Render times different for FLV?

I've set up a virtual cluster for Compressor and witnessed the astounding speed increase using Activity monitor. I'm wondering though- the render jobs were using an .mov being rendered to .mp4 (and other formats) and I was able to see all 8-cores being utilized. When I try to render to flv (using Quicktime Export Components) performance drops to ONE core, and takes much longer.....is flv just processed differently in Compressor or does Qmaster not process all jobs the same?

Only certain codecs can be segmented, h264 can, flv can not.

Similar Messages

Render times differing for same composition

I'm trying to render a composition and the first time I tried it said 7 hours. So I shut the Mac down, went home and set it going over night. But it only took 40 mins to create a Lossless render. Great, I thought.
So today I'm trying to export the same Comp and it's saying 4-5 hours! I've cleaned the Cache, enabled multiprocessing etc and it's still saying 4-5 hours.
Any idea why that would be?
All the settings are the same. Output codec, output location etc.
Stumped.

Welcome to the wonderful weird wacky world of After Effects renders. Do you have multiprocessing turned on? How much RAM do you have? How many cores? What's your source footage? What resolution & codec are you rendering to?
I've run into the same situation. Even though I have gobs of RAM (32GB), sometimes rendering with fewer cores and less assigned memory results in faster rendering. I'm sure Todd will be along to provide the link for tuning your RAM/Core settings (I don't have it offhand). I find usually my Mac is fastest after a fresh reboot and not running anything else.
So you are saying you got 40 minutes with multiprocessing turned *off*?
40 minutes vs 7 hours seems a little extreme. You sure it completed rendering? You did it at full rez at full frame rate?

Render time. Is 11 hours render time reasonable for a 90 minute file?

I started with a 120 minute AVI file, trimmed the last 30 minutes off and am rendering it to an identical format AVI file. Is 11 hours render time reasonable? Have I set something improperly?

More information needed for someone to help... please click below and provide the requested information
-Premiere Pro Video Editing Information FAQ http://forums.adobe.com/message/4200840
AVI is a wrapper, what is inside YOUR wrapper?
Exactly what is INSIDE the video you are editing?
Codec & Format information, with 2 links inside for you to read http://forums.adobe.com/thread/1270588
Report back with the codec details of your file, use the programs below... A screen shot works well to SHOW people what you are doing
http://forums.adobe.com/thread/592070?tstart=30 for screen shot instructions
Free programs to get file information for PC/Mac http://mediaarea.net/en/MediaInfo/Download
What effects have you applied to the video?

Post-Render time waste for H264 & ProRes Export on Windows?

There is one thing that bothers me all the time.
If i render to some 1 Gbit/s windows network location, everything is fine until multiplexing starts, and it goes like this:
1. Premiere CC creates .m4v file for video and .aac file for audio - thats OK.
2. Then multiplexing starts and some temp file like "s3r4." is created - no problem.
3. After that Premiere creates desired ".mp4" file with suffix ".mp4._00_" and copies "s3r4." to that file
4. Than it renames ".mp4._00_" to ".mp4" and render is finished
The thing i don't get here is why Premiere have to copy already multiplexed file "s3r4." to "_00_" file
and than rename it back to mp4? This procedure takes a lot of extra time if you have some huge file size (4GB? 8GB?)
The other thing: i have also installed Drastic MediaReactor for Direct ProRes export from Premiere CC on Windows.
Files are huge, no multiplexing in this case, but still after render Premiere starts
it's unneeded copy process of ".mov" file to ".mov._00_" and than back to ".mov".
I've tried to "end task" and just use the uncopied ".mov" which is available immediately
after render and it works fine and has no problems at all. I have also contaceted
Drastic Tech on this case and they said they have nothing to do with this copy/pasting.
That's why i don't know why Premiere has to copy/paste post-render files to the same location after render.
In case of ProRes they can reach ~20GB and you can imagine how long does it takes to copy/paste
such file especially to a network location.
Why instead of "copy" Premiere can't just use a "rename"? That could be blazing fast!

Try exporting to a local hard drive instead of one on a network.

Ending Frame of the StrobeMediaPlayer - differs for FLV and MP4

I am using the following HTML code to stream a FLV video from our Flash Media Server application, which is Flash Media Server 4.0- Interactive Edition. This code is taken directly from the videoplayer.html page that is delivered with the application.
<object width='640' height='377' id='StrobeMediaPlayback' name='StrobeMediaPlayback' type='application/x-shockwave-flash' classid='clsid:d27cdb6e-ae6d-11cf-96b8-444553540000' ><param name='movie' value='swfs/StrobeMediaPlayback.swf' /> <param name='quality' value='high' /> <param name='bgcolor' value='#000000' /> <param name='allowfullscreen' value='true' /> <param name='flashvars' value= '&src=rtmp://10.16.10.162/vod/timelapse/lakepoint1&autoHideControlBar=true&streamType=vod &autoPlay=true'/><embed src='swfs/StrobeMediaPlayback.swf' width='640' height='377' id='StrobeMediaPlayback' quality='high' bgcolor='#000000' name='StrobeMediaPlayback' allowfullscreen='true' pluginspage='http://www.adobe.com/go/getflashplayer' flashvars='&src=rtmp://10.16.10.162/vod/timelapse/lakepoint1&autoHideControlBar=true&stre amType=vod&autoPlay=true' type='application/x-shockwave-flash'> </embed></object>
When the FLV video completes, the player shows the image of the start of the video, instead of stopping on the last frame of the video. When using this same code to play a MP4 video, the player stays on the last frame of the video. Has anyone else noticed this inconsistency, or found a way to force the player to stay on the ending frame of the FLV video?
Thanks!

This video played very good from here (DSL). But i noticed
that it starts playing almost instantly when you click. This
probably means that you don't wait and buffer a part of your video
before playing it. With a slow internet connection, the playhead
will catch up with what is buffered and then the video lags. Maybe
just increase the bufferTime parameter of your FLV player. If you
put a value like 10 seconds it could help, but of course you will
have to wait those 10 seconds at the beginning. The best way is to
use a listener that will start playing your video only when say 40%
of your video is buffered.

EX cam footage render time?

I have EX1 footage captured as EDcamEX.
I placed it in a pro res time as recommended for faster renders.
The footage was an event- low light. The gain was high.
I applied a BCC Denoise filter.
The filter cleaned it up nice but,
The timeline is about 90 minutes and The render out is taking 11 hours (?)
Is this render time correct for this type of filter on this length timeline?

Hi -
Couple of things:
1) When dealing with XDCam and Ex footage, I routinely get very long estimates for renders that rapidly drop as the render progresses.
2) I am not sure what advantage you got by editing the Sequence with the Codec set differently from the source. I routinely edit Ex footage on sequences that have the same codec (XDCAM EX 1080p30) and the renders, while not zippy, are not outrageously long. If you think about it, now FCP has to convert all the footage used in the sequence to Pro Res as well as rendering any effects or transitions. This is probably quite CPU intensive.
Don't know if you want to consider this, but you might want to create a new sequence that has the same codec and frame size and frame rate as you source material, and copy all the clips from your ProRes Sequence and paste them to the new sequence and try a render there.
Hope this helps.

Average Render Times for HDV? 1 1/2 minute takes 30 minutes

Hello everyone.
I was curious if anyone has heavy experience with HDV render times or have a good resource to get average render times for HDV so I can make sure my times are close to everyone elses. I know HDV has a different GOP structure which makes render times longer, but it just seems to take for ever.
I have a Dual 3GHz Qaud-Core MacPro with 5GB RAM using Compressor 3.
1 1/2 minutes from FCP 6:
Preset 90min high quality DVD MPG 2- takes 30 minutes
Preset M4V for iPod/iPhone - takes 40 minutes
Any knowledge or resources would be much appreciated.
Thanks,
Deuce

To be honest, neither of your planned "upgrades" would improve performance all that much because you have too few drives and you will still have too little RAM. Premiere Pro CS5 still needs more than 12GB of RAM just to perform at its best. In addition, two drives (in your planned upgrade case, one SSD and one HDD) barely meet Adobe's stated minimum system requirements; Adobe strongly recommends more than three separate drives. (In your case, you'd need at least four completely matched hard drives in two (R)AID 0 arrays, with each matched pair in one array, in addition to the SSD.) And current AMD x4 CPU-based systems perform at least three to four times slower than a fast Intel-based system because the AMD CPUs do not fully support the SSE 4.x extensions that Adobe makes heavy use of.
As currently configured, you have a system that's at least 8 to 10 times slower than a fast, properly tuned Intel-based system. And even after you perform the so-called "AMD upgrade", your upgraded system will still be six to eight times slower than a fast mid-priced Intel-based system. And all that is due to the fact that you'd still have only two drives and too little RAM (IIRC you'd still have only 4GB of RAM and only a single hard drive for everything except the OS and programs even after the AMD-based "upgrade").

Much longer render time for a few small changes?

I bought a new Macbook pro last week with a 2.4 ghz intel i5 processor and 8 GB of ram. I got Adobe premiere pro yesterday and I made a video which combined taking footage shot from my flip video camera and audio commentary that I recorded across the full video. The video was rendered with the setting H.264 and the preset Youtube widescreen HD, the final video was about an 800 MB .mp4 file. It took about 30-40 minutes to render.
Today, I added a 17 minute mp3 file for background music, which was 17 mb. I also added 6 images and 2 titles which overlay the video at different points. The only other thing I added along with these two was a 16 second mov file which serves as an intro video to my series. With these 3 (what I consider to be fairly small) changes, my render time has jumped dramatically. With the same export settings, I see the render time build and build until it reaches 4+ hours and I decide to stop. I've tried a couple different formats, but I'm not sure what else to do. I feel that 8gb should be plenty to render a video like this fairly quickly (like I did yesterday) and I'm not sure why it's taking so long compared to yesterday. Is this normal? Any help to guide me with reducing render times would be much appreciated. Thanks!

The Export time will differ by what one is Exporting from and what they are Exporting to. In this case, the "to" has not changed.
What has changed?
Well, the stills for one. What is the format and what are the pixel x pixel dimensions of those?
The addition of the MOV is another place to look. What is the CODEC inside that MOV "wrapper?" Having to decode via the CODEC can impact the Export time.
The new Audio has changed. What is the format/CODEC of that music file?
Those are the first thoughts that I have on this, and good luck,
Hunt

Consolidating FX generators for better render time?

"There were no places results found for render time"
Wow. OK.
I have used a generator to create a collection of photos-- I used the same genertaor affect about 6 times in a row in end credits
And it seems this has created a render time of an additional 4 hours.
Overnight render time wasnormal-- BEFORE i added seveal of these photo generators in 40 min documentary.
Can anyone advise as to whether consolidating those generators will save on render time?
Also, it would be nice to not have to wait a day each time I want to se a rendered version of my work, in case anyone has created a list of pointers beyond the obvious stuff. I am on a macpro w 16 gigs RAM.
I have many event folders (15 ish). If I consolidate all of these (as advised on youtube tutorials), will their be a big difference in render time?
Event folders are spread out on 3 different drives. Is thgis truely adding to render times... or not? I do not want to risk loss of data or a technical nightmare after 6 mos of work.
I do not want to make any changes that jeoporodize my project. Havn't upgraded to 10.1 yet, as every slight change seems to bring with it more troubleshooting and tutorials to review.
In some disucssions, I am warned about the additional procesing power required by consolidated files. But if my render time is suddenly double after adding therse generators, I have to wonder if consolidating them would be the answer. I do NOT have another 24 hours to get simple answers to questions via test renders. Sometimes after an overnight render, I get a render failure error message, etc.
Thanks.
I am using Event Manager x

I am aware I am pushing this to the edge of it's dsign abilities. I get spinning ball for a few seconds when I do just about anything. Even ther simplest thing. Like chnaging location in timeline. I am on 10.7.5-- partly due to reason relating to my Avid Pro Tools HD recording setup.
The generator that seesm to be a processing hog is Prowall plugin by Pixel Films FX.
and I'll show you the way the plugins look (orange)-- about 6 generators back to back.Each holding about 8 pics. I found that even with consolidating, it still is a very slow render--- I attempted render only the 1 minute portion where the 6 generators happen (after consdolidating!) and after an hr it showed only 8% progress. So that little test would cost me a full work day. Argghhh!
Thanks for any advice you can offer!

Hi. I am using a time capsule for few PC s. I have made 5 different account to access time capsule. but in windows when i enter account name and password for one account, i cannot access other accounts, because windows saves username

Hi. I am using a time capsule for few PC s. I have made 5 different account to access time capsule. but in windows when I enter account name and password for one account, i cannot access other accounts, because windows saves username. how can i prevent this from happenning. I really need to access all my accounts and dont want it to save automaticlly.

Why have 5 accounts if you need to access all of them.. just have one account?
Sorry I cannot follow why you would even use the PC to control the Time Capsule. Apple have not kept the Windows version of the utility up to date.. so they keep making it harder and harder to run windows with apple routers.

Help with configuring Qmaster for optimal render times.

I have a 6 machine Qmaster cluster. All machines in the cluster are PowerPC based and I have set up the cluster manually through Qadministrator. I personally like to have everything set up manually so no changes can be made for any reason to this cluster. I have processed a 2.5GB video in the cluster and in an Intel based Mac Pro.
Using Compressor, my output for the file is H.264 for iPod video and iPhone 640x480. The source was placed on an external Firewire 800 drive and destination is to the local disk on the Intel machine.
For the cluster, the source, also from the external Firewire 800 drive, is connected to the cluster controller, which has full read write access to all nodes in the cluster and is not a node in the cluster itself, just set to be the cluster controller. ( I saw this setup on a Qmaster setup page here, http://www.eventdv.net/Articles/News/Feature/Distributed-Network-Encoding-with-A pple-Compressor-37923.htm ) Also, I am using the same output settings, H.264 for iPod video and iPhone 640x480, and output is being written to the cluster controller local disk.
Once the batch is completed, it takes 23 minutes to render the video in the cluster as opposed to 25 minutes on the Intel machine. Not the difference in render time that I hoped for.
Here are the specs for the machines,
Cluster nodes:
1.6 Ghz Powerpc g5 1.25 GB
1.8 Ghz PowerPC G5 4GB
Quad 2.5 Ghz PowerPC G5 2.5 GB RAM
Dual 2.0 Ghz PowerPC G5 2.5 GB RAM
Dual 2.0 Ghz PowerPC G5 6 GB RAM
Dual 2.7 Ghz PowerPC G5 2.5 GB RAM
Cluster Controller:
Dual 2.0 Ghz PowerPC G5 2.5 GB RAM
Intel Machine:
2 x 2.26 quad core Intel Xeon 6GB RAM
Now I know these machines in the cluster are old but all machines, except for the cluster controller, have gotten fresh installs of Leopard 10.5.8, all relevant updates applied and fresh installs of FCP6, Qmaster and Compressor. I still cannot figure out how to speed up rendering. Is there something in this setup that I am doing wrong or is it the limitations of the old hardware. In theory, it should break up this video into segments across all machines in the cluster but it doesn't. The chunk of segments are sent to the Quad G5 and yes I have enabled multiple cores in the Qmaster prefs pane. So I figured i'd only enable 2 instances on the quad core machine and then it started to send segments to other machines but render time is still slow.
Also, even though all machines had read/write access to each other, I would scroll through the log files and see that it could not read the segment from the source media, but I could clearly see the activity light flickering.
After all of this trial and error, I still cant get the render time under 23 mins for this video.
If there's anyone out there with some advice, it would be much appreciated. Also, I forgot to save a copy of the log file, so my fault on that one but I will post once I get a chance to get back to working on that cluster.
Thanks.

Update:
So I have been able to get the cluster working. I've used Compressor and have the source media located in the shared folder that the controller mounts as cluster storage on all of the machines. The problem that I can't figure out now is why the cluster won't read the source media from any other location. I've read through Qmaster documentation and the permissions are supposed to be bypassed when you are using Qmaster.
I have both the mac HD and clusterstorage drives mounted for each machine, on each machine. Still, the cluster won't transcode the file unless I drop the media into the shared folder first. Every sharing option I could think of has been enabled. I have enabled file sharing, appletalk, set read & write permissions for each drive in each machine for all users and It still won't work any other way.
It just seems impractical to have to first copy the source media to the shared folder mounted by the controller since all locations on each machine have been totally opened up.
If anyone has some advice about this issue, it would be greatly appreciated.

Presets For .flv Export Different in Pro CS5 vs. Elements 4

Hello:
I have been using the simple Premier Elements 4 to read in Flip Video 640X480 movies.
When I export them, I can choose a Flash Video preset of NTSC Flash Video 7 (320 X 240).
But, in Premier Pro CS5, that's not an option. The closest I can come is FLV Flash Video 8 And Higher, with different dimensions: 368 X 272. Unfortunately, the 368 X 272 will not fit in the space on my web page. It needs to be 320 X 240.
Does anyone know why they changed the dimensions? I can change the preset to a lower pixel dimension, but what problems will that create?
Regards,
Rich Locus
Logicwurks, LLC

One change that may be related is that we didn't include the Sorenson Spark codec with Premiere Pro CS5; we just included the On2 VP6 codec for FLV files.
The On2 VP6 codec was introduced with Flash Player 8, so (sometimes) "Flash Player 8" is a different way of indicating On2 VP6, and "Flash Player 7" is a way of indicating Sorenson Spark in preset names.
This shouldn't present any practical problems for you on export, but it may explain the difference that you're seeing. I can't be 100% sure, since I don't have Premiere Elements installed on this laptop, so I can't check what its presets contain.

Render time taking 1 hour for 4 seconds

I am trying to render a 4 second clip of 3d text doing a 360° turn, I export in H.264 and have okay computer specs.
3.04 GHz Quad core
12 gigs of ram
I'm new to this and want to start exporting more video for a project, but I can't do it with this render time.
Brandon

LilTunchiB8 wrote:
What would be a conventional, frame-based codec? Thank you so much for the help!
Uncompressed AVI is one, but the file sizes are going to surprise you. Quicktime with the PNG codec is usually smaller, but still huge. Quicktime with the PhotoJPEG codec is even smaller. Some popular choices that are even smaller are DNxHD and Cineform (either MOV or AVI). They're still going to be much larger than the H.264, but that's okay!
For many (most) professionals, the usual workflow is to render a lossless file (or nearly lossless) out of AE and then use the Adobe Media Encoder or Premiere to create the final deliverable. There are several reasons, but one is that AE is not very good at rendering h.264 files (and there are some very technical reasons for this). Anyway, the AE team deprecated the h.264 rendering in the latest version of After Effects because of it.
In your case, I would suggest rendering into a TIFF sequence (or something similar). This way, if there is a crash or something, you can pick your render up at the last frame it rendered rather than having to start at the beginning. I always do this when I'm doing 3d work and the render is going to be over an hour. Then you can bring that TIFF sequence into Premiere as if it's a video file, put your audio and everything to it, and send your final H.264 render from Premiere.

Is 12hrs a normal estimated render time for a 10min video?

I'm trying to make visuals for my music, which consists of 4 pictures, 4 adjustment layers (each with 1 evolution keyframe from start to finish on each), & 2 audio spectrums. Being brand new to AE, iss 12 hrs a normal render time upon exporting?
I'm running on an i7-4820k, Windows 7/64bit, 32GB of RAM, & a Geforce GTX 760.
Any answers/questions for me would be great!!

Ah, your lack of experience won't help, I'm afraid. You cannot intuitively decide how to optimize rendering without lots of experience. AE just doesn't work like Photoshop or FCP or Illustrator. You could, hover, transcode your mp3 to aan uncompressed codec and save some processing. You could copy your Audio and Audio Spectrum layers and put them in their own comp and render that out as prores422. Bring the rendered movie back in and replace all those layers with the movie. Adjustment layers are sexy but misleading. You probably are using them incorrectly, most beginners do. Fractal noise can be a rendering slug. Render that as a movie if you can.
Prerendering is basically just looking at your comp and deciding what effects/layers are not dependent on others. Precomposing is a process of grouping layers together to form a desired output that can be fed back into the main comp for a specific effect or filter. The classic example is drop shadow on a rotating object. If you just apply the drop shadow to the rotation object, the shadow is applied before the object is moved and the shadow moves with the object as it rotates, destroying the illusion. If you precompose the rotating object and apply the shadow to the precomp, the movement is rendered before the shadow is applied. It's complicated but you get the hang of it after a few epic rants.
Try to have fun. AE is unbelievably cool.

Same Computers/Settings - Different Render Times

Hello, great and powerful Adobe Community. My coworker and I have identical machines (Dell Precision T7600 - Xeon 2GB, 32 GB RAM, NVIDIA Quadro K5000, Windows 64 bit OS), both running Adobe CS6. Our render set-up for After Effects is to pull source footage from the main HD (2 TB), render to a second HD (150GB), with a third HD (150GB) set up for cache, media cache, and database folders. Both machines are set up to render multiple frames simultaneously. However, for some reason his machine is taking ridiculously long to render. For the same file, my machine finishes rendering in about 2hrs, while his is showing an estimated time of 17 hours. When left to run overnight, his machine still hasn't finished the next morning and shows 60+ hrs remaining. This is a recent problem and his computer has had reasonable render times up until recently. Software is up to date across the board, no new software has been installed on either machine lately, and any updates required by our system admin would have been applied to both.
Other than checking that the above listed conditiions are identical between our two towers, I'm not sure where to look for the problem. Can anyone suggest a reason for his machine's poor performance? Any insight would be appreciated. My thanks in advance for your help.

I wouldn't call it a bug.
Besides, although the effects you use in all these comps might all be precisely the same, you didn't breathe a word whether the codec of the footage used in these comps is precisely the same... and codecs can make a big difference.
Sure, you may have rendered to the same codec. Did you START with source footage that's all the same codec? If not, you're not comparing apples to apples.

Render times different for FLV?

Similar Messages

Maybe you are looking for