Hardware Encoding - Not Found


#1

Hello there,

We are starting to do some video editing, and we are mainly trying if we like it, before spend money in new rig.

We are trying to use our (quite old) Acer desktop and everything works flawlessy, but obviously the rendering time is quite long.

We would like to try the “Use hardware encoder” option during export, but we face some problems:

  • If I click the checkbox, the message “Nothing found” appears
  • If I open “Configure…”, and select one of the options (tried them all one by one), The resulting video is black (but with audio)

Now, I would like to know if I’m missing something somewhere (system settings, drivers, something stupid perhaps…) or what else.

My system configuration:

  • Windows 10 64 bit
  • shotcut 18.11.18
  • Acer Desktop with:
    8 GB RAM DDR-3 1333
    Nvidia GeForce GT 320 (1GB)
    SSD drive

Many thanks for the huge job: as a software developer, I know how much work there’s inside :slight_smile:

The software is really a joy to use. We tried some other free/open editors, but this one is the best mix between features and usability.


#2

I’ve already started a thread on this.


#3

Hallo Hudson thanks for the reply.

I’ve seen your topic, but my case looks like to be different as selecting any of the checkboxes the video will not export correctly (black screen).
My real question is how can I know if the (old) GT 320 should be supported or not.


#4

And there was another post concerning Shotcut version 18.11.13 (which has been replaced with version 18.11.18)

You didn’t specify which version you have, but the above post does similar to your issue.


#5

Again, thanks for finding this. I looked for in the forum, but didn’t see this second thread.
So it looks like I’m not alone…

I edited the original post adding the version used (18.11.18)


#6

NVENC is not supported on this card:

Do not forget to click “a complete list” on that page to see all.
You did not specify your Intel chip in order to check Quick Sync Video, but if that card is any indication of the overall age of your system, it is unlikely to be supported:


#7

Many thanks, the system is older than my son :slight_smile:

Will upgrade to a new system soon…


#8

I have a modest nvidia GT710 about one year old. I just Exported an NTSC SD project about 43 minutes long and it took just about 15 minutes! This new GPU Export support is wonderful! It should be worth the upgrade - even with a modest video card.

Update: Currently, $44 USD at Amazon - a bargain!
https://www.amazon.com/MSI-GT-710-2GD3H-LP/dp/B01AZHOX5K

-=Ken=-


#9

My 7-years-old Nvidia Geforce GTX 570 has no encoding chip. I’m generally not much interested in graphic cards, as I never played computer games. But now, my grandson will make fun of me: I’ll be interested in purchasing a graphic card for the firs time in my life.


#10

A bit more expensive yet still affordable on Amazon France: €49.49.


#11

Yes, Francois, a newer video card, even a modestly priced one will give these results. I was pleasantly surprised to find that Shotcut supported the GPU encoding - in Linux!

-=Ken=-


#12

I’ll throw out there that if you’re planing to get a card for something like this, and sticking to budget options that you’d be better off looking towards the 750.
The GT710 is a kepler class card, this arcitecture was introduced in 2012 and was replaced by Maxwell in 2014 then Pascal in 2016 and recently Turing in 2018.
Fermi(Keplers immediate predecessor) left support status earlier this year, nvidia seems to support cards for roughly 8 years meaning kepler is likely going to be moved to a similar status in 2020. A maxwell based card like the 750 will have 2 more years of support(also I hate split generations where more than one arch(7xx had 3) is included in a numerical card series.
Also skip the 1030, it looks good until you realise they stripped out the NVENC even though the little quadro P400(only 120 apparently on amazon https://amzn.to/2PfPsdC, cheap for a quadro) kept it Pascal gets you up to 8K HVEC encoding too interestingly enough.