Best Export Settings for High Quality Instagram Reels: Why One Perfect Preset Doesn't Exist

What are the best export settings for Instagram Reels?

You'll find a lot of very specific answers online: export at 1080x1920, use H.264, set your bitrate to exactly 8, 10, 12 or 15 Mbps, use 30 fps, and your Reel will supposedly look perfect. After testing this myself, I don't think it works like that.

I'm a software engineer with 15 years of experience and also a videographer. For this research I went through Meta's own engineering documentation and tested video compression myself using DaVinci Resolve, FFmpeg and VMAF.

And there's a technical reason why the same Instagram Reel can look sharp on one phone and worse on another.

The Instagram Reels export settings I use

If you just want my current settings, here they are.

For a normal 1080p Instagram Reel, I use:

• 1080x1920 resolution
• H.264 codec
• Bitrate around 15-20 Mbps
• AAC audio at 48 kHz

• 2160x3840 resolution
• H.264 codec
• Bitrate around 60-80 Mbps
• AAC audio at 48 kHz

If I actually finished the project in 4K vertical, I normally keep it at 2160x3840 instead of intentionally reducing it to 1080p before upload. But the important part is that 15-20 Mbps (or 60-80 Mbps for 4k) is my recommendation, not an official Instagram bitrate.

There is no single bitrate that guarantees maximum Instagram video quality for every Reel. My tests show why.

Why Instagram Reels can lose quality after upload

A common idea is that you upload a video, Instagram compresses it once, and everybody then watches that compressed file. That isn't how Meta describes its video system.

Instagram can create multiple versions of your uploaded video. Those versions can have different resolutions, bitrates, codecs, frame rates and quality levels. When somebody watches your Reel, Instagram can then choose an appropriate version for that playback.

In one example from Meta's engineering team, they tested a single 1080p source using seven different resolutions and five compression levels. That gives 35 possible encodes from one source. Meta also discusses using AVC/H.264, VP9 and AV1 inside its video pipeline.

This explains something I've seen people asking about for a long time: why does my Instagram Reel look good on my phone but blurry on another phone?

You may not always be watching exactly the same encoded version. That also makes many Instagram compression tests difficult to trust. If I upload one Reel at 1080p and another at 4K, then look at them on one phone, I still don't know exactly what Instagram generated or which version it decided to send to me.

So instead of trying to guess Instagram's internal settings, I started looking at something we actually control: the quality of the video we upload.

What is the best bitrate for Instagram Reels?

This is probably one of the most common questions about Instagram export settings. And this is where my own test became interesting.

I took two clips that were very different visually. The first was a close shot of a yellow car. It had large smooth areas, a soft background and relatively little fine detail. The other was a moving drone shot over a city with thousands of buildings and small details.

I exported both as H.264 at 1080p and 30 fps, then tested different bitrates. At only 2 Mbps, the car shot scored around 92.6 VMAF. The detailed city shot scored only 67.3 VMAF.

That's a huge difference from two videos exported with identical settings. At 4 Mbps, the car was around 93.5 while the city was still around 82.5. At higher bitrates the difference became much smaller.

Why does this happen?

Video complexity matters.

A soft background with large areas that don't change very much is relatively easy to compress. A moving cityscape, trees, grass, water, hair, crowds and other fine moving details are much harder. So when somebody says that 8 Mbps, 10 Mbps or 12 Mbps is the “best bitrate for Instagram Reels,” there's an important part missing.

The same bitrate can produce very different quality depending on the video.

That doesn't mean bitrate doesn't matter. It means there isn't one number that guarantees the same result for everything you upload.

Does noise affect Instagram video quality?

I also wanted to test something that isn't normally mentioned in Instagram export guides: sensor noise. I had a high-ISO S-Log3 shot with quite a lot of visible noise. I created an original version, a moderately denoised version and another version with added film grain. Then I compressed them using the same settings.

• The original scored 58.9 VMAF.
• The denoised version improved to 62.9.
• The version with film grain dropped to 43.9.

The idea here is fairly simple.

Noise creates random changing detail, and the encoder has to spend data trying to describe that detail.

This doesn't mean you should apply very strong noise reduction before uploading every Instagram Reel. Bad denoising can destroy textures and make skin look unnatural. But if your footage already has unwanted sensor noise, reducing some of it can also make the video easier to compress. Film grain is different because you may want it creatively. I use grain myself. Just remember that there can be a compression cost.

1080p vs 4K for Instagram Reels

Should you export Instagram Reels in 1080p or 4K?

I originally wanted to answer this by uploading the same Reel in 1080p and 4K and comparing them. After learning more about Meta's processing pipeline, I don't think one playback comparison can prove very much. If Instagram creates multiple versions of each upload and can deliver different versions to different viewers, looking at one phone doesn't tell us the whole story.

My approach now is simple. If my project is really 1080p, I export 1080x1920. If I genuinely shot, edited and finished it in 4K, I normally upload the 2160x3840 master. That does not mean I'm saying Instagram shows everyone a 4K Reel. I don't have evidence for that.

I simply prefer giving Instagram the better source instead of throwing away information before the platform even receives it.

How to upload high quality Instagram Reels

My approach now is different from trying to find a secret Instagram preset. I start with a clean source. I don't intentionally compress it down to Instagram's assumed bitrate before uploading. Instagram is going to process it anyway. If there is unwanted sensor noise, I may use moderate noise reduction. I'm careful with heavy grain and unnecessary extra detail because those things can make compression harder.

And inside Instagram I keep Upload at Highest Quality enabled. There are still things outside our control after that.

Instagram decides how the video is processed and what version is delivered to a viewer.

That's why two people can sometimes report different quality from the same Reel.

Why are my Instagram Reels blurry after uploading?

There isn't one possible cause. The source video may already be too compressed. Detailed or noisy footage may be more difficult to encode. Instagram will also process the uploaded video and create its own versions for playback. Network and playback conditions can affect which version is delivered as well. This is why increasing your export bitrate doesn't automatically fix every blurry Instagram Reel.

The goal is to avoid giving Instagram a bad source file in the first place.

Is H.264 good for Instagram Reels?

Yes. I currently use H.264 for my Instagram Reel exports because it is reliable, widely supported and gives me good results.

Instagram itself may use different codecs later when it processes and delivers the video. Meta has publicly discussed AVC/H.264, VP9 and AV1 in its video infrastructure. What Instagram eventually sends to a viewer doesn't need to use the same codec you uploaded.

Should you use 30 fps or 60 fps for Instagram Reels?

This one deserves its own test.

I created a real 59.94 fps Reel where every frame had its own number and then checked what different devices and Instagram accounts actually received. The result was much more complicated than simply saying “Instagram supports 60 fps” or “Instagram converts everything to 30 fps.” I'm covering that separately because there's enough there for a complete test on its own.

What are the best Instagram Reel export settings?

There are good starting settings. But I don't think there is one perfect Instagram export preset. For most of my 1080p Reels, I use 1080x1920, H.264, Rec.709, AAC 48 kHz and around 15-20 Mbps.

More important than finding one magic bitrate is understanding what happens before and after that export. A simple shot and a very detailed shot can react very differently to compression. Sensor noise and film grain can make encoding harder. And after you upload the file, Instagram creates its own versions and decides what to deliver to viewers.

So my goal is no longer to reverse-engineer Instagram's exact compression settings. I give Instagram a clean, high-quality source and let its own system make the compromises it needs.

If you want to see the actual moving examples, VMAF graphs, DaVinci Resolve tests and the differences between the compressed files, watch the video embedded above.

Sources and further reading

This research is based on Meta's own engineering documentation, Instagram documentation and my own compression tests.

• Meta Engineering — How Meta brought AV1 to Reels
• Meta Engineering — Reducing Instagram's basic video compute time by 94%
• Meta Engineering — FFmpeg at Meta: Media Processing at Scale
• Netflix — VMAF