Karaoke Forums

Ok, so I've touched on this before, but I want to chime in here and give you my 2 cents, take it or leave it. I have tracks that are anywhere from 128 to 320. Some of the tracks that are at 128 sound every bit as good as other higher bit rate tracks. What I have found is there are tracks that are overly compressed during mastering at the studio. I'll give you a couple of examples, both are Sound Choice , Deep Purple - Smoke on the Water & Matchbox 20 - Real World. Smoke On The Water sounds fine until the drums come in, the drums are so squashed by compression that they loose all of their punch, but the track was ripped at 320. On the other hand Real World is recorded at an overall lower volume than most tracks also ripped at 320. These are just two I can think of right off the top of my head. So what can you do about these tracks? Not much except find better ones or deal with what you have.

I have done a number of things in order to try to get the best sound out the speakers I possibly can with the tracks I have. I use Breakaway, a sound plugin I know several people on here have been using for years. An external sound card such as the Lexicon Alpha, also gone over here more times than I care to count. I do a bit of EQing on the channels that the karaoke tracks play on, done with a fully parametric EQ. A small boost in the lows peaking around 50hz, a cut in the mids, which btw is where most ear fatigue comes from, and a small boost in the highs, shelving that starts around 7khz. So with the slight cut in the mid range I have carved out a space for vocals to fit. Some tracks need more of something and some need less but I leave that to my ear to pick out something that just ain't right. EQing vocals as well as adjusting gain or fader on each singer is another novel I could write, but it all plays an important role in how everything fits together and sounds.

So where am I going with this? Well...compression! Were not dealing with a live band here, were dealing with recorded tracks mastered in a studio. So even though we have a live singer we have to try to match the level of compression on the singer to be somewhere close to the level of compression on the track or the vocals will stick out like a sore thumb. And of course we add some effects to gel it all together nicely. There are some tracks that are easier than others to make things fit but for the most part I can get it done no matter if they are 128 or 320. It's a totally different animal than mixing a live band and I've been behind the board for both as well as mixing in the studio, all three are different.

Bottom line for me is this, we are dealing with a bunch of vastly different tracks recorded by different producers, musicians & companies. I think it has a lot less to do with bit rate and a lot more to do with mastering. We as hosts have to recognize with our ear what it takes to make it sound right.

_________________
KLAM Entertainment..."Come Out Of Your Shell And Sing"

TheMechanic: Your post made me smile...thank you. Well Put.

I've noticed the compression issue as well and usually drop the track (rock) if I have something that feels better.
I do run a quad compressor (one comp per mic) at roughly 3:1. Helps a bit.
I also use an external card (Have to ground lift it due to laptop psu noise) and ext eq. Will check out breakaway.
The mid cut trick seems to escape many people including "Pros"

_________________
Purple Frog Karaoke Cuz we all feel odd and love to "croak"

I'm glad you enjoyed. Breakaway is great it really helps bring out the best in compressed music. I highly recommend it, for thirty buck you can't go wrong. You will have to experiment with the different settings. I start with a preset called "Reference Settings" and dial back a couple settings to give the music a bit more dynamic variation. Each setting is explained fairly well by hovering over the setting with your mouse.

This thread did get me looking at my Karaoke-Version files though. So I downloaded some of the files I've purchased in different formats and found the MP3+G files audio is at 160, MP4 is the worst at 128 & WMV the best at 260. So I guess WMV would be the way to go as long as your player supports it.

_________________
KLAM Entertainment..."Come Out Of Your Shell And Sing"

Screamer and Mechanic:

Where do you center the mid cut on backing tracks? I'm not using a parametric but I have a mid-sweep on the channel EQ. Would you give a corresponding boost at the same spot for the vocal to make it pop out more?

Depends on the track if I had the luxury of a sweep eq on my kj board... 1k-3k track mid.
You have to play with it to find "Your" sweet spot against the way you want your lead vocals to sound and the timbre of the singer. We all perceive our idea of a mix differently.
For the guys you may want to go as low as 500 to get the guitar out of the way!!
6.3k up a tad on the master eq (vocal definition), 4k dip on master eq ( a pain freq typically bad with titanium tweeters ). 16k up a tad (air).
I prefer rubber surround woofers for KJ use so I don't need to do much in the mid bass or sub bass region.

_________________
Purple Frog Karaoke Cuz we all feel odd and love to "croak"

If I were working with analog board with a mid sweep on the music channel I would probably center it somewhere around 600hz give or take. Probably cut it 3-5db. Since you are dealing with a semi-parametric EQ you can't adjust how narrow or how wide of a range you cut or boost. So that is going to depend on the board.

Ok, on vocals, I am working with a digital board so I can be a bit more precise, but I hardly ever boost anything on vocals except the highs a few db for intelligibility and breathiness. Everything else is cut, first the high pass filter is engaged, second I make a cut in the lows that ranges from where the high pass filter leaves off to roughly 400hz, centered on about 250hz. Third is where your mid range sweep comes in, this will vary from singer to singer. You are going to make a cut here based on the most dominant frequency in their voice, everybody has it but it varies from person to person. This is going to range from 400hz to about 1k. Whenever I try to explain to someone how to find that frequency I tell them to sweep it back and forth slowly until they find that sweet spot. You will know you found it when all of a sudden their voice clears up. It's actually the opposite though, that sweet spot means you found and cut out the most obnoxious frequency in their voice. I make a small cut around 2k this minimizes the consonant sounds like "C" & "K" this cut is not as necessary as the others in most cases and probably could be left mostly flat in this area. So on an analog board I would cut a few db on the LF knob, cut a few db from the MF and sweep to find each persons offending frequency. Boost the HF a bit. If you do it this way you will find the voice to sound much more clear and natural. You also should be able to get a bit more gain out of the mic channel and have a bit more headroom on the fader before feedback.

These EQ settings I have outlined are based on using a Sennheiser e835 microphone and should be very similar for a SM58, different mics have different coloration and may not need as much cut out of the bottom end, this is true with my beta 58a and especially my beta 87a.

I am attaching a picture that should illustrate what I am saying fairly well. Hope this helps.

_________________
KLAM Entertainment..."Come Out Of Your Shell And Sing"