Transients are your worst enemy!

[Plec]

I'm going to post a statement here that I think people will have quite different reactions to. But it would be interesting to ventilate this subject a bit since I haven't seen it around very much.

Ok, so transients are your worst enemy...
Most people usually agree that recording sounds rich in transients like drums, dynamic guitars, bass, percussion etc... sounds a lot "better" going to analog tape than to a digital multitrack. Why?

The deal with analog tape is...
Well, you throw away a big part of the transient enabling the harmonics that make up the body of your sounds to be recorded at a higher level, consequently making them more full, bigger more real and present.

Your digital signal is represented by bits and the most significant bits reside at the very top of your dynamic range. IOW the part of your recording that is represented most clearly is the fast transients of the signal and not the harmonics that make up the fabric of sound which reside in the middle of the dynamic envelope. As your signal level in digital goes down, your resolution goes down (increased distorsion) so the very resonance and body of your sounds are not represented by the most significant bits anymore.

The digital mixdown problem...
When you then get to mastering working off of a digital mixdown you are finally limiting/removing the transients that is the very part that digital represents the best, and you bring up the low level signals that sound grainy due to fact that they have been "encoded" at a lower resolution.

Limiting; will not really fix the issue since it's just compression and will squash down on the information below the transients that are being limited.

Analog Clipping (Tape Saturation); is where the answer lies since clipping cuts off the transient and doesn't push it down like compression/limiting.

Digtial Clipping; is a part of the answer since you can get away with a small amount both when tracking and mastering.

Now, if the above statement would be considered true, it means that you HAVE to clip/saturate your signal before converting to digital in order to gain the extra resolution. This would mean that you either need to record to analog tape and then transfer to digital OR use something like the (wonderful ) Empirical Labs FATSO to clip the signal before your A/D conversion. This would also mean that so many software manufacturers are missing the point when it comes to tape emulation in the digital domain.

Then there is of course the school of thought that since we're recording at 24bits and going to 16bits in the end you should be able to record your 24bit signal at -40dbFS and still end up with the same quality when dithering to 16bits.

My experience with all of this is that when I listen to my final productions I find my sounds to end up much fuller and bigger when I have tracked what I can with the FATSO, just throwing those transients out the door. When I mix down I usually make a number of different versions and the mixes that I find to sound the very best when I get to mastering (and after) are the ones where I mixed down with a couple of db compression and a little bit of FATSO clipping and maybe 1-2db digital clipping when going through the mixdown A/D.

I would love to hear everyones experiences and views on this subject. stike

[wilderness russ]

Nice. I'm sure i've heard alot of drum n bass / breakbeat producers clip the gains on their analogue desks on most things they do to achieve this effect

[Tim Farrant]

Well, it's all about how "spectacular" the sound you want. Preserving transients whilst recording to me is the utmost priority. Without them you don't have crisp snare and kick or forward acoustic guitar. If you decide to ditch them in the mix, then that's cool, but recording the transients in the first place is paramount. This is something digital is good at. So many times in the past I have got the perfect acoustic gat sound only to have it played back from the MTR90 (SR) sounding like a flat fish.

I say, praise digital, praise the analog path into it - and go make spectacular sounding records. When you speak of transients, you are also speaking of harmonics - the very backbone of music IMHO. Think square wave - a fundamental and a host of harmonics extending many octave beyond...

Cheers
Tim

[Relax]

Originally Posted by Tim Farrant

So many times in the past I have got the perfect acoustic gat sound only to have it played back from the MTR90 (SR) sounding like a flat fish.

thats gangsta

[jebjerome]

Quote:
Originally Posted by Tim Farrant
So many times in the past I have got the perfect acoustic gat sound only to have it played back from the MTR90 (SR) sounding like a flat fish.


thats gangsta

hahahahaha. nice.

[Tim Farrant]

Plec, forgive me if I offend but I also meant to ask you = have you actually used a 24 trk 2" tape machine to complete a finished record?

[headz]

Originally Posted by Plec

.... I find to sound the very best when I get to mastering (and after) are the ones where I mixed down with a couple of db compression and a little bit of FATSO clipping and maybe 1-2db digital clipping when going through the mixdown A/D.

I would love to hear everyones experiences and views on this subject. stike

really? i get the fatos clipping but you mean clipping on the A/D converter so your transients shoot 0dB???

[Jules]

Originally Posted by Plec

I'm going to post a statement here that I think people will have quite different reactions to. But it would be interesting to ventilate this subject a bit since I haven't seen it around very much.

Ok, so transients are your worst enemy...
Most people usually agree that recording sounds rich in transients like drums, dynamic guitars, bass, percussion etc... sounds a lot "better" going to analog tape than to a digital multitrack. Why?

The deal with analog tape is...
Well, you throw away a big part of the transient enabling the harmonics that make up the body of your sounds to be recorded at a higher level, consequently making them more full, bigger more real and present.

Your digital signal is represented by bits and the most significant bits reside at the very top of your dynamic range. IOW the part of your recording that is represented most clearly is the fast transients of the signal and not the harmonics that make up the fabric of sound which reside in the middle of the dynamic envelope. As your signal level in digital goes down, your resolution goes down (increased distorsion) so the very resonance and body of your sounds are not represented by the most significant bits anymore.

The digital mixdown problem...
When you then get to mastering working off of a digital mixdown you are finally limiting/removing the transients that is the very part that digital represents the best, and you bring up the low level signals that sound grainy due to fact that they have been "encoded" at a lower resolution.

Limiting; will not really fix the issue since it's just compression and will squash down on the information below the transients that are being limited.

Analog Clipping (Tape Saturation); is where the answer lies since clipping cuts off the transient and doesn't push it down like compression/limiting.

Digtial Clipping; is a part of the answer since you can get away with a small amount both when tracking and mastering.

Now, if the above statement would be considered true, it means that you HAVE to clip/saturate your signal before converting to digital in order to gain the extra resolution. This would mean that you either need to record to analog tape and then transfer to digital OR use something like the (wonderful ) Empirical Labs FATSO to clip the signal before your A/D conversion. This would also mean that so many software manufacturers are missing the point when it comes to tape emulation in the digital domain.

Then there is of course the school of thought that since we're recording at 24bits and going to 16bits in the end you should be able to record your 24bit signal at -40dbFS and still end up with the same quality when dithering to 16bits.

My experience with all of this is that when I listen to my final productions I find my sounds to end up much fuller and bigger when I have tracked what I can with the FATSO, just throwing those transients out the door. When I mix down I usually make a number of different versions and the mixes that I find to sound the very best when I get to mastering (and after) are the ones where I mixed down with a couple of db compression and a little bit of FATSO clipping and maybe 1-2db digital clipping when going through the mixdown A/D.

I would love to hear everyones experiences and views on this subject. stike

I agree - some transient crush - (I am not refering to compressions or limiting, read above) is a good thing - for pop music

The 'cold capture' of digital needs some assistance

[Fletcher]

Originally Posted by Plec

My experience with all of this is that when I listen to my final productions I find my sounds to end up much fuller and bigger when I have tracked what I can with the FATSO, just throwing those transients out the door.

Interesting.

It's been 1971 in my car for the last couple of weeks... and one of the things that has struck me about those recordings is how the dynamics of them really help the arrangement of the songs and the expression of the emotion of the performance.

In so many ways I have found that we have become hell bent on the "perfection" of the sonics often at the expense of the emotion of the performance. It's like we're using compression like Prozac... we're not letting things get too high or too low in the mix... everything is to take middle ground and be "heard"... great (sarcasism), not great (?), or is this just "modern" production techniques that will time stamp current recordings as "late 90's/early 2k" like gated reverb on drums did to the 80's and Kotex on the one headed rack toms did to the 70's?

Now I love my FATSO Jr. as much as the next piece of gear... but I did use it on overheads once and would really rather that I hadn't... it was murder trying to pull a drum sound out of the tracks where I used the FATSO Jr. on the overheads while it was a piece of cake to make everything work and play together when I didn't use the FATSO Jr. Maybe I just fukked up when I was setting up the FATSO Jr.... maybe I should have waited to apply the damn thing when I was mixing... the fact of the matter is that everytime I have ever applied any kind of dynamic processing to any kind of percussive instrument whilst recording that percussive instrument I have lived to regret the decision.

In a lot of ways we have lost our way in terms of sonic and musical arrangements... in many ways the stuff I'm hearing now isn't made a damn bit better by all the "layering" and the 'bazillion track' capability... and it certainly isn't being made to be a whole lot better by toning down the transients on things... especially on drums.

For some reason I had the bass turned up a bit higher than usual in the car the other day... for the first time I noticed that Charlie Watts was getting 'happy footed' on some of the tracks on "Exile on Main St." [you know, "announcing" a new section with a harder kik hit on the "1" of that new section]. I smiled. There is no way in hell I would have let that pass... but ya know what, it helped the damn song [especially when I bought the damn bass down on the car stereo to where it belonged... you could more "feel" the happy foot shit than hear it when the system was closer to properly balanced].

One of the things I've noticed about working in the digital domain as opposed to the analog domain is that even with 24 bit resolution it is still difficult to get the depth and width I was able to get in the analog domain. It's kinda like HDTV video vs. 35mm film. Film still looks better than even "Hi Def" video... though the two are getting closer. From my experience with my RADAR I think we're getting closer to the potential of the digital audio medium... but it's still a medium because it is indeed exceptionally rare that it's ever well done.

I like transients... I hate boring musical arrangments and pedestrian sounds that don't reinforce the musical statement / intention of the artist... and worse, artistic statements that have little or nothing to actually say. As the "technology" gets "better" there seems to be less and less "expression" [though with the power of revisionist history we do seem to forget that ever generation has been plagued by a ton of shitty music that should never have seen the light of day].

We have greater and more varied tools than at any time in history... but it seems that we're managing to do less with them than at any time in history. Could it be the over intellectualizing of the process... or our reliance on the tools more than the art/craft/skill of the recording process... or are we following a flawed production model in an attempt to sound like the other stuff on the radio?

Only time will have the answer to these and other questions... I just wonder how fukking dumbass we're all going to feel about the recordings of early the early 2k's when we get to view them in rhw historic perspective of 2k30... much like we kinda roll our eyes at much of the drivel that was spewed during the 80's.

[WidgetNinja]

Originally Posted by Fletcher

Interesting.
I like transients... I hate boring musical arrangments and pedestrian sounds that don't reinforce the musical statement / intention of the artist... and worse, artistic statements that have little or nothing to actually say.

Eh, I think we need some definitions. You describe dynamics in [b]my[b/] dictionary. Transients in my dictionary have a negative impact on our ability to mix digitally. Looking at waveform in my DAW (which represents the computer's ear) and comparing it to my own ears, there's a whole lot of transients that the human ear just glosses over, just like how my ears don't report phase issues when listening to more than one speaker. My ear doesn't hear 15 db spikes on my guitar when I use an API pre... but the damn CPU does. All my my plugins see those damn transients and it screws up their theoretically acheivable fidelity. Compressing out the transients ITB raises noise, and I tend to like light limiting going into the converters to smooth out transients.

Now, I really like dynamics, but I hate transients by and large (in theory anyways).

[peeceebee]

Notwithstanding Fletcher's foray into dynamics, I think the original issue brought up in this thread centers around what most people describe as "warm" sound verses "cold" digital sound.

The fabled "warm" sound is originally an artifact of older recording technology; tubes, transformers, tape heads, tape, paper capacitors, point-to-point wiring, primitive speakers at reproduction, etc. Each of those components that were ubiquitous to varying degrees in the fabled "tropical sound era" of the 50's & 60's can have tendencies to varying degrees to variably clip high frequency transients (and phase smear low frequencies, and everything else[?]) in audio signals passing through them; and "clean" digital technology does no such thing.

The questions are:

a) Is "warm" sound really better, or have many of us just been conditioned to think so-

b) If yes to (a), what is the best/easiest way to recreate that sound with current technology, and is it possible to do so without keping a storeroom full of tape machine spares and a full time tech on staff to rebuild your aging 16-track every few days?

[Ruphus]

Originally Posted by WidgetNinja

Compressing out the transients ITB raises noise,

With the stuff I recorded so far you just can´t get that presence of the mix. You end up needing a ton of compression to have it yet not even close as loud like ( great sounding ) tracks out there. ...

I´m planning to limit and or compress before convertors at next tracking session. But still ... I doubt to be coming even close that way to what my reference tracks provide. That bold and clear, yet smooth sound that kicks your homey crap in the trash with its volume that yet is being far from distorting and fizzling, not enough could even take a couple more dB of limiting and would still sound better than what you´d dream of.

Ok, million dollar studios with highly educated engineers in there might not be the right address to compare myself to, but damn, I´d still like to know how they do it.

As we know the answer can´t entirely be top notch analog chains and tape machines, there are engineers who prove that there can be done stellar work digitally meanwhile.

Only how do they get the digital stuff to work out so bold and round?

Ruphus

[kats]

I hear what your saying Fletcher. And it's almost to the point where I have a hard time listening to most current R&R productions without getting bored out of my skull. IMO it's not so much the technology in in itself as much as it is what it has allowed us to do.

I think it comes down to the basic level that more and more, live on the floor recording is becoming rarer and rarer. 2/3rds of the emotion of music is felt through dynamics and tempo (the other part is the tension and resolution of the music itself - IE the song writing). Four guys recording live on the floor will feed off of eachothers emotion and the dynamic of the band will have a signature in the swing of tempo and swing of dynamics. Something that the listener will feed off of.

Can you imagine some of the great symphonies recorded to a click track or BD'd, AND THEN compressed to remove most of the dynamics??? There's a reason they have a conductor!

Anyhow, I don't think Charlie Watts would have got as excited on his kick if he were playing to a click, and I think we would have been less excited listening as well. IMO, this is the saving grace of a pro studio. While the market is moving to smaller spaces and affordable gear - nothing will replace a good room large enough to record live off the floor. We're just still infactuated with all the tricks these new toys allow us, but it ain't gonna last. We're just in transition, that's all.

[Doublehelix]

What kills me is to work so hard to get a great drum sound with lots of snare crack and kick umph, and than have the ME take away all the dynamics and squash the hell out of all the transients.

[eligit]

slight tangent:

has anyone used either of the saturation modes while tracking w/ the lavry blue A/D?? if so, what do you think?

naturally you will end up with some hot levels (which leads to a whole other discussion when it comes to mixing...) but how does it sound?

[Ruphus]

You guys sound like any loud sounding record could only be crap which is BS. There seem ways to get the stuff loud and still full with dynamics, expression and all that you need to feel performance and yet be sonically excited.

And don´t come with that purist sound like I have heard it a couple of times. Totally without compression, but far away, thin, weak and boring.

Listen to Black Crowes´"Shake your money maker", "Sheryl Crows "Sheryl Crow", or David Lindley´s "Al Rayo X". Stuff that blasts your speakers in a wonderful way. Definitly compressed, but magically so.
I want that.


Even those old recordings obviously are compressed. Charly Parkers "Remember Mr. Banks" which I have posted once here is just too present to be not squeezed somewhat somewhere.

Anyway, digitally there is no way without compression.

Ruphus

[bunnerabb]

This is a lot of food for thought and the main reason I use these boards.

Thank you.

thumbsup

[Ruphus]

Originally Posted by eligit

slight tangent:

has anyone used either of the saturation modes while tracking w/ the lavry blue A/D?? if so, what do you think?

naturally you will end up with some hot levels (which leads to a whole other discussion when it comes to mixing...) but how does it sound?

The digital compression is amazingly subtle for the amount it does ( + 6dB ).
When the unit came in I compared it to same none-compressed material and found the uncompressed a tad more open. So, I chalked it up under "degrading" the material, however seing how I had to abuse uncompressed tracks afterwards to get them a bit upfront ... that has been ten times more harming than the Lavrys procedure could ever have been instead.

Maybe I should try it for the next recordings, might well help me away a bit from the dilemma of ending up with too soft crap. At least for LF sources it must be a good idea to use it, indeed.

Ruphus

[eligit]

Originally Posted by Ruphus

The digital compression is amazingly subtle for the amount it does ( + 6dB ).
When the unit came in I compared it to same none-compressed material and found the uncompressed a tad more open. So, I chalked it up under "degrading" the material, however seing how I had to abuse uncompressed tracks afterwards to get them a bit upfront ... that has been ten times more harming than the Lavrys procedure could ever have been instead.

Maybe I should try it for the next recordings, might well help me away a bit from the dilemma of ending up with too soft crap. At least for LF sources it must be a good idea to use it, indeed.

Ruphus

well, i will be hooking up a lavry ad/da over here this week for tracking....i will at least experiment w/ the various settings to see what's what...

[not_so_new]

Originally Posted by kats

I think it comes down to the basic level that more and more, live on the floor recording is becoming rarer and rarer. 2/3rds of the emotion of music is felt through dynamics and tempo (the other part is the tension and resolution of the music itself - IE the song writing). Four guys recording live on the floor will feed off of eachothers emotion and the dynamic of the band will have a signature in the swing of tempo and swing of dynamics. Something that the listener will feed off of.

Can you imagine some of the great symphonies recorded to a click track or BD'd, AND THEN compressed to remove most of the dynamics??? There's a reason they have a conductor!

Anyhow, I don't think Charlie Watts would have got as excited on his kick if he were playing to a click, and I think we would have been less excited listening as well. IMO, this is the saving grace of a pro studio. While the market is moving to smaller spaces and affordable gear - nothing will replace a good room large enough to record live off the floor. We're just still infactuated with all the tricks these new toys allow us, but it ain't gonna last. We're just in transition, that's all.

I agree that this is just a transition, this too shale pass.

I disagree that you can't make meaningful emotional music to a click track while the musicians play in separate rooms or at separate times. You may like Steely Dan, you may hate them but I happen to love them and I feel they are very emotional and musical. Many of their tracks were done as overdubs from what I understand including drum tracks all played to a click. Some of the newly released Hendrix stuff had drummers playing 25 years after Hendrix died yet they are still emotional and powerfull songs (songs being the key word).

Emotion is different for everyone, what you think is emotional may not be for others and what I think is emotional may not be for you.

On the whole I agree that most of today's music is sterile and boring but I personally don't think this has much to do with the production.

Let me put that another way, the production is not helping the cause but in the end if you have a good song no production is going to hide it and if you have a bad song no production is going to save it.

The "get naked test" applies here. If you have an acoustic rendition of a song, striped of all it's production elements and it is still interesting and exciting then you have a good song. The Beatles "Hey Jude" would not be helped by today's production but it would still be a GREAT song. Now take your pick of a song from today, no amount of "cool" 60's production is going to save it. It all comes back to the song, the song is king the rest just supports the song.

[not_so_new]

As to the original thread about Transients being our worst enemy, I think you make some good points but I think a good part of this is what you listen to and when you started listening.

Most people on this board in the professional scene are a bit older and tend to listen with "tainted" ears. When I say tainted I don't mean that in a bad way, I just mean we get used to what we get used to and most of us are used to analog recordings.

Who knows what the future will hold, maybe in 20 years people our age then will wonder how we could ever have listened to analog tape because it was all compressed and lost all of it's transients? I am not saying this will happen and yes you can point out kids who hear good analog recordings now and how they love them yadda yadda yadda.....

[rlnyc]

Originally Posted by WidgetNinja

Transients in my dictionary have a negative impact on our ability to mix digitally. Looking at waveform in my DAW (which represents the computer's ear)

stop looking.

is it evolution when music gets looked at rather than listened to?

in the east, suicide is committed by cutting the belly (hara-kiri "cutting the hara")
in the west, it is by shooting oneself in the head, or by hanging.

in eastern culture, the "I" is in the gut, in the west, it is in the head brain.

now we are moving the experience of music from the ears to the eyes. maybe in the future people will evolve so they don't need ears at all, and through synesthesia they will derive pleasure from music through the eyes. after all, light is a higher bandwidth of sound anyway, in the 40 or so octaves or elecromagnetic radiation that exists in the cosmos. when the record (a diamond scaping through grooves in a pancake of plastic) was replaced by the CD (digital information being read by a laser on a piece of plastic), it was that kind of progress - an evolution of form.

all this discussion about transients as evil code in the digital realm is making me ill. even your every breath has a transient edge. you want your music to have emphysema?

not expecting a good result from this post,
i remain,
friendly yours,
rlnyc.

[joeq]

I don't see transients as the enemy. Maybe as a drummer I am biased. With good players who are in control of their transients, they give life to the music.

If I could deliver my 24 bit mixes to the public, I would probably use half the compression and limiting that I use today.

[kats]

It all comes back to the song, the song is king the rest just supports the song.

A poorly performed song is not music, it's just a bunch of notes. The performance makes or breaks it. So IMO, compromising the perfomance in any way for the sake of efficient production is a negative. The idea is to avoid as much of that that is reasonbly possible.

[Darius van H]

Peeceebee said it well...the older technologies were not good at capturing/retaining transients and dynamic range in general.........that was a big part of the so called "warm" sound of the days of yore........if you're aware of this it's fairly easy to recreate that sound with the gear we have available these days, although doing it with tape may become increasingly troublesome (or is there going to be a cheap and reliable supply in the future?)

[not_so_new]

Originally Posted by joeq

I don't see transients as the enemy. Maybe as a drummer I am biased. With good players who are in control of their transients, they give life to the music.

If I could deliver my 24 bit mixes to the public, I would probably use half the compression and limiting that I use today.

Maybe we have an issue on this thread with the definition of the word "transient".

I think some people are getting "transient" confused with "dynamics" they are related but they are not the same thing. A good player can and should use dynamics but no player can control the transient nature of his or her instrument without fundamentally changing the instrument it's self.

In my book a transient is the leading edge of a sound source, the rise time of the sound in question. An electric guitar has a medium fast transient time, a synth can have a very slow transient time. A snare drum (most all drums really) have a very fast transient time. There is a strong argument to be made that the initial transient of an instrument is what clues the ear to the instrument making the sound, not that you would confuse a snare drum with a bass guitar but if you get rid of enough of the transient attack you might be confused between an acoustic guitar and a violin.

Anyway there is no way a player can control the transient nature of his instrument, a snare drum can be played loudly or softly but it's transient nature will always be very fast, that is part of what a snare drum is. You can change the sound of the drum by how loudly you hit it or where you hit it, i.e. a rim shot vs. a standard hit but both are going to be very fast transient events. Same thing with playing it louder or softer, the transient is going to be the same, fast. Other things will change the louder or softer you hit, overtones may stand out more or less, sympatric vibrations may get set up within the room, the decay of the instrument will be longer etc. but the transient nature of the instrument will not change.

[joeq]

Originally Posted by not_so_new

Maybe we have an issue on this thread with the definition of the word "transient".

I think some people are getting "transient" confused with "dynamics" they are related but they are not the same thing. A good player can and should use dynamics but no player can control the transient nature of his or her instrument without fundamentally changing the instrument it's self..

Personally, I meant transient, not dynamics. I think the player _can control the amount of transient relative to the amount of other tones coming out his instrument. I would not expect the _speed of the transient to change, but the proportion and even the peak level is under the musician's control.

When I said a good player can control his transients I was thinking specifically of some bass players and guitarists I know, but even in drums, stick selection, grip and "digging in" can change the nature of that leading edge.

[vtone]

Originally Posted by Fletcher

I smiled. There is no way in hell I would have let that pass... but ya know what, it helped the damn song

word.

it's refreshing to read this on this particular forum...

thanks for that fletcher.

thumbsup

[opentune]

Transients?? ...uhmm... they´re just passing by... (man, i´m a philosopher )

[Fletcher]

Transients are part of dynamics... erase the transients and you're well on your way to erasing the dynamics of the song... only quicker with the attack of the dynamics removal device.

I was at a friend's shop last summer when someone sent their limiter back to him to be converted to being an RMS limiter from a peak limiter... in other words they wanted an analog "L-2"... which is a sad fukking comment on the state of our industry.

As for intial transients... I contend that if you're listening to music at 100-104 db and a snare hit doesn't make you blink, then you've fukked up the front end of the drum sound... but maybe that's just me.