Dynamic Range not only a Problem in Mastering.

[Mr.HOLMES]

Originally Posted by Paul Frindle

musically!

All I did to your track was try to recover the transients as asked, what happens to it by the time it gets mixed and mastered is another matter entirely. If transients are deemed to be 80's or 90's and therefore passe' then they will surely not make it into the final product anyway... IN this case even if you are the musician - and - the engineer, you will still end up with little or no choice. And if your tracks need percussive transients to have artistic impact, the chances are they will be deemed inappropriate and dropped :-(

This I tried already to explain to the client the transients will go down during mastering.
But what can I do nothing he wants it like this and he told me he does not want to hear engineering talk.

Originally Posted by domokunrox

Sounds to me like you're confused a bit, OP.

From what I understand reading what you're saying is that the drum compression just isn't sufficient.

You are trying to keep your drum dynamics. Trust your ears, not your eyes. Just because you see yourself squaring off peaks, doesn't mean you're doing anything bad. You need to take the tinfoil hat off. Somewhat of Hyper compressed drums is a mixing choice and style of production if you want it to be and everyone else wants it like that.

Look, peaks aren't pleasant, either. You need to just sit back, think about the song, and ask yourself, "Is it supposed to be louder here?" and if it is, then whats the problem? The answer to your question, it is in the song, not here on Gearslutz. If you need to automate, then automate. Don't be lazy and let a compressor handle what you can do with your fingers.

Still, compression is perfectly normal to use.

Think of it this way. Does live music from the band come out of speakers naturally? Well, duh, of course it doesn't. Most consumer speakers and/or headphones don't apply any natural compression at point the drivers push the air, this the design you're working with and cannot avoid. You have to apply compression to some things, otherwise they won't cut through the mix.

Here let me make it easy for anyone to understand because even I could confuse myself.

A person talks to you and their throat naturally compresses the sound they make and obviously gives its directionality and intensity, and the sound goes to the listeners ear and THEIR ears ALSO naturally compress the sound they hear.

This is a live situation.

What if its not and you're capturing it in recording medium? The singer speaks into a transducer that doesn't naturally compress the sound they make. The low frequency and mid range frequency drivers that consumers have don't compress the sound naturally. You have to manipulate it, otherwise it won't be heard or present in the mix without these nasty obnoxious peaks that take up the oh so valuable headroom real estate we try to conserve for the mastering engineer to use.

Compression, use it. Its not Satan's asshole until you make it that. Use it, but don't obviously over do it. You don't even have to try to get anyway with all the compression you possibly could.

This has been a Compression 101 lecture by domokunrox.

Don`t get me wrong here I never said that compression or limiting is Satan`s tools.

Yes I was confused because my client wanted to have strong transient sounding drums and delivered files which had no transients.
As Paul said their where taken away by fast LIMITING.... did not came to my mind as well thanks again PAUL.

So the client got one fair mix for a bargain amount of money.
When the client was here he also hugged me and told me how great it is sounding.

So a few days later the client send me an e-mail that the mix is bad because of a famous producer who told him "it would not be the same song anymore".

So I was confused but I had this situation not for the first time that clients change their opinion faster than I can say "What is wrong"
So I went with my mix to other studios and checked and it translated well everywhere.

So I asked other engineers about the drums and they all said nearly the same:
They could be stronger and the snare could have more height but overall not bad mix.

One engineer stated just:

It is all a taste thing.

So yes I am very confused about this client.
And I wrote him in an e-mail that I do not understand his 360° change in opinion.

So from there on confusion started.

Anyway all over all If find while mixing the hardest decisions are compressing.
When and why and does it changes the groove?
There is still one conflict in my mind I just can explain in an example.

If I compress something e.g. a Pop Vox so that it is up front in my face.
Ok nice now the VOX is in my face and it is just right for this stile of music.

But compression also takes away many things which have been nice before.
The transients the dynamic feel the live in this VOX track.

So do not get me wrong I think compression is one of the hardest decisions in mixing.

[Paul Frindle]

Originally Posted by Mr.HOLMES

One engineer stated just:

It is all a taste thing.

So yes I am very confused about this client.
And I wrote him in an e-mail that I do not understand his 360° change in opinion.

So from there on confusion started.

Anyway all over all If find while mixing the hardest decisions are compressing.
When and why and does it changes the groove?
There is still one conflict in my mind I just can explain in an example.

If I compress something e.g. a Pop Vox so that it is up front in my face.
Ok nice now the VOX is in my face and it is just right for this stile of music.

But compression also takes away many things which have been nice before.
The transients the dynamic feel the live in this VOX track.

So do not get me wrong I think compression is one of the hardest decisions in mixing.

You are right - it IS one of the hardest things to get right and (sadly) it can make or break a production in the current production climate.

In my experience, if you know that mastering will forcibly take out all the transients anyway in attempts to reach the required level of loudness 'density', either by direct compression or by over driving other pieces of 'character gear' to fill up the harmonic spectrum, you have to aim for other ways of imparting the 'rhythmic needs' of the track that do not rely on dynamic range.

Several ways exist to do this which involve a concept of 'making room' in the continuous modulation so that some of the percussion can make it out of the whole process and still form a valid part of the result the user hears.

If you think about it, if you have to make a result that will end up at full modulation whatever happens all the time, there are only two things you can do to 'slot' in the percussion;

You can dynamically reduce everything else for a minimum period during the percussion events - like in deliberate compression where the percussion peaks reduce everything else momentarily and the 'pumping' of the compression can even form part of the rhythmic 'groove' of the sound.

And/or somehow make holes in the frequency ranges of everything else to make a place to 'slot' the percussion events into. This is largely achieved by static EQ, as long as the sounds of things do not get too ridiculous and unbelievable.

You can even hard limit (or clip) the results to give a blast of extra harmonics across the spectrum during loud percussion events, that fill parts of the spectrum not already fully modulated by other instruments. This kind of thing was quite popular during the hard drum sound era of the 1980s to produce the impression of attack where tape media had limited dynamic range.

Both might be achieved at the same time to some extent with multi band compression, which will to some degree produce 'holes' in one part of the freq range responding to percussion peaks, whilst leaving the rest less affected. But the problem with this is that a 3 band (or even 5 band) multi comp may not be enough to do this effectively - and what you hear is a constantly changing timbre.

Something like the DSM can help much more in that there are many more active bands, so it can squash in some ranges whilst leaving enough unaffected bands to retain the timbre of other instruments.

But of course all this produces a significant restriction on the sound palette you have available for the various instruments in the track - and so in the end decisions have to be made about the content and sounds of the contributions to the track, constantly bearing in mind how it will all end up when the dynamic range is removed.


One useful way to do this kind of engineering is to go whole hog and stick a load of compression on your buss and mix into it as you go - so that you make better mixing decisions for something that will have no dynamic range in the end anyway.

You can even take it all the way and produce a result that has no dynamic range in itself at all and the frequency ranges are 'ready stuffed', so that mastering can reduce it no further (the DSM can achieve this at the limit). But in this case you will take a far greater responsibility for the end product, as the mastering engineer will have very little room to manoeuvre. If you go to these lengths you and the clients must be very sure what they actually want.......

[Mr.HOLMES]

Paul you make my brain

Originally Posted by Paul Frindle

You can dynamically reduce everything else for a minimum period during the percussion events - like in deliberate compression where the percussion peaks reduce everything else momentarily and the 'pumping' of the compression can even form part of the rhythmic 'groove' of the sound.

And/or somehow make holes in the frequency ranges of everything else to make a place to 'slot' the percussion events into. This is largely achieved by static EQ, as long as the sounds of things do not get too ridiculous and unbelievable.

Sorry for my bad English but do you mean e.g. sidechain compression?
Couldnt I achive this goal with fader automation as well?

Originally Posted by Paul Frindle

You can even hard limit (or clip) the results to give a blast of extra harmonics across the spectrum during loud percussion events, that fill parts of the spectrum not already fully modulated by other instruments. This kind of thing was quite popular during the hard drum sound era of the 1980s to produce the impression of attack where tape media had limited dynamic range.

Limiting brings also up harmonics?
I thought harmonics where generated onyl by gear e.g with tubes,tapes transformers so limiting gives harmonic impact... WOW or is my English too bad here?

Originally Posted by Paul Frindle

Both might be achieved at the same time to some extent with multi band compression, which will to some degree produce 'holes' in one part of the freq range responding to percussion peaks, whilst leaving the rest less affected. But the problem with this is that a 3 band (or even 5 band) multi comp may not be enough to do this effectively - and what you hear is a constantly changing timbre.

There you said it!!
In the last 6 years since I am doing mixing now I only had one good result with MB compression.

Originally Posted by Paul Frindle

Something like the DSM can help much more in that there are many more active bands, so it can squash in some ranges whilst leaving enough unaffected bands to retain the timbre of other instruments.

The DSM is an dynamic EQ?

Originally Posted by Paul Frindle

But of course all this produces a significant restriction on the sound palette you have available for the various instruments in the track - and so in the end decisions have to be made about the content and sounds of the contributions to the track, constantly bearing in mind how it will all end up when the dynamic range is removed.

That is too much for me.
Till today I thougt it is the Mastering House approach to make my track shine more and not to destroy it. If I listen to up do date productions I have the feeling waht for the hard mixing work?

One useful way to do this kind of engineering is to go whole hog and stick a load of compression on your buss and mix into it as you go - so that you make better mixing decisions for something that will have no dynamic range in the end anyway.

I mixed with 2 Bus comp plugs as well as HW for 2 years rock stuff and it was very fun to read that user MHB thinks the same as me a project studio owner: It takes away base range.... I dont like it.

You can even take it all the way and produce a result that has no dynamic range in itself at all and the frequency ranges are 'ready stuffed', so that mastering can reduce it no further (the DSM can achieve this at the limit). But in this case you will take a far greater responsibility for the end product, as the mastering engineer will have very little room to manoeuvre. If you go to these lengths you and the clients must be very sure what they actually want......

Thats my Problem with over compression in mixing and limiting in mastering.
In the end the music is sucked out of the recording but they paid big bucks for great studio musicians ..... these msucians had maybe an idea in mind how to accentuate the music..... it is smashed away.

Thanks for the ball you played back PAUL!!!

Andreas

[DSPaudio]

Paul's DSM is a new process. There is no accepted industry term for it yet, (like parallel compression, sidechain compression or convolution). I would probably call it Spectral Compression.

It's easiest to just check the demo and try presets. It's one of the most immediately rewarding plugins out there.

It has some small design similarity to the Matching EQ but is not so limited or "useless" IMO. Matching EQ's are nice for training beginners but ignore the personality of the mix which is what the engineer is supposed to be working with! We should not attach the term "Matching" to DSM, because this will cheapen peoples ideas of what is going on under the hood and this plugin is anything but cheap (technically).

The DSM is a force multiplier for the engineer. Imagine having 512(or more) compressors operating simultaneously across the spectrum of a track. Then think of sculpting a compression profile that intelligently directs the compressors to work harder or softer based not only on frequency but also over time.

Hey Paul, the captured compression profiles are also time aware no?

Anyway, the results speak for themselves and my technical guesses about the engine are only that. The code behind making a plugin like this is an art in itself.

The rules Paul & Team set up are very musical and allow for a wide range of compressions including restorative, utility and the holy grail (loudness without losing subjective dynamic range).

+1

As said before I don't have any affiliation with the makers of DSM. It's just one of the few plugins I've seen come along that's really groundbreaking.

Anthony

[Mr.HOLMES]

Originally Posted by DSPaudio

Paul's DSM is a new process. There is no accepted industry term for it yet, (like parallel compression, sidechain compression or convolution). I would probably call it Spectral Compression.

It's easiest to just check the demo and try presets. It's one of the most immediately rewarding plugins out there.

It has some small design similarity to the Matching EQ but is not so limited or "useless" IMO. Matching EQ's are nice for training beginners but ignore the personality of the mix which is what the engineer is supposed to be working with! We should not attach the term "Matching" to DSM, because this will cheapen peoples ideas of what is going on under the hood and this plugin is anything but cheap (technically).

The DSM is a force multiplier for the engineer. Imagine having 512(or more) compressors operating simultaneously across the spectrum of a track. Then think of sculpting a compression profile that intelligently directs the compressors to work harder or softer based not only on frequency but also over time.

Hey Paul, the captured compression profiles are also time aware no?

Anyway, the results speak for themselves and my technical guesses about the engine are only that. The code behind making a plugin like this is an art in itself.

The rules Paul & Team set up are very musical and allow for a wide range of compressions including restoration and impossible loudness without losing subjective dynamic range.

+1

As said before I don't have any affiliation with the makers of DSM. It's just one of the few plugins I've seen come along that's really groundbreaking.

Anthony

Interesting.
I just took the demo but at the moment I do not get what the difference to a multi band compressor is.... may working with it will show me the difference.

andreas

[Mr.HOLMES]

Originally Posted by Mr.HOLMES

Interesting.
I just took the demo but at the moment I do not get what the difference to a multi band compressor is.... may working with it will show me the difference.

andreas

Mhhhhh

I tried it today but I don`t get the logic behind it.

it is not a multi and not a whole band compressor what the heck is it?

[Paul Frindle]

Originally Posted by DSPaudio

Paul's DSM is a new process. There is no accepted industry term for it yet, (like parallel compression, sidechain compression or convolution). I would probably call it Spectral Compression.

It's easiest to just check the demo and try presets. It's one of the most immediately rewarding plugins out there.

It has some small design similarity to the Matching EQ but is not so limited or "useless" IMO. Matching EQ's are nice for training beginners but ignore the personality of the mix which is what the engineer is supposed to be working with! We should not attach the term "Matching" to DSM, because this will cheapen peoples ideas of what is going on under the hood and this plugin is anything but cheap (technically).

The DSM is a force multiplier for the engineer. Imagine having 512(or more) compressors operating simultaneously across the spectrum of a track. Then think of sculpting a compression profile that intelligently directs the compressors to work harder or softer based not only on frequency but also over time.

Hey Paul, the captured compression profiles are also time aware no?

Anyway, the results speak for themselves and my technical guesses about the engine are only that. The code behind making a plugin like this is an art in itself.

The rules Paul & Team set up are very musical and allow for a wide range of compressions including restorative, utility and the holy grail (loudness without losing subjective dynamic range).

+1

As said before I don't have any affiliation with the makers of DSM. It's just one of the few plugins I've seen come along that's really groundbreaking.

Anthony

As promised I have uploaded a zip file of the session I used for the drum loop stuff. Please note that the Oxford limiter was not used even though it's in the channel. I tried it but found the DSM better in this case..

LOL @ Spectral Compressor! This was precisely the name we used for it when the idea was first proposed!! Amazing :-)

I don't really want to hijack this thread with off topic stuff about the DSM, as I used it only very lightly on the drum loop just to correct the louder than average snare hits, before I did the transient expansion and EQ

It's best to look at the videos I did to explain how the DSM works and some of what can be done with it. They are here Dynamic Spectrum Mapper introduction on Vimeo. Obviously I am doing circus acts on those videos to be sure people can hear it working with the bad quality sound of the video coding etc. But it can be used for extremely subtle stuff as well, like the way I used it on the drum loop track.

But briefly, you can think of it as a compressor which has a valuable extra dimension of being able to work with a knowledge of the spectrum of the music itself :-)

It is indeed similar to a large scale multi-band Comp, but with a different way to split and combine freq bands that produces less artefacts (IMVHO). Because there are so many bands we can do the other stuff with it, like matching sounds and character, and very large scale compression without many of the usual artefacts.

The capture button primes all the thresholds in each band with the levels in those bands from the music itself.

Because the threshold of the compressors are set from the track itself (or in fact any other track you have captured and stored) it's action is more true to a believable version of the music, which means you can do much more compression before things start sounding awful.

The EQ style controls allow you to tweak the captured responses yourself - and the limiter is there to deal with peaks and allow you to maximise the loudness of the results without causing overloads.

BTW the captures are in fact static frequency responses - and are not time aware..