Software compression

Discussion in 'Tracking / Mixing / Editing' started by Unregistered, Apr 5, 2011.

  1. Unregistered

    Unregistered Guest

    Is there such a thing as software compression that runs while you are recording?
     
  2. TheJackAttack

    TheJackAttack Distinguished Member

    There are compressor plugins. There is no inherent compression in digital recording like there was in reel tape.
     
  3. RemyRAD

    RemyRAD Well-Known Member

    Unfortunately this nomenclature can be confusing. When talking about audio, there really is 2 vastly different kinds of compression. Here's the lowdown:

    When many of us record certain vocals and/or instruments, some of us choose to utilize dynamic range compression on certain sound sources. Some people pooh-pooh this Method of processed recording to leave this process for after the recording has been made. That doesn't stop me as I came from analog tape days back in the 1970s and that's what we did to keep the noise level of the tape from becoming too obtrusive.

    Then there is the style of data compression. This is where formats such as MP3, MP2, WMA (Windows media audio), QuickTime, MPEG 4, etc.. These are delivery formats that compresses the data down by more than a factor of 10. This has no effect on the dynamic range whatsoever.

    If you want to record with dynamic range compression while recording, there is only a couple of pieces of software out there capable of doing that and is generally not the accepted practice except by folks like, well, me. But I don't utilize the software when I want dynamic range compression, I utilize hardware devices such as DBX, Universal Audio, Orban compressor/limiters which then get recorded to uncompressed .wav files. I actually have software that will allow for real-time data compressed recording at MP 2 and/or MP 3 but that is something you absolutely don't want to do. Data compression such as that is " lossy". And with that, it has a tendency to impart its own audible anomalies/artifacting. Great if you like that underwater kind of sound. NOT

    So there is a crash course in compression and/or, compression.

    So, what kind of compression do you think you might be asking about?
    Mx. Remy Ann David
     
  4. Unregistered

    Unregistered Guest

    I guess I could be completely wrong with the terminology (still pretty noob-like) but what I'm looking for is a piece of software that will cap the volume @ a certain DB while tracking to avoid peaking.

    I thought compression was the right word, but I could be totally wrong.
     
  5. TheJackAttack

    TheJackAttack Distinguished Member

    Software (or hardware) that limits to a certain dB is called a limiter. This is a compressor with a specific function. Now a compressor is one of the most abused pieces of gear/plug out there. If you crush something by hitting the limiter too hard then you will have ruined the track. That is not to say don't hit it or even in some instances hit it hard, but a limiter for laying down tracks is a safety net. In 24 bit recording it is better advice to not record hot enough to hit your limiter at all and bring the tracks up within the DAW itself. Use the limiter for a safety if you wish but know that too much is not good and tracks that are too hot are hard to mix.
     
  6. bouldersound

    bouldersound Real guitars are for old people. Well-Known Member

    Generally speaking software compressors or any other DSP is applied after recording. I can't think of an application where it would be necessary or useful to process the signal after the converters but before it's written to disk. Why exactly to you think you need to do it that way?
     
  7. Big K

    Big K Well-Known Member

    But you can record without any compressor od FX and still listen to the recordings compressed and with fx ( just not too many and no slow ones) by inserting the FX on the Tracks playback channel.
    If you own a reasonably powerful system and a decent DAW & sound card ( good drivers!!) you still can achieve latencies as small as 1 millisec, which is even fast enough for headphone mixes.

    Most, if not all, DAWs can insert a compressor in the recording channel.
    If you know what to do, there is no problem with that. Haven't we all recorded to tape using compressors or gates in the front end? Be adviced with DAWs it is not necessary and actually rarely done, though. Just leave a safe headroom and off you go... 24 bit is your friend...
     
  8. RemyRAD

    RemyRAD Well-Known Member

    Using a limiter to prevent overloading 0 dbfs (Decibels, full scale) generally has to be done with a hardware limiter post preamp, pre-converter. Numerous Multitrack audio interfaces already have these built-in. While it can keep you safe, they can also keep you from making good recordings. You'll get lazy with your level setting leaving the limiters to make up for bad engineering technique. Even at 16 bit 44.1 kHz recording, you have 96 DB worth of range. That's nearly the full available dynamic range of the microphone preamp you're using. If you can't set your levels a little lower to prevent overload, you shouldn't be allowed to drive either. You don't drive with your eyes closed do you? You'd definitely get some crunchy sound that way. Not including the time it takes to fix it or replace it with costs a lot more money in the end. Recording is the same way. This is part of your learning curve. You guys have it so much easier with 96 DB worth of range compared to our 45-65 DB worth of range in analog tape. And we still managed to make good recordings that weren't distorted unless we wanted them to be (as in tape saturation, which is nonlinear when it starts getting good N' crappy. Digital doesn't do that. Digital as you know just gets crappy but there are caveats to that too) . Here's a good example. You can overload digital in very precise & calculated ways. Because digital is linear, so is the clipping. Pure odd harmonics which are dissonant & not considered musical as they can not be created in real life. Extremely short duration drum peak transients can be utilized as a small method of creating a more definite peak sounding dynamic. But this procedure is better left to deliberately cause that in postproduction as opposed to your initial tracking session. Drum peaks can be slightly " over normalized" in the software. Then you reduce the level overall from -.6 db to -1 db. This keeps the converters from "clicking" from exceeding the Converter & output amplifiers capabilities. This can only be done with very short duration peak transients. This can actually give the drum set a little harder smack to its sound. But that's not like overloading converters or amplifiers. That's done as an enhancement trick. Many Aural Exciters actually had controls to vary between 2nd & 3rd harmonic distortion characteristics. When you do what I've described, you only get the 3rd, odd order enhancement. And that's just plain everyday free to those that know how to do it without a plug-in. It works great on drums, I don't do it to much anything else as that's where clipping just becomes clipping.

    I was clipped over 10 years ago
    Mx. Remy Ann David
     
  9. bouldersound

    bouldersound Real guitars are for old people. Well-Known Member

    I don't use a lot of different DAWs, but I'm not aware of any that can apply effects to the audio before it hits the disk. Am I missing something?
     
  10. Big K

    Big K Well-Known Member

    For example the recording/input channels of Cubase and Nuendo.
    Those are full, pre-harddrive, recording channels with EQs and FX inserts and automation, but no sends.
    Inserts are nice to have, but I don't use them.
     
  11. Unregistered

    Unregistered Guest

    Thanks for all the responses!

    Picked up Cubase and it does a wonderful job. So many options to mess with but hopefully I'll figure it out.

    I pretty much needed something to keep my vocal line from peaking during recording. Besides getting better hardware and software I had tried everything i could think of but nothing was really working. It seemed like without the limiter in Cubase I was stuck with three extremes.

    1) Peak every time I vocalized (big voice with very high resonance frequencies, recording art songs/music theatre/operatic arias)
    2) stand so far away from the microphone that I peak less (but still peak)
    3) Turn down gain and mic volume so much that it hardly sounds like anything (at least it didn't peak)

    With Cubase my vocal lines sound a lot better! I need to try and figure out a way to try and keep some depth to the sound though. It doesn't sound tinny, but its not at rich as I would like.

    Anyhow, thanks again!
     
  12. TheJackAttack

    TheJackAttack Distinguished Member

    What is the issue with 3? All you do is normalize it once it's in the can. That is what a professional studio would do. If a pro studio used a limiter for a vocal-which happens on some occasions-it would only hit the limiter very very few times. What you are describing is too much gain. Too much peaking. Turn the gain down. When you are done, select the whole wave and normalize to -3dB. Now use your VST FX and run your faders as necessary within Cubase.
     
  13. Big K

    Big K Well-Known Member

    What can you do:
    Look at the mic, if there is a pad button to lower the sensitivity of the mic > lower output.
    turn down the mic pre-amp. Check if there is any limiter in your frontend and if so, test if it sounds any good. Increasing he distance to the mic is only good to a certain point. When starts to sound bad and you pick up too much room, you'll have to use a c ompressor / limiter unit between mic-pre and A/D stage. If you record with a pre-amp featuring a digital out, already, you might want to buy another pre-amp, a small mixer or a mic with less output.

    I know the problem well, because I have a rather loud voice,too. The first time we went in a studio for recording it was an engineer's nightmare. Using a U97, -10dB pad in, more distance to the mic, pre-amp on the console's all the way down, the console attenuation pad in and the meter on the desk still clipped happily away. Finally, we used a SM57 and it did the job, eventhough we did not get the Neumann sound, which would have been really nice.

    These days, my voice is not as loud, anymore, and I have learned to sing just as well with a more controlled volume.
     
  14. BobRogers

    BobRogers Well-Known Member

    When I got started with digital recording I always tried to track too hot. From the posts we get around here it's a very common problem. With 24 bits there is plenty of room to get an accurate picture of a signal with a lot of dynamic range. You say it "doesn't sound like anything," but the sound is really there. You just have to turn it up. Now yes, you might have to compress or limit it a bit so it doesn't make the mix clip. But it's far easier to a good job of this in the box than when you are tracking (where you have to get it right the first time).
     
  15. bouldersound

    bouldersound Real guitars are for old people. Well-Known Member

    Interesting, but I've never really felt the need for another layer of adjustment. That could be because I usually have a mixer front end with effects, eq and compression if i want it. If all I had were preamps into the interface I might want more control before the signal hit the drive.
     
  16. Big K

    Big K Well-Known Member

    With a decent DAW, there is no circumstance I can think of to compress ITB before the harddrive.
    Direct listening is not influenced and the important bit is not to clip the front end ( pre-amp & A/D stage )
    As you already said, it is all better done later, when mixing. Bringing a recorded signal up by 10, 15 dB is no problem once recorded properly.
     

Share This Page