What Sample Rate Should You Record At? Video, Music, Podcasts and More

[@HyperSuperBothi]

The picture at 3:52 is extremely misleading. Digitally sampled audio is not connecting samples as dots with straight lines. This picture gives the idea for the average that you have to sample a sine reaaally high sample rate to represent it accurately. Digital audio DOES NOT alias like that. It's important to emphasize that you can capture a freq PERFECTLY with a sample rate higher than said frequency (low pass limitations apply), and while natural sound is not a sine wave, it is composed of many and often-changing sine waves. Which means there is NO AUDIBLE DIFFERENCE between 48k and 96k recordings (excluding DAC hiccups, latency, in normal speed and in the human hearing range).

[@PW-qy8mf]

Hi Alex! Thanks for all the effort you put in to produce such amazing and helpful videos! I‘ve been missing your videos lately... Are you okay? (only busy, I‘d hope...) Cheers, Matthias

[@rhoten8039]

I’ve never thought about the pitch down concept. Thank you

[@jonathanhall5670]

Dude absolutely amazing content!!! The amount of knowledge/wisdom you have of all things sound and the ability to clearly convey that to others is priceless. We truly appreciate it.

[@jogeirstrand1]

Very useful info. Didn’t ever think about the side effects of pitching down the audio. Great video again, Alex! 😀👍

[@soundvoyager457]

Great and useful representation! Yo people, let's give this channel some hype! As of now, I think this is the most explanatory video regarding sample rate haha :P

[@nicolasmak2508]

Thank you very much ! The pitch down part is my favourite

[@savethefails]

see you still hanging onto that 208 ;) Such a great mic that is under the radar. I've been using mine for pro VO and it pays bills without any questions from clients. Solid. My 103/416 combo are just sitting in a box.

[@axk]

@lautenaudio really got it right! Love the LS208, it’s an awesome mic.

[@TheREAPERBlog]

The images around 4:00 are a bit misleading. A pure sine of low frequency doesn’t need more samples to capture accurately. But real life sounds are not pure sines. Camera frame rate is a bad analogy since audio is not stuttering with a low sample rate, it is just darker.

[@tisbonus]

I'm glad you said it. I was vacillating and dubious at the same time listing to the "only the peak and trough" statement. Rate is rate. Sine is completely recreated up to 22.05K. Above that and aliasing becomes a problem but not apparent. Love the "real life sounds are not pure sines" quote!

[@highfivepictures7310]

Great video, thank you Alex!

[@morten1]

I have had it at 44.1 khz and 24bit for years now. Very happy. Set it and forghetti

[@rosskinghorn]

Man you're stuff's so helpful, thank you!

[@axk]

Glad the info is helping!

[@PerEng2405]

Brilliantly described. Much appreciated.

[@brycefunk]

Fantastic explanation! Thanks, Alex!

[@RodolfoVanzini]

Hi Alex, when I first watched this video a few days ago your subscriber number was 44.1 k. Anything to do with that sample rate?! 😅 Great video as always. Thank you.

[@axk]

I’m trying to get to 48k so I can post videos at a higher frequency!!

[@ClassicCyclingCC]

Thanks for the Video. Very interesting and clearly explained.

[@ChrisSpiegl]

I don't do crazy sound design, but this gave me a better understanding as to why one may want or need to record over 48khz. Thank you 🌸.

[@geckocheese]

your videos are so cool keep it up please i love these

[@GCGFilmsYT]

Perfectly explained. Thank you for the knowledge! Currently recording my stuff at 128-bit 48kHz and the compatibility of the frame rate when using 24fps or 23.976.

[@filnn]

Great, informative video - Just like the recent Sennheiser video hosted by Alex! I sense a running theme here...

[@axk]

Haha, I heard somewhere that consistency might be important… 😂

[@ralphbullis40]

Great as usual.

[@AndrewKnightMIT]

great explanation!

[@RobyDavis]

Not confusing at all, this was perfect... I was able to kind of map the ideas from sound to video, not one to one but there is enough overlap to make it make sense. Your example of frame rates helped frame the ideas, especially when you introduce slowing video or audio down in post. Slowing down 24fps can look like crap vs 60 of 120, so hearing the difference between a higher sample rate helped a lot. Would you say sample rate is to FPS, as frequency is to light.. or is that taking it too far?

[@Steveoaudioandstuff]

Brilliant!

[@ShinilPayamal]

Thanks!

[@hashirama_-senju]

Thank you.

[@ali32]

really appreciated, thanks a lot

[@kristicifci5389]

Hi Alex we miss your videos

[@AndreasYoutube-z1u]

I love how low your audio is on a audio channel, can´t hear a thing without blastig my speakers haha..

[@michaelanctil4972]

Hey! I really enjoy the videos you make. I tried to buy sound effects from your site, but it fails every time I add a sfx library to the cart.

[@dqstudios]

I’m a fan and subbed, but you need to be consistent - in your 32-bit float video you argued it wasn’t potentially the best option for recording since delivery standard is still 24-bit. In this video you remind us of your Source Quality Rule where you suggest recording at 96khz even though delivery standard is 48khz. 32-bit float recording has a workflow problem, but is still the best format to record in if you take your own advice or when dynamic range can be large or unpredictable.

[@HairSuitGentleman]

Alex, would you record in 44.1khz if you know your final output (e.g. a podcast) will use that sample rate? I understand that the quality of anti-aliasing in many DAWs and plugins is not great. Would the nyquist distortion from downsampling from 48 to 44.1 outweigh oversampling benefits?

[@aseomg]

Good explanation here, but the confusion of this topic seems related to software when working with audio. As an example; recording audio to a hardware recording device with a high sample rate (above 48kHz) for certain reasons, as mentioned in the video...is straightforward. But on the software side; like recording or editing in a DAW or any digital audio software using high sample rates, then it's not straightforward. Because people have different workflows, editing requirements, using various audio plugins, or audio that's produce entirely in the digital domain, etc. etc. etc., and that's where the confusion seems to happen around sample rates. With audio hardware; we record. With audio software; we can record, create & edit.

[@stevieelder2422]

Excellent tutorial Alex. Learning a lot from your videos. Thanks. Whats wrong with the website link i cant seem to get to it. Just comes up with error.

[@zeoverlord]

Honestly Nyquist theory only makes sense if you can record the velocity vector of each sample as well as amplitude, without it if you sample a sine wave at its Nyquist frequency you might end up sampling the mid levels in every sample. better be safe and add more samples if its important to capture certain frequencies.

[@radiozelaza]

to capture audio at 96kHz and more, your microphone has to have that capability. Few have. So there's no benefit and no frequencies beyond 20kHz when you pitch down.

[@rodprod8522]

Is any "Hollywood" dialogue recorded at 96khz? Or do they just stick to 48khz for convention an delivery?

[@axk]

Totally depends! I’ve never seen production/location sound recorded in 96kHz at the “major” level unless there was a specific request to do so (there rarely is). Now, if the intent is to record sound effects, or dialog for video games, or music…that’s a different story. But production audio for film/tv/video is almost always going to be 48kHz.

[@rodprod8522]

@@axk thank you!

[@meck10101]

On the other hand lower sampling rates introduce less unwanted sounds and disturbances

[@radiozelaza]

find me an adult who can hear 20 000 Hz

[@MarkKelly2262]

Good video, One little thing you did not mention is the higher bit rate like 192k has significantly less latency.

[@Energine1]

Alex, great show but I have to ask. Are you ok its been a while by the looks of things...