PDA

View Full Version : Classical Music Compression


orpheus_2
02-02-2008, 05:48 PM
Hi all,

here I am to bother again with question for our little string quartet production.

We would like to add some more volume power in terms of loudness to some soft movements. It doesn't work properly with Gain, it is still too weak - or tends to lose the color. Should we give a try to compressor? Some folks say it's a no-no in classical music. Anybody has maybe some template with slight compressor presets that could work for us.

Cheers!

sowby
02-02-2008, 06:21 PM
Try using upward compression, ie. uncompressor. The Waves rcomp has this setting called "uncompressor". It can work well for this situation. Also, who said compression is a no-no in classical? I have worked with one of the best classical engineers in the country, and he definitely uses compression, skillfully, on all his albums. Also, try using some "bottom up" compression, as opposed to "top down". This can bring out the low level passages without squashing the loud ones.

orpheus_2
02-02-2008, 06:38 PM
Try using upward compression, ie. uncompressor. The Waves rcomp has this setting called "uncompressor". It can work well for this situation. Also, who said compression is a no-no in classical? I have worked with one of the best classical engineers in the country, and he definitely uses compression, skillfully, on all his albums. Also, try using some "bottom up" compression, as opposed to "top down". This can bring out the low level passages without squashing the loud ones.



Thanks, I have also thought that there shouldn't be ideological rejections of any means to enhancing the music. Also, I listen to the music in the car a lot, and miss often the soft passages, which makes me unhappy. Is there some inexpensive plugin we can try before getting the expensive ones?

orpheus_2
02-03-2008, 09:21 AM
Are there some "safe values" when using PT's own plugins (Compressor-Limiter) etc. which would enhance the overall loudness of soft passages, but deform the timber under no circumstances (different player, radio situation, iPod compressed file etc.)

Thanks,

ORP

M.Brane
02-03-2008, 05:47 PM
Some subtle parallel compression might do the trick. Worth a shot anyway. It's a good way to bring up the low-level stuff without lopping the tops off the transients.

Finnur Hansen
02-04-2008, 01:42 AM
Parallel compression works good for classic!
They call it New York compression "over there".
Be careful with the phase though.
The DAW delay compensation is REALLY important here!
Usually the compensation works fine but keep an ear on it.
Still - do not over-do the compression in classic even though it is parallel!

chris110000100101001
02-04-2008, 04:23 AM
This can be a tough one, since you want to compress the quiet stuff and leave the louder parts alone. The parallel compression method is good. Something else to try is a combination of volume automation and bus compression. Let the louder parts remain just below threshold and push the level for the quiet parts so they trigger compression. Try a couple of low-ratio compressors in series instead of slamming just one, the results are more transparent.


Some folks say it's a no-no in classical music



Those folks are clearly not present at the mastering session. What is the dynamic range of vinyl again?

It reminds me of an acoustic act I was mixing a few months back. Their Violinist came up to me before the show an said "We dont want any eq or effects and NO COMPRESSION. Just flat channels please." Right. After the show they praised me on the lovely, natural sound. Yep, that would be the reverb, delay, eq and compression...

Naagzh
02-04-2008, 07:57 AM
Parallel compression works good for classic!
They call it New York compression "over there".
Be careful with the phase though.
The DAW delay compensation is REALLY important here!
Usually the compensation works fine but keep an ear on it.
Still - do not over-do the compression in classic even though it is parallel!



I've been referring alot to a recent article in TapeOp, which demonstrates "2-Stage Compression" (basically a more sophisticated, and arguably more transparent, version of parallel compression). I've used it with much success. And to echo: careful with that phase! Delay compensation might not take care of everything with this one...

It is quite applicable in your case, as its goal is to compress somewhat aggressively, and yet preserve the balance of loud and soft passages within a song. It took time to get used to it, but this is the best trick I've ever learned. I feel like a pro.

orpheus_2
02-04-2008, 09:51 AM
Surely, me gotta ask: how do I do that parallel compression in terms of having a plugin open? What do I mess with? If there is some good link describing the procedure, i.e. giving a recipe, I would appreciate.

Also, I know this "natural sound" phenomenon. It does require messing with EQ and reverb and other things. However, Compressor is the feature that I came across most controversial info about - at least as a performer. Also, there are very many CDs (especially with recent classical music) one can't even think of listening in the car, as some passages are of very low volume.

orpheus_2
02-04-2008, 02:14 PM
I think I found the article describing the "parallel compression". There is also a video on YouTube

http://www.youtube.com/watch?v=NvHlx3XHpBA

It seems it would be a bit of an overkill.

We'll try something simple.

Thanks a lot for the input!

Niljam
02-04-2008, 02:56 PM
I use the Sony Oxford Inflator plugin on classical, acoustic music (as well as on anything really).

It's the only "mastering-plugin" on the market, that I know of, that can really push up the gain without losing dynamics or distorting the sound. It's also the only software that the leading mastering-engineer in Sweden would recommend. (He uses only high-end hardware limiters etc).

Nope, I'm not employed by SonyOxford… just a happy user.

stoneinapond
02-04-2008, 03:06 PM
I use the Sony Oxford Inflator plugin on classical, acoustic music (as well as on anything really).

It's the only "mastering-plugin" on the market, that I know of, that can really push up the gain without losing dynamics or distorting the sound. It's also the only software that the leading mastering-engineer in Sweden would recommend. (He uses only high-end hardware limiters etc).

Nope, I'm not employed by SonyOxford… just a happy user.



Ditto for the "Sonnox" Oxford Inflator

Niljam
02-04-2008, 04:49 PM
Ditto for the "Sonnox" Oxford Inflator



It's the same thing with a new name, right? Great stuff anyway...

stoneinapond
02-04-2008, 06:02 PM
It's the same thing with a new name, right? Great stuff anyway...



Yes and yes.

orpheus_2
02-05-2008, 09:11 AM
I understand this NY compression or OxfordSonnoxSony thingy must work miracles, but you are all audio freaks. How are we supposed to invest so much cash in advance for a slight improvement (our target audience is composed by no means audiofiles), if the composer of the most of the works we are introducing drives bus???

Should somebody be sympathetic, please send some values for native PT plugins which surely wouldn't mess up the playing. I think low dynamics parts need help, the loud ones are more than OK.

Sorry for being slightly negative - we would like to leave all the audio work up to experts. But - in our quartet for instance - second violinist is raising three kids, has double Master's (violin and education), and does a PhD in conducting (costs her also some cash). That's indie classical musician's life in 2008. She surely could have gone the sleazy way, instead of doing the things that are worthwile.On the other hand, we pay all the royalties and commissions to composers, and they are all a lot worse off than us.

That's 2008

stoneinapond
02-05-2008, 09:41 AM
...Should somebody be sympathetic, please send some values for native PT plugins which surely wouldn't mess up the playing. I think low dynamics parts need help, the loud ones are more than OK...



While I have all the sympathy for your personal situations, you have to understand that it is difficult to work on or suggest something for material that has not been heard.

It would also help if you specified exactly what you have in terms of plug-ins and equipment.

But in the end you have to know what it is you want and how to get there. You also shouldn't be afraid to experiment. The beauty of PT is that it is non-destructive, so slap a compressor on your track and start moving controls. Take a limiter and do the same. Put both on, one after another and listen, listen and keep listening.

While the following article does not cover string quartets per se, reading it may provide you with some starting points.

Compression Article (http://www.tcelectronic.com/media/droney_massey_2001_compres.pdf)

Remember, that compression and limiting will to some degree change the "sound" of the original material. To what extent depends on what you do with the processes. You might also consider some gentle EQ'ing prior to compression/limiting, also bearing in mind that what you do at one stage affects what comes later.

I also had to drive busses for living at one point in my life. Pursuing a dream doesn't always come easy. So don't feel sorry for yourself, roll up your sleeves and next time be very clear about what it is you need.

orpheus_2
02-05-2008, 09:57 AM
Hi fella,

thanks, we all never worked with that, so if people say "never do the hard knee", that's already help. We are here pretty basic, and we admit it. Also, it is not that any of us is sorry for him/herself - we believe it is shame (and not our job) to have to do it on our own. Other than that, we figured that this project would probably NEVER see the light of the day if we didn't do it on our own. We've been postponing it (trying to fundraise etc.) since 2001! We are all actually ashamed of the all "Googling and foruming", seeking the chippo advice...at least we legally own PT and hardware

Guess what - three years ago we did expensive mastering, and the "master of the mastering" was busy, hard to get, late for the appointment etc. The CD turned good though, but just covered the costs (we got only 5-6 specialized concerts and 2 reviews. This year we are getting lots of unsolicited promo for recording, mastering, repli- dupli- etc. It seems though we are less able than ever to commit to professional audio...

Hope the end is in sight even if I don't see it

Naagzh
02-05-2008, 10:05 AM
I understand this NY compression or OxfordSonnoxSony thingy must work miracles, but you are all audio freaks. How are we supposed to invest so much cash in advance for a slight improvement (our target audience is composed by no means audiofiles), if the composer of the most of the works we are introducing drives bus???

Should somebody be sympathetic, please send some values for native PT plugins which surely wouldn't mess up the playing. I think low dynamics parts need help, the loud ones are more than OK.

Sorry for being slightly negative - we would like to leave all the audio work up to experts. But - in our quartet for instance - second violinist is raising three kids, has double Master's (violin and education), and does a PhD in conducting (costs her also some cash). That's indie classical musician's life in 2008. She surely could have gone the sleazy way, instead of doing the things that are worthwile.On the other hand, we pay all the royalties and commissions to composers, and they are all a lot worse off than us.

That's 2008



Well, if your audience doesn't mind (they're not audiophiles), and you can't afford to invest in your mixes, then there's really no problem. Leave the mixes as they are, and continue to make the music you find "worthwhile".

But at the same time, you want to leave the audio work up to experts, yet not being able to pay them for their work leaves you with no other option than to tap the potential of a non-professional (who may have lots of experience). There are plenty to be found here on this forum. By the by, it will NOT help to describe their non-classical endeavors with words like "sleazy".

I'm sure many here wouldn't mind taking the job of mixing your music for a few dollars, maybe even send you the session file so you can learn from what they did. I'd be happy to do it, and I think I'll do a very good job.

But don't think that a specific plug-in will magically transform your mixes. I've accomplished 2-stage compression using only the stock Digi Comp/Dyn III plug. Expertise is what you should seek and develop foremost, before you run out and buy a bunch of plugs.

Sending you plug-in values will not work, because how you've recorded the piece is unique, the players and instuments are unique, the performances themselves are unique, and so on. Different recordings require different settings and approaches, often within the same song.

Shawn Simpson
02-05-2008, 10:07 AM
Do any of these universities where you poor souls are studying have an audio department? You might solve your problem very quickly by simply asking for a student of audio engineering to spend a little time with you. The student would gain from the experience and you would most certainly benefit from their knowledge.

In case you didn't know, we audio freaks do what we do because we love it too. To suffer for your art is not the strict territory of the tortured string player.

stoneinapond
02-05-2008, 10:13 AM
If there is some way that I can communicate with you I may be able to help.

It seems you are not in the US, correct?

orpheus_2
02-05-2008, 12:03 PM
Dear all, I see some folks got offended, so it is probably really better to experiment for ourselves, and keep quiet.

I am sorry if anybody thought we considered any non-classical music as inferior. When I meant "sleazy" I though of Canon in D endlessly being played, same as Vivaldi's Seasons - all great pieces, but much too often played to appease always same (low) level of classical music culture. There is lots of music killed by such "sleazy" programming, and that's surely one of the reasons why all that is original has such a hard time.

"Audio freak" was not meant as offense. We are contemporary music freaks and string quartet freaks. We actually would appreciate some good "audio freak" at our side, but one we can afford AND trust. Here we come to stories of the experience we made we student recording engineers. Contrary to popular belief, they weren't after the quality, but rather after money - even if the amounts were small. However, there was no repetitive procedure they wouldn't mess up to charge us more hours for fixing it...not again, thanks.

But - to say the truth - classical music acoustically and technically IS a completely other beast. I can ascribe some unnecessarily harsh reactions to lack of dealing with us classical folks.

M.Brane
02-05-2008, 06:03 PM
Being a listener of many different kinds of music (classical included) my suggestion of parallel compression (not 2-stage compression like that video you linked) was not just a shot in the dark.

Don't knock it until you've tried it. I think if done properly it will get you the results you want even with the stock Dyn III comp. No high-dollar plug-ins required.

orpheus_2
02-05-2008, 06:23 PM
Hi M.Brane, do you have some other internet "recipe" to link me to?

M.Brane
02-05-2008, 08:46 PM
Hi M.Brane, do you have some other internet "recipe" to link me to?



No, but I do have this quote from Bob Katz:



This technique is designed to compress with minimal or no effect on transient peaks. Mix the source with a "parallel compressor" (which some people call a side compressor, but this term is confusing). The source must be delayed to match the delay of the parallel compressor and avoid comb-filter effects. To determine the delay, invert the polarity of the source or the output of the compressor, set the compressor's threshold to 0 dBFS and makeup gain to unity. Mix the two and adjust the delay for complete cancellation. This delay will usually be a few samples.

The parallel compressor should have a threshold of -50 dBFS (that's right!), a ratio of 2:1, the fastest attack that is possible, and a medium release, somewhere around 800 mS to start. Adding this to the (delay compensated) "dry" signal produces the upward compression. As the signal gets higher and higher, the parallel compressor is less and less in the circuit, until at levels above about -20 dBFS, the parallel compressor is at least 20 dB down. You can see how transients will be little affected by this more-transparent technique.

You adjust the amount of compression not by changing the settings on the parallel compressor, but rather by raising or lowering the output of the parallel compressor (or makeup gain), and listening by ear. You can also test with test tones how much the gain goes up at low levels to judge the amount of compression. It's so gentle and so natural that it is almost impossible to tell this compressor is in the signal path.




Hope this helps. I know it's helped me more than once.

BTW the easy way to do delay compensation for this in LE is simply to put the same plug on both tracks, but have the first one set so it has no effect on the track.

Neale
02-05-2008, 10:14 PM
My humble advice is: save your time typing and go try stuff. If it sounds good, it IS good. There are no rules here, just stuff that MAY work better than others SOME of the time.
No one should claim that they can give you numbers that will work in a plug-in that WILL make your goal happen. Someone previously posted something to the effect that this won't happen because none of us can hear your music. Someone MAY be able to suggest some guidelines that will point you in the right direction, but it may not be the be-all and end-all of what you wanted.
I think the most instructive post was from the person who had the classical client who didn't want any effects at all, just flat signal, then complimented them on the "natural" quality of the effected sound that they were actually hearing. Anyone can take any given piece of equipment and have it sound passable, excellent, or poor; anyone can take any methodology, whether it's regarding EQ'ing, compression, or reverb, and have it sound passable, excellent, or poor. One compressor may get along well with classical programs, another may not, hence, maybe you use the useful one all the time ("you CAN use compression on classical music") or you may choose to never use the not-so-useful one ("you CANNOT use compression on classical music), leading to blanket statements that may or may not be applicable in your given situation. On top of that, the same program material may alternately sound passable, excellent, or poor to different people on different days on different systems... On and on.
You've received some helpful hints so far. If you don't want to buy the Sonnox stuff or invest in anything beyond what you have, then go play with what you have. There's nothing wrong with it, and it doesn't matter WHAT brand it is. The plug-ins you have JUST DO STUFF. Whether you think the stuff they do is GOOD or BAD, that's up to you. You have to be able to rationalize your decision(s) to yourself in the end, and no amount of asking other people for THEIR advice will allow you to do that for yourself.
So...
Bottom line is, take one copy of the unaltered program, then take the one you mangled. Flip back and forth. If you like YOUR version better, good for you. If not, go back and do it again. Get the other people in the quartet's opinions, see if you can get some general consensus on good and bad stuff; if there's a lot of good, leave it alone, and if there's a lot of bad, go back and do it again.

orpheus_2
02-06-2008, 07:16 AM
Thanks Neale, that's generally the way we go now.
I think we'd give a try to the parallel compression though.

orpheus_2
02-11-2008, 05:26 PM
Hi,

I managed to catch some time and get back to our favorite project (with no cash attached for the next several dozen years ).

I can tell you what I did:



Hi M.Brane, do you have some other internet "recipe" to link me to?



No, but I do have this quote from Bob Katz:



This technique is designed to compress with minimal or no effect on transient peaks. Mix the source with a "parallel compressor" (which some people call a side compressor, but this term is confusing). The source must be delayed to match the delay of the parallel compressor and avoid comb-filter effects. To determine the delay, invert the polarity of the source or the output of the compressor, set the compressor's threshold to 0 dBFS and makeup gain to unity. Mix the two and adjust the delay for complete cancellation. This delay will usually be a few samples.


I skipped this part, as I don't understand it at all...


The parallel compressor should have a threshold of -50 dBFS (that's right!), a ratio of 2:1, the fastest attack that is possible, and a medium release, somewhere around 800 mS to start. Adding this to the (delay compensated) "dry" signal produces the upward compression. As the signal gets higher and higher, the parallel compressor is less and less in the circuit, until at levels above about -20 dBFS, the parallel compressor is at least 20 dB down. You can see how transients will be little affected by this more-transparent technique.

You adjust the amount of compression not by changing the settings on the parallel compressor, but rather by raising or lowering the output of the parallel compressor (or makeup gain), and listening by ear. You can also test with test tones how much the gain goes up at low levels to judge the amount of compression. It's so gentle and so natural that it is almost impossible to tell this compressor is in the signal path.




We have L and R track in the stereo mix. Luckily, in this recording, they are mostly identical. I put same reverb and EQ on both. Only on the R I did put this compression. Sounds nice to me (haven't tested it in all environments), but now it seems we have basically one compressed and one uncompressed loudspeaker. Is it OK?


Hope this helps. I know it's helped me more than once.

BTW the easy way to do delay compensation for this in LE is simply to put the same plug on both tracks, but have the first one set so it has no effect on the track.

M.Brane
02-11-2008, 10:00 PM
Um, OK.

The reason why you didn't understand the first part of that procedure is that it assumes that you know in order to run the compression in parallel you need to have 2 identical signals (that's 4 mono tracks if you are in stereo) one (pair) with compression, and one (pair) without.

Obviously if you are running one track through some kind of processing in parallel with another identical dry track you need to make sure the tracks are time-aligned aka phase coherent since processing of any kind takes time. When working in the box this would be known as sample delay. The first part of that quote describes how to do that regardless of working environment, and I've also explained how to do it the easy way ITB (in the box) in PTLE.

If you still don't understand maybe you should reconsider mixing. That's not meant to be an insult, but the ability to understand basic signal flow/phase/timing issues is a requirement of work in AE.

It's not one-sided compression.

orpheus_2
02-11-2008, 11:46 PM
If you still don't understand maybe you should reconsider mixing. That's not meant to be an insult, but the ability to understand basic signal flow/phase/timing issues is a requirement of work in AE.

It's not one-sided compression.



Hi - no insult, don't worry. We are learning in this process, and we admitted not to be audio- but rather contemporary music-freaks.

Sure would I be able to understand how to overcome this delay, if there were somewhere a step-by-step guidance thru the process. Seems that we got the process itself right, but not the preparation.

orpheus_2
02-12-2008, 08:35 AM
Um, OK.

The reason why you didn't understand the first part of that procedure is that it assumes that you know in order to run the compression in parallel you need to have 2 identical signals (that's 4 mono tracks if you are in stereo) one (pair) with compression, and one (pair) without.


BTW, why would one need to work with two pairs of tracks in stereo? wouldn't it just be possible to make one compressed, and leave the other dry, but send each to both buses?

I think we did just that, and like the sound, but for some reason the intonation seems a bit down the drain.

Obviously if you are running one track through some kind of processing in parallel with another identical dry track you need to make sure the tracks are time-aligned aka phase coherent since processing of any kind takes time. When working in the box this would be known as sample delay. The first part of that quote describes how to do that regardless of working environment, and I've also explained how to do it the easy way ITB (in the box) in PTLE.

If you still don't understand maybe you should reconsider mixing. That's not meant to be an insult, but the ability to understand basic signal flow/phase/timing issues is a requirement of work in AE.

It's not one-sided compression.

M.Brane
02-12-2008, 09:33 AM
Yes you could do it either way: duplicate tracks or aux busses. Whichever suits your workflow better.

However you do it you need to make sure that the tracks are perfectly time-aligned or you're gonna have comb-filtering. To do this with busses you would simply put the same compressor plug on both busses, but on the first one you'd set the threshold as high as possible, the ratio low, and the attack slow to make sure it doesn't touch the signal.

orpheus_2
02-12-2008, 10:05 AM
Yes you could do it either way: duplicate tracks or aux busses. Whichever suits your workflow better.

However you do it you need to make sure that the tracks are perfectly time-aligned or you're gonna have comb-filtering. To do this with busses you would simply put the same compressor plug on both busses, but on the first one you'd set the threshold as high as possible, the ratio low, and the attack slow to make sure it doesn't touch the signal.



I don't notice any comb filtering, or anything really affected. However, the intonation really worsened.

M.Brane
02-12-2008, 12:30 PM
That's odd. This process should have no effect on pitch. I've done it many times without problems. Could it be emphasizing a particular problem/frequency that was being masked before?

This is where it becomes difficult for me to know what's going on over there since I can't hear it.

orpheus_2
02-12-2008, 02:16 PM
I don't think there were any frequencies masked, maybe the intonation problem is the way this "comb" is actually affecting string quartet sound/ Anyway, I found DigiDesign's own plugin "TIme Adjuster" - I'll let it run tonite to see if that is a fix. As I said, sound seemed fine, though we didn't run the CD in the car or on different systems. Apart from intonation problem, we noticed that it seemed there was less reverb than without this compression. Also, the audio picture did appear more coherent, with nicer instrument spacing, and there was some certain physical beauty of the sound at the low range, which did appear to be a result of digital calculation, but still contained certain mellowness...gotta go, I'll see how we feel later about it.