Activity › Forums › Creative Community Conversations › Background processes are an impediment to starting the edit
-
Background processes are an impediment to starting the edit
Bill Davis replied 8 years, 10 months ago 13 Members · 76 Replies
-
Michael Gissing
September 24, 2017 at 5:52 amWhy would drawing waveforms slow an NLE? It doesn’t on DAWs. 25 years ago the software people writing DAW software realised the bleeding obvious. Accurate waveforms are really important so draw them once but only as a background task. Fairlight for example draws waveforms blindingly fast but always in the background. Drop a polywav into the timeline and play it. Waveforms pop up and quickly but never affecting real time playback.
Reaper is another example of how to draw waveforms quickly without affecting playback and edit performance. Any NLE that can’t prioritise performance over waveforms is primitive compared to what has been normal in the DAW world for a couple of decades. And drawing a zero waveform cannot take longer or slow a system unless it is a bug.
-
Oliver Peters
September 24, 2017 at 12:57 pm[Bill Davis] “To me, that means a 100% negative post ABOUT X, whether that one post is factual or imaginary – it should be met with one that’s 100% positive for balance.”
Isn’t it more helpful to discuss the actual technical issues involved, rather than merely to throw out random positive, but unrelated comments?
– Oliver
Oliver Peters – oliverpeters.com
-
Oliver Peters
September 24, 2017 at 1:00 pm[Michael Gissing] “Why would drawing waveforms slow an NLE? It doesn’t on DAWs.”
That’s a very good question. Media Composer used to have huge issues with drawn waveforms (not part of the default timeline view). But that was fixed a couple of years ago as the code was rewritten. However, still not like a DAW.
– Oliver
Oliver Peters – oliverpeters.com
-
Jeremy Garchow
September 24, 2017 at 2:08 pm[Oliver Peters] “Sounds like some sort of RAM leak or something. Does it vary with model of Mac – for example, Mac Pro versus iMac?
“Every time this comes up, someone blames the hardware.
Unless all the macs are total pieces of shite, it’s not the hardware. Yes, this happens regardless of vintage or processor type, GPU, or RAM.
-
Jeremy Garchow
September 24, 2017 at 2:15 pm[Michael Gissing] “Why would drawing waveforms slow an NLE”
I don’t know. Whatever FCPX is doing, some sort of caching, slows it down.
When you reduce the amount of waveforms on the interface, it speeds up. Closing the inspector (which has to load and report a decent amount of data/information every time the play head stops) helps.
The weird thing is , FCPX gets confused on empty audio tracks in files. There is literally no waveform, it should be blank. But for whatever reason, FCPX seems to want to not want to play those back easily. I don’t know why that would be. It has literally been a problem since the release of FCPX.
-
Oliver Peters
September 24, 2017 at 3:07 pm[Jeremy Garchow] “Every time this comes up, someone blames the hardware. “
I’m not exactly sure of what causes RAM leaks in general, but my suggestion was rather that this was software-related.
– Oliver
Oliver Peters – oliverpeters.com
-
Andrew Kimery
September 24, 2017 at 3:34 pm[Oliver Peters] “[Jeremy Garchow] “Every time this comes up, someone blames the hardware. ”
I’m not exactly sure of what causes RAM leaks in general, but my suggestion was rather that this was software-related.”
Yeah, RAM leaks are caused by software mismanaging memory allocations, not bad hardware.
-
Michael Gissing
September 24, 2017 at 9:46 pm[Jeremy Garchow] “The weird thing is , FCPX gets confused on empty audio tracks in files. There is literally no waveform, it should be blank. ”
Empty tracks are still carrying PCM data and it will have a waveform. It will be a straight line but silence does still contain samples. However generating a straight line from this data should be blindingly fast and no different from a data stream containing varying sample data. Carrying a bug like this for six years is not good for a task that is fairly fundamental.
-
Jeremy Garchow
September 24, 2017 at 11:36 pm[Michael Gissing] “Empty tracks are still carrying PCM data and it will have a waveform. It will be a straight line but silence does still contain samples”
Yes, it’s still a file with information, and FCPX has ‘rectified’ waveforms so it looks blank rather than a straight line in the middle of the clip.
Still, why would FCPX playback an 8 channel audio file with 4 active channels and 4 empty channels worse than an 8 active channel audio file? What is it about the empty channels that causes FCPX to not play back instantly? And when the empty channels are muted, the file then plays back instantly?
-
Michael Gissing
September 24, 2017 at 11:42 pm“And when the empty channels are muted, the file then plays back instantly?”
Sounds like a bug to me.
Reply to this Discussion! Login or Sign Up