OK, I'm trying to compress some vocals. The first thing I tried was Tracktion's built-in compressor/limiter plug-in, with the vocal preset. It made no audible difference. I tried adjusting the compression ratio to see whether I could detect any audible difference, but no matter what, I couldn't hear any compression going on. I also tried adjusting the "threshold", which definitely did _something_, but it sounded more like EQ than compression.
So next I tried switching to the bundled Acuma mono compressor, on the vox preset. Now this one I could hear a difference. The signal was clipping and sounded weird, so I tried turning off the "soft clip" feature, adjusted the gain (or boost compensation or whatever it was called), and also lowered the attack time. Sounds pretty good to my ear, but since I tweaked things pretty randomly, it disturbs me that I don't have a good underlying mental model of what's going on. I simply don't understand what all the buttons and knobs on this compressor do.
So questions:
1. Is there any documentation on how these plug-ins work?
2. Any advice on how to get good results from the built-in compressor/limiter plug-in?
3. Can anyone explain the features of the bundled mono compressor?
4. How does the bundled compressor differ from the native Tracktion compressor?
Thanks,
Mark
Compressor question
-
- KVRAF
- 6490 posts since 14 Jun, 2004 from Rochester, NY
-
- KVRist
- 418 posts since 8 May, 2003 from Somewhere in the triangle...
links please?rpc9943 wrote:answer:
get buzfree comps
try grancomp. its a good one.
RonC
-
- KVRian
- 591 posts since 5 Feb, 2004 from southwest england
-
- KVRAF
- 6740 posts since 25 Mar, 2002 from sheffield, england
I think there is a Final Mix manual somewhere: the dynamics sections work the same way in the full band versions.mle wrote: 1. Is there any documentation on how these plug-ins work?
Good monitors and well trained ears.mle wrote: 2. Any advice on how to get good results from the built-in compressor/limiter plug-in?
Compressors are simply automatic volume controls. The native compressor is a simple design offering only downward compression with a hard knee: you choose a threshold, and when the signal exceeds this threshold the gain is turned down according to the ratio setting. (a ratio of 2:1 means that a signal entering the compressor 6dB higher than the threshold will leave only 3dB higher). The attack and release controls allow the gain modulation to be smoothed to suit the material. Finally, the make-up gain is simply a static volume control designed to restore the compressed peaks to their original level, thereby increasing the level of the uncompressed parts.mle wrote: 3. Can anyone explain the features of the bundled mono compressor?
4. How does the bundled compressor differ from the native Tracktion compressor?
The Mackie compressors are vastly more powerful and flexible: really they shouldn't be called compressors as they are capable of expansion also.. however, the default setting has just one node enabled and is set up to apply downward compression in the same way as the native version. When you start turning on extra nodes it becomes possible to set up "soft-knee" responses, or apply upward compression and expansion:
-
- KVRian
- 831 posts since 7 Sep, 2004
mle,
use percussive sounds to learn about compression. Not vocals.
use percussive sounds to learn about compression. Not vocals.
-
- KVRAF
- 2327 posts since 13 Apr, 2004 from Vancouver, Canada
Compression is a funny topic. MLE, compression shouldn't be thought of as an 'effect' that one would hear as soon as applied (unless intentional). Think of it more as a tool.
Nothing sounds worse than poorly used or over-used compression. For vocals, rather than soloing them, apply compression with all the other instruments in place. Essentially, you'd be listening for how uniform and well behaved the overall volume of the vocal track is in comparison to the other instruments. Soloing them to apply compression won't help much in that department!
Unless, of course, the volume of the vox is wildly dynamic, and all-over-the-place. In that case, it's almost better to re-record them then to try fixing with compression.
Just an opinion.
Nothing sounds worse than poorly used or over-used compression. For vocals, rather than soloing them, apply compression with all the other instruments in place. Essentially, you'd be listening for how uniform and well behaved the overall volume of the vocal track is in comparison to the other instruments. Soloing them to apply compression won't help much in that department!
Unless, of course, the volume of the vox is wildly dynamic, and all-over-the-place. In that case, it's almost better to re-record them then to try fixing with compression.
Just an opinion.
-
- KVRAF
- 6740 posts since 25 Mar, 2002 from sheffield, england
Nah, not if thats the only thing wrong: thats what volume automation is for.. or alternatively, split the vocal take into individual lines (or words even) and set the gain for each.advaya wrote:In that case, it's almost better to re-record them then to try fixing with compression.
-
- KVRist
- 195 posts since 31 Dec, 2004
I do alot of my gain setting at the clip level. Not just for vocals either. I find it faster than automating, more controlable, and easier. Just split out the word, drum hit, whatever, and adjust the gain. Or stretch it out and cross fade, overlay fade, etc. Most of my stuff doesnt even use the automation.IIRs wrote:Nah, not if thats the only thing wrong: thats what volume automation is for.. or alternatively, split the vocal take into individual lines (or words even) and set the gain for each.advaya wrote:In that case, it's almost better to re-record them then to try fixing with compression.
Sorry for going OT.
-
- KVRAF
- 1615 posts since 28 Mar, 2005
because compressors can change gain very quickly they can effect the tonal balance and character of sounds as well as loudness. And of course they emphasize different parts of the envelope depending how they are set up.
advice: pay attention to the attack and release times: for vocals these should probably not be too short.
advice: pay attention to the attack and release times: for vocals these should probably not be too short.
-
- KVRist
- Topic Starter
- 63 posts since 1 Jan, 2005
1. I think I now understand what I was doing wrong with the native compressor. Probably, the vocals were all below the default threshold. Since the native compressor doesn't show you any sort of graph or meters of the incoming vs. outgoing signals, it was just hard to realize what was going wrong.
2. I think the reason I'm used to "hearing" the effect is that I usually use a hardware combo device that integrates compression, expansion, noise gate, des-essing, exciting, and limiting. I have a pretty good understanding of what that "sounds" like. In particular, I know what expansion+compression sounds like. The expansion makes the breaths louder, for example. I didn't realize that these compressor plug-ins are only for "downward compression". It's basically just making everything quieter to varying degrees, and it didn't sound like what I expect.
3. It sounds like the Mackie Compressor is a pretty powerful tool once I understand it. So can someone please explain, or point me to a reference for, each little feature of the interface. For example, once you're done placing all the nodes, what does the graph really mean? What is the difference between soft clip and not. What does auto makeup do?
Thanks,
Mark
2. I think the reason I'm used to "hearing" the effect is that I usually use a hardware combo device that integrates compression, expansion, noise gate, des-essing, exciting, and limiting. I have a pretty good understanding of what that "sounds" like. In particular, I know what expansion+compression sounds like. The expansion makes the breaths louder, for example. I didn't realize that these compressor plug-ins are only for "downward compression". It's basically just making everything quieter to varying degrees, and it didn't sound like what I expect.
3. It sounds like the Mackie Compressor is a pretty powerful tool once I understand it. So can someone please explain, or point me to a reference for, each little feature of the interface. For example, once you're done placing all the nodes, what does the graph really mean? What is the difference between soft clip and not. What does auto makeup do?
Thanks,
Mark
-
- KVRAF
- 6740 posts since 25 Mar, 2002 from sheffield, england
The graph represents input level (x axis) against output level (y axis).
The default preset shows a diagional line up to the first threshold node, ie: x = y, no modulation is taking place. Above the threshold the line drops towards the horizontal indicating that an increase in x (input) results in a smaller increase in y (output).
If the line you create is below the diagonal x = y line, then you are applying downward processing, and will likely need some positive make-up gain to compensate. Above the line means you are modulating the gain upwards, and any gain compensation should be negative. Auto make-up gain calculates how much gain modulation would be applied to a fixed refernce input, and applies the opposite amount of make-up gain automagically.. if you turn off the option the slider will jump to the actual amount applied rather than turn off the make-up gain which is a nice feature IMO (no big jumps in level: you just get to see the real value and know it won't change if you make further tweaks to the transfer curve)
The default preset shows a diagional line up to the first threshold node, ie: x = y, no modulation is taking place. Above the threshold the line drops towards the horizontal indicating that an increase in x (input) results in a smaller increase in y (output).
If the line you create is below the diagonal x = y line, then you are applying downward processing, and will likely need some positive make-up gain to compensate. Above the line means you are modulating the gain upwards, and any gain compensation should be negative. Auto make-up gain calculates how much gain modulation would be applied to a fixed refernce input, and applies the opposite amount of make-up gain automagically.. if you turn off the option the slider will jump to the actual amount applied rather than turn off the make-up gain which is a nice feature IMO (no big jumps in level: you just get to see the real value and know it won't change if you make further tweaks to the transfer curve)