Home Theater Forum and Systems banner

First Graphs...

Tags
graphs
7K views 48 replies 5 participants last post by  brucek 
#1 · (Edited)
Allright,
I have calibrated and measured and I think I have this almost under control. I am going to post my graphs to make sure everything is looking alright. The first picture is my soundcard calibration with my mic cal (I think that is the to line). The second graph is my left channel response, the third is my right channel reponse, and the last is my left channel response with the impulse response controls. Not sure what the last graph is supposed to look like. Everytime I applied the Impulse Response Control and set the pre ref to 125, my graph turned into a bunch of very close vertical lines. I will let you respond to this before I ask any more questions. Curious if the soundcard looks okay and how bad the response graphs are? Also, it doesn't matter if I run the left or right channel measurement, all four subs fire. Is that weird?
Roly
 

Attachments

See less See more
5
#27 ·
Okay, I think I am getting confused again. Raising the gain levels on my amp appeared to increase my measured response to be more similar to the target range. I thought this is what I wanted? I realize the graph looks similar, maybe identicle, only raised to a higher SPL level. Since I now have more of the response line closer to the target line, won't this be easier to EQ, or am I just not understanding something major. My first time. I will get better, promise. So if I don't use the gain on my amps and keep my first measure, I am uncertain how you suggest I raise the input on my subs. Can you give me a quick step by on how to do this? Also, what would I use the gain on my amps for or when would I use them? Is that for after the whole process is over and I just want some more volume? I have never been able to choose my Input Volume. I just use the default, which I believe is 75. Not sure if that matters. Thanks.
Roly
 
#28 ·
Hey Roly ,

From one nub to another I think brucek is saying just set the spl meter (plugged into your computer)and bump it up 5-10 db and then retake you readings and you wil get a truer reading of were to make your cuts and gain .
If I am reading what brucek said.You are doing reat by the way can't wait till I get time to run my first test now that I have my sound card calibrated :jump:
good luck roly hope everything goes well and sounds even better.
Ron.
 
#29 ·
I thought this is what I wanted?
Yep, it is. I just didn't feel it was high enough to do what we wanted. But we're really going in circles here - so I think we'll accept that everything is good and carry on with creating filters (which of course is what all this is leading toward). :)

So take your reading at 80dB as you did before and then carry on with the step by step filters stuff I wrote up (combined with following the HELP files).

Please, please save the raw measurement as a mdat file and attach it so I can have it...

brucek

ps
I am uncertain how you suggest I raise the input on my subs
Turn up your processor or subwoofer amplifier if it has a gain control

Also, what would I use the gain on my amps for or when would I use them? Is that for after the whole process is over and I just want some more volume?
Yes. Normally before you begin the whole equalization process you set the processor trims to zero and then do a simple processor speaker level setup and balance the level of all speakers with the processor trims. You can adjust the gain of the subwoofer amp (if your sub amp has a gain control) along with the processor subwoofer output trim to set the desired level of the subwoofer for the speaker test.
After equalization you may need to tweak the sub gain up further with the trim or the amp volume.

I have never been able to choose my Input Volume. I just use the default, which I believe is 75
75dB? The input volume is set up by pressing the input volume button as discussed. If REW can't obtain a high enough level then adjust the gain of your mic preamp gain higher. I don't see any reason to adjust the input volume - REW does it for you.

brucek
 
#32 ·
Did you try a filter at 23Hz with some gain and a fairly wide bandwidth of about 15. Sometimes gain works magic and sometimes it does nothing. If it does nothing remove the filter.

You could also add a cut filter around 56Hz and 100Hz.

It doesn't look too bad though.

brucek
 
#33 ·
Are there any no-no's when doing any of this? I can play around with it, I understand, and as long as I have saved my original data I am okay. But are there things to avoid. My first thought is to boost that low end, which I know may depend on my room and the trial and error thing. Are there certain sized bandwidths to avoid? How large can I make them? Also, I really don't care much about the frequencies above 120 or so, since my crossover is at 80. Why would I waste time on those high frequencies around 200? Just asking because I don't understand. Thanks for all the help. This is actually fun, so far. Can't wait until it is all finished.
Roly
 
#35 ·
You've done a good job. Hopefully the room will co-operate and the predicted will become the actual.

My first thought is to boost that low end, which I know may depend on my room and the trial and error thing. Are there certain sized bandwidths to avoid?
Bandwidths can be any size, whatever works. Gain is the trouble maker. Sometimes you need to apply two filters side by side to get a result. If you have the filters available it's better to divide a large gain or cut in two and divide it between two filters. Two at +5db is better than one at +10db at the same or close frequency. The filters with gain shouldn't be too high, but there is nothing wrong with some gain if it works - you have a fair bit of headroom with your sub amp.

Note you can save your filter settings in the BFD as a req file under FILE / Save Channel Filter Settings. Then you can recall the raw mdat file and recall the filters and play with them off line - at work perhaps... uhoh :devil:

Also, I really don't care much about the frequencies above 120 or so, since my crossover is at 80. Why would I waste time on those high frequencies around 200
Exactly. If you run out of filters you can steal filters from the higher frequencies since they have minimal effect. For example, your filters # 3, 4, 5 ,6. While they do a great job of making the response at that point close to the target, they are working in a frequency area where the signal is 20dB down and the mains will be overwhelming that frequency with their own signal. You may be able to use a single wider bandwidth filter there (that doesn't do quite as good a job) and steal three of the filters if you need them for lower frequencies.

Next. Since you are using an IB, you can't move it around to obtain a better response, but I wonder if you took a few other sweep measurements at a few different spots to see if you happen to be in a null where you're measuring. Some people have to average out their final EQ'd signal to satisfy a larger listening areas and some EQ for a single sweet spot.

Next step. Once you feel you're satisfied with your response, it's time to check the integration with the mains. This where you'll see the interaction at the crossover that sometimes creates a problem. You may have to adjust some of your filters to modify the results.

Turn on your two mains and do your sweep (with the BFD filters on of course) exactly the way you have been doing it. Now you'll get two pieces of information.

One is that you'll find what the level balance between the mains and sub are. If you find the sub too low in relation to the mains, here is the time to turn up that sub amp (or the processor sub trim if you feel it needs a boost to get more level into the BFD evidenced by the BFD LED's). Some people like their sub 10dB higher than the mains and some like it flat - whatever you think. Adjust the level of the sub and redo the sweep until it matches where you want it.

The next information this test reveals is the crossover area interaction. You may find you have a bad dip now at the crossover. You might have to completely or at least modify filters in that area to smooth this overall response. A regular sub has a phase control to play with. This has quite a bit of an effect on the crossover area. An IB doesn't usually have that convenience, so you rely on filters to do the job.

Anyway, you have to get the sub completely EQ'd before you test with the mains, but now you know the next step anyway. Then you're done except for listening tests where you might adjust the overall sub gain up or down a little to suit.

brucek
 
#36 ·
I've attached my attempt at your filters with an REW target of 75. I only needed 10 filters and only 2 have a small +5dB gain. The rest are cuts. I just wanted to show that you can do this offline.

I suspect with the small amount of gain used in the filters that your actual will be fairly close to the graph I've shown.

Note, because you have an IB, I set the Graph Axis Limits down to 10Hz to observe extension. See how nice your response extends down very low. You can save and try my filter attempt or use your own.....

RAW DATA FROM YOUR LAST MDAT FILE

Text Blue White Line Plot


RESULT WITH 10 FILTERS

Text Blue White Line Pattern


View attachment filtersR.req


brucek
 
#37 ·
Thanks Brucek,
Is there any advantage to using less/more filters, besides the time consumed creating them? Also, I noticed you only have one filter around 200Hz. I was curious why we bother with such high frequencies when we cut the sub so low? Does it just give a better slope overall? Also, I have heard a lot about house curves and that perhaps they sound better. Is this something I should play with or wait until I get the rest of my equipment setup? And last question, looking at my sub response, is there any reason why I couldn't use some M&K s-150's for my speakers. They are monitors and only have a frequency response of 82-20KHz +/- 2 dB. The sound great for theater, but I would hate to have a gap in my frequency. Here is a web site if you are interested or if it helps you to look at the stats. If you go to the very bottom of the page it lists the stats.
http://www.mksound.com/satellite_speakers.php
Thanks for taking the time on all this. I will play around with your graph as well. Oh, and just thinking, if someone where going to spec my frequency response,what would it be, based on our theoretical REW EQ?
Roly
 
#38 ·
Is there any advantage to using less/more filters
Less would be considered better overall because it leaves unused filters that you may need later to adjust around your crossover, and also each filter introduces phase shift which may or may not be good. As I said before though, if you had a filter with a 10dB gain, I would divide it into a couple at 5dB. The filter file I provided allows two spare and no filters with gain above 5dB......

Also, I noticed you only have one filter around 200Hz. I was curious why we bother with such high frequencies when we cut the sub so low?
Yeah, I answered that above in this post. Filters at 200Hz would really be considered useless and could be removed once you see the crossover with the mains is good and that when removing the ~200Hz filter had no effect on the sub + mains response.

Also, I have heard a lot about house curves and that perhaps they sound better. Is this something I should play with or wait
Wait on that and see how it sounds.

They are monitors and only have a frequency response of 82-20KHz +/- 2 dB.
Down 2dB at 82Hz..........I wouldn't recommend them for your system.

Oh, and just thinking, if someone where going to spec my frequency response,what would it be, based on our theoretical REW EQ?
Well, hard to say. Looks like that IB is still going strong down to 10Hz. Who knows where it drops off. Safe to say it goes quite low and would be a step up from most retail box subs...

brucek
 
#39 ·
Okay, another dumb question. Here is a picture of the Axiom sub. A big sub, for sure. Here is what I can't understand. This sub appears to be kicking my subs butt, at least from 15-100 Hz. 90dB, straight across. Mabye mine goes better to 10, not sure. My first question is this. Is Axiom using the same type of test that I am? I understand this is a professionally built sub, but come on. I have some big honking drivers and a huge box. It makes me feel as if this whole DIY was kind of a waste. Don't get me wrong. I have learned a lot, which is worth a great deal, but I seem to have an end product that is barely superior, or even inferior, to a sub that might be around the same price, and with no stress involved with the creation of it. Let me restate this. I really enjoyed this process. I have enjoyed the learning aspect. I am eager to do more. I just seemed to be under the impression that this sub would be phenominal, not an average- nice sub. With the extra cost for materials, amps, BFD, etc, I must be nearing $1500. And, I can't use the speakers I like because the sub won't hit 80 Hz. Just kind of bewildered. Sorry to vent. Thanks for the help. I re-read some of your earlier statements. They are making more sense now. People can tell you what they know, but if you aren't at the right time in your education, it just kind of goes over your head. Thanks a ton Brucek
Roly
:dizzy:
 

Attachments

#40 ·
This sub appears to be kicking my subs butt, at least from 15-100 Hz. 90dB, straight across
Thats a $1780USD Axiom EP600 in an anechoic chamber measurement and it's response is down -3dB at 17Hz and -25dB at 10Hz.

Your sub will eat this sub for lunch and still be hungry.

A room versus an anechoic chamber measurement are completely different.

You want 90db, turn your sub up. Go ahead, use REW as a simple sine wave generator by pressing the PLAY button and move the frequency down to between and 10Hz and 15Hz and crank your processor volume - now you'll have everyones attention in the house.
You don't even have an SPL mic that will measure how low your sub goes.

The difference between 10Hz and 20Hz is where the real subs stand up. Seems to me your sub is extending down to 10Hz and still no sign of dropping off. Do you realize how hard that is to get with a commercial sub? Your sub kicks butt.

And, I can't use the speakers I like because the sub won't hit 80 Hz
Huh, I'm really missing the point here. Your sub will easily hit 80Hz - but it's rolled off there by your processors crossover. The weakness is not in your sub, it's in the M&K S150 which is a small 23 pound monitor that is usually purchased as part of their M&K S150 THX Ultra System. It integrates well into their system where they'll have the sub crossed up to a higher level to allow the smaller satellites.
You require a floor stander where the lower end extends to about 60Hz to integrate into your 80Hz crossover.

brucek
 
#41 ·
Man, you get up early BruceK. You on the east coast? Your reply time says 0427.
As always, thanks for the reply. Okay, I am getting a better understanding. So if I brought home the Axiom and tested in my room with the same conditions, the graph would look completely different than the one they posted. Good. Thanks for making me feel better. I really was set on those speakers, so now I will be on another quest for some towers. I'll start a forum about it.
I would like to push my subs a bit, just to see how they will sound and perform. Anything to be careful of? I don't want to max them, just push them. Your a great guy BruceK. Think my wife is getting jealous. :hail:
Roly
 
#42 ·
I was just going to get working on my BFD setup, was reading the manual, and see it is a little different than the 1124. Probably not a big deal. I have the FBQ2496. I think I can figure out which buttons do the same as on the 1124. One question though is that I see my FBD has 20 channels per side. I know we were using 12 before, but I assume I can use the full 20 channels. Correct? Any major differences I have to be careful of with the 2496? Thanks.
Roly
 
#43 ·
Man, you get up early BruceK. You on the east coast? Your reply time says 0427.
Yep, I had to go out of town that day, and so was up quite early.

I know we were using 12 before, but I assume I can use the full 20 channels. Correct? Any major differences I have to be careful of with the 2496? Thanks.
From my understanding, the 2496 indeed has 20 filters, but only one program - no other real differences. The 1124 that most use here has multiple programs. For example, I have a movie program and a two channel program that I use depending on what I'm listening to at the time.

No big deal - set up your single program and you'll be fine.

I really doubt that you'll need 20 filters though. If you use the filter.req file I created, you'll only need ten....

brucek
 
#44 ·
I really doubt that you'll need 20 filters though. If you use the filter.req file I created, you'll only need ten....

brucek
I tried to load that file and it wouldn't fully load everything. It said something to the effect that all of the filters wouldn't carry across. Maybe I did something wrong. Can you explain again, briefly, what is the difference between setting my target at 75, or 80, or any other number. The response looks the same. I see it creates a difference between target baseline and the peak of peaks. Are we just trying to manipulate the graph so less EQ has to be done, or is it something else?:reading:
Roly
 
#45 ·
I tried to load that file and it wouldn't fully load everything. It said something to the effect that all of the filters wouldn't carry across
EDIT: I posted here earlier that it worked for me when I tried it, but then I remembered you had a FBQ2496 and not an 1124.
This means you should select that model in REW. Under EQUALIZER pull down, you should select FBQ2496.

I converted my filter file that I made to a FBQ2496 and attached it to this post. Just click and save it to a directory in your computer.

I also attached the jpeg of the filters that you can just enter into your FBQ to test them if you want or you can load your raw mdat file and the filter file into REW and see the result.

To load a filter file:

Start REW and Select SUB tab.

Select File / Load Channel Filter Settings the file filtersFBQ.req from the directory where you saved it.

That's it - the filters will show in the filters area of REW...

Here's a pic of the ten filters in the filter file.

Text Line Font Screenshot Parallel


View attachment filtersFBQ.req



Can you explain again, briefly, what is the difference between setting my target at 75, or 80, or any other number. The response looks the same.
The target is a value you can manually set at any time that has zero effect on the response and is only the level for the REW calculations to work with for creating the finding of peaks above that target value...that's it.


brucek
 
#46 ·
Okay,
Played around with my BFD. A little different than the instructional page, but I used the owners manual and it seemed to work. Here is the response. Everything was calibrated. When I ran the BFD filters a message came up stating "the highest level in the measurement is
-24.2" said it should be -10.0 and to turn up the processor volume. If I just turn up the volume on my processor, won't that mess up my measurement. If so, do I set up the entire thing again at a higher level, then retest with the filters. Also, at one point, it said clipping was detected on the input line. Not sure what is going on. First it is clipping, then not loud enough. Did I just overcorrect the levels? Anyway, how is the line looking? Also, the measurements I have been using have been for a target of 80. Can I use your line for 75 and just have a higher curve Brucek?
Roly
 

Attachments

#47 ·
When I ran the BFD filters a message came up stating "the highest level in the measurement is
-24.2" said it should be -10.0 and to turn up the processor volume.
Sounds like what I've been saying all along. hehehehe. Even REW is telling you that now. :)

If I just turn up the volume on my processor, won't that mess up my measurement. If so, do I set up the entire thing again at a higher level, then retest with the filters
Yes, takes about two seconds.....

Also, at one point, it said clipping was detected on the input line. Not sure what is going on. First it is clipping, then not loud enough. Did I just overcorrect the levels?
Likely....

Anyway, how is the line looking?
I think it looks real good. :T

Can I use your line for 75 and just have a higher curve Brucek?
Of course. My target of 75 was only because that's what I used given the level you had done your raw measurement at the time. Once the filters are entered into the BFD, you can then remeasure at any **** level you want. You can also tweak any filters you want if you like it any different.

So, now it's time to do the exact same measurement, except turn on the mains (when they're ready) and set up the relative level between the mains and sub. Also at that time you may need to tweak a filter or two around the crossover area if there's any funny stuff. :)

brucek
 
#48 ·
Er, um, I don't have any mains. No funds yet. My plan was to get my subs most of the way calibrated and running, then start adding to my HT, bit by bit. Tell you the truth, putting in that data to the BFD was really easy, even just using the manual to do it. Maybe I have not done it correctly, but I was figuring it would take me months to set up my amps, run measurements, work and re-work the BFD and add room treatments. I realize I have more work to do, but it really only took like 15 mins to read the manual and throw the numbers in. Let me know what I am doing wrong! :duh: Also, I would like to add any room acoustics if needed to help the bass. What would you recommend along those lines? How can I tell from my graph what types of room treatments would help and where to place them? Trial and error? I had planned to use the M&K s150's as mains, but I have been getting mixed signals on if they will work with my subs. They have frequency of 80Hz-up? Not sure if I can cross these subs at 80 and make it work. Have a thread going suggesting some other speakers that sound similar. One last thing. When my automatic meaure tone starts, I can hear it begin, then kind of dip or pause. Is that just the dip in my waveform that I am hearing, or is that the clipping issue, or both?
Thanks Brucek.
 
#49 ·
Let me know what I am doing wrong!
Nothing. Entering the data manually is a breeze. That's the reason why I don't have a midi cable to do it automatically. It just seems too easy to enter by hand. Most of the work is in understanding REW... and once you get that down, setting up a sub is simple.

I would like to add any room acoustics if needed to help the bass. What would you recommend along those lines?
I'm afraid I know very little about that stuff. Maybe post in another section for that help.

They have frequency of 80Hz-up? Not sure if I can cross these subs at 80 and make it work.
The theory is that mains should have a specification of being only -3dB down at an octave below your crossover to blend properly. That means that an 80Hz crossover require a set of mains with a bottom end of 40Hz +-3dB. That's the reason most people can't get away with a 60Hz crossover. Not many mains are capable enough.

That said, lots of people use mains that don't qualify, but still sound OK. I'm sceptical of a set that has an 82Hz bottom end though.... that's just not good enough. But you can test them if you're convince you want them. Doesn't hurt to try.

When my automatic meaure tone starts, I can hear it begin, then kind of dip or pause. Is that just the dip in my waveform that I am hearing, or is that the clipping issue, or both?
Neither, that's just exactly how it sounds...funny eh

My plan was to get my subs most of the way calibrated and running, then start adding to my HT, bit by bit.
Good idea, keeps you from rushing. But it must be painful waiting.... :sweat:

brucek
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top