Jump to content

Is it ever ok to use clipped sound?


Recommended Posts

I'm working on a project that includes a cut of a muscle car screaming through a tunnel. We have several sound recordings to choose from. Some are just under 0db, and some are clipped. Honestly, it's the clipped ones that sound the best to me. I don't hear distortion, just BIG reverberating sound. And since playback will mostly be on mobile phones and computers, I want to make sure the sound IS big (it's a video for an automotive exhaust company, and sound is a big part of it). So my question is, is it ever ok to use a clipped sound? Recently I asked an audio engineer if the human hearing system has a limit and clips sound. He felt this was true. I was in that tunnel recording that car, and for me, the sound was so big and deafening that the only way I can reproduce the excitement of that moment in a video is to use one of the clipped recordings. I also suspect, without any proof, that if I were to analyze the soundtrack of a blockbuster movie DVD, I just might find parts here and there that are clipped. But I'm no expert and I welcome the opinions of those here. Appreciate your input. 

Link to post
Share on other sites

If it sounds good, it sounds good. Hard to know without hearing it but if you don't hear that ugly clipping/popping sound you sometimes get with digital clipping, it might be fine. 

 

Personally, if I had a clipped piece of audio like that, I would try to declip it with something like iZotope RX4, and if required then try and add crunchiness in other ways (like plugins that introduce simulated analogue tape overdrive etc).

Link to post
Share on other sites

If you don't have the tools as described above, then I'd go with the first bit of advice. If it sounds the way you want it to sound, then it's not "wrong". But make sure in the distribution of the clip that compression doesn't further distort to the clip to a point that you're unhappy.

On my first union feature, I was using a Mackie which didn't hold onto some of the shouts and screams (horror movie). I was recording into 744T, and they didn't "clip", it was just a bit crunchy from the board. I kinda liked it. So I asked the director (also a successful editor) who said he loved it too. The minor distortion added character to the shouts and screams in the way a voice might distort naturally. Although I could have adjusted gain settings to protect for the shouting and screaming a bit better, we left it as is. Not "correct" necessarily, but our choice. The movie was mixed by academy award winning RRM who also loved the sound as a whole.

To quote JW's quote from Ray Charles, "How does it sound, baby?"

Link to post
Share on other sites

I don't see why it would be a problem. If it sounds good, use it.

People are happy to watch crappy pictures shot on dinky cams, phones, CCTV cameras so why would a little bit of distortion on car sound effects be an issue. I imagine it gives a throaty sound to the exhaust which a lot of cars do naturally anyway.

If its on dialogue, then it's a different story.

Regards

Link to post
Share on other sites

Clipping is just another type of compression. People use compressors aesthetically all the time (guitar distortion for example).

 

The common wisdom is that digital clipping is less pleasant than limiters or overloading a preamp. This may be because digital clipping adds odd harmonics to the original signal whereas other causes of compression tend to add more even harmonics. There are a few other parameters to compression as well such as slew rate, attack and decay times, and frequency response. They all change the sound in different ways. They're all technically distortion of the original signal, but depending on how and when you use them, they can sound pretty cool. Next time you may want to try overloading your preamp instead and see which sound you prefer.

 

There is one other consideration which is that if you record a distorted signal instead of the original, you can't get the original back (though you may be able to hide the offending material). You can always distort the original signal later. Sometimes though, you know that they're not going to spend the time trying out different types of compression, and you can reasonably decide to add your aesthetic to the final product.

Link to post
Share on other sites

Thanks everyone, appreciate all the feedback & advice. My ears aren't exactly what they used to be, but I don't think I detect any pops, clicks, etc., just really loud "car music" that sounds like a guitar on distortion. But that said, I'm going to do further research based on the suggestions here. Thank you.

Link to post
Share on other sites

Once you've mixed it -- assuming you're not mixing all the way up to 0 dBFS -- it's not clipped any more. There might be flat-topped waves and vertical rises with infinite harmonics, but that's now part of the sound

 

Since you're mixing for phones and other portables, you definitely don't want your mix to top out at 0 dB. That can cause unpredictable results, particularly with low voltage battery devices. 

 

And since this will have large distribution OTT, there'll be plenty of compression that you can't predict. All those flat waves and vertical rises can wreak momentary havoc with the compression algorithm, stealing bits from the midrange... even if the mix level of your now-clipped waveforms is controlled. So be careful if there's dialog (or music) during any of these screaming fx. 

 

But other than that... mix for the track, mix at a reasonable level, and make sure there's a good spread of volume levels over the course of the scene. Don't worry about specific clipped sources: they're already clipped, and that distortion is baked into the sound. 

Link to post
Share on other sites

When there us a gun shot in a scene or a very fast door slam, I usually let it clip. I even switch all limiters off because to me (and nobody ever complained) a very short clipping sounds more natural than a limiter pumping.

 

"Longer" sounds like an airplane launching or a scream are something different, you would hear the clipping, but "it depends".

Link to post
Share on other sites

I still think recording dialog at -20dB (as opposed to say -40dB) is an artifact of old technology that no one questions because that's all anyone's ever known. Today's audio systems get well over 100dB dynamic range. Much more than anyone can comfortably listen to.

http://www.audiocheck.net/audiotests_dynamiccheck.php

Who's recording dialogue at -20?
Link to post
Share on other sites

I always thought of -20 dBFS as a reference/calibration level)  As an absolute program peak, that would be low in most PSM's book. With program peaks @ -40 dBFS, I'd be receiving threatening phone calls (with numerous expletives) and likely never work again.

Link to post
Share on other sites

I don't mean -20/-40dB peaks. That would be silly. I mean baseline level, where your dialog should at least hit. Is that not how people think about levels?

Personally I just much prefer low levels over the sound of compression and limiters.

Link to post
Share on other sites
  • 2 weeks later...

Here's an example of what I'm talking about. At the high end of the revs around the 5:18 mark, the exhaust sound gets a bit clipped. Past experience has been that if I use the limiter, it can sound almost corrupted or something — you can tell the limiter is struggling with it. If I don't let it clip, the 744t doesn't capture the meat of the mid-frequencies as the rev winds down. This is about what I consider just right. But I'm completely open to disagreeing opinions:

 

 

Link to post
Share on other sites

My argument would be that the mid-frequencies aren't 'captured' by the clipping, they are merely emphasized by it (as fast attack compression tends to do). It is possible to duplicate that effect (even by reproducing the clipping) were it recorded without clipping.

Link to post
Share on other sites

My argument would be that the mid-frequencies aren't 'captured' by the clipping, they are merely emphasized by it (as fast attack compression tends to do). It is possible to duplicate that effect (even by reproducing the clipping) were it recorded without clipping.

​Thanks, can you please elaborate on that? Are you saying I should record without clipping then boost the sound in post? When I've tried that before, the texture of the rumbling roar in the mid-freqs seems to lose impact. Sorry but I'm just hacking at this and trying to learn.

Link to post
Share on other sites

Two possibilities with your experience with post:
- DAW software has a much higher dynamic range to work with and may not clip the sound depending on what kind of gain you choose

- When clipping your ADC, you're also overdriving your preamp which adds its own limiting effect

Clearly you can get the effect you want by clipping while recording, so if you and everyone else in the chain like it, then use it. The disadvantage is that you can't get back the undistorted audio after the fact. 

Recorded cleanly, it may take more experimentation with effects than you'd like to spend time on. But you could try different types of compression/limiters with fast attack/release times, high ratio, and a hard knee, or if you have them, there are overdrive/distortion effects that you can add that would give you a similar effect.

Link to post
Share on other sites

One option with this kind of thing could be to record one channel with clipping for whatever that brings to a certain part of the effect while having a clean(er) track for capturing a different part of the effect...mixing between the tracks in post to get the best of both worlds as desired. You might come up with something even better in experimenting, combining. Kind of a bastardized HDR process for SFX capture/creation. I would be curious what you could get here with a very slight "doubling" of the track..very slight. Sometimes you can get a larger sound that doesn't have a blatant slapback or phase.

Personally, I don't care for the clipping in the example...but I haven't heard the alternative track..

Link to post
Share on other sites

One option with this kind of thing could be to record one channel with clipping for whatever that brings to a certain part of the effect while having a clean(er) track for capturing a different part of the effect...mixing between the tracks in post to get the best of both worlds as desired. You might come up with something even better in experimenting, combining. Kind of a bastardized HDR process for SFX capture/creation. I would be curious what you could get here with a very slight "doubling" of the track..very slight. Sometimes you can get a larger sound that doesn't have a blatant slapback or phase.

Personally, I don't care for the clipping in the example...but I haven't heard the alternative track..

​Thanks Rich, yes actually of the two mics in the example I posted, one is set higher level than the other, and I do exactly what you describe above in post. But to take it step further, I need to find out if the 744t can make two recordings from the same mic, with one set at a lower db level. I've been using this feature on the Tascam DR60D to record "drive-by" exhaust sounds and really like it.

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...