Quantized Drums

thegummy

I have heard so many times from so many places that quantized drums don't sound as good to people as drums that aren't exactly in time, such was historically true with human drummers and is the intention of "humanize" functions on sequencers.

I have always questioned this and think it's theoretical or even a Ludd-esque response from drummers. I don't know if it's true that people listening to music would notice on any level, even subconsciously, if the drums are perfectly in time (particularly when the rest of the instruments are played live and not quantized).

The only bit of science I've seen on the topic was a study involving music students listening to, and rating, clips and the result was actually that the more quantized the clip was, the more groove it was rated as having.

So for anyone who does believe in the idea of perfectly timed drums being worse - are there any examples of records you can give that suffer because of the drums being strictly in time? I don't mean sequenced synth music like Kraftwerk or minimal techno, I mean a rock or pop record where the drums have been rigidly programmed (or quantized after the fact).

Just to be clear - I'm not talking about when drummers deliberately rush or drag a specific section of the song, I mean just the natural imperfection of the human body when they're trying to stay in time as much as possible; i.e. the thing that the "humanize" function is made to simulate.

mixolyd

Rick Beato has done more than one vid on this topic, here’s the latest.

https://youtu.be/orJggsV9JHM

spaceman_inorbit

I've just watched Rick Beato's youtube vid on quantizing John Bonham -  How would John Bonham sound today https://www.youtube.com/watch?v=hT4fFolyZYU&t=328s and How Computers Ruined Rock Music, https://www.youtube.com/watch?v=AFaRIW-wZlw . It makes for interesting viewing.

thegummy

Can't stand that guy and it doesn't surprise me in the slightest he believes in it lol

Roland

I believe in it. Not so much timing as attack, tone and decay changing from verse to chorus, and from bar to bar. After all, we do similar things on guitar.

thegummy said:

The only bit of science I've seen on the topic was a study involving music students listening to, and rating, clips and the result was actually that the more quantized the clip was, the more groove it was rated as having.

Reactions are going to depend on what people are used to hearing. A student age sample is more likely to have been immersed in quantised music than you or I. There was similar research a few years ago about sound quality. Those who’d grown up with mp3s preferred them to CDs or vinyl, and vice versa.

bbill335

oh yeah, that's why everyone regards over 30years of dance/electronica with sequenced drums to be unlistenable. 

it's just more rockism and the arguments have never got any better than “disco sucks”

Roland

@bbill335 Were you commenting on me or Beato (which I haven’t got time to watch)?

LastMantra

The basic humanisers that I've seen aren't very natural imo. Just randomly having one hit before the click, one after, isn't what a real drummer does usually. 

Maybe there are better ones these days?

thegummy

bbill335 said:

oh yeah, that's why everyone regards over 30years of dance/electronica with sequenced drums to be unlistenable. 

it's just more rockism and the arguments have never got any better than “disco sucks”

For the reason you say there's absolutely no doubt that people do like listening to rigidly sequenced music in its own right.

But what I'm interested in is if you take a rock song and secretly quantize the drums on it - would the listeners (not necessarily just Joe Public, musicians also) hear it as being distinctly different and sound like it was slightly crossing over in to electronica sounding music, or would they not even notice there being a difference (or even find it better subconsciously)?

Conversely, if the song was made with programmed drums and everything else played live as normal then a second version was printed after the "humanize" function was performed on the drums, would people hear that differently?

When it's a video (like above) of someone trying to prove their point then it doesn't really say much to me, I'm interested if anyone actually knows of a song that either sounds bad or just sounds different to most other band-based songs purely because the drums are perfectly in time.

steven70

thegummy said:

 I'm interested if anyone actually knows of a song that either sounds bad or just sounds different to most other band-based songs purely because the drums are perfectly in time.

I think you're gonna be hard put to find an example because songs which have quantized drums most likely have any number of other things also done to them which make the song sound different.

You'd have to take a song with non electronic instruments from the pre digital era and quantize the drums. 
Then you could do some A/B/ tests Be interesting.

(OK, is that what the chap in the video did? I  didn't watch cos I don't like the look of him, seems like a TV evangelist)

I suspect quantizing drums is the least of it when it comes to the sound of most music made post 2000...moan, moan, I'm an old man moan, but true.

bbill335

Roland said:

@bbill335 Were you commenting on me or Beato (which I haven’t got time to watch)?

just responding to the notion that “ quantized drums don't sound as good to people as drums that aren't exactly in time”

thegummy

steven70 said:

thegummy said:

 I'm interested if anyone actually knows of a song that either sounds bad or just sounds different to most other band-based songs purely because the drums are perfectly in time.

I think you're gonna be hard put to find an example because songs which have quantized drums most likely have any number of other things also done to them which make the song sound different.

You'd have to take a song with non electronic instruments from the pre digital era and quantize the drums. 
Then you could do some A/B/ tests Be interesting.

(OK, is that what the chap in the video did? I  didn't watch cos I don't like the look of him, seems like a TV evangelist)

I suspect quantizing drums is the least of it when it comes to the sound of most music made post 2000...moan, moan, I'm an old man moan, but true.

I'm not talking about comparing old music to new music in any other way, just the drum quantization thing.

steven70

thegummy said:

steven70 said:

thegummy said:

 I'm interested if anyone actually knows of a song that either sounds bad or just sounds different to most other band-based songs purely because the drums are perfectly in time.

I think you're gonna be hard put to find an example because songs which have quantized drums most likely have any number of other things also done to them which make the song sound different.

You'd have to take a song with non electronic instruments from the pre digital era and quantize the drums. 
Then you could do some A/B/ tests Be interesting.

(OK, is that what the chap in the video did? I  didn't watch cos I don't like the look of him, seems like a TV evangelist)

I suspect quantizing drums is the least of it when it comes to the sound of most music made post 2000...moan, moan, I'm an old man moan, but true.

I'm not talking about comparing old music to new music in any other way, just the drum quantization thing.

Yeah, but how do you intend to do that? How would it be possible to do an A/B using just that criteria for comparison or just isolate that single thing as the cause of whether a song is good or bad. Reckon it's going to get subjective. Just interested likes...cos its a good point.

PolarityMan

Roland said:

I believe in it. Not so much timing as attack, tone and decay changing from verse to chorus, and from bar to bar. After all, we do similar things on guitar.

thegummy said:

The only bit of science I've seen on the topic was a study involving music students listening to, and rating, clips and the result was actually that the more quantized the clip was, the more groove it was rated as having.

Reactions are going to depend on what people are used to hearing. A student age sample is more likely to have been immersed in quantised music than you or I. There was similar research a few years ago about sound quality. Those who’d grown up with mp3s preferred them to CDs or vinyl, and vice versa.

The modern approach is to grid existing audio so that you end up with all those human variations in attack and hit but not the timing variations.

PolarityMan

steven70 said:

thegummy said:

 I'm interested if anyone actually knows of a song that either sounds bad or just sounds different to most other band-based songs purely because the drums are perfectly in time.

I think you're gonna be hard put to find an example because songs which have quantized drums most likely have any number of other things also done to them which make the song sound different.

You'd have to take a song with non electronic instruments from the pre digital era and quantize the drums. 
Then you could do some A/B/ tests Be interesting.

(OK, is that what the chap in the video did? I  didn't watch cos I don't like the look of him, seems like a TV evangelist)

I suspect quantizing drums is the least of it when it comes to the sound of most music made post 2000...moan, moan, I'm an old man moan, but true.

You cant really do that as the other instruments will compensate for the shitty timed drums in the original. What you would have to do is record a drum track, record all the instruments around it, then grid the drums and re-record the other instruments then A/B the 2.

PolarityMan

Oh and Beato also did a pretty poor job of the quantizing too. there was already a thread on that exact vid some time ago 

thegummy

steven70 said:

Yeah, but how do you intend to do that? How would it be possible to do an A/B using just that criteria for comparison or just isolate that single thing as the cause of whether a song is good or bad. Reckon it's going to get subjective. Just interested likes...

Your suggestion would only be limiting if there was a certain point in history where before that point no drums were quantized and after that point all drums were quantized as well as other changes.

But that's not the case so it would just be someone pointing out track X by band Y sounds like whatever because it has perfect drum timings. Then it could be compared to a similar band who don't have quantized drums.

Not really sure why you're asking that unless you do think that after a certain point all drums are quantized?

thegummy

PolarityMan said:

Oh and Beato also did a pretty poor job of the quantizing too. there was already a thread on that exact vid some time ago 

I wouldn't expect him to be setting up a test to learn from, I'd be expecting him to put together something that proved what he thought was right.

That's the main reason I can't stand him, he seems to me to be the type who thinks their way is the right way and better than any other way.

@PolarityMan I can tell from your posts here that you totally get the idea of what I mean (not saying other repliers don't).

The thing is, if people who think that quantizing drums is a bad thing can't actually name examples of it being bad then how did they come to that decision in the first place, just based on theoretical thinking? Maybe it started when drum machines came out and drummers were thinking up reasons for them to be inferior rather than trying them with good intentions but then finding a problem naturally?

PolarityMan

104% of modern rock music is quantized anyway. I assert that the vast majority of people couldn't tell when it is done well. 

Danny1969

The biggest difference is songs that have drums put on the grid lines, like using beat detective etc are songs where the drums were done first, THEN other instruments added to the corrected drum track to build the song. This means your then hearing a song that's built up part by part ..... your not hearing a band playing together as a band and the extra energy that can come from that. Now generally things like the lead vocal and BV's are always added later but it's the energy in that original drums bass and guitar backing track that survives the multitrack process.  

You can't record a band as a live unit then correct the drums .... so the outcome in each case is a different feel

Personally I don't like dodgy drummers who's base tempo is all over the place but there's loads of great music from Kate Bush, Van Halen, Supergrass, Beatles where the tempo creeps up and down and it adds to the feel for me. But then after years of correcting timing issues in the studio I have developed a liking for bands recorded more or less live, warts and all 

octatonic

As Danny says you never track over uncorrected drums and then correct them.
The whole idea with Pro Tools and Beat Detective is to allow you to grid things up as to allow moving sections about and spot correct mistakes.
It obviously went too far and sounded robotic in certain cases but not all.

Go and listen to any Clutch record and tell me they don't groove.
You don't need to time correct a drummer like JP Gaster, but they track into Pro Tools.
You don't have to edit.

I mixed an album that Darby Todd played on a few years ago- I did not need to correct a single note because the playing was brilliant.
Was it perfectly gridded?
Not exactly, but gridding it wouldn't have made it better.

As far as humanising midi programmed drums- the best way is to learn to think like a drummer and then edit the part to suit.
I no longer use electronic kits, they all suck-
If I need to program some drums and can't record an acoustic kit then I work out my parts on the drum kit, which are recorded, transcribe the parts and input them into Logic or Pro Tools.
It is a bit more labour intensive but it yields a better result.

Or just hire a drummer to record for you- it really isn't that expensive.