"Software Only" stuck on encode settings panel.... Premiere Pro CC 2018 latest update

Hello there! i know im 2 years late but if anyone is still having this issue, just watch this vid to enable Hardware Encoding.

[Moderator note: It is Adobe policy to not allow comments about the Console on their forums. That is for engineer work, and users can seriously screw up their installation by messing there. The link has been removed.]

Ok so since u seem to not wanna listen to others how do u explain after updating to 2020 my videos now take 2 hrs and never once finish, it always freezes half way now which is why I'm here. I notice ld before update my 1080ti was selected and encoding took 20 to 30 mins easy. For the past 3 days I have been trying to finish exporting my work I have buyer my ass on and pay for a software that crashes. My PC is perfect and top of the line hardware and worked in previous builds now freezes on them all 

Here the fix thank me later https://youtu.be/obWeExQMxsg

But if that's the case, why does my render time get cut in half when I choose hardware rather than software encoding?

Hello

I managed to activate hardware decoding for Premiere Pro, but in Media Encoder this is still software only.

I tried all kind of things to add sequence to queue, never got to see the hardware accelerated encoding, only CUDA, not Intel iGPU 630.

Any thoughts?

Thanks

Great work Neil, but can I jump in here?  It's a great thing that "there's nothing to fix", but the "WHY" of your answer is a bit confusing.  Let's go into that a little more, shall we?  Again, nice work on the find with intel's built in codec.  I loved that part.

For the rest:

I've always used a single pass encode, mainly because it takes a shorter time, and gives me the result I want.  I was using a mac laptop with an intel core 2 duo topping out at 2.16ghz and an nvidia geforce with 256mb of vram, all attached to 4gb of internal ram.  Now I'm working an acer predator with 16gb ram, quad i7 gen 6, nvidia gforce gtx980 with 4gb vram, and optimus set up through the nvidia toolsets provided by some developers (windows apparently removed it from their fileset, quoted some kind of redundancy and then apps started breaking; redundancy?  no.  Dependence.  I guess they don't know the difference).  I fly through those single pass clips with ease.  For every minute of video, it takes about 30s to process an ingest operation that keeps them all in line, and thats on battery power.

Like all of you, I entered the realm of 4k thinking, hmm, 2 passes can decrease file size and slightly increase quality, why not use all my power.  Here's the RUB:  The standard H.264 encoder uses an older 32bit algorithm to render out the second pass.  This slows things down immensely on 64bit systems.  By running a built in encoder with your intel processor, you are making use of a more accurate, and much better algorithm, and it's lightning fast.  The downside?  Nobody can render across two different cards at once anymore.  This used to work just fine when windows allowed optimus usage in the system itself.  Renders were fast, and this "BUG" did not appear, as the app could see the intel card, but with a cuda header attached as a secondary renderer, and it allowed the app to offload to cuda when effects were used.  NOW: you are running an older algorithm that is less accurate, slower and clunkier, on a cuda card of course, but still clunky.  It takes longer for the 2 passes, as it has to run the algorithm in a software layer on top of your card.  THUS a mesage about software rendering in your settings box, even though you have CUDA enabled in your main renderer.

THE FIX:

2pass VBR is useful mostly for space saving.  Since you only get about 10% space and a small bump in quality, you can estimate the quality effect, and use a single pass, taking up a little more space, but retaining the quality.  IF you are on your final output, use the 2Pass method instead, as it will make your video deliverable, with excellent quality.

IF you are handling 3k and higher video, you are better off with HEVC codecs that are meant for higher depth.  They are just as fast, if not faster, and they compress a 4k to about the same dataspace as a 1080p (2k).  The other difference is that sites who take a deliverable of HEVC are more likely to incorporate fault tolerance with VFR issues.  They run it as a split codec, using two streams, interleaved on two separate clocks, which are then put back together on your end to play the video.  This means video and audio are separately streamed, and their clocks can alter midstream without affecting one another.  Don't bank on it though.  Always check the delivery format needed, and look for VFR notations.  If they don't show anything about it, or they note only specific context of audio and video together, chances are, you'll need to fix the VFR yourself.

Have we found a fix for this yet. I am having the same issue. Despite my GPU being selected it refuses to use it. It will only use the CPU's. All my renders also say Software only. I have 2 GPU's in my workstation and both just sit idle while the Media Encoder runs through renders.

why you so mad ?

Premiere Pro supports hardware-accelerated H.264 encoding on Windows 10 with 6^th Generation (or later) Intel® Core™ processors and Intel Graphics enabled, and hardware-accelerated H.264 and HEVC encoding on Mac OS 10.13 systems with supported hardware.

On supported systems, you can select Hardware Encoding in the Encoding Settings section of the Export Settings dialog. If the specified combination of Export Settings is not supported by the hardware, Premiere Pro falls back to Software Encoding.

Hope this helps

What happened in the 2018 update? I'm getting Software Only in the export settings for H264. Running a 4970K and a 1080ti. Cmon adobe.