Rendering

Rendering (also called export, output, or delivery) is the process of converting your timeline — with all its effects, grades, transitions, and audio mixes — into a self-contained file in a specific codec and container. The codec determines how the image data is compressed (or not), the container (MOV, MXF, MP4) wraps the video and audio streams together with metadata, and the delivery specs determine the codec, resolution, frame rate, bit depth, and color space requirements. Choosing the right render settings is not optional — it is the difference between your grade looking correct on the audience's screen and looking like a different edit entirely. The three primary delivery targets and their correct settings. YouTube/Vimeo/web: H.264 High Profile, Level 4.2 (for 1080p) or Level 5.1 (for 4K), variable bitrate targeting 15-20 Mbps for 1080p and 35-50 Mbps for 4K. Render at the native timeline resolution — do not upscale before upload. YouTube re-encodes everything to their own specs, so giving them a higher-bitrate file than necessary just wastes upload time without improving quality. Color space must be tagged as Rec.709 with BT.1886 gamma in the file's metadata. Festivals/theatrical/DCP: ProRes 422 HQ or ProRes 4444 (if you need alpha channels) as the master file, then a DCP (Digital Cinema Package) conversion using a tool like OpenDCP or Resolve's DCP rendering. DCP uses JPEG2000 compression in an MXF container, XYZ color space, and a maximum data rate of 250 Mbps for both 2K and 4K per the DCI specification. Broadcast: DNxHR HQX (10-bit) or ProRes 422 HQ in an MXF container, meeting the specific broadcaster's delivery specs (every network has their own requirements document — get it before you render). Common broadcast specs include 10-bit 4:2:2 chroma subsampling, specific audio loudness targets (-24 LKFS for US broadcast per the ATSC A/85 recommended practice, -23 LUFS integrated for EBU R128 in Europe). Render cache in DaVinci Resolve is a separate concept from final rendering. The render cache stores processed frames on disk so that playback is smooth during editing and grading — it does not affect the final export quality. You can set the render cache to User mode (you decide which nodes to cache) or Smart mode (Resolve caches anything that causes dropped frames). Clear the render cache before final delivery to ensure stale cached frames do not persist.

The single most common rendering mistake: selecting "Same as Source" or "Original Settings" in the export dialog. This preserves whatever codec and resolution the footage was shot in — but your timeline is not a single source file. If you shot ARRI Alexa at 2.8K ProRes 4444 and edited in a 4K DCI timeline with Resolve effects, rendering "Same as Source" would output a 2.8K ProRes 4444 file with the wrong resolution, wrong color space tagging, and none of the delivery specs the platform requires. Always configure export settings explicitly for the delivery target. Bitrate selection matters enormously for lossy codecs. H.264 and H.265 are lossy — they throw away data to achieve compression. Too low a bitrate and you get visible macroblocking in dark gradients, color banding in skies, and mosquito noise around high-contrast edges. For 1080p YouTube delivery, 15-20 Mbps VBR is the sweet spot — below 10 Mbps and compression artifacts become visible even after YouTube's own re-encoding. For 4K, target 35-50 Mbps. For ProRes, the bitrate is fixed by the codec: ProRes 422 HQ runs at ~176 Mbps for 1080p24, ProRes 4444 at ~330 Mbps. ProRes is visually lossless — you will not see compression artifacts at any reasonable viewing distance, but the files are 5-10x larger than equivalent H.264. Audio rendering is often treated as an afterthought but carries its own pitfalls. For broadcast delivery, audio must meet loudness standards: -24 LKFS per the ATSC A/85 recommended practice for the US, and -23 LUFS integrated with a true peak not exceeding -1 dBTP per the EBU R128 standard for Europe. YouTube normalizes all audio to approximately -14 LUFS integrated loudness — if your mix is significantly louder, YouTube will turn it down and may introduce pumping artifacts. For theatrical DCP, audio is 24-bit 48kHz PCM in the MXF container, typically 5.1 or 7.1 channel layout. DaVinci Resolve's Fairlight page includes a Loudness meter that measures integrated LUFS, true peak, and loudness range — use it before every render.