be there. ;-)
--
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
R Pickett * This line under construction -- pardon our dust.*