- Posts: 21
- Thank you received: 1
3 Post Moderation Rule for New Boarders (25 Jan 2021)
New boarders will have their posts moderated - Don't worry if you cannot see your post immediately.
'new mac pro's & fcpx
- MetalJesusRocks
-
- Offline
- Junior Member
-
I've confirmed this with another big YouTube gaming channel who also edits with FCPX. He sticks with FCP 7 and the performance is fine...and only uses X for green screen and multicam stuff.
Edit: Question - if there any benefit to using Compressor on the gameplay footage first? I haven't used Compressor much, so I don't know where even to begin.
Please Log in to join the conversation.
I can understand mixing of footage could slow things down, but all my footage is usually transcoded into 422 before edits.. sometimes on small projects I dont bother with doing that. Now, that being said, if FCPX can work with various footage, that various footage no matter the codec should work fine as well on the timeline, because FCP 'supports' it.
An argument can be made either way for that position, understandably..
Please Log in to join the conversation.
Please Log in to join the conversation.
What you are describing is pretty much what I am seeing also. It's not necessarily happening at the start of a new project but eventually some or all of those things will be happening. Restarting does not always help either.
I noticed a difference in timeline performance when I updated my laptop from 10.0.9 to 10.1.4. The interface definitely "slowed down".
I have been through all of the troubleshooting and maintenance that others have mentioned and it didn't make a difference. I think it's curious that others not seeing this. If the next update does not fix the issue I am going to completely strip everything out of my system and start over from scratch again. If that doesn't fix it I will be talking to Apple.
Regardless of the issue though, I can still get things done faster and (IMHO) better in FCP X than other systems I've worked on. If they can fix this I'll be very happy.
Please Log in to join the conversation.
Doing this rules out any 3rd party stuff, and other OS apps (that may or may not be running while FCPX is open)
So now I am left with, 'is a Mac pro 5,1 with gtx 680 (tried both Nvidia and apple drivers) Raid SSD, 32gigs RAM & a 3.33 Ghz CPU not compatible with FCPX?" By compatible I mean a fluid working application and an app where I am not using 25% of my time dealing with application issues.
Please Log in to join the conversation.
Don't let us down Apple! Looking forward to the next update!
Please Log in to join the conversation.
You saved me a few grand I can use elsewhere..
I will put off upgrading to the cylinder Mac Pro trash can until FCPX actually is fixed and works correctly..
thanks again
Kurt
Please Log in to join the conversation.
Redifer wrote: ....FCP7 is always 100% instantaneously responsive. It always has been. FCPX NEEDS to be equally responsive in the timeline at all times. FCPX must currently have very poor or unoptimized code or something. They need to fix it pronto....
Unfortunately I agree with this. Even on small projects I periodically see inexplicable slowdowns where the UI becomes sluggish and laggy. Restarting FCP X temporarily fixes it, which indicates it's a problem within the process: memory fragmentation, memory leak, handle/object leak, etc.
This doesn't mean the problem is within the FCP X code itself -- it could be certain plugins or codecs. But for the customer it's still a problem.
In software design there's always a debate about whether plugins should be handled "in process" or "out of process". Loading them into the address space of the main process gives better performance but any misbehaving 3rd party code can derange the entire process. Loading them out of process is more resilient to failure but requires inter-process communication which is slower.
This issue was recently played out in tabbed browser design where initially most browsers ran each tab in process. They found a single malfunctioning tab would take down the entire browser, so most now use process-per-tab design.
The architect of the core product must use design features which protect stability and integrity, not just assume all parties write perfect code. The steps may include special validation suites, out-of-process design for add-ons, instrumented code which can be activated at the customer site to rapidly gather more information, or using live remote debugging. These items should have preeminent position in development priority, else the reputation of the product becomes so tarnished that no new feature will save it.
As customers our best step is communicate with each other, share notes and try to isolate the procedural, codec and plugin scenario causing the problem.
Please Log in to join the conversation.
This will cause long delays if 'all' is selected when switching inside the effects browser (switching between filters, transitions, titles, etc..)
So this happens when the plugin dont need to be actually used in the timeline.
If it is used, maybe other delays?
If the plugins could alias somehow, that may fix that issue.
But with my tests I mentioned at the start of the post, No plugins were installed.
Clean OS install with no other apps other than a new install of FCPX
Please Log in to join the conversation.
One good thing out of this test is that FCPX did not take that bad of a hit when tested with a million plugins and a million other apps on the OS... vs a clean OS and no plugins.
performance within FCPX was better on a clean OS and no plugins, but not that much.
This is a benefit of mac OS's as compared to a window OS, the apps (for the most part) remain within there own bundle and there is not too many 'dll's' (from my windows days) scattered throughout the OS
Kurt
Please Log in to join the conversation.
dkag7 wrote: ...with my tests I mentioned at the start of the post, No plugins were installed. Clean OS install with no other apps other than a new install of FCPX
That is good from the standpoint of a simplified, hopefully reproducible procedure. A portable, consistent replication scenario greatly facilitates getting bugs fixed. Even if the reproducibility is only 1 in 5 times, provided the overall procedure is straightforward and doesn't require a huge amount of data, that is still good.
From a software engineering standpoint, even a semi-reliable replication scenario is vastly better than trying to deduce a failure from looking at a crash dump, custom instrumentation code or live remote debugging.
Please Log in to join the conversation.
but the items on the timeline.
you always said you have no problems
please upload a video when fcpx's timeline has over 3000 items on it
(about 1000 video items and 2000 audio items)
then zoom in an out the timeline, and edit something on it, then playback
you will find the performance is very bad, it's very lag
but fcp7 is not
or you can upload the video (with a clock on the screen plz )to show it's not
Please Log in to join the conversation.
we've got 10 varicam35, and avid is going to support avc ultra at once
hahahaha
what about avid running on the ultimate mac pro cylinder?
Please Log in to join the conversation.
you will still feel very lag when controlling the over 1500 items timeline!
(and maybe over 900 audio items attach on the video!)
and it will always read the waveform whenever you have read it or not if you click the audio in the inspector!
Please do not read the waveform when I don't need it!
If fcpx can resolve these 2 problems, it will be faster enough!
Please Log in to join the conversation.




