« March 2007 | Main | May 2007 »

April 18, 2007

compressor does not dance

while mixing christa's vocals i encountered this problem that came up many times at my last place, namely that after working with the track for a bit i begin to notice a subtle harshness in the quality. on my yamaha monitors and on certain stereo systems, this harshness sticks out like a sore thumb, particularly when vocals are involved and particularly during louder portions. i spent hours last night fighting with the mix and using every trick i could think of to fix this problem--eq, compression, multiband compression, more eq, less eq, different eq, less compression, different compression....but no matter what i did i just couldn't get rid of it.

i started running through possible problems in my head. maybe it's the new preamp. but my other pres exhibited the same problem in the old place. maybe it's the room--the response is getting pretty good, but there are still those problematic frequencies i haven't tackled yet. but notching out those frequencies or their harmonics only fixed the resonant ringing and not this harshness. both my old and new audio interfaces exhibited the same problem, and same with both of my mixers. maybe something was clipping somewhere, but i was watching all the meters on all the gear like a hawk on saturday and was deliberately not running things too hot, since i wasn't sure yet how far i could push the new pre or the new audio interface.

i began to focus on the mic. it's a tube mic that i've been using as my go-to mic for vocals. it's probably getting to the point where i should start thinking about replacing the tube. but here again, i've had the same problem with other mics as well. but without anything else to really go on i figured hey, can't hurt to at least try replacing the tube to see if that has any effect.

then later on i wanted to work out some guitar parts and record some rough demos direct, so i grabbed my strat and plugged it in to the great river pre. figured i'd throw something other than vocals at it for a bit to get a feel for how it behaves. i dialed up a good starting gain and put a little bit of compression on my dbx 1066, which has been my default compressor for a while now. whenever i played more loudly i would start to hear clipping/distortion artifacts. so i pulled back the gain on the pre even though i had gobs of headroom according to the meters. things still clipped a bit on the louder parts. i pulled back the fader a bit on the mixer. same problem. i checked my cables to make sure they didn't have bad plugs or anything. but they all seemed to check out. i made sure everything was set to -10dBV. check.

at this point i was like, well, about the only other thing i haven't reallly eliminated is the compressor. i hit the bypass switch and lo and behold, the problem went away. i engaged the compressor again, problem came back. the problem seemed to get worse the more the compressor kicked in. but i dialed the ratio back to 1:1 and even when there was no compression it was still emitting these weird distorted-sounding artifacts. i know the dbx compressors are all famous for having that special "dbx color," but this just seems to have gone beyond the realm of color and into feces.

i had very similar settings to those used on saturday, so i'm pretty sure now that this is the culprit for the problems i'm having in the mix. i'm not sure yet whether this is a feature of the 1066 or if i just have a bad VCA or something--need to do more experimentation with that when i get some free time. if it's been like this for a while, it could very well account for problems i've had in other mixes and was never quite able to fix. my shitty alesis 3630's had the same problem but to a much worse extent, which is why i stopped using them ages ago and finally sold them off about a year ago. the problem on the 1066 is more subtle, which is why i haven't really picked up on it in tracking. but once you start setting up the mix and throwing in eq and compression it becomes more obvious. and leads to the "oh, it's a problem with the way i'm mixing" red herring.

so where am i now? well, i picked up a pair of dbx 160A's last year and never got around to using them except for when i did sound for sunil's live show. these are better quality than the 1066s, so i figured i'd do some a/b tests. i definitely like the ease of use in setting these up and the HUGE level meters are a big win for people like me who get off on blinkenlights. you can ride them pretty hard without getting any pumping/breathing artifacts. i was also able to slam it hard with my guitar and not hear any clipping, distorting, or any other sort of mangling. the only thing i really noticed was that when the compressor kicks in more than a few dB, a bit of the top seems to go away. but the resultant sound is very pleasant and easy to work with.

i also pulled out my presonus ACP-22 that i use here and there. it's an ok unit, but one thing that's always annoyed me is the transformer vibrates, thus generating a bit of noise that's just loud enough to make you go "what the hell's that?" if you're within a few feet of it. i also don't care for the level meters much--the attenuation LEDs are all green, which i hate. they should be red, dammit, the way god intended. no distortion artifacts, but lots of pumping/breathing when you hit it hard, and it colors the sound in a way i'm not sure i like, at least not on guitar.

so for the time being i'm going to unplug the 1066 and ACP-22 and just stick with the 160As, as they sound the best out of everything i have right now. but this whole ordeal got me realizing that i've been leaning more towards dbx compressors than perhaps i should, at least from a variety standpoint. i'm down to 3 now, but i used to have 2 166XLs as well. everything in my arsenal right now colors the sound one way or another, so i figured a first step would be to get something a little more transparent. not wanting to spend a lot of dough, i decided to buy a really nice compressor and see if it really is as nice as everyone says it is and as the name implies. they're only $200 new and i got this one for like 2/3 that price, so it's not like i'll be out a lot of money if i don't like it.

April 16, 2007

resonator

i finished building the helmholtz resonator tonight. here are some photos of the process. first i built the frame out of 2x4s, then i put up some 1x2's on either side of the middle stud and secured straps at 2" deep to hold the insulation in place. then i cut the rigid insulation and put it in place. next came the fabric, which is there for no other purpose than to make the gaps between the slats look nice, and that's partly why the installation is so sloppy (the other reason is i was lacking a decent means of cutting this fabric, which is a pain to cut insofar as fabrics go). finally, i put the slats up after staining them...

IMG_0412.jpg

IMG_0413.jpg

IMG_0414.jpg

IMG_0425.jpg

IMG_0428.jpg

the left-hand side of the resonator was an absolute bitch to do, owing to the really tight angle (24 degress on the inside, which meant having to cut everything at its complement of 66 degrees). since things that are capable of cutting/ripping at that angle are few and far between, i had to get creative. mostly it involved rotating the wood 90 degrees before putting it in the pathway of the saw. this sounds easy enough on paper, until you find yourself having to do things like stand a 3-foot piece of wood vertical and pass it along a table saw, or cut pretty much diagonally through a 2x4, which a standard 10" saw can't do. for the frame i had to do the rip cut using the band saw instead. which, as usual, came out very sloppy and i had to go back and plane the hell out of it. i finally got tired of planing and said, "fuck it, it's good enough."

this is what the angle cuts look like on the slats:

IMG_0422.jpg

i had to pick up a tenoning jig so that i could secure the pieces vertically to run over the table saw. yet another semi-expensive tool that i wsh i didn't have to buy, but it did do an awesome job and if i ever do use it for making tenon joints in woodworking, it has all sorts of knobs and stuff to tweak it out the wazoo.

haven't done another round of room analysis yet, but i do know already that it is not helping with the ringing at 320Hz, or at least not significantly. after walking all around the room and singing that pitch, it sounds like there's only one particular area in the room where the sound energy at that frequency is disproportionately high, so i'm going to tackle the problem by building a smaller, wall-mount helmoltz to go on the wall in the middle of that area.

April 14, 2007

minimalist recording setup

i set up a bare-bones recording setup so we can get moving on recording the rest of christa's songs while i continue building the studio. setup consists of a mic, pre, compressor, small mixer, computer audio interface, and monitors. here's a pic i snapped of the setup:

IMG_0416.jpg

christa came over this afternoon and we laid down vox for a tune whose guitar part we recorded about a year ago. this was my first chance to do a serious recording in the room as well as being able to give my new great river preamp and RME audio interface a whirl.

the good:
i love the hardware and software interfaces for the RME stuff. they're much easier to work with than my delta 1010's. all the important stuff on the box (word clock, sync source, -10dBV/+4dBu switching, etc.) is accessible on the front panel. the software seems very well suited to handling multiple interfaces, and it's much nicer being able to return to hardware that'll let me have up to 3 interfaces per card rather than having to deal with the one-card-per-interface crap. the clip lights on the box are dope, especially since they kick in a few dB below 0 dBFS, allowing me time to quickly adjust levels before ruining an otherwise good take with digital clipping. other converters i've used either don't have any level lights on the box or the clip lights are of the "if you're seeing this, you're hosed" variety. oh, and of course the converters sound quite nice and are a welcome change from the slightly-harsher-sounding deltas.

the great river pre is...well...great. lots of headroom. christa's a very dynamic singer, and it was able to handle her full dynamic range without me having to be constantly riding the gain or really holding it back. the 6-segment level indicators are a great feature. the sound is great and very easy to work with, at least for vocals--haven't tried anything else yet.

while i was setting things up i was listening to a bunch of mixes, both my own and those on commercial cd's. i'm amazed at the bass response in the room. the bass traps have given an extremely tight and clear low end. i heard some things in my own mixes that i had never heard before and, remembering some of the challenges i had when mixing the low end on those songs in my old place, i had a few "ahhhh! so that's what i wasn't able to hear accurately" moments.

working on actual music in that room after spending a year doing almost nothing but construction work in that space was a very surreal experience. this is going to sound kinda hokey, but instead of having a sort of "i'm sitting here and here's a desk and there's a person over there singing" sense of distinction, i felt myself and everything around me to be more...idunno...liquid, i guess? the room, the equipment, the singing, me...it all just kind of fused together. it's weird having a space of my own where i don't feel like the room is the enemy of the sound. here the room really embraces the sound.

the bad:
those frequency response spikes in the room around 200-350Hz are quite noticeable--they show up when recording and are exacerbated on playback. i started building a helmholtz resonator yesterday but haven't finished it yet. i really hope that will help tame those frequencies, as that will be critical to getting good recordings and good monitoring.

the console desk sits about 6" too high for my liking. taking the casters off will help at least somewhat, but it still won't leave me at an optimal vantage point once i install the console. and i'm definitely going to have to remove those rack shelves along the very top and knock the overall height down to provide a single surface on which to place both of my computer monitors and aux monitor speakers.

i'm not overly thrilled with the speaker stands i got last week. the main reason i got them was so that i could set them slightly higher than 48" and then angle them down towards me with MoPADs. however, they get really wobbly under the weight of my mackies. the connection between the stand and the bass isn't quite what i would call strong. i need to play around with them for a bit in the coming days and see if i can make them more sturdier. once i get the desk height down i'll also be able to drop the monitor height several inches, which should help at least somewhat.

so overall i'm quite pleased so far, although there is still much much work left to be done.

April 09, 2007

more on lighting

after a bunch of searching, i've settled upon the following lights for the ceiling:
Vision Small Flush Wall/Ceiling Light

i think they look really dope, plus they're really slim--only 3.5" deep.

i'm also going to install 4 wall uplights throughout the room to help illuminate the ceiling and provide more general illuminance. choosing these is a bit trickier, seeing as how most wall lights are designed to be hard-wired, but i never had the walls wired for lighting, only the ceiling. i've read that at least some types of lighting can be converted from hard-wired to plug with the use of these kits you can buy at home depot and lowes. i'm gonna go check those out tomorrow and see if they'll work.

related to this, over the weekend i found a cool lighting estimator page from GE. actually, i found that page several months ago, but at the time wasn't sure how to use it since i wasn't versed in things like footcandles. but thanks to an email message from tommy last week with a bunch of useful lighting info, i got on the right track and rediscovered that page. the original, less-technical web page i found on this topic suggested an amount of lighting that was about half what GE's estimator says. which very conveniently matches up with my subjective observation that the room is only about half as bright as it should be.

i should clarify in all this talk about lighting that the amount of lighting for any given session will be completely flexible; the ceiling lights will be on a dimmer, and the wall lights will be individually switchable. i also have a floor torchiere i'll be putting over by the console.

April 06, 2007

next round of room analysis

i did a bunch more measurements today after installing the bass traps. below are the waterfall plots from 7 different mic locations in the room. here are the results from the original analaysis for comparison. frequency and dB ranges are the same, but be aware that the time scale is different. the traps have significantly decreased the reverb time of the room and brought it more in line with what would be appropriate for that size room. so the previous time scale went from 0-1000ms while the new one goes from 0-500ms.

04-06-wf1.jpg

04-06-wf2.jpg

04-06-wf2.jpg

04-06-wf2.jpg

04-06-wf2.jpg

04-06-wf2.jpg

04-06-wf2.jpg

the area around 60-70hz (particularly right around 60hz) has been MASSIVELY improved by the traps. the region just above 300Hz no longer rings out like it used to, but it could still stand to be tamed a bit with a helmholtz on the wall. there are still dips leading up to 100Hz, but they're not as bad as before. don't know what's up with the spikes @50hz. i need to look into both of those some more.

overall i'm extremely happy with what i'm seeing.

not my week for electricity

so the electricians have done nearly all of the work in the studio. the outlets are all wired and the lights are all in. all that remains to be done is to install the dimmers. i had to order some special dimmers that won't generate any interference when used. i ended up going with the lutron nova line of dimmers, which is what other studio peeps have been using when they don't want to go all hard core with a variac transformer. unfortunately, the dimmers are on backorder and i don't know yet when they're expected in. and i've learned this week that for some reason electricians have a real aversion to the color black for plugs and switches. jerry and dave tried repeatedly to talk me out of using black plugs and to use dark brown instead, but i stood my ground. and i'm glad i did, because black plugs in metal boxes with metal face plates is HAWT.

the booth sconces look fucking awesome:

IMG_0403.jpg


the lights in the main room are a different story. the lighting is too directional, leading to problems with shadows and being blinded from intense brightness if you happen to be standing in the wrong place. the fixtures are also too domineering. when you walk into the room, your attention is pulled right towards them, and not in a good way. and overall, even though it's 800W of lighting total, it's still very dark in the room. michelle and i are going to spend some time this weekend exploring other lighting options. bah. i thought i was *finally* done with this crap. oh well, these things happen, no biggie.

in other electricity-gone-bad news, i received the RME PCI card today. it's actually a main card and a daughtercard, and the person i bought it from decided to put both cards in the same static bag when packing the box. this is generally a bad idea, and the damage the main card received as a result speaks to that. check out the damage it sustained.

superchunk!

no, not the indie rock band, but the bass trap of the same name where you pile a corner from floor to ceiling with wedges of rigid insulation in order to smooth out a room's bass response. while disco was over this week cutting the insulation, i was building the frames for the two traps. here's how one of them looks (sorry for the shitty quality):

IMG_0398.jpg

notice the access panel. in retrospect, i really should have put that panel somewhere else, but at the time it didn't seem like it would be a big issue to put it there. no biggie, i just had to make sure the trap dimensions were such that it completely covers the panel.

next came stacking all the wedges inside the frame:

IMG_0400.jpg

note that the bottom 10 or so wedges are in two pieces. this is so if i ever have to get into that panel i can just pull out the small pieces in front of the panel and pop a short 2x4 or something in the corner to keep the pieces above propped up while i go into the wall.

here's how the wall looks with both traps installed. the pieces that disco cut are on the left. the pieces that i cut are on the right. you can kind of tell that my frame of mind was, "i don't care how these look, i just want to cut them as quickly as i possibly can").

IMG_0401.jpg

eventually these bad boys will be covered in black fabric so they look all sleek and stuff. but the next step is to do another round of room analysis to see how much these have helped the low-frequency response problems.

April 03, 2007

mmmm.....all-nighters...

lately i've been staying up until 3 or 4 in the morning working on the studio. but since the electricians are coming a little after 8, i figured i may as well just stay up all night and plow through a bunch of stuff.

disco came over tonight to help building the bass traps. he spent most of the evening cutting out wedge-shaped pieces of rigid insulation while i started building the frames. i'll try to have pix up tomorrow so everyone can see how i'm building them.

i built a new access panel to replace the one i messed up over the weekend. overall it turned out ok, but i'm still not thrilled with the way it's engaging the seal when closed. i made it 2" thick this time around instead of 2 1/4", which helped a bit, but i still can't get the panel to go flush with the jamb. i suspect i'll have to go back to my original plan of angling in the door edges so it tucks into the seal better. i had disco crawl into the access panel and then i closed it behind him and had him check for any light seepage. everything checked out there, but i'm still hearing a little more sound transmission through the panel than i'd like. i need to experiment with a few things to see if i can improve it. it's not like horrible or anything right now, i just think i could do better.

i'm almost done building the wooden boxes to encapsulate the lighting junction boxes and make the lighting assemblies look all nice. i just need to cut out the holes for the junction boxes and do a little more gluing and sanding and stuff and then they'll be ready to be painted.

on the mudding/taping front, everything is taped now. well, save for the outside corner on the iso booth bulkhead, but we won't talk about that. i'm typing this up while i wait for the most recent mud coat in the vox booth to dry so i can slap on another coat.

on the gear front, i purchased the PCI card for the RME audio interface a few mins ago. i'd been looking on ebay for the past week but wasn't turning up much, so i decided to just give up and buy a new one from somewhere. i found a place that was selling them for $570, which is about $30 below the average. just as i was getting ready to place the order, i decided to check ebay one last time just for the hell of it. i'm glad i did--i found someone had put one up with a buy-it-now of $340 just a few hours ago. needless to say, i snatched that little bugger right up.

i talked with dave at work today about how one might go about building the metal frame to hold the absorption cloud above the mix area. he suggested i get a machine shop to fabricate the frame for me out of welded aluminum. i'm now completely convinced that this is the right way to do it and will have the most professional-looking finish. only downside is i'm probably looking at around $100-150 for materials and then anywhere from $100-200 for the labor, depending on some of the joint particulars. i'm gonna try and get an estimate this week or next.

man, even after 3 energy drinks i still feel like crashing. i'm gettin old. can't handle these all-nighters like i used to.

April 02, 2007

status report

mudding and taping in the booths is very far along. but i've been doing it in between other tasks, so it'll probably be another couple of days before it's completed.

electricians come on tuesday to do the first round of wiring stuff up. all lighting has been purchased, although they won't be hooked up in the first round. more on that later.

met with michelle today to finalize remaining color choice stuff and discuss other design things. all fabric colors for the acoustic treatments have been finalized. capeting will be picked out tuesday night. i've narrowed the door jamb stain down to 2 choices.

framing for the bass traps has started. disco's coming over tomorrow night to help cut all the insulation for the traps while i finish the construction/installation of the frames. i'll also be making a new access panel in the main room due to the aforementioned goofup when cutting it.

as soon as we finalize a stain color for the door jambs, i'm going to start staining them and move on to installing the stops and seals and all that fun business.

sometime next week i will put in a *very* minimal recording setup down in the studio to begin finishing up recording for christa's cd. this will be an exciting opportunity to get a feel for how the room sounds for a real session.

i can finally start thinking about gear again

now that the weeping and gnashing of teeth known as drywalling is nearing completion, i'm able to devote more time to more interesting things like determining gear to buy for the studio as i rev it up.

one thing i will definitely be ordering this week is the PCI card for my RME 8 in/8 out analog audio interface. i bought the interface over a year ago for an insanely good price and never got around to buying the PCI card needed to interface with it, so it's just been sitting in a box unopened all this time. which is really quite a shame because RME makes some really nice converters, a definite step up from the delta 1010's i'm using right now, although those
guys have served me well the past few years. once i start wiring up everything in the studio i'll need to get another RME interface so i can have 16 ins and outs. 8 just doesn't cut it for lots of projects.

i'm going to ditch my avalon preamp. that thing has been nothing but a major disappointment for my tastes. fortunately, these units sell pretty well, so i should be able to get a decent price for it.

in its place i'm going to get a great river mp-2nv. people talk about this unit over and over again on gearslutz and elsewhere and of the 50 or so times i've seen it come up in discussions, i've heard absolutely nothing but good things about it. it's quite rare that i see this kind of unanimous support for a piece of gear, so i'm pretty stoked about it. i just gotta get off my butt and sell my avalon so i can actually afford it.

April 01, 2007

my first door seal

the corner bass traps in the mix area are such that one of them is going to completely cover the access panel in the wall. this means that before i can get moving with those i first need to complete the panel assembly. this is basically a small door but without hinges and made out of plywood, so it was a good thing for me to practice on in preparation for doing the real doors.

first i had to create the jamb in the opening and seal it with acoustic caulk:

IMG_0383.jpg

next came the panel. i took 3 pieces of 3/4" plywood and glued them together. i had to make it this thick since plywood isn't as dense as drywall and i wanted to make sure i have a mass in that space comparable to 2 layers of 5/8" drywall.

IMG_0387.jpg

IMG_0388.jpg

the rubber seal that goes around the doorway is kind of a weird shape. the longer side of the "V" shape gets recessed in the door stop...

IMG_0389.jpg

here's the door stop ripped to create a cavity for the seal. note to self: rip fence at 3/8", blade at 22.5 deg.

IMG_0385.jpg

next i cut the stops for each side and mounted on the jamb:

IMG_0390.jpg

and finally i glued the seal in. they come all coiled up, so it was a real bitch getting the pieces to stay in place until the glue started to set--they kept curling and popping back out of the stop. i eventually managed to get them to stay put just long enough for me to shove the panel in the opening, prop a concrete block up against it, and call it a night.

IMG_0391.jpg

overall things came out ok. in the future, i need to do a better job on the carpentry of the stops--one corner came up a hair too shy. i also need to find a better way to get the gaskets to stay in place, as well as doing a better job with the corners--i kind of cut them haphazardly and sort of eyeballed the angle. dumb idea.

my initial tests indicated that the seal could be compressed up to about 3/8" thick, so i set the stop back a distance of the panel thickness + 3/8. after everything was in place, i found that under normal circumstances the seal won't actually compress that much and i should have recessed it another 1/8-1/4" to compensate, as the panel won't go flush with the jamb. this morning i had the bright idea to knock off the inside edges of the panel a bit, just enough to allow the panel to fit all the way in but still fit snugly in the seal. but in my haste i didn't double-check where i had the rip fence set and midway through the cut i realized i was ripping too deep into the panel and now it fits too loosely inside the seal. so it looks like i'll be making me another panel today. fortunately, they're relatively easy to make.