higher sample rates and eq plugins

[loaf]

could someone enlighten me as to why, when I use 96khz sample rate and look at a frequency analyser of the signal on an eq the bass end is much higher than using 48? Is it that the higher sample rate means that you record a more accurate signal, i.e shows you what's really going on in a bad room? I'm talking about recording with mics, vocals, guitars etc. Apologies if this is a bit of a newbie question, but I couldn't find a forum called 'confused'

[dropblacksky]

I think something else is going on....check for other variables. The higher sample rate shouldn't make a massive difference in the frequency spectrum.

[loaf]

doesn't seem to be anything else going on, just record at a higher sample rate and the frequency analyser in logic channel eq is totally different

[amishsixstringe]

EQ and frequency response are two different things, so don't confuse those. Something else must be going on. Can you HEAR a difference in your signals? Is it a multed signal or different takes? Are you sure your spectrum analyzer doesn't default to a different setting when a different sample rate is used? High frequency content is MINORLY altered at different rates, but the low frequencies usually aren't much if at all different.

Neil

[sonicdefault]

I've encountered that type of anomaly, where a frequency display shifts in a peculiar way when changing sample rates. What I've found is that it doesn't seem to be a real problem (except that perhaps the accuracy of the display is compromised). Sorry to be vauge, but if everything sounds ok, it's probably fine. Also, it doesn't have anything to do with HD capturing more of the source sound as you speculated. You ought to try a different Freq. analyser and see what happens.

[loaf]

thanks for the replies, I'll do some more testing - thing is, I've begun to use the frequency analyser a lot to see what eq adjustments might work, and this has totally thrown me

[Dave Derr]

When looking at a short piece of a source on a digital spectrum analyzer, you CAN get different looking responses, depending on the actual analyzer. Most spectrum analyzers in the digital domain use discrete frequency "bins" which are then measured for amplitude. By "bin", I mean a range of frequencies such as 40Hz - 100Hz. Theres usually a dozen or more bins. In a 31 band spectrum analyzer, you could say there are 31 "bins". If a harmonic or note of the short snippet falls on the edge of a "bin", and then you change the sample rate, the Bin itself moves, and therefore may cause the note to jump from one bin, into another, or possibly in between bins. This will change the way your response looks. In the low frequencies, most DSP based analyzers have less bins (actually, its just less resolution to our ears), and therefore more chance of tones and frequencies "jumping bins". This may be what you are seeing.

I bet if you swept the frequency of a constant amplitude sine wave into your analyzer, you would get a humpy response that goes up and down and up and down, as opposed to an even movement at constant amplitude.

Also, if your analyzer is FFT based, you may have an option to change "windowing", which is how the bins segway into each other. By experimenting with the windowing, you may find you can minimize the difference between the measurements at different sample rates. Examples of different windowing types are Hanning, Hamming, Rectangular, Exponentilal, and Flattop, among others.

In DSP, you are at the mercy of the guy who did the math. If he is experienced and knows the things to look out for, his software will probably be pretty solid. It's even better if he has good beta testers with different hardware and studios, and has had thousands of problems in the past that he had to agonize over, and fix.

[loaf]

interesting stuff - I'll see if I can get another analyser and see how it compares