________________________________________________________
WHEN REPLYING, PLEASE QUOTE ONLY WHAT YOU NEED. Thanks!
Change your preferences or unsubscribe here:
http://lists.figure53.com/listinfo.cgi/qlab-figure53.com
Follow Figure 53 on Twitter here: http://twitter.com/Figure53
On 2/26/11 at 5:42 AM, stephen....@gmail.com (Stephen
Pruitt) wrote:
>Second - How much RAM does the system have? You're going to
>need a minimum of 8GB, and things will probably work better
>with 16.
Because QLab v2 relies on QuickTime 7, it's stuck in 32-bit
land, which means it can only access the first 4 GB of RAM.
Having 8 GB rather than 4 or less will definitely make a
difference, because most of the OS X 10.5 processes are 64-bit
and can use the higher memory space, out of QLab's way. But
unless you're running other memory-intensive applications at the
same time, you may not actually see much of a difference between
8 GB and 16.
We're hoping this can change in the future, and that QLab v3 can
be 64-bit. No promises yet, though -- Apple has been letting
QuickTime X languish half-finished for a while now. But finding
a way to make it work is high priority.
Cheers,
Sean
It's not important for all screens to be the same resolution.
It's best if the media can match the resolution of the screen it will be displayed on. That minimizes the amount of work required, since no scaling is needed.
-C
Thanks for your help,
Jonathan
If you're connecting to the monitors with VGA or DVI, they'll likely represent themselves to the computer as 1366x768 - the 16:9 partner to 1024x768. If you connected with an HDMI cable, you'd probably be able to do at least 1080i, if the monitors support that. Oddly, for a 720p tv, 1366x768 is its native pixel depth most of the time. Often, 1080p tv's that are less than 50" or so actually have the same number of pixels, or so I gather. Remember, HD is a marketing term more than a standard.
Anyway, the point is, if your movies are scaled before they're played back, that's less computation in the moment of playback, and so likely better performance. I think the biggest issue you'll have with this one is hard disk access. Three movie files at the same time is a lot to ask. Add a hard drive or three, as Stephen suggested.
luckydave
luck...@figure53.com
i bet a whole dollar that it's either the hard disk or the graphics card, and lucky for us all there is a super easy way to figure it out: make several versions of your videos, some of which are highly compressed to have small file sizes, and others which are gently compressed to have large files sizes. if the large-file-size videos play back smoothly, then your problem was overtaxing the video cards. if your small-file-size videos play back smoothly, then your problem was overtaxing the hard disk.
i recently did a show with lots of video that required transparency and so all my cues were encoded with the animation codec which uses a lot of disk space, but apparently is fairly light on the GPU. i ran that show off a pre-unibody macbook pro with the cues on an external 7200 RPM hard disk connected via firewire 800. no problem.
another trick i've learned is to encode a bunch of versions of videos with progressively lower data rates... it's amazing how low the data rate can get for some videos before you see artifacts. the lower the data rate, the happier the hard drive.
cheerio
sam
Thanks for this info.
I'm doing my editing in Final Cut Studio. As far as compression goes, do you notice any difference if you use a codec like animation or photo jpeg in the sequence settings themselves, versus using a higher quality codec in the sequence, and then compressing the file via compressor or another program?
Thanks,
Jonathan
On Mar 6, 2011, at 3:02 PM, Jonathan Wald wrote:
>
> Any more ideas?
Did you try assigning each cue to only a single screen/projector?
-C
J.
One note on the following:
On Mar 6, 2011, at 3:02 PM, Jonathan Wald wrote:
>
> I wasn't able to upgrade the graphics cards unfortunately - but my understanding from a previous poster was that because shrinking the file size helped, it was likely not a graphics card issue.
I don't recall what the cards were, and they're probably fine, but FWIW I'd hesitate to say that any component of the rendering chain can be easily counted out as an issue unless it is explicitly tested.
I haven't ever worked with long cable runs using video, and I don't know why it would matter, but out of curiosity have you tried connecting the machine up to screens with "normal" length cabling to see if there was any difference?
Also, I forget, did you send us a copy of the workspace to look at? If not, try sending a copy to sup...@figure53.com and I'd like to take a look at it.
Best,
Chris
Er, duh. Sorry! You told me what they were in your last post. Anyway, point still stands.
it was i who suggested this logic, my thinking being that a version of the video with reduced file size with equal picture quality necessarily makes the graphics card work harder while making the hard disk work less. ergo, if that scenario results in smoother playback, it can be a good indicator that the hard disk was a bottleneck and the video card was not.
not proof, mind you.
i would spend good money on a video card performance monitor. sigh...
regarding the OP, this is seriously vexing to me because your description of your setup sounds quite reasonable to me and i feel like i've had success with very similar setups. i'll let you know if i think of anything...
cheerio
sam
> i would spend good money on a video card performance monitor. sigh...
http://www.atpurpose.com/atMonitor/
Includes some stats about the graphics cards, such as available vram, current FPS for screens, etc.
On Mar 6, 2011, at 5:13 PM, sam kusnetz wrote:
>
> regarding the OP, this is seriously vexing to me because your description of your setup sounds quite reasonable to me
My thought as well. :/
-C
> I am only running VGA in my cable path, so I would have thought I'd be okay.
Your lengths sound fine to me.
> [...] and I have a number of cues that I put on infinite loop, though never many at once.
>
> I thought about using the "reset" command instead of the "stop" command, but haven't tried that yet - the problem occurs immediately in the show run in any case, so I wouldn't have thought anything was being held over.....
Two things here:
First:
In my experience (well, I'm newish to QLab, but have been doing video for ages) Quicktime doesn't handle infinite video loops very well (there is always a bit of a lag at the loop point) and QLab seems to handle them even less well (very laggy at the loop point). In a recent production (last weekend in fact) I had a 10 second clip that loops seamlessly but never played back correctly at the loop point (no matter what player I used, QLab included). To fix it, I created a 15 minute clip (which more than covered the period the talent needed) in Final Cut which just contained the 10 second clip over and over again. That way I didn't need to rely on the looping feature and all worked well in QLab.
Second:
When you say "occurs immediately in the show run" can I just make sure that you are pre-loading the cues before hitting go? If your problem is right at the start then perhaps QLab is just busy collecting and buffering all the media. If you haven't already done so, you may want to put all these cues in a group so that the entire group is pre-loaded before you execute the go.
Regards,
Keith.
I found the same thing with the loops - and have been using the same solution, creating longer video files so that the break in the loop happens less often. And trying to balance that with not making the file size too long!
With pre-loading, I have noticed - sometimes, not consistently - that it does help avoid lag at the very start of a cue. However, this problem is bigger than that. This is a case of, I load cue one on the first monitor, which is a 7-minute infinitely looping clip of approximately 300mb - getting much smaller seriously compromised quality. Then I try and load (with or without pre-load) another much smaller infinitely looping clip on monitor 2, and sometimes it plays fine; but more often both of them start seriously freezing, or jumping, or both. And it's not just those two cues, or I'd think it was a problem with the files....
J.
>
> First:
> In my experience (well, I'm newish to QLab, but have been doing video for ages) Quicktime doesn't handle infinite video loops very well (there is always a bit of a lag at the loop point) and QLab seems to handle them even less well (very laggy at the loop point). In a recent production (last weekend in fact) I had a 10 second clip that loops seamlessly but never played back correctly at the loop point (no matter what player I used, QLab included). To fix it, I created a 15 minute clip (which more than covered the period the talent needed) in Final Cut which just contained the 10 second clip over and over again. That way I didn't need to rely on the looping feature and all worked well in QLab.
>
> Second:
> When you say "occurs immediately in the show run" can I just make sure that you are pre-loading the cues before hitting go? If your problem is right at the start then perhaps QLab is just busy collecting and buffering all the media. If you haven't already done so, you may want to put all these cues in a group so that the entire group is pre-loaded before you execute the go.
>
>
> Regards,
> Keith.________________________________________________________
Back in the fall I was running a show with 2 panasonic dw6300 projectors image blended and my qlab monitor. I was running pretty much raw video (approx 1gb of data per min of video) outputting at the projectors native resolution of 1280x800 (I believe).
I don't remember the exact system specs but it was about a dual core 2.8ghz processor, 6gb of ram, 2 hard drives (one for the system drive running apps, the other with all the show files)
I had 2 graphics cards, I believe they are the geforce cards, each card would have 1 projector and 1 display screen for the operator. So a setup similar to what you are running.
As pervious people have stated I think there could be an issue using the infinite loop. You might be better off creating your loops in final cut and then exporting.
Not sure if this helps but just letting you know what I've had success with in the past.
Cheers
Geoff Hollingshead
Head of Sound
Arts Club Theatre
Sent from my iPhone
>
------------------------------ IMPORTANT NOTICE ------------------------------
This email transmission and any accompanying attachments contain confidential
information intended only for the use of the individual or entity named above.
Any dissemination, distribution, copying or action taken in reliance on the
contents of this email by anyone other than the intended recipient is strictly
prohibited. If you have received this email in error please immediately delete
it and notify sender at the above email address.
------------------------------ IMPORTANT NOTICE ------------------------------
oh my god! atmonitor! where have you been all my life?
thank you, this is excellent.
regarding your issues with stops causing stutter in playing cues, do you mostly have several stops firing at once? if you haven't already, try spacing them out.
cheerio
sam
I do still have some media files on the drive running the apps - as I have three different screens to go to, and only three hard drives, I left some on the main drive so that the other drives never had to share. Could that be an issue?
I'll have a look at getting rid of the infinite loops entirely - a huge amount of work though, may not be able to do it for this show! Would be interesting to know if that's the issue....
Thanks,
Jonathan
On 07/03/2011, at 1:12 PM, Geoff Hollingshead wrote:
> I don't remember the exact system specs but it was about a dual core 2.8ghz processor, 6gb of ram, 2 hard drives (one for the system drive running apps, the other with all the show files)
>
> As pervious people have stated I think there could be an issue using the infinite loop. You might be better off creating your loops in final cut and then exporting.
>
> Not sure if this helps but just letting you know what I've had success with in the past.
>
> Cheers
>
> Geoff Hollingshead
> Head of Sound
> Arts Club Theatre
________________________________________________________
I do not know a reason that setting a video on an infinite loop would cause any issues. (Unless there is a memory leak in that circumstance that I do not know about -- which of course I definitely want to know about if people are seeing it.)
The loop point of a video is indeed currently susceptible to a delay, because QLab does not currently double-buffer video (i.e. have a second video ready and waiting at the start time). At the loop point it "rewinds" as quickly as possible to the start time and then proceeds to continue playing the file.
To my knowledge, this would not account for the kind of drops you are seeing.
Would love to see a copy of the workspace file, if possible. May not lead anywhere, but could possibly provide more clues.
Best,
Chris
What kind of copy? The file itself, or a screen cap?
J.
> Hey Chris,
>
> What kind of copy? The file itself, or a screen cap?
The QLab ".cues" file itself. You can send a copy to sup...@figure53.com. (Not the media files, just the QLab workspace file.)
Is it possible the culprit could be related to the OS or to whatever version of QuickTime is on the machine?
Mark V.
Sent from my iPod. Pardon the typos.
Stephen Pruitt
Fluxion Scenic and Light
To: "Discussion and support for QLab users." <ql...@lists.figure53.com>
Subject: Re: [QLab] Performance issues Redux
Message-ID: <856870D1-F256-4596...@yahoo.com>
Content-Type: text/plain; charset="us-ascii"
Hey thanks everyone for your continuing feedback.
I am only running VGA in my cable path, so I would have thought I'd be okay.
I don't know if the video lengths are relevant, but some are quite short; and I have a number of cues that I put on infinite loop, though never many at once.
I thought about using the "reset" command instead of the "stop" command, but haven't tried that yet - the problem occurs immediately in the show run in any case, so I wouldn't have thought anything was being held over.....
Jonathan