[Ipg-smz] What happens when Linux keeps getting better and better?

[Swapnil “Swap” Bhartiya]

I, on the other hand shoot with 3 cameras in 4K and record sound from 3 different microphones (uncompressed) - MixPre 6 - all recorded separately. Since I use 3 camera and record in LOG- I have to color correct them and I also do grading. Then I have to process the audio. On top of that I have to use multi-cam feature to be able to switch between cameras. I use many external tools from iZotopes and video plugins. None will work on Linux. Let alone Linux able to handle 4K HVEC format. Plus Linux doesn’t get optimization for GPU. Same video on a 6Core system with 64GB of ram takes 3 hours on Window with Adobe and real time on macOS. I can’t afford to spend 5 hours on one video (and probably 2-3 days on Linux) Don’t even ask me to try this on Linux. So those or us who use it from professional work, do know the real value of using macOS. 

But as it was said many times, its’ all about workloads. If you don’t have any professional multi-media work Linux desktop will do just fine. That said if I buy a 6core machine (roughly $5000) I am NOT going to use Linux on it; it’s waste of resources. 

Swap


> On Dec 20, 2018, at 7:05 PM, Stephen Satchell <ipg at satchell.net> wrote:
> 
> What little audio work I've done recently, I've been quite happy with
> Audacity on Linux Mint for multi-channel mixdown.  Still have difficulty
> getting more than stereo input (courtesy of a Focusrite Scarlett 2i4),
> but I've not needed more than two channels at a time.
> 
> I used to do audio work using Syntrillium CoolEdit on Windows, with a
> professional-grade sound card.  Then the sound card went unsupported,
> and Syntrillium was sold to Adobe...and so much for doing audio work on
> Windows.  (But then Windoz was banished anyway.)
> 
> My original usage was a little unusual:  I would hook the pro sound card
> inputs to a telco line simulator's audio side outputs, have two modems
> establish a connection, capture the two channels of audio, then use a
> custom program to create traveling FFTs of the two channels.  This was
> used to analyze the conformance of a V.32 modem (later V.32 bis) to the
> standards.  The system replaced a $40,000 audio spectrum analyzer, which
> meant that I could have more than one testing station checking this
> stuff without breaking the budget.
> 
> I used a similar setup to analyze a COM-10 cable system's telephone
> channel, to find out why fax wasn't working.  Using a software
> implementation of the industry-standard 23-tone test for telephony, I
> recorded an hour of transmission, then took a slew of samples of various
> parameters in the playback and plotted them.  Found the problem: their
> timing chain was unstable, so the phase jitter in the DS0 channel would
> exceed the ability of V.29 (and V.27ter) to establish a fax connection.
> 
> I've also done a lot of audio work over the years, but not digitally.
> The reason: never found a digital multi-channel (8 and 16 channel)
> recorder with a decent price.  If I did, I would have fed it with
> per-channel output from a Ramsa mixing board, and remix using the line
> inputs to start, then move to digital remix when that became available.
> The recording solutions I did find just couldn't keep up with 44-khz
> sampling at 16 bits on more than two channels.  That was the fault of
> the computer (and Windows).
> 
> Things are better now with regards to stand-alone digital recorders.
> (Sweetwater, anyone?)  My need for multi-channel recording has passed --
> I don't do on-location recording anymore.  (Call me stay-at-home...)
> 
> 
> 
> -- 
> Ipg-smz mailing list
> Ipg-smz at netpress.org
> http://netpress.org/mailman/listinfo/ipg-smz_netpress.org