Take the 2-minute tour ×
Sound Design Stack Exchange is a question and answer site for sound engineers, producers, editors, and enthusiasts. It's 100% free, no registration required.

In my opinion, I hear a drop in quality going from 96K to 48K.K.

When do you think it will be standard to mix a film at 96K or possibly 192K and have it stay at that sample-rate?

What are your guys' thoughts on 48K as opposed to 96K? Am I alone when I say I can hear a difference in quality between 48K and 96K?

Also, when you record dialogue or effects or foley, do you record straight to 48K or do you record at 192 or 96 and then downconvert? And how do you do the downconverting?

share|improve this question
"I hear a drop in quality going from 96K to 48K." In a double-blind test, with absolutely no other difference between the two mixes? –  endolith Jun 22 '10 at 11:17
on a film dub stage with the X Curve? –  user49 Jun 23 '10 at 7:31

8 Answers 8

Actually, I don't feel the need of 96kHz mixes. Something that sometime I really need (and it's already here with digital cinema) is uncompressed audio. More than one time I found the (not so bad, for the time when it was born) Dolby data compression (the ac3 like compression that the DMU do for the MODisk) to be really annoying and changing a lot the sound. But in the end it really depends from the kind of movie; and actually, the real big deal is the sound system of cinema theater: 8 times out of 10, the cinema has wrong level, or outdated ampli and speakers, or wrong sound correction. Just my 2cent

share|improve this answer
True true. I guess it's hopeless! =-( –  Utopia Jun 22 '10 at 17:54

96 kHz and 32-bit floating point aren't for the final mix. They're for intermediate processing. In the final mix you can't hear a difference between 48 kHz and 96 kHz (unless you're Batman or the mix was downsampled poorly or the 96 kHz mix contains ultrasound that's distorting in your speaker). Likewise, 20-bit is adequate to represent all of the 120 dB dynamic range you can ever get out of a real system (blame physics).

BUT, when putting a waveform through a lot of processing in a DAW, you want lots of headroom in both frequency and amplitude, to prevent unnecessary computation noise, aliasing, etc. Like photocopying a photocopy, the better the quality of each copy, the less degradation there will be in the final product.

share|improve this answer
@endolith, well said. –  Jay Jennings Jan 4 '11 at 7:27

Good question, i'd like to A-B some 96k vs 48k myself.

Although from a theoretical point of view, a sample rate of 96kHz means the signal can carry frequencies up to 48kHz (http://en.wikipedia.org/wiki/Nyquist_rate), and we can't really hear above 20kHz (most of us are lucky to hear up to 17kHz!). Whereas a sample rate of 48kHz can, theoretically, carry signal up to 24kHz, which is still beyond our range of hearing. Also, i don't think any widely used speakers are capable of reproducing any frequencies over 20kHz.

As to whether a higher sample rate affects the frequencies we can hear, i can't say. It may be possible, but i think the benefits to a typical audience would be negligible.

On the upside, FX recorded at high sample rates won't lose all their high frequencies when they're pitched down! (as far as i know...)

share|improve this answer
That's true. I suppose it is adequate for theater speakers/TV systems... –  Utopia Jun 22 '10 at 17:55
Someone mentioned Nyquist! Awesome. This would probably be the best reason why. Also, after the age of around 30, we significantly lose a lot of our high end frequency reception. So, I don't think I'd appreciate the difference much anyay –  Hubert Campbell Jun 23 '10 at 1:02
Wow. I'm 25 and I take extra special care of my ears. It's an age thing? Really? You just scared me... –  Utopia Jun 23 '10 at 2:34
Yeah, the little hairs that turn vibrations into neural signals get a bit beaten up as we get older. I'm 27 and can hear 17kHz, which is meant to be pretty good. By all means take care of your ears, and i guess take comfort in the fact that as you slowly lose the higher frequencies, your knowledge and critical listening gets better! LIke i say; a 5 year old can hear a lot better than i can, but you don't see any 5 year old mixers... –  Roger Middenway Jun 23 '10 at 4:21
Ya, nothing you can do to prevent the loss in upper frequencies. Every human being will suffer this. There was an article on CNN a few months back about high school students who had their cell phone ring at a very high frequency that their teachers cannot hear. This is the classic example of that high frequency loss. Adults can't hear their students' cell phones ring but, of course, they can. –  Hubert Campbell Jun 24 '10 at 15:09

"I'm a total stickler when it comes to quality."


you must have noticed the X Curve in every screening you have ever been to then?

and if thats such a worry do you also notice the sync difference between the front row and the back row in a theatre?

share|improve this answer
@Ryan why do you equate 'poor audio work' with a technical spec? My point is; If you so 'hate' films mixed at 48k where is the evidence in the actual final film soundtracks? Obsessing over hearing the subtle difference between 96k and 48k in your edit room on near field speakers is maybe not the best use of your time & energy... –  user49 Jun 27 '10 at 1:27

I'd say its pretty impossible in the near term and pretty unlikely in the long term.

First off, sample rate depth has serious consequences for the way that DAW and digitial processing work, and mixing is a processor intense activity. In order for a process like an eq or summing mixer or distortion to work correctly, it has to do its process at double or 4x the rate it would at 48k. Ever actually try mixing something complex at 192k? Give it a spin and let us know how your computer holds up.

Also, have you done a double-blind listening test between 48k and 96k? Are you sure you can hear a difference? I can certainly set one up for you. What if I use 48k source sample rate material recorded outdoors and then encode each into AAC? Because another issue is the production process: Most audio for films is not captured at super high sample rates. Its captured at 48k, so unless you're asking your location sound dept to run at 192k then you're just upsampling the dialogue and pfx anyway.

Actually, that first point is the biggest one. You're crippling your computing power and you are doing it for what will amount to zero sonic benefit. This is doubly true when you consider that location source audio is captured at 48k, and the final distribution master is encoded in a lossy format.

In the end quality is about content. Mixing decisions are about story, not sample rates. I'm not trying to be harsh, I'm just trying to steer you away from the placebo mentality that makes Monster Cables so much money. Trust your ears, but only when your eyes aren't involved.

share|improve this answer
@Rene, you are my kindred spirit! Well said, indeed. –  Jay Jennings Jan 6 '11 at 17:28
cool man, but have you tried a complex mix at that sample rate? How quickly do you max out dsp there? Just asking. I know my rig can get chokey at the higher rates when I try complex processing. I also know that I can't lock to black burst very well at those rates. The Sync I/O really kind of hates it. I'd like to hear about your listening tests though... –  Rene Jan 6 '11 at 18:25
Wow. I must have struck a chord here. Thanks for the answer, but I disagree on the part about forwarding and upgrading our profession. Short answer is: Computing power is not an issue for me. All my dialogue/VO is shot at 96K, which is where I think we hit head-on because I do see your point that upsampled 48K has no point, but I've always started and stayed at 96 the whole way through. True, I've never worked a large dub-stage at a top film studio but what I've worked out and set up for myself works well and keeps my clients happy. –  Utopia Jan 6 '11 at 18:25
Depends but my music mixes get pretty complex - nearly always maxes the voices and I use quite a bit of instances of Oxford and Waves plug-ins. The most complex mixing though was for a gigantic multicast audiobook I did in 2008 - way more involved than any film I've done. What about the tests would you like to know? –  Utopia Jan 6 '11 at 18:33
A lot of the viability of test results depend on the methodology and techniques applied when setting them up. For example, did you do the tests as double blind? Were just testing to identify differences or were you testing for aesthetic? Did you include others in the test? What was the nature of the source audio and the recording of said source audio? Were the lower res samples simply downsampled or were the recorded at a the lower sample rate but using the same mic and performance? etc. –  Rene Jan 6 '11 at 18:53

Some Blu-Ray titles are offered at 24/96. So far, they mainly seem to be live concert recordings.

Given that major movie theaters are playing back basically mp3 quality audio, and that the film festival circuit is even worse (Sundance only supports Stereo LoRo or LtRt encoding for a phasey LCRS mix!), I have resigned myself to the fact that 48kHz is just the standard.

share|improve this answer
Film festivals such as Sundance play DCP which is up to 24/96 for 16 discreet channels. It sounds like you are talking about SR (matrixed LCRS) and SR-D (AC-3) tracks on film, and film fests and first run theatres don't play film any more. A lot has changed in the 3 years since this post. –  Evan Jerred Jan 24 '14 at 1:42

If i can I will always record at 192 kHz, and I can still hear 17 kHz despite being considerably older than 27. I have students who are considerably younger than 27 who struggle to hear 13 kHz.

The higher sampling rate does make a difference when it comes to reverberation, as it works in a similar manner to an LFO (Low Frequency Oscillator), so inaudible frequencies affect audible frequencies.

However, all of the cinemas I ever go to are lucky if they have the projector focused regularly, never mind having the audio properly calibrated.

We will move to 192 kHz eventually, but there is so much that has to catch up first.

The first thing that I would love is a loudspeaker with a flat frequency response even up to 17 kHz, even my favourite Genelecs colour the sound.

share|improve this answer

In my subjective tests, I can't hear much/any difference between 96 and 48k. 24 and 16 bit however is a different story.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.