Its an unpaid job, Mike, I do it because I want to. I like to do the best I can, within reasonable time constraints of course.
Wavelab can do exactly what I want in half an hour tops, but not with such large files - the technology is almost decades old. I was hoping to find its successor. I'm not sure why something apparently so trivial is so difficult to achieve.
>>> I am dealing with long continuous live recordings, >>> (using Audition) ,where the levels go up and down a >>> lot, and need a good way to even them out afterwards. I don't want to >>> use any sort of plugin compressor or >>> limiter, I would rather manually adjust things.
>> Sounds like a church service. Been there, done that >> weekly for about 6 years.
>>> I have Audacity and Audition but can't seem to find any >>> way to do the same. What you can do is to hard select a >>> region and adjust the gain. If you do this, there is >>> obviously going to be a step change in gain at the >>> beginning and end of this region which is totally >>> unacceptable. Trying to fade these in afterwards is >>> going to be a nightmare. There are no gaps in the >>> recording that makes any of this possible. >> I break the recording up into zones where the levels are >> consistent within the zone. The zones are bounded by >> small regions basically background noise level. I >> adjust the zones by selecting each one and applying an >> appropriate gain correction. After doing this for 6 >> years I kinda-sorta know that works for each kind of >> zone. > I've tried this, and do it when I think I can get away > with it, but cant adjust by any more than about 2dB > without it being too noticable - I have no gaps at all.
If you make an envelope in Cool Edit/Audition that starts out with 0 dB gain and slowly ramps up to the desired amount of attenuation or gain, you can use it to apply gain changes that are less noticeable.
> As the event progresses, the faders gradually get turned > up higher and higher, so the overall waveform is like a > wedge of cheese, but with short term volume fluctuations > I want to keep intact, thus no compression.
Pretty strange. The usual goal of gain riding is consistent levels.
What can I say - the recordings I make tend to be uniform, with the caveat that there are inherent differences between mixing live sounds for the room, and mixing live sounds for a recording that need to be dealt with when reformatting a "board tape" (actually a digital recording these days) for distribution as a recording.
> Arny Krueger wrote: >> Sounds like a church service. Been there, done that >> weekly for about 6 years. > But 8 hours long? (unless I have the wrong thread in > mind) That's a church I wouldn't want to attend, ever!
In certain denominations and churches, the services run more-or-less continuously for the better part of the designated day, and people come and go during the service time as desired or needed.
I imagine that some really devout people (or people badly needing lives) do the whole enchilada.
It could also be a music festival, although I record the ones I do on a group-by-group basis. Of course there's a safety recording that runs for a number of groups (like after lunch break to start of dinner break) but I rarely need to actually use it.
>>>> I am dealing with long continuous live recordings, >>>> (using Audition) ,where the levels go up and down a >>>> lot, and need a good way to even them out afterwards. I don't want to >>>> use any sort of plugin compressor or >>>> limiter, I would rather manually adjust things.
>>> Sounds like a church service. Been there, done that >>> weekly for about 6 years.
>>>> I have Audacity and Audition but can't seem to find any >>>> way to do the same. What you can do is to hard select a >>>> region and adjust the gain. If you do this, there is >>>> obviously going to be a step change in gain at the >>>> beginning and end of this region which is totally >>>> unacceptable. Trying to fade these in afterwards is >>>> going to be a nightmare. There are no gaps in the >>>> recording that makes any of this possible.
>>> I break the recording up into zones where the levels are >>> consistent within the zone. The zones are bounded by >>> small regions basically background noise level. I >>> adjust the zones by selecting each one and applying an >>> appropriate gain correction. After doing this for 6 >>> years I kinda-sorta know that works for each kind of >>> zone.
>> I've tried this, and do it when I think I can get away >> with it, but cant adjust by any more than about 2dB >> without it being too noticable - I have no gaps at all.
> If you make an envelope in Cool Edit/Audition that starts out with 0 dB > gain and slowly ramps up to the desired amount of attenuation or gain, you > can use it to apply gain changes that are less noticeable.
>> As the event progresses, the faders gradually get turned >> up higher and higher, so the overall waveform is like a >> wedge of cheese, but with short term volume fluctuations >> I want to keep intact, thus no compression.
> Pretty strange. The usual goal of gain riding is consistent levels.
Not really. When you walk into a venue at opening time you do not expect the volume to be the same as when things are really kicking off with a full audience. Nobody would even enter the room.
> What can I say - the recordings I make tend to be uniform, with the caveat > that there are inherent differences between mixing live sounds for the > room, and mixing live sounds for a recording that need to be dealt with > when reformatting a "board tape" (actually a digital recording these days) > for distribution as a recording.
>>> In this instance it is a recording from a Live PA system feed. It >>> starts quite chilled out and quiet, and ends full on and maxed out. >>> You might find something similar with an Orchestra with soloists, or as >>> someone else mentioned, a Theatre show with dialogue and loud musical >>> numbers.
>> This is what we call "dynamic range." Enjoy it while you can.
> Yes, I am trying to reatain as much as possible of this dynamic range, I > don't like the current trend of Everything Can Be Sorted By A Compressor.
Well, unless you can train everybody to speak at a similar level, at a similar distance to the microphone, or 'own' the job and do the volume envelope thing for the full length, then lacking divine intervention compression/limiting is about all you can do.
Maybe pray hard for 8 hours and it will just 'happen', else just do the work.
> Its an unpaid job, Mike, I do it because I want to. I like to do the best > I can, within reasonable time constraints of course.
> Wavelab can do exactly what I want in half an hour tops, but not with such > large files - the technology is almost decades old. I was hoping to find > its successor. I'm not sure why something apparently so trivial is so > difficult to achieve.
>>>>> I am dealing with long continuous live recordings, >>>>> (using Audition) ,where the levels go up and down a >>>>> lot, and need a good way to even them out afterwards. >>>>> I don't want to use any sort of plugin compressor or >>>>> limiter, I would rather manually adjust things.
>>>> Sounds like a church service. Been there, done that >>>> weekly for about 6 years.
>>>>> I have Audacity and Audition but can't seem to find >>>>> any way to do the same. What you can do is to hard select >>>>> a region and adjust the gain. If you do this, there is >>>>> obviously going to be a step change in gain at the >>>>> beginning and end of this region which is totally >>>>> unacceptable. Trying to fade these in afterwards is >>>>> going to be a nightmare. There are no gaps in the >>>>> recording that makes any of this possible.
>>>> I break the recording up into zones where the levels >>>> are consistent within the zone. The zones are bounded >>>> by small regions basically background noise level. I >>>> adjust the zones by selecting each one and applying an >>>> appropriate gain correction. After doing this for 6 >>>> years I kinda-sorta know that works for each kind of >>>> zone.
>>> I've tried this, and do it when I think I can get away >>> with it, but cant adjust by any more than about 2dB >>> without it being too noticable - I have no gaps at all.
>> If you make an envelope in Cool Edit/Audition that >> starts out with 0 dB gain and slowly ramps up to the >> desired amount of attenuation or gain, you can use it to >> apply gain changes that are less noticeable. >>> As the event progresses, the faders gradually get turned >>> up higher and higher, so the overall waveform is like a >>> wedge of cheese, but with short term volume fluctuations >>> I want to keep intact, thus no compression.
>> Pretty strange. The usual goal of gain riding is >> consistent levels.
> Not really. When you walk into a venue at opening time > you do not expect the volume to be the same as when > things are really kicking off with a full audience. Nobody would even > enter the room.
By all means. But in the gigs I work, there is a pretty clear boundary line between the prelude and the main show. I guess it is not written in stone that it *has* to be that way.
Gareth Magennis wrote: > Its an unpaid job, Mike, I do it because I want to. I like to do the > best I can, within reasonable time constraints of course.
It's the unpaid work that always has the greatest time requirements. The reason is simple - the "customer" doesn't want to pay what it would cost to do the job.
Gareth Magennis wrote: > Not really. When you walk into a venue at opening time you do not > expect the volume to be the same as when things are really kicking off > with a full audience. Nobody would even enter the room.
Please, what IS this project? Who will be listening to the recording, and why? It sure doesn't sound like something that anyone would want to listen to for 8 hours straight. Why can't you break it up into logical pieces that you can manage?
Is this simply a product of nobody paying attention to the recording? And if so, why aren't the loud parts horribly distorted? That's usually what happens with a set-and-forget recording.
Gareth Magennis wrote: > Its an unpaid job, Mike, I do it because I want to. I like to do the > best I can, within reasonable time constraints of course.
> Wavelab can do exactly what I want in half an hour tops, but not with > such large files - the technology is almost decades old. I was > hoping to find its successor. I'm not sure why something apparently > so trivial is so difficult to achieve.
It's not trivial.
But it is relatively easy to achieve 'roughly', with say a hand drawn volume envelope on the displayed waveform. Dunno if your software can do that.
But if you want quality results, you sill have to put the effort in addressing each section individually..
|
