Rerurn to Romy the Cat's Site

Audio Discussions
Topic: Introductory RTA sources on the internet

Page 1 of 1 (11 items)


Posted by zanon on 06-10-2010
fiogf49gjkf0d
Hello:

I would like to ask this forum for advise.

I picked up a very simple and cheap RTA and decided to use it to listen to my speakers. It is 1/3 octave RTA, so I know Romy will say it is useless. It may not be very useful, but I start with nothing so I think of it as a first step.

I play pink noise from my stereo (2.1) and see a big hump in bass (also a dip from 125-500). I turn off the stereo and just play the subwoofer -- I see a big hump in bass only and nothing on the top. I play just the R and left channel, and I see relative flatness, except of course for the dip from 125-500.

This is my quandry. I am in a very large and open room, and without the subwoofer bass is truly in audible. The double bass opening from "So What" by Miles Davis really can hardly be heard. With the subwoofer, it is easy to hear, and I think very well integrated. I took great care to position the sub, time align it with RL channel, set the volume and phase correctly etc. so I get nice integrate bass extension but also you do not really know that a subwoofer is playing, the presentation is pretty seamless.

But it looks terrible and bloated in the RTA, while the "flatter" graph sounds thin.

Is this all an artefact of my cheap, low quality RTA and I'll see something dramatically different with more sophisticated measurement? Or is it that the bass is being lost in the large room, and this "extra" bass is what is needed to provide a better integration between low and high frequencies?

HF is very directional btw and the presentation does not "charge the room" as Paul S says, but instead sounds OK in a small sweeet spot.

Posted by Romy the Cat on 06-11-2010
fiogf49gjkf0d
Zanon, 1/3 octave RTA is a bit too wide, I would say 1/6 octave RTA I would consider more or less acceptable. There are tones of them available out there and for free. The pink noise is not a good thing to measure as noises a linear only in time domain – you need to run time summation to get true picture.

To learn to interpret RTA is a bit tricky question – what are you trying to do? A measurement itself is an abstract visualization. To tight it to the applied reality you need to observe a measurement in context to something.  I know it is not what you would like to hear but I am not come up with simple manuals. You need to understand what you do and then to observe how your measurements correlate with auditable experience.  Then your measurements become meaningful. I wrote a lot about it before.

The Cat

Posted by jessie.dazzle on 06-11-2010
fiogf49gjkf0d
Absolutely flat response will usually sound as you've described it; listen to your ears.

If you are hearing the opening bass line of "So What" via the sub, you must be running the sub pretty high in frequency to compensate for lack of mid-bass. From this one might guess that your RTA is telling the truth, and you may indeed be dealing with a mid/upper-bass suck-out.

You can confirm your RTA measurements as follows:

Find a means to play specific frequencies via your audio system; if you are able to play sound from your computer out through your audio system, downlaod a tone generator for the computer (I paid $25 a few years back; you can probably now find one for $0); if not, buy a CD with pre-recorded test tones.

Buy a cheap SPL meter of the sort Radio Shack sells:

http://www.radioshack.com/product/index.jsp?productId=2103667

Play a series of tones from 20-20KHz and compare the results with those of the RTA.

If this test reveals discrepancies, get a calibrated mic for the RTA (can be had for about $80):

http://www.cross-spectrum.com/measurement/calibrated_behringer.html

The mic will likely require a +48V Power supply:

http://www.behringer.de/EN/Products/MIC100.aspx

The above pre-amp/power supply will allow you to adjust gain and can be connected to your computer via its normal mic input (likely an 1/8th" mini-plug, so you might have to make up an adapter, as the pre-amp/power supply looks like it outputs via a 1/4" jack, of the sort you'd find on an electric guitar).

If there are still discrepancies, spend some cash on a more serious RTA.

Large rooms can be very difficult to load; you may want to wall off part of the space. This would almost certainly alter the situation in the event you really do have a 125-500Hz suck-out (though deficiencies and surplusses over limited frequency ranges are not a simple function of cubic volume).

My current "ideal" room would measure 23 x 20 x 8; such a space should in theory not "cramp" frequencies down to about 35Hz. At least that's what I'd like to think. A straight line from the lower front left corner to the upper rear right corner measures just under 32 ft, or about the dimension of a 35Hz wave, but in fact in the midst of writing this I'm thinking its ridiculous to assume the waves are going to orient themselves according to the longest dimension in the room (the diagonal) without some form of perhaps audible protest.

Here's a nice animation showing natural orientation of a wave as it leaves the source:

http://dev.physicslab.org/Document.aspx?doctype=3&filename=WavesSound_IntroSound.xml

Now take into account the following (from physicslab.org):

"...When two or more waves travel through the same medium at the same time, their waves actions will interfere with each other the same as any other types of waves..."

The only thing that can be said with any certainty is that large cubic volumes are more difficult to load. The best room may well be the one having just enough space to allow breathing room on all sides for both the speakers and listener, as well as adequate distance between speakers and listener.

jd*


Posted by scooter on 06-11-2010
fiogf49gjkf0d
You might check out the following website for a relatively easy to use and free RTA program for your computer. The site also has an active forum that can help get you up to speed with respect to getting the software working, mic sorted, and your initial readings:

http://www.hometheatershack.com/roomeq/

Don't expect miracles as it is a home theater site so people look for powerful explosions not timpani crescendos. They can however provide some basic assistance on some lower fq measurement and room issues.

Over time I found the iterative process of measuring, moving stuff around, and listening were very helpful in better understanding my goals and optimizing my systems. Romy has some interesting discussions about flat/boosted bass on his site that are worth contemplating.

One note of caution: some people use digital equalization to "fix" things.  Digital equalization has some visual appeal (e.g. pretty charts) but there are some major fundamental flaws related to things like trashing bits, time, mic placement, and room interaction. Unless you are looking to degrade your system, I don't think this is the correct path. Romy has wasted time here as well and posted his more developed views (use search function). 

(BTW I am not fully convinced eq is a terrible tool for major low fq problems below say 60-70hz range that just can't be solved via all the other tools at your disposal but tempting as it may be I am staying away)


Posted by Paul S on 06-11-2010
fiogf49gjkf0d
Zanon, it sounds like you are dealing with very typical "big room issues", among which are a large-ish disparity between the amounts of "charge" you experience from different frequencies at a given point in your room. As you get into this, you should find out what part or parts of the "irregularities" actually bother(s) you, and whether you want to try to "balance the charge" in the room-at-large or do it right at your ears, +/- in a "sweet spot". Generally, the latter strategy is easier to implement in a large room, taking into account the inevitable differences in the parts of the spectrum the room itself invariably amplifies, attenuates or cancels, outright. This quest leads promptly either to room treatment, in the broadest sense; to directing the axes of the transducers relative to one's ears; to electronic amplification or attenuation of certain frequencies. Some "blending-type" speakers are not really suitable for large rooms, and some, like the Legacy Focus, or most "pro" "monitors", are literally designed to blast straight at you. Remember: Ultimately, it takes big speakers and big power to really charge a big room. Also, no matter the strategy going in, bass often needs final, narrow-band "tailoring" to put the finishing touches on a system.

Best regards,
Paul S

Posted by zanon on 06-11-2010
fiogf49gjkf0d
Thank you all for your ideas and suggestions on RTA.

Romy -- you are correct, measurements always need to be made in context of sonic experience and more importantly musical intention. For me, though, to begin the process, I need to unlearn much of what is written about measurements. For example -- I have anechoic reading of my system (san subwoofer) from magazine, and my in-room, seating measurement looks completely different. So what is point of anechoic reading if it does not match real life situations?? Is targeting such a graph the wrong approach? I now feel that it is, especially for LF. I do not know what "time summation" is -- I will do research to find out what you mean.

Jessie -- yes. It took me forever to find a way so that I could run my sub high enough to get my mid-bass, while still keeping it integrated with the overal presentation. The most interesting thing I am learning from his is how to understand my sound in the context of the room. You need to know which string to push on!

Scooter -- yes. I read Romy's piece on his straight line but tilted towards bass slightly to make up for his old small room. I thought that since my room was larger, I would want to run flatter line but I was totally wrong. My line is more tilted than Romy's, it is the only way to bring back musical information at LF.

Your warnings in digital EQ are well taken. I also know that Jessie has had good results using dEQ at very low frequencies. I also like Romy's test where you introduce dEQ at very slight level, and then hear what it does to the content. All of this "tweaking" is within the context of objectives, and mine are not the same as Romy's, the degradation which comes from dEQ may or may not be worth it depending on whether I am getting closer or farther from what I am trying to say. Nevertheless, the big revelation to me has been how, if one was to use dEQ, it should not be to "flatten the line" so the picture is more pretty, but to see if you can, net of degradation, move closer to your objective irregardless of how the line looks (and I think the line is a good place to get ideas etc.) Of course it may be a complete dead end and degradation is so complete that you decide you cannot get there from here and pursue another avenue.

Paul S -- the ear targetting was how I managed to get HF improved. I don't love it because it seems to be opposite to DPols and "playing the room" but of everything I tried it had success. That was how I began my journey on room interaction and ignoring hardware (at least for now). That and focus on CP vs JS Bach.

When you talk of "electronic attentuation or amplification of certain frequencies" do you mean dEQ?

Posted by unicon on 06-11-2010
fiogf49gjkf0d
zanon
Rta is a handy tool to start on.
it shows you all ...

now you need to know what you gonna do
i suggest that you better get more help from your ears.
if you cant be fucked learning physics and acoustics you end up with nothing especial ...
it can be really a pain if u cant spend enough cash or hand it to someone who knows how to fix it all.

btw i say for first step you can start with huge movements of speakers and check channel left right symmetry frequency response...
and there maybe you can show us your rta pictures there maybe some help here.

good luck.




Posted by Paul S on 06-11-2010
fiogf49gjkf0d
Zanon, I do not mean DEQ, but that is one way to do it.  D[SET] is another way; and it can be done in real time, by a couple of means.

If you actually settle on the mid-field monitor approach, you will want that sort of speakers installed before you get too involved, since the sound field you'll get will be different than "normal speakers". You haven't said much about your present speakers, but what you've said so far suggests they are stretched thin, at least the way you are using them, without a high pass, and without proper mid/upper bass.

One way or the other, you will always have to keep thinking things through to meet the RTA more than 1/2 way.

Best regards,
Paul S

Posted by zanon on 06-14-2010
fiogf49gjkf0d
Paul S:

DSET is a very high power solution to a question that I only have a very primitive understanding of. It is not right for me at this time.

I would prefer if the model etc. of my speakers not be involved in this discussion. I know I am committing audio heresy, but I do not want to talk about gear, i want to understand how to better use RTA to support and interpret what I hear. My speakers are 2 way floorstanders, reasonably high efficiency, 10 inch woofer, horn loaded tweeter, and are ABSOLUTELY stretched thin.

It is certainly possible that my speakers are not right for this room. As the room is so large, like ROmy's new room actually, it may require a dramatically different gear to load it properly. That is not in the cards for right now, and is also a tangential issue from me getting the most out of what I have now, and even more tangential to understanding this new tool.

For an experiment, I set my RTA and looked at octave level measurement. By magic it became very flat -- the big dip I see at 1/3 vanishes. I am afraid at what I will now see at 1/6 given how much information was lost at octave level. It has also made me curious as to sonic effect of these kinds of dips and notches. Using dEQ (also terrible, I know) to crudely boost upper bass had little effect on opera (Soave, from Puccini La Boheme) but moved vocals on your typical rock/pop track dramatically to the front and, to my ears on some tracks through them out of balance. They were no longer as integrated in the past, even though they sounded more recessed. I will experiment more, and also try higher detail RTA as suggested by Jessie.

Unicorn: Left/Right is a good suggestion. I had not considered that at all, and have been just using toe-in to integrate two channels (my HF is very directional).



Posted by Paul S on 06-14-2010
fiogf49gjkf0d

Zanon, you seem to feel that RTA has some intrinsic value.  In any case, in terms of results, I think you already have a sense of where this is going. If you are bound to use your present speakers in a room you yourself deem too large for them, then I "recommend" that you basically forget about "loading the room" and tune the system so you can +/- comfortably wear your speakers like headphones. With some tweeters, one needs a layer of tissue, fine silk, or... to pull this off, and some attention to upstream matters may also be required.

Apart from DSET (only an option, not a recommendation), fully buffered, real time, multi-band equalizers are available on the 2ndary market (also an option rather than a recommendation).

Best regards,
Paul S


Posted by zanon on 06-14-2010
fiogf49gjkf0d
Paul S:

You are correct, I feel that RTA *might* have some intrinsic value, but I am not sure, and *if* it does have some intrinsic value, I am looking for help and guidance on how to unlock it. Or to be told that I am wasting my time.

For example, I have been so surprised at how different octave reading is from 1/3 octave reading. I have been so surprised at how non-flat curve sound better to me than flat curve. I am also being surprised at how flattening the curve sometimes create performance with more vocal emphasis (and "wetter" sound) but also how it can disconnect vocal, making it sound like the singer is with some other band, maybe in some other room, maybe even singing another song. All of this is giving me new reasons to discard standard industry output (like anechoic 1M readings) etc.

I am seening what new ideas L/R analysis and 1/6 octave can give me.

thank you for your help.

Page 1 of 1 (11 items)