Amateur alert...
By studying different rap/hip hop waveforms I see that the bass drum/snare are usually the peaks of the waveform.
That's why I noticed that most of the instrumentals I produce are not that loud.
When normalizing to 0dB, there's only like one sample that peaks at 0dB which is like a digital click/pop or whatever. The easy thing for me to do is clip at the level of the drums then normalize, but that isn't very smart. All I'm doing now is compressing the finished mix with the threshold being the drums level [which I usually get through guess and check...]
I usually compress drum tracks.
How can I get the drums to be the peak - eliminating those couple of samples that throw it off?
[Yes I know someone's gonna say get better at mixing, but let's pretend I was very good at it and am still having this problem...]
Comments
You're really gonna think this sounds a bit different but it wor
You're really gonna think this sounds a bit different but it worked for me. When I first got my system together I invited a drummer friend over to do some mike placement tests. In the process of listening to some stuff we had recorded after we finished, I noticed a heavy click on some of the kick hits. Turns out that a mic I had tried couldn't handle the SPL so I ended up with this pop. It wasn't all that important because we were just "screwin' around" but I thought "What if something like this happens for real?" What can I do about it without using punches or re-recording the entire track? I tried compressing, limiting, EQ, all of the usual stuff. None of it got rid of the pop without taking away from the sound of the kick. Then I had an idea, these clicks are in a high hz area around vocal range. The software I use has a de-esser tool. I ran the de-esser on the track which removed most of the click and made it almost inaudible without really killing the sound of the kick. When I added the rest of the drums back into the mix the masking effect took care of the rest of the pop.
As for not getting the pop in the first place. Use a mic that can handle the SPL for one (a big DUH on my part!) and secondly, record at a level where you're sure you won't get digital clipping and then normalize to 0db. If your software works well it should see the highest peak and then adjust itself accordingly so that the highest peak represents 0db without producing any digital clipping.
As for Normalizing, I only use it on tracks that aren't loud enough to get a decent mix with and stay away from it if at all possible. :w:
I also never normalize. Especially before mixing, as it takes aw
I also never normalize. Especially before mixing, as it takes away headroom one might need after adding other processing like EQ. The only time I might consider it is if I got a particular track that was recorded at an extremely low level, but I still wouldn't normalize it to zero, just boost it to a reasonable level.
On a finished mix I'll use a compresser or L1+ to boost overall level if necessary. But I still will try to leave room for the mastering engineer to do his/her stuff!
Try not going all the way to zero. -0.5dB would be fine. I did a
Try not going all the way to zero. -0.5dB would be fine. I did a test recently on consumer cd players. On over 50% of them, the output circuitry of the CD player (analog out) was clipping with peaks of zero. Zero should mean 0.775Volts of output...but it does not always work that way with consumer equipment. Excuse me if I am wrong...but it is consumers we are trying to get our music out there to...and you can get a really fat, powerful sounding CD without running to the limit every time. Now certain peaks could go to zero for a 100th frame...but the last 2 mastering jobs I got were so "scrunched" that the top of the wavefile looked like a flat line (not clipped but scrunched) when using a non magnified setting and they sounded like ass with no dynamics. When you mix, you should do so for mastering purposes. I would rather receive a non normalized track at -15 peak and let the mastering engineer (me) go for the higher level..than to compress the crap out of your mix and have to create dynamics on the backend. Remember...it does not have to be the loudest CD out there. when digital made it's debut in 1975...-17 was the zero standard. It is because of those who started running the gain all the way to zero is why the recording arts suffer. We do not have to compete with Britney levels to make a fine product. Most jazz recordings are done with peaks of -1 and average of -20. I have run across some hip hop that was done at -2dB peaks and sounded killer. Plenty of gain. I shoot for (hip hop and Rock) good peak levels and a cool -15 average. (depending on material) I have yet to get a complaint from a client that it was not "loud enough" By having the extra headroom, the output circuitry of your CD player (analog out) has less distortion and my CD's appear just as loud..if not louder and larger than the scrunched to death 0dB peaks and -5dB average. Try it and see what you think. I do have some sophisticated software that lets me go to +6 on the backside without digital clipping...but what is the use if the CD player is not calibrated for those levels. Use common sense. My biggest pet peeve of this industry is the "gain race" to see whose product is the loudest. It is unneccessary and yields poor craftmanship for the mastering engineer to deal with. 2 Months ago , I was x fering some analog open reel to digital. On hard drive recording...I did not go past -10. Once I mastered it..I knew all my peaks would be proportional...giving the live feel and vibe. They make weighted keys on instruments for a reason!
If you have the software, easist solution is to draw down the co
If you have the software, easist solution is to draw down the couple of offending waveforms with the pencil tool.
Sometimes digital pops are of such short duration you might be able to cut them out and slide the waveforms back together to cover the gap. But that only works if they are REALLY short.