bass issue

Today, I realized that I have a serious problem with my mixes: when I analyze them with various spectrum analyzers, it looks as if I had a high pass filter set at around 200 Hz. The frequency spectrum looks pretty much fine, except a ~6dB drop at 200Hz. After finding this, I analyzed several of my mixes, also some older ones, and all of these show the same behaviour (some more, some less).
Together with a colleague we checked every single plugin used in the mix, and I am positive that there is no such HPF anywhere.

I use Sonar X1c 64bit on a Win 7 64bit machine, a Metric Halo ULN-2 (standalone) run via Spdif through a Focusrite Saffire Pro 40. Apart from that, nothing fancy: Stillwell plugins, Voxengo, Redline, PSP, DDMF, Satson - no crappy stuff.

The room I work in is basically treated, though I am still fighting with some bass frequencies.

I have totally no clue what the source of this problem might be. Not sure if it is only conencted to my listening situation, though I doubt it. Can someone help with any hints what might cause this, or any advice on how I could find the source?

Here's a mix I am currently working on that clearly shows the problem.

Any chance that this is Sonar X1 issue?? So far, I've had all sorts of weird problems with X1 that this would not surprise me.

Describe your listening environment.

Nice song... the bass is there, but I'm not listening on speakers that have a good response in the lower freqs.... It doesn't sound unpleasant.

If room problems are getting in the way to a good mix..... take the room out of the equation with this>>> http://www.ikmultimedia.com/arc/features/

I have heard some really good things about this. I personally do not yet have this, but it is on my list and most likely the next software addition to my studio. 

One of the biggest problems home, garage, and basement studios face is the room acoustics that come into play and make a huge difference in the mixing and mastering process. 

ARC listens to the room, and corrects the sound coming out to ensure the mix is closer to true and not as colored by the room by compensating in advance knowing what the room will do to the sound. Pretty cool concept. 

And it's not all that expensive. 

Today, I realized that I have a serious problem with my mixes: when I analyze them with various spectrum analyzers, it looks as if I had a high pass filter set at around 200 Hz. The frequency spectrum looks pretty much fine, except a ~6dB drop at 200Hz. After finding this, I analyzed several of my mixes, also some older ones, and all of these show the same behaviour (some more, some less). Together with a colleague we checked every single plugin used in the mix, and I am positive that there is no such HPF anywhere.

While -6dB drop in the 200 range may seem abnormal, the track sounds fine and right.   So perhaps this is nothing to worry about.

Frankly a lot of pop mixes (as one example) exhibit a dip in the 160-250 range often (though not necessarily 6dB).

There is often a lot of frequency "ganging" that goes on in and around that range due to the fact most instruments fundamentals or 1st and 2nd harmonics fall in that range.   So it's not unusual to actually attenuate some elements in that range in order to get some breathing room in the track.

So - unless you really think you have a systemic problem in your mixing environment (speakers, processors, whatever), I wouldn't worry too much - and just continue to use your ears.

You might want to compare your mix (in a spectrum) to "reference" mixes you like in the same genre to see how they exhibit the dips in the same spectrum analyzer.

It sounds like the problem has been there for a while, since older projects exhibit the same roll off. You don't say how old is "older", but if you're like most folks you've changed software tools over time. What's most likely stayed constant are your speakers and your room, so that's where I'd look first.

The speakers and the room they're in constitute a single system. The room in fact has a greater influence on what you hear than the speakers themselves. It's said that whatever your monitoring environment is strong at is where your mixes will be weakest. So if your room/speaker combination produces too much bass, even if it's only at certain frequencies, your mixes will tend to be bass-light. Conversely, if your speakers don't have enough bass extension, your mixes will tend to be bass-heavy.

The most likely explanation for your problem is a room resonance around 130Hz, the lowest point in your bass spectrum. That resonance is causing constructive interference (a resonant peak) at the mix position, so that you hear much more at 130Hz than what's really in the mix. You are - perhaps unconsciously - compensating for that in the mix, so that the end result is a mix specifically tailored to your room that won't translate well to other playback environments.

The standard suggestion for mitigating this problem is acoustical treatment, but that may not be the best advice for you. 130Hz is pretty low, low enough to make treatment via bass traps difficult. To get enough absorption below 200Hz you need massive traps, which may not be practical in a small room. But advising on specific treatments would be premature without first analyzing your room.

The first thing I'd do is take some measurements to verify that the problem is really what I'm suggesting it might be. Get a copy of Ethan Winer's stepped sine test project. This is a SONAR project consisting of a series of 1-second sine waves starting at 40Hz and incrementing by 1Hz up to 300Hz. Place a microphone where your ears would normally be while mixing. Use an omnidirectional mic if you have one. Play back the project while recording it.

The resulting wave file that you record will show you how the room is affecting frequency response. Instead of being flat, there will be significant hills and valleys. The hills are resonant peaks, the result of constructive interference. The valleys are resonant nulls, the result of destructive interference.

Don't be surprised if the difference between the highest peak and the lowest null is huge - this is typical in small, untreated rooms. It can be as much as 30 or 40db! Even the best rooms with the right dimensions and lots of acoustical absorption and the speakers in just the right spot will still have 6db variances.

I won't write a book on the subject here, but take a snapshot of your room and report back. There are steps you can take to mitigate this problem, but you have to start with objective measurements. There are lots of other measurements you can also take, such as waterfall plots of reverberation time, but Ethan's test is a good starting point and very easy to perform.

thanks. I did Ethan's test, using a AKG 414 in omni mode, and this is my result:

does that indicate that 110Hz is problematic in my room?

Bass sounds good to me.

Listened on my headphones sounds great. Must be your room. Great song and mix. Wish I could do as well. Well done.

sounds fine here too. i understand your concerns, but remember not to listen with your eyes. ;-)

native


thanks. I did Ethan's test, using a AKG 414 in omni mode, and this is my result:

does that indicate that 110Hz is problematic in my room?

What it shows is that your microphone was positioned within a resonant peak at 110Hz. It does not necessarily mean that 110Hz is a specific problem in your room, only that it's a problem right there in that spot.

If you were to re-test with the microphone in different locations, you might see that 110Hz peak diminish or disappear altogether - and some other frequency pop up as a peak. If you could see all the peaks and nulls in the room at once, it would look like a three-dimensional topographical map, with peaks at different frequencies scattered throughout the room.

Unfortunately, this particular resonant peak occurs right where your ears are while mixing. It's going to throw off your perception of the bass guitar, making some notes sound louder than others. 110Hz happens to fall squarely on a standard pitch: A2 (one octave up from an open A string on a bass guitar). That note will always sound louder to your ears than it really is, causing you to instinctively compensate with EQ, which in turn hurts translation.

Now, 110Hz is a pretty low frequency, which means a long wavelength (314cm, or a little over 10 feet), and treatment with absorptive bass traps would be difficult. You could get fancy and design a Helmholtz resonator tuned to 110Hz, but that's not easy and even a resonator would be quite bulky.

A better approach, at least initially, would be to experiment with moving your speakers and mix chair to different locations. Although the wavelength is 10', you may only have to move by 2' to get out of the worst of it. Experimentation is key, because you could also move out of one peak and into another, or worse, into a null.

Another possible approach is room equalization, which can sometimes mitigate the problem at the mix position. You could, for example, insert a parametric equalizer into your master bus, set to dip at 110Hz with bandwidth and gain tailored to your measurements. This is a band-aid approach and you'd have to remember to a) replicate the EQ in every project and b) remember to bypass it before you export your project. But it costs nothing.