In this video tutorial, Peter Emery shows you how to check if DaVinci Resolve supports your footage for hardware-accelerated decoding. DaVinci Resolve leverages hardware to process (decode) demanding codecs and footage for improved timeline playback. Check out this article to see which flavors of H.264 & H.265 codecs Resolve supports for hardware-accelerated decoding.
Recognizing what footage Resolve supports for hardware decoding can help you optimize your editing experience by ensuring compatibility between your hardware, software, and footage. Even if your hardware supports specific footage and codecs, DaVinci Resolve might not support all variations for hardware-accelerated decoding and encoding. You can explore resources from NVIDIA, AMD, and Intel to determine the media formats your hardware supports.
Video Tutorial
Video Transcription
00:00:00:00 – 00:00:27:21
Peter Emery
Welcome to our video tutorial that shows you how to check and see if your footage is supported for hardware decoding in DaVinci Resolve. DaVinci Resolve provides quite a bit of flexibility in determining how the application utilizes hardware for processing your media. Depending on the type of media, Resolve can utilize the GPU for a lot of your processing, but also supports hardware decoding and encoding from both integrated graphics on your CPU or a dedicated GPU.
00:00:27:22 – 00:00:53:17
As long as the codec format is supported by both DaVinci Resolve and the hardware manufacturer. Before we jump in, Puget Systems has a helpful article that breaks down which hardware supports hardware decoding for H264 and H265 at different bitrates and chroma subsampling formats in DaVinci Resolve. Check the link below to quickly find out and see if your hardware is supported for hardware-accelerated decoding and DaVinci Resolve.
00:00:53:19 – 00:01:13:22
Understanding how codecs work can be very complex. Even if you know the base codec, bit depth, and chroma subsampling of your footage, you may still want to consider running some specific tests to see if hardware-accelerated decoding is supported and working. To do that, let’s move on to the first step. Step one, check your footage with MediaInfo.
Step 1
00:01:14:00 – 00:01:31:22
If you’re not sure about the codec details of your footage, download a program called MediaInfo. I’ve posted a link to MediaInfo below so you can download the program. You can skip this step if you already know the specs of your footage, but if you’re unsure, this is a great way to double check. MediaInfo will show you the codec, bit depth, and chroma subsampling of your media.
00:01:32:03 – 00:01:50:19
Comparing this and the hardware you have to the tables on our website, will give you a good idea about whether hardware decoding is or isn’t supported. Unfortunately, there is no way to directly tell if Resolve is using hardware decoding when processing your media. Instead, we have to run some performance comparisons with the hardware-accelerated decoding feature enabled and disabled.
Step 2
00:01:50:19 – 00:02:22:08
To do this, let’s move on to step two. This is a two part process where we check results with the hardware-accelerated decoding feature on and off. Let’s launch Resolve manually. Go to preferences Decode options and change the settings so that hardware-accelerated decoding is off. Let’s close Resolve and relaunch to make sure the changes take effect. Next, let’s import your footage into Resolve and we’re going to create a timeline for each clip.
00:02:22:10 – 00:02:49:15
For this tutorial, I’ll be testing two clips from two different cameras, but for your own testing, if you have different media types from different recording sources, you’ll need to create a timeline for each. Once our timelines are created, we can go directly to the deliver page. And let’s set our export specs to DNxHR LB. Exact settings don’t matter, but we want something easy for the system to encode too, so that decoding of the media is the main bottleneck.
00:02:49:20 – 00:02:59:05
Let’s also make sure we name this file decode off and save the project so that we don’t have to do this again later.
00:02:59:07 – 00:03:26:17
Next step is to open up Task Manager. Go to details and find Resolve.exe. Right click and set affinity. Finally we’re going to uncheck all of the processors except for CPU 1. This makes Resolve run only on a single CPU core with the idea being to make the performance without hardware decoding very poor, so that it will be very obvious if hardware-accelerated decoding is supported or not.
00:03:26:19 – 00:03:45:00
Now let’s go back to Resolve and render out the footage. What we are keeping an eye out for is the total time it takes to export this footage. Make sure you log this somewhere for reference. I’ll be recording with a screenshot so I can visually show you the difference after we get through step three.
Step 3
00:03:45:01 – 00:04:15:21
After the first export finishes, we’re going to go back to preferences and now turn on the hardware-accelerated decoding feature. If you want to test with a specific device like Intel, Nvidia or AMD, you can select just that one. Or if you don’t care which is used, you can turn them all on. Now let’s close Resolve to let the effects take place. Once Resolve relaunches, open the project you saved earlier. And before we move forward, let’s go back to the Task Manager and again set affinity of Resolve.exe to CPU 1.
00:04:16:02 – 00:04:45:23
Disabling the use of all other cores. Let’s go back into Resolve. And let’s first start off by changing the name of the file to decode on. We’re going to check our export specs to make sure they’re consistent to our first set of exports. Add this to the queue and hit render. Log our expert time. And now let’s move on to the final step to compare the results with hardware accelerated decoding of.
Step 4
00:04645:00- 00:05:06:15
Now if you see a notable time decrease in your decode on export from your decode off export, then congratulations! Your footage is supported for hardware-accelerated decoding in Resolve. However, if your footage exports at roughly the same amount of time, then with high probability, we can assume that your footage is not supported for hardware-accelerated decoding.
00:05:06:18 – 00:05:28:23
This could be for a number of reasons. It could be that your hardware doesn’t have support, or that it does, but Blackmagic has an added support on the software side. Or it could just not be possible with the hardware currently available. We hope this tutorial was helpful for you and if you really did enjoy it, please give it a thumbs up, or just leave a comment down below and we’ll get back to you.
00:05:29:01 – 00:05:50:19
And if you have any other questions, please feel free to reach out to one of our consultants and they can talk a little bit more about your system and Resolve. And if you’re still unsure how your hardware is supposed to be performing in DaVinci Resolve. Go to pugetsystems.com and try PugetBench for Creators. It allows you to run a benchmark and compare your system’s hardware specs to others with similar specs.
00:05:50:19 – 00:06:06:00
To see if your performance problems are due to your system or your workflow. Lastly, make sure to subscribe to our channel as we’ll be posting more content to help you optimize your workflows in DaVinci Resolve.