a cut off from a recording - how to separate some part?

If you are looking for advice on how to use Har-Bal best, or you have some tips of your own, post them here!
Post Reply
petr
Posts: 17
Joined: Thu Jan 17, 2008 8:31 am

a cut off from a recording - how to separate some part?

Post by petr »

Hello Friends,
can you advise me how to cut off some parts of my recording from a rock concert? I have a recording made by a camcorder and there are some parts where the public is for example clapping or singing with the singer....there is the perfect atmosphere in it.
I have also the original tracks from the mixpult and I mostly use only them as the basis of my final audio but sometimes I want to mix it with the public singing, clapping...how would you separate only the proper parts with the people, with the public? Can I somehow separate only some frequencies of the audio with what I want? Which ones would you recommend? I want only the part of the the audio with the ambience only. Any proved ideas?
HarBal
Site Admin
Posts: 761
Joined: Mon Apr 19, 2004 8:18 pm
Contact:

Post by HarBal »

What you are asking to do is generally very difficult to do, if not impossible. For the crowed noise to be separable from the live music it needs to be separated in time and or frequency. Something like a hand clap is broad band so separation in frequency (filtering) won't work. Time separability is no doubt just as problematic.

Joint time frequency signal processing may be able to do something but you'd need special gear for that. CEDAR by Cambridge Audio may be able to help you here but it is not cheep to buy.

Cheers,


Paavo.
petr
Posts: 17
Joined: Thu Jan 17, 2008 8:31 am

Post by petr »

HarBal wrote:What you are asking to do is generally very difficult to do, if not impossible. For the crowed noise to be separable from the live music it needs to be separated in time and or frequency. Something like a hand clap is broad band so separation in frequency (filtering) won't work. Time separability is no doubt just as problematic.

Joint time frequency signal processing may be able to do something but you'd need special gear for that. CEDAR by Cambridge Audio may be able to help you here but it is not cheep to buy.
Cheers,
Paavo.


Paavo, my questions are allways unanwerable, lol. Anyway, thank you for replying.
Can you at least advise me how to recort the crowds? I am no professional and have no experience with that. Anyway I have two mono microphones and a small portable minidisc recorder Sony. Shall I record the audience from the stage or from the rear? I think that from the rear there still will be the band very loud and will distrort the sound I want? And the best atmospfere is very close from the stage mostly.
HarBal
Site Admin
Posts: 761
Joined: Mon Apr 19, 2004 8:18 pm
Contact:

Post by HarBal »

Depends on when you want the crowd to be heard on the track. If only at the ends of the songs then you can only bring up the crowd recording in level where needed to avoid colouring the recorded music.

I'm not sure about the best position for the crowd mic but I'd guess a lot of live recordings do it with a stage mic facing the crowd that has a directional polar pattern (cardioid say). Earle will probably have a few suggestions.

Cheers,


Paaavo.
petr
Posts: 17
Joined: Thu Jan 17, 2008 8:31 am

Use of a compressor

Post by petr »

Paavo, Earle, please answer me one more question.
I did buy a book aboud sound processing but I did not find the requested answer ony my following question there.
My question is:
What is the goal of using the compressors (multiband ones too). I did come to the following answers myself.
1.To improve the sound by correcting the sound in the separate bands.
(For example I can lower too strong drums in a recording by compressing only the proper band?)
2. To get to the requested RMS in a natural way.
3. Combination 1 and 2 - that means improving the sound a bit where it needs to be improved and at the same time going to reach the requested final RMS.
Is there any other answer WHY I should use it? As reading all you write here I would think that compressing leads to loosing the quality, so I do not understand WHY I SHOULD USE IT?
Thanks for your reply in advance.
Petr
HarBal
Site Admin
Posts: 761
Joined: Mon Apr 19, 2004 8:18 pm
Contact:

Post by HarBal »

Multi-band compression can work well because it can mitigate the issue of one instruments envelope modulating another but only if they occupy different parts of the spectrum. I would say it is probably most helpful in the case where one or two instruments dominate the remainder and you wish to control that to some extent (your case 1).

The case of raising the RMS levels is another point entirely. I'd argue to never use that as a reason for multiband compression though I'm sure there are plenty who would disagree with me. Like I've said previously, if I had my way I'd have the music industry standardise on playback levels of around -13 to -11dB re full scale sine rather than the ridiculous -8 to -6dB levels that many aim for these days. In my view, -11dB is around the level at which obvious degradation in dynamics cut in irrespective of what processing is used so I wouldn't want to go beyond that. As for that level, it is pretty easy to achieve without MB compression so there isn't a need for it on that grounds.

On a point of recording quality I would say my favourite sounding CD's originate from the mid 90's with RMS levels around those I'd like to see as standard. Most stuff these days, I find, sound obviously damaged by over compression as a result of chasing stupid levels, annoyingly so. That disappointment has resulted in me losing interest in new CD's and as such I buy very few. They just don't sound good enough to me. I leave it up to you to decide upon the levels you wish to chase.

Cheers,


Paavo.
petr
Posts: 17
Joined: Thu Jan 17, 2008 8:31 am

Post by petr »

HarBal wrote:Multi-band compression can work well because it can mitigate the issue of one instruments envelope modulating another but only if they occupy different parts of the spectrum. I would say it is probably most helpful in the case where one or two instruments dominate the remainder and you wish to control that to some extent (your case 1).

The case of raising the RMS levels is another point entirely. I'd argue to never use that as a reason for multiband compression though I'm sure there are plenty who would disagree with me. Like I've said previously, if I had my way I'd have the music industry standardise on playback levels of around -13 to -11dB re full scale sine rather than the ridiculous -8 to -6dB levels that many aim for these days. In my view, -11dB is around the level at which obvious degradation in dynamics cut in irrespective of what processing is used so I wouldn't want to go beyond that. As for that level, it is pretty easy to achieve without MB compression so there isn't a need for it on that grounds.

On a point of recording quality I would say my favourite sounding CD's originate from the mid 90's with RMS levels around those I'd like to see as standard. Most stuff these days, I find, sound obviously damaged by over compression as a result of chasing stupid levels, annoyingly so. That disappointment has resulted in me losing interest in new CD's and as such I buy very few. They just don't sound good enough to me. I leave it up to you to decide upon the levels you wish to chase.

Cheers,


Paavo.


Paavo, thanks for your kind reply but please tell me very simply WHY I SHOULD USE THE COMPRESSOR (MULTIBAND) ON THE FINAL MIX? WHAT IS THE BENEFIT OF IT IF I ALREADY HAVE THE RMS OK AFTER HAR-BAL AND DO NOT WANT TO COMPRESS THE SONG'S RMS ANYMORE?
har-bal
Site Admin
Posts: 647
Joined: Wed Dec 31, 1969 8:00 pm
Location: Atlanta/Australia
Contact:

Post by har-bal »

petr wrote:
HarBal wrote:Multi-band compression can work well because it can mitigate the issue of one instruments envelope modulating another but only if they occupy different parts of the spectrum. I would say it is probably most helpful in the case where one or two instruments dominate the remainder and you wish to control that to some extent (your case 1).

The case of raising the RMS levels is another point entirely. I'd argue to never use that as a reason for multiband compression though I'm sure there are plenty who would disagree with me. Like I've said previously, if I had my way I'd have the music industry standardise on playback levels of around -13 to -11dB re full scale sine rather than the ridiculous -8 to -6dB levels that many aim for these days. In my view, -11dB is around the level at which obvious degradation in dynamics cut in irrespective of what processing is used so I wouldn't want to go beyond that. As for that level, it is pretty easy to achieve without MB compression so there isn't a need for it on that grounds.

On a point of recording quality I would say my favourite sounding CD's originate from the mid 90's with RMS levels around those I'd like to see as standard. Most stuff these days, I find, sound obviously damaged by over compression as a result of chasing stupid levels, annoyingly so. That disappointment has resulted in me losing interest in new CD's and as such I buy very few. They just don't sound good enough to me. I leave it up to you to decide upon the levels you wish to chase.

Cheers,


Paavo.


Paavo, thanks for your kind reply but please tell me very simply WHY I SHOULD USE THE COMPRESSOR (MULTIBAND) ON THE FINAL MIX? WHAT IS THE BENEFIT OF IT IF I ALREADY HAVE THE RMS OK AFTER HAR-BAL AND DO NOT WANT TO COMPRESS THE SONG'S RMS ANYMORE?


Petr

The only benefit of the multiband compressor is to help glue your mix and compress the separate parts of the spectrum that may have too much energy.

My suggestion is always this:

If you need you use a multiband compressor during the mastering phase, go back and remix and individually compress the problem track.

If you are unable to remix the track........you will need to use a multiband compressor.

If you already have the track sounding good as you indicated above and the loudness is comparable to commercial tracks and it sounds balanced...leave it alone

Cheers

Earle Holder
www.har-bal.com
HarBal
Site Admin
Posts: 761
Joined: Mon Apr 19, 2004 8:18 pm
Contact:

Post by HarBal »

I would ask this question of your tracks. After EQ'ing are there any points in the track where one instrument sounds too strident or harsh or brittle (say a loud guitar solo for example)? If there is then you have good grounds for using MB compression.

One way of seeing if this is the case is by comparing the average spectrum trace with the peak spectrum trace. If after making the average more uniform you find the peak trace is strong in one or more areas (ie. the space between the peak and average trace is bigger) then it is telling you that there is a short section of you track where an instrument is dominant and may require taming. I say may because you should base this decision on listening. You can use the spectrum information Har-Bal shows you to help set up you MB compression (ie. the frequency limits of you compression band).

My current redevelopment of Har-Bal is aiming to solve this issue. If it is successful then perhaps there won't be a need for MB compression at all. As I don't have a working prototype as yet I can't say if it will but I'm optimistic that it will work.

Cheers,


Paavo.
petr
Posts: 17
Joined: Thu Jan 17, 2008 8:31 am

Thanks Friends!

Post by petr »

My current redevelopment of Har-Bal is aiming to solve this issue. If it is successful then perhaps there won't be a need for MB compression at all. As I don't have a working prototype as yet I can't say if it will but I'm optimistic that it will work.
Cheers,
Paavo.[/quote]

Thnaks again! You are great and help me very much. I will buy any upgrade from you in the future as you help me very much!
Post Reply