what would cause HF rolloff in an audio circuit?

[Svart]

yes i know that was a broad question but I'll explain..

I posted a thread some time ago about my console and needing to heavily cut 2khz-3khz and 300hz-500hz. I've been playing around and worked on studio acoustics which helped some to a degree, however I've come to two conclusions.

1. I deadened the room so much that you cannot hear the room at all and i'm still getting the annoying frequencies, but now with such a dead sound that it hurts.. :sad:

2. i said screw it and ran a snake OUTSIDE and recorded some drums. the SAME annoying frequencies abound leading me to believe that maybe it's something in the circuit?

I notice that ALL my mics sound this way, some more than others but ALL have a very good bottom end with a rather boomy low mid and pronounced high mid. the HF is there, just seems considerably rolled off. I've tried a number of things including changing opamps in entire channels to much higher qualilty (tl08X to OPA stuff) and changing the phantom blocking caps to polyesters which helped the detail and definition considerably but the highs still seem much lower in volume than the rest of the frequencies. in contrast to the boomy sound, If I boost around 10k and then boost the 12k shelf the sound is much more like i would expect it to however the Q of the 10k boost starts to sound unnatural and the 12k boost starts to sound too *sparkly* (sorry, the only thing that comes to mind to explain what i hear), not harsh but not smooth.

I have some schematics of what is supposed to be the preamp and EQ sections but the preamp is different on the pcb than the schematics. I've traced and cannot seem to find an issue that stands out as something likely. the console has a hard EQ bypass(switch).

I haven't finished my GreenPres yet so i don't have anything to compare the preamps to, however the preamps are simply a pair of transistors before an opamp then to a section of another opamp. according to the schematics, the second half of the 5532 is being used as a servo but i haven't double checked this on the pcb.

2n4403-5532-tl084
now:
2n4403-OPA2604-OPA404

:thumb:

 

[Gold]

Have you listened to the mixes in other places? Make sure it isn't the room first. Have you tried recording directly to a DAT,CD.cassette to see if you get the same thing? If everything seems broken in the same way I would be suspicious.

 

[Svart]

i've taken a direct copy and listened to it in other places like the car and other people's homes and it does transfer from one place to another.

however there is one GREAT point to this problem. the annoying frequencies tend to be GREAT for guitar and bass. also great for classical guitars too.. no eq needed for those! go figure.

 

[PRR]

Try different speakers. They are by far the unflattest part of audio.

Even if you have two sets that seem to "sound the same", get a third set, as different as possible. "Cheap" 1-way mini-speakers are sometimes very revealing of sound lost in the complexity of highly-tuned many-way speakers.

Listen to known-good commercial CDs. If Sinatra and Phish also need 400Hz and 3KHz cuts, it is probably "you" (or your gear), not them.

As Gold says, try another room too. Or record outside, then bring player and speakers outside and listen.

 

[Svart]

true true. Very good advice but, I have listened to many cds in the room(for about 7 years) on the very same monitoring system and it sounds true, or at least similar to other systems, in other words I don't hear the problem from prerecorded material. only stuff that I record. that does bring up an interesting thought though. I did do some one take tracking (no multitrack recorder in the loop, MX-2424) directly to computer and i remember that the same problem exists doing it that way.. it's either ALL my mics or the console..

:?: it puzzles me.. Could I just be worrying about it too much? :green:

 

[Svart]

it's an analog console, MX-2424, and echo mia (OPA2604s for the old crap)

 

[Svart]

AKM converters in both the MX-2424 and the MIA. which ones do you mean? I tracked directly to the computer bypassing the MX-2424 completely with the same results.

 

[Svart]

my laptop i intended for audiotester testing was hit by lightning(phone line) so I don't have a dedicated testing rig right now.. but I DO intend to do that type of testing as soon as i can, however I was wondering about biasing and feedback network issues that I might look for in the circuit itself. I'm just wondering if the Preamp is actually optimally designed but again I don't have much to compare to.

the preamp is very similar to the green pre without the later current stage.

I read of people fighting over using EQ and how many people don't even EQ this or that and wonder why I have to EQ to get what i want.. :guinness:

EDIT: OK i've dug up enough parts to build a testing computer will give this a try when i get some time!

 

[Svart]

and i have a pair of the B*hringer ecm8000 measurement mics... currently being modified... :green:

I layed out a PCB based on Marik's 3x2sk170 design. Got his blessing to post it if i can find a host.. :thumb:

 

[Svart]

it may be just one resistor or one cap per channel.

that's what I hope. It's a simple design with the audio passing through few components.

also this reminds me, I did take a channel and gut the preamp out of it and replace it with a ssm2019 setup. it worked ok however I only used it with a dynamic mic. I need to test it with my condensers( where the problem really shows because of their expected frequency responce) but i did notice that it's not as detailed as the regular preamp when it comes to clarity.

 

[Svart]

directs to the tascam and back, then mixdown through the busses and master channel. both the busses and master channels have opa2604/opa404s and panasonic FCs, which didn't make much difference to the annoying frequencies but did make an astounding difference in the clarity of the audio.

line ins from keyboards and the like seem normal. the guitars are going through the preamp via FET based DI. they sound good without eq. Male voice sounds pretty good without eq too, however anything I need to have a nice top end like overheads just sound boomy. the closer I get to the cymbals the more high end i get but run into cymbal wash too which is even worse.. :?

I posted a topic a while ago about impedence in which i changed from the 3.3k to 600R and gained a drop in the mids of a few db but everyone seemed to conclude that this was wrong, so I put the 3.3k's back. I'm still feeling that this could be part of the issue...?

EDIT: Also, I'm going through 50 ft of snake from the mics to the console, fully balanced of course.

 

[Svart]

true. I have a schematic that alesis sent me but it's different than what's actually on the board and it's unfinished, guess all they could find was a proto schematic.. I'll have to trace it and lay out a schematic.

 

[Gold]

So things recorded through a DI are okay as well as things recorded very close to the mic capsule? Sounds like the recording room is the problem to me. How does the recording room sound when you stand in it? Is it deadened with something terrible like carpet? Try putting a couple of plywood sheets at angles above and behind the drum kit. It might help.

 

[Svart]

done that, it changes the sound but i still get boom. i've recorded outside too with some boom noted.

 

[Svart]

it feels like the HF is rolled off, because the low end and the mids are of the same volume but the highs are much lower as frequency goes up above 5k

 

[jrmintz]

Svart,

Don't take this the wrong way, but have you had your hearing tested?

 

[NewYorkDave]

Here's a quick test you can do at home to get a very rough idea of whether or not you're going deaf at higher frequencies.

Can you hear if a TV set is turned on in the house, even if the volume is off? If you can, then you can hear up to about 16kHz. But that doesn't mean that your hearing is flat up to 16kHz :wink:

 

[PRR]

> Don't take this the wrong way, but have you had your hearing tested?

Most of us who have been around drums have some hearing loss.

But if he is hearing his own work different from recordings by others, that suggests that it isn't his hearing.

 

[Gold]

I'll take a chance and suggest the obvious. Have you done a frequency sweep through the signal path in question? From mic pre through stereo buss with a mic like source impedance from the oscillator?

 

[Svart]

no offense taken, you are perfectly right, I have had my hearing tested recently and it's the same as many years ago when i had it tested. I wear earplugs whenever I play the drums or go to a concert too.

The DI does sound ok but I think it has more to do with the frequency bumps being right for the guitar sound I am after.

the bumps also sound good on male voices too! the only thing it doesn't sound good on is snare and overheads where i need good HF response.

strange isn't it? I've scratched my head at it a lot recently. :?

EDIT: I still need to do the frequency sweeps, I have to either fix my laptop or build another test machine. I still would like to gather ideas so i can tackle this once i get it all together.