The Song Fight Bulletin Board

Thanks Spud. The only links that didn't work, jb posted them above, so I heard them all. The truth is, the original wasn't too bad, just needed to be a little fatter and wider, which is what jb did in his re-mix.
I listened to each mix before I read what was done, to see if I could hear for myself what was done. The simplest mixes seem to be my favorite.

We used the SMU remix on our last CD.

That was a cool mix too, because it was so fast compared to the rest. But it was way over produced in my opinion. But I guess you wouldn't have put it on your CD if you didn't like the way it sounded. But can you do it live just the way it sounds on your CD? I guess I'm always thinking along those lines.

Sven wrote:If you ask someone like blue, he'll tell you that louder is better. This is acheived by using T-Racks, or Izotope Ozone, or any of several other multiband compression/mastering apps/plugins. I personally think these things squash the hell out of the dynamics, but I use them myself, so there you go.

lies and slander! your tune should be as loud as is reasonable - no one wants to sit there and ride the damn volume because some motard couldn't even figure out how to normalize their shit.

I can tell you right away that your guitar tones are a big part of why this mix is assy. They are all drippy and incoherent, there's no definition to anything. Listen to Brad Sucks or Smalltown Mike or Gert for a better example of some good rock power chord sounds. And yeah, too much reverb.

I think most bad mixes come from bad sounds, and most of your sounds are clearly not pristine recordings. It's really freakin' hard, if not impossible, to make bad things sound really good after the fact.

Upload VBR mp3s per track, mono if possible. I'm sure lots of ppl would love to remix it, and then you can ask specific questions about specific sounds. Holler if you need hosting space.

Plus I will erase your crappy solo and be able to listen to the tune in peace.

Billy's Little Trip wrote:How would I go about giving you the tracks?

blue gave the gist, and either him or I can give you hosting space.
basically you just need to bounce down each track individually. high bitrate mp3s or wav files or fine, but I don't know your bandwidth situation.

the thing is they must all be from the start of the track, so we can just blindly drop them on top of each other and not have to line things up.

I for one am not interested in 10 drum tracks. One way to go is to give the kick drum track and the snare track and then a 3rd with all the other drum parts.

When you get that together, we can find out how you want to distribute them. email is a bad idea.

-craig

Billy's Little Trip wrote:Where is the in between point to sound good in any situation?

The times when reverb is needed on the master bus is outnumbered by the times when it isn't by 2 or 3 orders of magnitude. When in doubt, leave the reverb off.

The problem you run into is pre-delay: We're used to hearing reverb reflections arrive at different times based on the sound source's distance from us, and distance from any reflecting surfaces. When you use the same reverb for the vocals and drums, our brains get confused, because the vocals are supposed to be much closer than the drums, yet the reflections arrive at the same time.

Billy's Little Trip wrote:How would I go about giving you the tracks? Email mp3's one at a time? There are something like 10 tracks just for the drums alone. I'm using Cubase, is that an issue?

It's easiest to use MP3s, as blue suggested (though for a "real" project, you'd use uncompressed WAV files.)

Export the tracks one at a time, to separate files, from Cubase. And this is really important: DISABLE ALL EFFECTS when you export. You want to leave the tracks raw, and the let mixing person decide what treatment to use.

Give the MP3 files useful names.

drums-snare.mp3
drums-kick.mp3
guitar-electric1.mp3
vocals-lead.mp3

That kind of thing.

Zip the MP3s into a single archive, and find a place to host it. (Blue offered space ... or you could place it on YouSendIt or one of the similar large-file sharing sites.)

Billy's Little Trip wrote:How would you line up the tracks?

Every file you export should start at 0:00. Pretend you're sending individual tracks that you recorded to tape. For each, you'd have to rewind the tape before starting the export.

This means that for some of the tracks, like the guitar solo, there'll be long sections of silence at the start. That's why blue suggested using VBR encoding for the MP3s: It does a much better job of compressing files with silence in them.

As long as all the MP3 files start at the same place, lining them up is trivial.

One other thing: Your guitar solo sounds like you connected the guitar directly to your PC, and ran it through an amp simulator. Is that the case? If so, export the raw version of the track along with the "amp'd" version, because the person remixing will probably want to re-amp it.


edit: makes more sense now

blue wrote:I think most bad mixes come from bad sounds, and most of your sounds are clearly not pristine recordings. It's really freakin' hard, if not impossible, to make bad things sound really good after the fact.

This is called "Cooking with shit."
Or maybe "directing Paul Walker."

It doesn't matter how good a cook you are, it's still a movie starring Paul Walker.

deshead wrote:

Billy's Little Trip wrote:Where is the in between point to sound good in any situation?

The times when reverb is needed on the master bus outnumber the times when it isn't by 2 or 3 orders of magnitude. When in doubt, leave the reverb off.

The problem you run into is pre-delay: We're used to hearing reverb reflections arrive at different times based on the sound source's distance from us, and distance from any reflecting surfaces. When you use the same reverb for the vocals and drums, our brains get confused, because the vocals are supposed to be much closer than the drums, yet the reflections arrive at the same time.

Des, please go into further detail. This is new information to me. (Maybe it shouldn't be, but it is.)

Note: My first impulse was to type DESHEAD FOR THE LOVE OF GOD EXPLAIN THIS SOME MORE OR I WILL KILL YOU AND NO ONE WILL HAVE THIS INFORMATION. But I decided a calm, medicated response would be best.

I would also like to emphasize what somebody said about preferably exporting your tracks in mono. Exporting in mono is very easy in Cubase, it's just a checkbox in the export dialog, then it will ask you if you're sure you want to do that and you'll click OK. Make sure to center all your tracks, disable effects (as mentioned), disable eq, etc. Get them back as close as possible to the state they were in right after you recorded them, before you did anything to them. Tell people the bpm of the track, too, so they don't have to figure it out. As somebody said, for the drums, limit the number of tracks. At most you should do tracks for kick, snare, overheads, and toms, if not just tracks for kick, snare, all else, as suggested. Etc.

jb wrote:Des, please go into further detail. This is new information to me.

Picture a room 20 feet deep. Ignore the side walls. You're standing at one end of the room, Bjam is 5 feet in front of you (so 15 feet from the back wall) and blue and his drum kit are 5 feet behind that (so 10 feet from the back wall.)

Sound travels about 1 foot per millisecond. So Bjam's voice will reach you directly in 5ms. Her voice will also reflect off the back wall, and that echo will reach you in 35ms (i.e. 15 feet to the back wall, then 20 forward to your position.)

The sound from blue's drums will reach you in 10ms. The drums will also reflect off the back wall, and the sound will reach you in 30ms.

So the reflections of Bjam's voice arrive 30ms after the direct sound where the reflections of the drums arrive only 20ms later.

Here's a sketch to illustrate. Note the reflected sound has further to travel the closer the source is to you.


Our brains use these differences in arrival time between intial sound and reflected sound to determine dimension and position in a room. In practice, it's much more complicated than my examples, because you also need to consider the side walls, and the floor and ceiling, and the sheer girth of blue's drum kit.

But in general, with a reverb unit you control how a listener's brain determines distance by manipulating the time between the intial sound, and the first reflections. On most units, there's a knob marked "pre-delay" for setting this.

As it relates to my comment to BLT: When you put reverb on a mix, rather than on the individual elements in the mix, you're effectively telling the listener's brain that ALL the sounds in the mix originated at the same position. But your other mix decisions, like volume, panning, and EQ, probably betray this. So the result is a confusing, washed-out sound.

Here's the best article I could find on the principles: http://www.soundonsound.com/sos/1997_ar ... asics.html

So in practical terms, do reverb units/plugins have a standard measurement? What does "pre-delay" mean... is the number I set that to supposed to be the distance from the "instrument" to the "listener"?


You mentioned room size and such. I've noticed that my reverb doohickeys include a setting for room length and width, but I don't know how to use those settings. I just fiddle until it sounds good. What's the difference between width and length (as in "width means how clangy the sound is, length means how flurfle it is)? I guess a nice couple of examples would do well to illustrate the difference, if you or anyone has a spare minute. I know I'd find them useful.

Lunkhead wrote:I guess the lesson is, if other folks remix your tune, save a backup copy. ;)

Save a backup copy? I still listen to these remixes! Here they are:

• My original mix (submitted to the fight)
  Blue's raw mix (just levels and panning)
  Blue's final mix
  Bell Green's mix
  and (a must hear) Future Boy's remix

Here's some of what I learned from the experience:

• Listen to each track seperately before mixing anything together. Nowadays, I bus my tracks and add a little EQ and compression immediately after recording.
  Think about the panning before panning anything. I used to move sounds around willy-nilly, putting piles of sound in the most available free-space, like papers on my messy desk. Now I plan my pans before I start recording. I don't record something if I don't know where I'm going to put it.
  Go easy on the EQ. I used to think of the frequency space the same as the panning space: instruments have their preferred place in the spectrum, but EQ could push them "up" or "down" to make room for other instruments. No matter how mushy things got, I was convinced I could sort out the sound with EQ, and the harder I worked at it, the more convinced I became that I could do it. Problem is, when I put my tired ears to bed and listened again all fresh, it sounded like crap. When I took the mix to any other sound system, it sounded like crap. When I played it for my wife, it sounded like crap. EQing is like staring into a mirror: you feel like you're learning something, but you're really just losing touch with reality.
  Bass is special. Only one instrument should carry the bass. Instruments can share the high frequencies really well, but fight like betas over the bass. I know an awful lot about human hearing and for some reason this is still hard to put into practice.
  Compression sucks. I love to compress. I still do it. But it sucks. It flattens sound and squashes timbre. Always compress your vocals, but think twice before compressing anything else.
  Don't be afraid. Probably the most important thing. Blue's mixes made me cringe at first, because he didn't hide sounds I was embarrassed about. But that's just it, half of my mixing process was concerned with covering up things I didn't like (turning down wonky vocals, putting chorus on poorly-tuned instruments, EQing the midrange out of percussion that sounded wimpy, etc.) Stop it. No one else cares. Mix with love, not fear.

I still love this song. The percussion is a soda can, a beer bottle, and a big tupperware pot. I'm kind of proud of that.

I've thanked Blue, Bell Green, and Future Boy extensively for their work already, but it doesn't seem enough. That anyone would choose to spend an afternoon pushing my sounds around their heads still astounds me. Thanks again, guys.

Damn, this thread is gold. I was a little concerned that the guru's would start talking over my head, but surprisingly, I understand everything that is being said here. I have been recording and playing around with sound for years, so maybe that's why. I just need to learn some terms, like VBR. I've seen it, but had no clue, so I looked it up real quick. Then realize the reason I didn't know that VBR means variable bit rate is because it's a digital term (for other novices that are making the switch to full digi like me). The "computer" recording and mixing is part of my new quest as well as learning how to properly mix sound and master.

By the way, handing over all my tracks raw and dry is going to be like standing naked in front of a large group of people.............on a cold day.

No problem on the drums. I can easily break them down to 4 tracks. I just remembered the reason I have so many drum tracks is because I add fills later and put them on seperat tracks until I decide if I like them. I can easily mix in the fills on the tom and snare tracks.

starfinger wrote: I for one am not interested in 10 drum tracks. One way to go is to give the kick drum track and the snare track and then a 3rd with all the other drum parts.

I, on the other hand, would much prefer each part to be on it's own track. So I guess the question is, are you reasonably happy with how the drums hang together? If you are, then there's perhaps no point in anyone putting effort into the drum side of things and having them all on one track would make sense. If you think there's a problem with the drums, then it'll be difficult to do anything with them if they're lumped together.

jb wrote:So in practical terms, do reverb units/plugins have a standard measurement? What does "pre-delay" mean... is the number I set that to supposed to be the distance from the "instrument" to the "listener"?

Pre-delay tends to be measured in milliseconds, and is the length of time to wait before adding the 'verb.

jb wrote:
You mentioned room size and such. I've noticed that my reverb doohickeys include a setting for room length and width, but I don't know how to use those settings. I just fiddle until it sounds good. What's the difference between width and length (as in "width means how clangy the sound is, length means how flurfle it is)? I guess a nice couple of examples would do well to illustrate the difference, if you or anyone has a spare minute. I know I'd find them useful.

I dunno what reverb you're using that has seperate width and length (they mustly just specify size). But I would assume that if you had a large length but small width (stop sniggering at the back), half of the reflections would arrive quickly (from the sides) and half of them would turn up later (from the front/back). Just like if you were to stand in a long narrow hall and sing.

I almost feel ike I'm deshead's official opposition here, but I feel compelled to point out that the thing about not using the same 'verb on the same tracks only matters if you care about having your mix sound 'realistic' (ie, like something a live band might have once played).

obscurity wrote:I dunno what reverb you're using that has seperate width and length (they mustly just specify size).

width is probably the stereo width, and the other is the reverb time.

-craig

Yeah, that makes more sense.

Mostess wrote:Don't be afraid. Probably the most important thing. Blue's mixes made me cringe at first, because he didn't hide sounds I was embarrassed about. But that's just it, half of my mixing process was concerned with covering up things I didn't like (turning down wonky vocals, putting chorus on poorly-tuned instruments, EQing the midrange out of percussion that sounded wimpy, etc.) Stop it. No one else cares. Mix with love, not fear.

I think that's an important point for a lot of SF! folks who are new-ish to the recording thing. Don't try to cover up those aspects of the performances with post production techniques. You should either just live with them, or fix them by doing more takes. (Tune your guitar and try again, sing that one line over another 50 times, etc. etc.) That might actually make your performances improve over time, so that eventually you'll be able to get better takes sooner and you'll have less to do in post to "fix" things. Fudging things with EQ and effects isn't going to fool anybody anyway, and it's definitely not going to help you improve as a musician. At least, that's my opinion, as off topic as it may be...

Sven wrote:Read this.

The article Sven linked to wrote:Tip: Never let the drummer in the control room, except under extreme sedation, unless you want all your mixes to sound like Led Zepplin.

Blue (at Spud's remix project site) wrote:...and put my patented rock-improving move on it: Turn The Damn Drums Up.

I'm loving this thread.

I would very much like to read the following articles in SongFight! Magazine:

1) "How to post-process vocals" by Jon Benjamin
2) "How and when to use guitar" by Blue Lang

Page me when they are written.

That article Sven referenced is really good, though it's a bit overstated (i.e. the drummer prejudice cited above). This one stuck out for me:

The article Sven referenced wrote:On a digital board (or software mixer) you never want to go over 0db, anywhere, ever.

This seems entirely wrong to me. I switched from an old Tascam 4-track to a PC with Audition. The old analog me was used to detecting clipping with my ears: the 4-track has little red LED's that light, voltage dependent, on the amount of >0db, but the onset is late and it won't bother for very fast peaks. But my new computer would light the red box if a sound hit +0.01dB for any of the 44,000 samples per second. It took me over a year to realize that my "fear-the-red" rule from my analog days was way too conservative for the wolf-crying digital world.

That said, I have yet to find the right way to intentionally use clipping, but it comes in handy. Especially since I hard-limit the full mix as part of my "mastering" process, all noise from clipping on individual tracks or even the full mix gets absorbed by the sound, rather than sounding like an artifact of numbskull mixing. So assuming you have your ears open, worrying about going into the red is kind of silly.

Caveat: if you ask others to master a track that you've already saturated at 0dB, they will get mad that you didn't leave them any "head room". If you pay them, though, they will probably just grumble quietly to themselves and do the best they can.

Lunkhead wrote:

Mostess wrote:Don't be afraid. Probably the most important thing. Blue's mixes made me cringe at first, because he didn't hide sounds I was embarrassed about. But that's just it, half of my mixing process was concerned with covering up things I didn't like (turning down wonky vocals, putting chorus on poorly-tuned instruments, EQing the midrange out of percussion that sounded wimpy, etc.) Stop it. No one else cares. Mix with love, not fear.

I think that's an important point for a lot of SF! folks who are new-ish to the recording thing. Don't try to cover up those aspects of the performances with post production techniques. You should either just live with them, or fix them by doing more takes. (Tune your guitar and try again, sing that one line over another 50 times, etc. etc.) That might actually make your performances improve over time, so that eventually you'll be able to get better takes sooner and you'll have less to do in post to "fix" things. Fudging things with EQ and effects isn't going to fool anybody anyway, and it's definitely not going to help you improve as a musician. At least, that's my opinion, as off topic as it may be...

This is really true in my case. I have no problem letting my guitar be heard because I'm comfortable with my strings, but I'm constantly hiding my vocals. I don't know if this is normal, but I have to put reverb and delay on it while I'm laying down my vocal tracks to get in the moment. It helps me to bring them out more.