The Number That Matters More Than Resolution
Resolution is the number of pixels in the frame. 1920x1080 is 1080p. 3840x2160 is 4K. More pixels means more potential detail. But potential is the key word. Whether those pixels contain actual detail depends on how much data was used to encode them.
Bitrate is that amount of data. It determines how much information is available to describe each frame of video. A high-resolution video encoded at a low bitrate will look worse than a lower-resolution video encoded at a high bitrate. This is not an edge case. It happens constantly.
Think of it this way. Resolution is the canvas size. Bitrate is the paint. A huge canvas with almost no paint is just a big blank smear. A smaller canvas with rich, generous paint can look extraordinary. The combination matters, not the canvas size alone.
The reason resolution gets all the attention is marketing. "4K" is a number consumers understand as meaning "better." Bitrate is a technical measurement that varies per file and requires checking. Nobody puts "encoded at 45 Mbps average bitrate with VBR and H.265" on a product listing. So people buy resolution and wonder why their "4K" video looks soft.
What Bitrate Actually Is
Bitrate is the amount of data transmitted or processed per unit of time. For video, it is measured in megabits per second (Mbps) or kilobits per second (kbps).
One megabit is 1,000,000 bits. Not bytes. Bits. One byte equals eight bits. So 8 Mbps is 1 megabyte per second. This distinction matters when you are calculating download times and file sizes.
Higher bitrate means more data is being used to represent each frame of video. That data captures more detail in fast-moving areas, preserves fine texture in complex scenes (grass, fabric, foliage), and reduces compression artifacts (the blocky, smeared quality that appears in over-compressed video).
Lower bitrate means the encoder has less data to work with and must make aggressive choices about what to discard. It keeps broad shapes and dominant colors but sacrifices fine detail. Fast motion looks smeared. Edges get blocky. Dark areas fill with noise.
The encoder (the software converting raw video to a compressed format) has a hard job. It must fit a meaningful amount of detail into the available bitrate. Smart encoders (H.265, VP9, AV1) do this more efficiently than older ones (H.264), which is why a 4K video in H.265 at 20 Mbps can look as good as the same video in H.264 at 35 Mbps. The compression algorithm matters, not just the bitrate number.
The File Size Formula
This is genuinely useful to know. File size in video is deterministic. It is not a mystery.
Formula: File size (MB) = (Bitrate in Mbps x Duration in seconds) / 8
The division by 8 converts bits to bytes (since storage is measured in bytes, not bits).
Practical examples:
- A 10-minute video (600 seconds) at 8 Mbps: (8 x 600) / 8 = 600 MB
- The same 10-minute video at 20 Mbps: (20 x 600) / 8 = 1,500 MB = 1.5 GB
- A 60-second clip at 12 Mbps: (12 x 60) / 8 = 90 MB
- A 60-second clip at 3 Mbps: (3 x 60) / 8 = 22.5 MB
This formula works for constant bitrate (CBR) encoding. For variable bitrate (VBR) encoding, the average bitrate fluctuates, so the formula gives you an approximation based on the target average bitrate.
The audio bitrate adds to this total. A 192kbps audio track in a 10-minute video adds: (0.192 x 600) / 8 = 14.4 MB. Small compared to the video track but not nothing.
You can work this formula backwards. If someone tells you a file is 2.4 GB and 20 minutes long, the average bitrate is: (2400 MB x 8) / 1200 seconds = 16 Mbps average. That tells you roughly what quality level to expect.
How Resolution and Bitrate Interact
A 4K video needs roughly four times the bitrate of a 1080p video to look the same quality. That is because 4K has four times the pixels (8,294,400 vs 2,073,600). Each pixel needs its share of the available data budget.
When the bitrate does not scale up with the resolution, the encoder must compress each pixel more aggressively. The result is a video that is technically 4K (the frame dimensions are 3840x2160) but looks no better, and often noticeably worse, than a properly encoded 1080p version of the same content.
This happens more often than you might think. Streaming platforms serving 4K content to mobile users on limited data plans often throttle bitrates to levels that cannot actually represent 4K quality. You are downloading 4K-labeled content that is delivering 1080p-equivalent visual quality, at four times the file size. The worst of both worlds.
Framerate multiplies the data requirement further. 60fps video has twice as many frames per second as 30fps video, and therefore needs roughly twice the bitrate to maintain the same per-frame quality. A 1080p60 video needs about the same bitrate as a 4K30 video to look comparable in quality terms.
For practical decision-making: always check the bitrate of what you are downloading or receiving, not just the resolution label. A 1080p file at 12 Mbps will look significantly better than a 4K file at 8 Mbps for most content types.
Recommended Bitrate Ranges by Resolution and Framerate
These are the ranges used by major platforms (YouTube, Netflix, Apple) as encoding targets, and by professional editors as delivery targets. They represent the minimum adequate, good, and excellent quality thresholds for each resolution and codec combination.
| Resolution | H.264 Bitrate | H.265 Bitrate | VP9 Bitrate |
|---|---|---|---|
| 720p30 | 4-6 Mbps | 2.5-4 Mbps | 2.5-4 Mbps |
| 720p60 | 6-9 Mbps | 4-6 Mbps | 4-6 Mbps |
| 1080p30 | 8-12 Mbps | 6-9 Mbps | 6-9 Mbps |
| 1080p60 | 12-20 Mbps | 8-13 Mbps | 8-13 Mbps |
| 4K30 | 35-45 Mbps | 20-35 Mbps | 20-35 Mbps |
| 4K60 | 53-85 Mbps | 30-50 Mbps | 30-50 Mbps |
These are not minimums for acceptable viewing. They are ranges for quality that a critical viewer would call "good" to "excellent." The lower end of each range is adequate for straightforward content (talking head, static backgrounds). The upper end is necessary for complex content (sports, nature footage, fast-panning action).
Content complexity affects how well a given bitrate holds up. A video of a person talking in front of a white wall is easy to encode efficiently. A video of a forest in the wind with dozens of moving leaves, dappled light, and detailed bark texture is extremely hard to encode and requires more bits to represent without artifacts.
CRF (Constant Rate Factor) Explained
CRF is a quality-based encoding mode rather than a bitrate-based one. Instead of telling the encoder "use exactly 8 Mbps," you tell it "maintain this quality level and use however many bits that requires."
The CRF scale for H.264 runs from 0 (lossless, enormous files) to 51 (lowest possible quality, tiny files). The mathematically correct lossless option is CRF 0, but "lossless" at that scale produces files so large they are impractical for anything other than archival master files.
The meaningful range in practice is roughly CRF 15 to CRF 30.
CRF 15-18: Near-lossless. The video looks indistinguishable from the original source to virtually any viewer under any conditions. Used for master files and archival copies. File sizes are very large. A 10-minute video at CRF 17 in 1080p might be 3-5 GB.
CRF 18-22: Excellent quality. Most viewers cannot distinguish this from the original. All fine detail is preserved. Fast motion looks clean. File sizes are large but manageable for storage. This range is ideal for anything you plan to edit further, re-upload, or archive long-term.
CRF 23: The H.264 default. "Good enough" for most viewing situations. Subtle quality loss is present but requires direct comparison with the original to notice in most content. File sizes are reasonable.
CRF 24-28: Visible compression. A discerning viewer will notice softness, some blockiness in fast motion, and reduced texture detail in complex areas. Acceptable for casual viewing on small screens. Not suitable for editing or professional delivery.
CRF 28-35: Significant compression artifacts. Blockiness is visible, edges are poorly defined, and the video has a generally "processed" look. Only suitable for the smallest possible file sizes where quality is secondary to storage constraints.
For H.265, the CRF scale is numerically similar but different in its quality mapping. H.265 CRF 28 is roughly equivalent to H.264 CRF 23 in visual quality, because H.265 is simply a more efficient codec that achieves the same visual result with fewer bits.
CBR vs VBR: Which Is Better for Downloads
CBR is Constant Bit Rate. Every second of the video uses exactly the specified amount of data, regardless of whether that second contains a static talking head or an explosion with fast motion and complex detail.
VBR is Variable Bit Rate. The encoder allocates more data to complex frames and less to simple frames, using the bitrate budget efficiently across the entire video.
For video downloads, VBR is almost always better. Here is why.
Consider a 10-minute interview video. Half of it is a single person talking against a plain background. The other half has occasional cuts to b-roll footage with complex textures. With CBR at 8 Mbps, both the simple talking-head frames and the complex b-roll frames get the same 8 Mbps budget. The talking-head segments waste data (they look fine at 2 Mbps). The b-roll segments may look compressed even with 8 Mbps because all the extra detail needs more budget than simple frames.
With VBR at a target of 8 Mbps average, the encoder gives the talking-head frames 2-3 Mbps and redirects the saved bits to the complex b-roll frames, which might get 14-16 Mbps. Both segments look better. The average file size is the same. The quality distribution is smarter.
CBR exists for specific legitimate use cases: live streaming (CBR gives a predictable bandwidth requirement for stream servers), broadcast delivery (many TV broadcast standards require CBR), and certain hardware encoder limitations. For stored video files that you download and watch, VBR is superior in virtually every scenario.
When evaluating downloaded video quality, check if the bitrate is consistent throughout or varies. Most media players (VLC, MPV) let you view bitrate in real time. A VBR-encoded video will show bitrate variation. A CBR video stays constant. Neither is a quality indicator by itself, but VBR indicates that the encoder was working smarter rather than just pushing data.
Codec Efficiency Comparison
A codec (encoder/decoder) is the algorithm used to compress and decompress video. Different codecs achieve different levels of compression efficiency for the same visual quality. More efficient codecs produce smaller files at the same quality, or better quality at the same file size.
| Codec | Efficiency vs H.264 | Compatibility | Encode Speed |
|---|---|---|---|
| H.264 | Baseline | Universal (every device) | Fast |
| H.265 (HEVC) | 40-50% better (same quality, smaller file) | Good (most modern devices, may need codec on older Windows) | 2-4x slower than H.264 |
| VP9 | 30-50% better | Good in browsers (Chrome, Firefox); limited on some smart TVs | 3-5x slower than H.264 |
| AV1 | 50-60% better | Growing (Chrome, Firefox, newer Android/iOS; limited TV support) | 10-20x slower than H.264 (software encode) |
H.264 remains the safest choice for maximum compatibility. If you are downloading a clip to use anywhere, H.264 in an MP4 container will play on everything: every phone, every TV, every computer, every editing app, every browser. No codec installation needed, no compatibility questions.
H.265 is the sweet spot for storage-efficient high-quality video. If you have a modern device (post-2018 for most platforms), H.265 files play natively. Half the file size for the same quality is a meaningful benefit for a large clip library.
VP9 and AV1 are primarily streaming codecs. YouTube and Netflix use them for delivery efficiency (lower bandwidth costs). They are less commonly used for locally stored files because the encoding time is much longer and the compatibility gaps still exist on certain devices.
How YouTube Compresses Video
When you upload a video to YouTube, the platform re-encodes it into multiple quality levels for adaptive streaming. Understanding this process explains why downloaded YouTube video quality can vary significantly from what you uploaded.
YouTube's encoding pipeline as of 2026 works roughly as follows:
Primary codec: VP9. YouTube delivers most content in VP9, especially on Chrome and other Chromium-based browsers. VP9 is a Google codec (YouTube is Google), so they have strong incentive to use it. It offers good compression efficiency without the patent licensing costs of H.265.
AV1 for newer content. YouTube has been rolling out AV1 encoding for newer content, particularly 4K and HDR. AV1 offers better efficiency than VP9 for complex content. As of 2026, AV1 delivery is primarily on Chrome, Firefox, and newer Android devices.
H.264 as compatibility fallback. Older devices, Safari on some configurations, and embedded players often receive H.264 streams. YouTube maintains H.264 versions of all content for this reason.
YouTube's target bitrates for each quality tier (approximate, subject to change):
- 1080p30 VP9: 8 Mbps video + 192 kbps audio
- 1080p60 VP9: 12-15 Mbps video
- 4K30 VP9: 35-45 Mbps video
- 4K30 AV1: 18-25 Mbps video (much smaller file, similar quality due to AV1 efficiency)
The critical detail: YouTube's re-encoding always introduces some quality loss from the original upload. This is unavoidable. The upload is re-encoded, not stored as-is. If you upload a pristine 1080p60 file at 50 Mbps, YouTube will deliver it at roughly 15 Mbps VP9. The delivered bitrate is determined by YouTube's target levels, not your upload quality. Uploading a higher-quality source does help slightly, because the encoder has more source data to work with, but the delivered bitrate cap applies regardless.
What YTCut Actually Encodes
When you use YTCut to download a clip, you are getting YouTube's served video stream, which has already been encoded at YouTube's quality levels. YTCut then applies a second encoding pass for the clip you have selected, specifically for the output format you chose.
For MP4 output, YTCut encodes using H.264 at CRF 21. This is a deliberate choice that sits in the "excellent quality" range of the CRF scale (one notch better than the H.264 default of CRF 23). The rationale: most people downloading clips will use them for viewing, presentation, or light editing. CRF 21 produces files that look excellent on screen while remaining reasonable in file size. A 60-second 1080p clip at CRF 21 is typically 50-120 MB depending on content complexity.
For WebM output, YTCut uses VP9 encoding at CRF 32. VP9 CRF 32 maps to roughly the same visual quality as H.264 CRF 21, because VP9 is a more efficient codec. The WebM files tend to be 30-40% smaller than the equivalent MP4 at comparable visual quality.
For MP3 output (audio only), YTCut extracts the audio track and encodes at 192 kbps AAC or MP3. This is above the "good enough for music" threshold of 128 kbps and below the "transparent for audiophiles" threshold of 320 kbps. For speech content, podcasts, or music listening on standard headphones, 192 kbps is clean and clear.
There is a generation loss consideration: the clip you download from YTCut has been encoded twice. Once by YouTube (original upload to YouTube's format), and once by YTCut's encoder for your output format. Each encode introduces some additional compression loss. For most practical use cases, this double-encode is imperceptible at CRF 21. For professional editing workflows where quality preservation is critical, you want the source video downloaded without re-encoding, which requires a different approach (yt-dlp with -c copy and format selection for original streams).
File Size Reference Table
Applying the formula to real scenarios. All estimates assume average content complexity. High-motion content (sports, action) will run larger; simple content (lectures, interviews) will run smaller.
| Resolution | Bitrate | 1-min clip | 10-min video | 60-min video |
|---|---|---|---|---|
| 1080p30 | 5 Mbps | 37.5 MB | 375 MB | 2.25 GB |
| 1080p30 | 8 Mbps | 60 MB | 600 MB | 3.6 GB |
| 1080p30 | 12 Mbps | 90 MB | 900 MB | 5.4 GB |
| 1080p60 | 20 Mbps | 150 MB | 1.5 GB | 9 GB |
| 4K30 | 35 Mbps | 262 MB | 2.6 GB | 15.75 GB |
| 4K60 | 68 Mbps | 510 MB | 5.1 GB | 30.6 GB |
These numbers make the storage implications of quality choices concrete. Downloading full-resolution 4K60 videos for archive is a serious storage commitment. A library of 100 one-hour videos at 4K60 would require over 3 terabytes. The same library at 1080p30 at 8 Mbps would be 360 GB. Same content, fraction of the storage requirement.
What CRF 21 vs CRF 23 vs CRF 28 Looks Like in Practice
Since you cannot see a side-by-side comparison here, descriptions of what to look for when comparing CRF settings.
CRF 21 vs CRF 23: In most content, the difference is subtle and requires direct comparison to notice. In a talking-head video with a static background, you would be hard-pressed to identify which is which. The gap becomes visible in complex scenes: forest foliage where individual leaves have fine detail, fabric textures with tight weaves, or dark scenes with subtle gradients. At CRF 21, these retain more fine detail. At CRF 23, fine detail in complex areas begins to smooth out slightly, and dark gradients may show minor banding. For a typical viewer watching on a laptop at arm's length, this difference is negligible.
CRF 23 vs CRF 28: This difference is noticeable and does not require a trained eye. At CRF 28, fast-motion sequences (someone moving their hands quickly while speaking, rapid camera pans, sports footage) will show visible blurring and some macro-blocking (rectangular patch artifacts in the moving area). Static scenes look reasonable but complex textures are noticeably softer. Skin texture in close-up face shots becomes smooth in a slightly unnatural way. Dark areas may show visible compression noise.
Where CRF 28 is acceptable: Delivery to low-bandwidth situations where file size is a hard constraint. Reference clips you will watch once and delete. Preview renders during an edit session. Small thumbnails or low-resolution proxies. Not for anything you intend to present, publish, or archive.
The codec also changes how each CRF level looks in practice. H.265 CRF 28 looks better than H.264 CRF 28 because H.265 is more efficient at any given CRF value. If someone quotes CRF without specifying the codec, the number alone is not enough information.
When to Choose 4K vs 1080p Download
This comes down to four factors: what you are watching on, what you are doing with the file, your storage budget, and what quality the original was actually shot in.
What screen are you watching on? A 1080p monitor cannot display 4K resolution. It will display the video at 1080p regardless of whether the file is 4K or 1080p. The benefit of downloading 4K for display on a 1080p screen is zero from a visual standpoint, and the storage cost is 4x higher. A 27-inch or larger 4K monitor, a 4K TV, or a modern iPad Pro (2732x2048 display) can show the actual benefit of 4K content. Phones, even flagship phones with high-resolution screens, are too small for 4K to look meaningfully different from 1080p at normal viewing distance.
What are you doing with the file? If you are editing the clip in a video editor and plan to output the final edit at 4K, download 4K source footage. You want to edit with the highest available resolution so you can reframe, zoom, and grade without quality loss. If you are just watching the clip or using it in a presentation, 1080p is enough for any screen smaller than a 4K TV.
Storage budget. 4K files are 4-8x larger than equivalent 1080p files. If storage is limited, 1080p is the sensible choice for most casual use. The visual improvement of 4K on appropriate screens is real but not dramatic for most content types. Talking-head content, tutorials, and interviews that were shot on a standard camera look similar at 1080p and 4K on most displays. Nature footage, architectural detail, and sports footage show the most meaningful 4K benefit.
What resolution was the original shot in? Not every "4K" YouTube video was actually shot in 4K. Some creators upscale 1080p footage to 4K for upload to get access to YouTube's higher-bitrate 4K encoding tier (which actually improves the delivered 1080p quality when the viewer downscales). If the source is genuinely 1080p that was upscaled, downloading the 4K version produces a large file with 1080p actual detail. Check the video description or comments for production information if quality matters.
Audio Bitrate Matters Too
Video bitrate gets all the attention and audio bitrate is treated as an afterthought. For certain use cases, audio quality is actually more noticeable than video quality.
Audio bitrate for compressed formats (MP3, AAC, Opus) follows a similar principle to video: more bits per second means more detail preserved, cleaner high frequencies, less compression artifact noise.
128 kbps: The old "good enough" standard. For speech content (podcasts, lectures, tutorials, talking-head videos), 128 kbps is genuinely fine. The information bandwidth of the human voice fits easily within 128 kbps. Most listeners cannot identify 128 kbps speech as compressed if the recording was clean at the source.
For music, 128 kbps is noticeably compressed. High-frequency detail (cymbals, sustained strings, airy vocals) loses sparkle and clarity. The classic test is a pair of headphones and a cymbal-heavy recording. Compression artifacts in cymbals are unmistakable at 128 kbps.
192 kbps: The practical sweet spot for music content. For most music on most headphones and speakers, 192 kbps AAC or 192 kbps MP3 is perceptually transparent or very close to it. High frequencies are clean. The overall sound is full and detailed. File sizes are manageable.
320 kbps: Overkill for MP3 and AAC. At 320 kbps, the compression artifacts are inaudible on essentially all playback systems. The benefit over 192 kbps is marginal for virtually all listeners on virtually all equipment. Use 320 kbps if you are archiving music you care about and storage is not a concern, or if you need the maximum quality ceiling for future re-encoding purposes.
Opus at 128 kbps: Opus is a modern audio codec significantly more efficient than MP3 or AAC. Opus at 128 kbps sounds better than MP3 at 192 kbps. YouTube's WebM streams use Opus for audio. If you download a WebM file from YTCut, the Opus audio at 128-160 kbps in that file will sound better than the MP3 128 kbps you might compare it to.
For video files, the audio track adds a relatively small amount to the total file size but a disproportionate amount to the perceived quality of the experience. A video with crisp audio and decent video often feels higher quality than the same video with excellent video but muddy, compressed audio. Do not neglect the audio channel when evaluating download quality.
FAQ
Is there a way to check the bitrate of a video file without special software?
On Windows, right-click the file, select Properties, then the Details tab. The "Total bitrate" field shows the combined audio and video bitrate in kbps. On Mac, select the file in Finder, press Command-I for Get Info, which shows file size and duration (calculate bitrate using the formula). VLC's Media Information window (Tools, Media Information, Codec tab) shows bitrate in real time during playback and is the most comprehensive free option on any platform.
Why does the same 1080p video look better from YouTube than from a streaming TV service sometimes?
YouTube's 1080p bitrate targets (8-12 Mbps VP9) are actually higher than some streaming services' 1080p targets, especially on constrained bandwidth plans or on platforms that throttle based on device type. Netflix, for example, served at around 4-5 Mbps for 1080p at standard quality tiers, which is well below YouTube's typical delivery. The "1080p" label tells you the frame resolution. It does not tell you the bitrate or quality level. Always compare bitrates when evaluating "same resolution" content from different sources.
Does uploading a 4K video to YouTube improve the 1080p version people watch?
Yes, modestly. YouTube applies a higher average bitrate to videos uploaded at 4K resolution, even for the 1080p delivery tier. This is a quirk of YouTube's encoding pipeline that has been confirmed by various creator experiments. Uploading at 4K gets your video into a different (higher quality) encoding queue. The 1080p stream delivered from a 4K upload often has better detail and fewer compression artifacts than the 1080p stream from a native 1080p upload. The improvement is real but not dramatic.
What is the smallest acceptable file size for a 1-minute clip I plan to use in a presentation?
For presentation use on a 1080p projector or screen, a 1-minute 1080p clip at 5 Mbps (about 37 MB) is the minimum where quality is not going to embarrass you. At 8 Mbps (60 MB), the quality is comfortable for presentation. There is rarely a reason to use a lower bitrate than 5 Mbps for any professional presentation context unless you are working with strict email attachment limits.
Does variable bitrate mean the quality changes during playback?
No. VBR means the amount of data used changes, not the perceived quality. The encoder allocates more bits to complex frames so they look as good as simple frames. Ideally, the visual quality is consistent throughout. The bitrate number goes up and down; the quality stays stable. Poorly implemented VBR can occasionally produce inconsistencies, but a properly encoded VBR file should look uniform throughout even though its bitrate graph looks like a hilly terrain rather than a flat line.