Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How To Interpret Graphs - Page 148 - AVS Forum
First ... 146  147  148 149  150  ... Last
Audio Theory, Setup, and Chat > Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How To Interpret Graphs
sdrucker's Avatar sdrucker 07:26 PM 08-16-2013
Quote:
Originally Posted by sdurani View Post

We're hobbyists, we can do better!

That's only partially in jest. Most calibrators walk into a room and are stuck having to work with what's already there (speaker/sub placement, seating locations, room treatments). In this thread, we've gone way beyond measuring by discussing where to move speakers & subs and where to locate seating so that some of the problems that pro calibrators have to deal with can be eliminated prior to calibration/EQ.

We're hobbyists - if we can't do better, we'd be consumers smile.gif. Who wants to be a puppet dancing at the end of someone else's string? cool.gif

Even if it's Chris K's....

kbarnes701's Avatar kbarnes701 03:54 AM 08-17-2013
Quote:
Originally Posted by sdurani View Post

No, we're not stuck in an endless loop. Seems to be agreement that smooth is a better goal than flat. Seem to be agreement that ±3dB from target is fine for the "better" column (as you originally had). So that's a couple of steps in the right direction. Maybe ±6dB for the "good" column?

 

Good points.


kbarnes701's Avatar kbarnes701 03:56 AM 08-17-2013
Quote:
Originally Posted by sdurani View Post

Quote:
Originally Posted by kbarnes701 View Post

I've been surprised (and heartened) by reading on more than one occasion that many pros consider +/-6dB to be "very good".
We're hobbyists, we can do better!

That's only partially in jest. Most calibrators walk into a room and are stuck having to work with what's already there (speaker/sub placement, seating locations, room treatments). In this thread, we've gone way beyond measuring by discussing where to move speakers & subs and where to locate seating so that some of the problems that pro calibrators have to deal with can be eliminated prior to calibration/EQ.

 

Oh yes - concurred fully. Similarly, I wouldn't want either my PQ or my AQ to be 'as good as' a typical local cinema. I want it to be better than that. And it is!


kbarnes701's Avatar kbarnes701 04:02 AM 08-17-2013
Quote:
Originally Posted by sdrucker View Post

Quote:
Originally Posted by kbarnes701 View Post

I think the problem is that preference comes into it so much. Sanjay, for example, prefers a reflection-rich environment. I, OTOH, prefer to eliminate the lot and rely on the multichannel content for all the ambience etc that I want. Neither of us is right or wrong - it's just two different POV.  What is important is that once we know what we want, we are able to use the tools at our disposal to achieve it. 

I think, on reflection, it might be a hiding to nothing to produce the guide you described. Maybe some info on deciding what sort of model is preferred and then some advice on how to achieve it?  Hang on - that's what Jason said way, way back isn’t it: first choose your model??

I always thought a brief discussion of models wasn't a bad idea for a guide, so that even if you depart from the conventional thinking of this thread (which is LEDE, aside from Keith's preference for a Non-Environmental room, correct?), you had a basis for properly selecting the kind of room you wanted to achieve. Maybe a paragraph or so per leading model and some links for further information?
 

 

I don't think I actually have a NE room but I am much closer to wanting that than I am any of the other models. The kind of room I like is the kind I currently have, so I am pretty happy ATM.  But remember that I use the HT solely for movies with multichannel soundtracks where all the ambience, environment, spaciousness etc etc is already 'built-in' to the soundtrack. If I listened to music in there I may well want something else. HST, I do use music in there to evaluate the bass subjectively and I am not in any way disappointed with the sound. But it's jazz or jazz fusion not orchestral so, again, that probably makes a difference. 


giomania's Avatar giomania 01:53 PM 08-17-2013
I will throw in my opinion here. I think we should at minimum point people to where they can research the various "model" types (not sure if that is an accurate description), but it would be better to summarize them somehow. I will probably do that for my own edification, when I get to researching them in the future. Then we can offer the "Summary of Acoustical Standards" from the Mellor / Hedback paper as one example of how to analyze the measurements to achieve a certain goal. You note they may be happy with what we have provided, or they may prefer a different sound model, and more research may be required to achieve their ideal sound preference.

To use an analogy from school: I think if you give everyone the cliff notes to read, then some still may want to read the novel, and some may just take the test. To each his own.

Mark
sdrucker's Avatar sdrucker 09:47 AM 08-19-2013
Quick follow-up to the 'science experiment':
Oppo 103 and and R-972 are ordered, but thanks to something that ss9001 (Steve, to those of you that know him from the old Pioneer threads) said in a PM, it got me motivated to do a three-way comparison, since I still have that old Pioneer SC-27 on the island of abandoned electronics. All would have REW connected via Oppo 103 as the HDMI input source to forward the test signal for measurement - remember, the 103 has HDMI inputs - and all would use the Denon 4311 as my amp, either digital (XT32) or on MULTI CH IN (MCACC and Trinnov). I'd have to be careful about level matching analog vs. digital, but since nobody's ever done this to my knowledge, we could do a "real" A/B/C/X test. I would be one of the few people to have experience sooner or later with all three RCs, let alone be able to measure them.

Anybody up for plots of MCACC, XT32 and Trinnov? I even have that old AS-EQ1 that I haven't quite sold on the Classifieds (but I've gotten inquires this month), so this could be a unique REW comparison. WIth the baby likelly coming around the week after Labor Day, it's probably "now or never". At least for measurements.

Even if the results are generalizable to a sample of 1 (me), it might be worth the trouble simply to know what the tradeoffs might be in more quantifiable terms between RCs.

This is what happens when you do research for a living....eek.gif
LastButNotLeast's Avatar LastButNotLeast 09:51 AM 08-19-2013
Quote:
Originally Posted by sdrucker View Post

Anybody up for plots of MCACC, XT32 and Audyssey?

Absolutely! Now would be good. wink.gif


sdrucker's Avatar sdrucker 10:00 AM 08-19-2013
Quote:
Originally Posted by LastButNotLeast View Post

Absolutely! Now would be good. wink.gif

This falls into the "do it because you can" level....I'm really just Trinnov curious compared to Audyssey, but AFAIK there aren't too many people that have used MCACC and XT32 within the last 18 months, and still have the equipment to make it a "fair fight", as well have a way to test three RCs on the same platform, so to speak.
cuzed2's Avatar cuzed2 10:13 AM 08-19-2013
Sorry to resurrect this previous discussion; Some of you may recall that I have just gotten back into analyzing my set-up with REW. This quest has gotten so much easier thanks to the excellent guide (and the new USB mics)! Last month I used REW to help analyze and recalibrate my 3 cylinder subs, I used REW to optimize the phase relationship between these subs. They are all playing nice together now, as per the following comparison:

Before


After - analyzing and adjusting the phase for the fronts (against the back sub)


However; I still have some issues with my mains rolling off too high? Audyssey likes to set my mains at 150Hz.

Sdurani and others suggested I check my mains and verify what they are capable of?
The following is a close mic'd measurement 1" from the woofer on my SEOS mains.


I think I can conclude that the reason for the 150Hz roll-off is not a limitation with the mains, but probably an issue with the room, placement of the mains, or possibly the need for additional rm treatments.

Here is the LF, Center, and RF (all set to large) response when measured from the MLP


Next - I varied the position of the RF main, by:
#1) Moving it forward, Violet = closer to my seating - 40" from front wall,
#2) RED =pushed back into the RF corner, 12" from the RF bass trap,
#3) Blue was the original mid-point positon




So; I decided to move the mains to position #1, what looked to me like the best of the 3 positions, and re-run Audssey
Despite the re-positionining of the mains; Audyssey still set my mains at the same 150Hz XO setting.

Center w-wo subs


LF w-wo subs


RF w-wo subs


The following are the final results, after bumping sub levels 5db hot :-)

RF


LF


This is where I will leave it until I decide what to try next (perhaps some more sweeps while moving around some temporary absorption to see if I have floor-ceiling bounce). I guess it also possible this is just a null, as per my room dimensions? If so I will leave it as-is.

Having the 3 subs in separate locations & picking up everything below 150Hz sounds fine to my ears (I close my eyes and can't seem to localize the subs). However; I may also only be half-way down this rabbit hole, Suggestions anyone?
LastButNotLeast's Avatar LastButNotLeast 11:21 AM 08-19-2013

Another sub, obviously.

 

wink.gif biggrin.gif

 

It looks like those are smoothed. Can we see a sub graph not smoothed and the full graph 1/6?


cuzed2's Avatar cuzed2 11:30 AM 08-19-2013
Yep - everything was done with 1/6 smoothing.
I'll revisit / re-post the lower end stuff with no smoothing
jkasanic's Avatar jkasanic 12:32 PM 08-19-2013

@cuzed2:  Could you also restrict your frequency range on the bass region plots (i.e. 0-300Hz instead of full range)?  These should also be posted unsmoothed.


AustinJerry's Avatar AustinJerry 12:39 PM 08-19-2013
And finally, since you will be posting more graphs, can we see bass response to 300Hz, unsmoothed, with left+right+subs? I don't see any reasons to be concerned with the response in the graphs you showed.
IgorZep's Avatar IgorZep 04:24 PM 08-19-2013
From the graphs posted it looks like Audyssey is doing everything right. But... it is really interesting, and I've never seen that, in room response of speaker does not have peaks boosted by the room modes, it looks like there are only dips... The dip between 120 and 300 Hz is quite common in many rooms, and often unavoidable, but usually it is accompanied by a number of the heavy peaks below 100 Hz. Rooms usually help to boost bass, not the opposite. It really worth to look at the reasons why such a thing happens and what is cancelling the bass from mains.
cuzed2's Avatar cuzed2 05:01 PM 08-19-2013
Quote:
Originally Posted by AustinJerry View Post

And finally, since you will be posting more graphs, can we see bass response to 300Hz, unsmoothed, with left+right+subs? I don't see any reasons to be concerned with the response in the graphs you showed.

How about these:
.
That 60 Hz dip on the sub is either a 60Hz line artifact, OR perhaps more phasing tweaks are needed between these 3 subs (sweeping them one at a time and then combined). The mains are SEOS12's on stands about 40" forward of the front wall, about 9' from the MLP


And LR + Subs

The 160 and 280 Hz dips are certainly troublesome, floor-bounce, or...??



Center

Interesting how the center is smoother (it is on the floor under the screen, against the wall and tilted up directly towards MLP, about 12' from the MLP)
I gotta wonder if the dip at 85Hz (crossover area) could be improved by tweaking the sub distances? Audyssey nailed the sub distances to tape measure accurate (probably not optimized)

jkasanic's Avatar jkasanic 05:30 PM 08-19-2013
Seems pretty clear from the first graph why Audyssey is setting your XO to 150Hz! I'll take a deeper look when I get home with respect to those dips and your room layout (is this info posted elsewhere?) but with no smoothing applied it's likely the one at 280 Hz won't be noticeable during playback of normal content.
cuzed2's Avatar cuzed2 05:53 PM 08-19-2013
Jkasanic,

Thanks!

Several pages back I put a diagram up.
And I just realized DUH - I made these last graphs with the 150 hz XO in place. Forgot again to set mains to large for those sweeps
:-(
jkasanic's Avatar jkasanic 10:03 PM 08-19-2013
Ok, I had a look back at some posts you made about a month ago. Seems you got some good advice from Sanjay (sdurani) during that time and your response looked pretty good. Since most of what I know about room modes comes from him, I won't profess to be able to offer more insight on your room (plus I saw where he even read your build thread!). Does the bass response above represent what you showed in post 4247 "After" plot (sorry no link...iPad + editor = no play nice)? That response "appears" (I say appears because I'm trying to compare a full range plot to the one above so it could just be the graph axes or smoothing you applied) to be flatter above 60 Hz and below 200 Hz. If not, do you know what is different or has changed?
cuzed2's Avatar cuzed2 04:51 AM 08-20-2013
Actually my earlier graphs had smoothing and these most recent ones do not.
The only other thing that has changed is that I pulled the mains about 18"' closer to the seating.
jkasanic's Avatar jkasanic 06:49 AM 08-20-2013
Do you still have the mdat file for those measurements? I'm just curious how the unsmoothed response looks with similar axes?
pdxrealtor's Avatar pdxrealtor 01:43 PM 08-20-2013
Question-

When running sub graphs only there is a difference, aside from output, between timing reference set to

'HDMI Audio Out' 1 and 2.

Selecting 2 ups the output, that's obvious, but does REW intend for you to set timing reference to 1 or 2 when testing subs only?

I'm sure it wouldn't matter I have two subs since they are a mono out from my receiver, but none the less FYI I do have 2 subs and use 2 sub outs on my receiver.
pdxrealtor's Avatar pdxrealtor 01:46 PM 08-20-2013
FYI-

A while back in this thread I was having noise floor issues with my UMIK-1.

Just thought I would let everyone know that I believe that it was the Redmere HDMI cable I was using.

I've not had one issue since using a standard cable.
AustinJerry's Avatar AustinJerry 03:12 PM 08-20-2013
Quote:
Originally Posted by pdxrealtor View Post

Question-

When running sub graphs only there is a difference, aside from output, between timing reference set to

'HDMI Audio Out' 1 and 2.

Selecting 2 ups the output, that's obvious, but does REW intend for you to set timing reference to 1 or 2 when testing subs only?

I'm sure it wouldn't matter I have two subs since they are a mono out from my receiver, but none the less FYI I do have 2 subs and use 2 sub outs on my receiver.

 

When testing subs only, set the REW output to HDMI Out 1, and set HDMI Out 2 on the timing reference channel.  This sends a mono signal to both the left and right inputs.  On the AVR, select either Stereo or PLII Cinema (doesn't matter which).  Turn off the mains and run your frequency sweep, which will measure the subs you have connected, assuming of course that your mains are set to "small" and you are using bass management.  If you are using both sub out connections from the AVR, you can measure sub1, sub2 or sub1+sub2 by selectively powering off the subs.


pdxrealtor's Avatar pdxrealtor 03:31 PM 08-20-2013
OK....

It's odd how that works. I'm pretty sure when timing ref. is on two, that I actually get a higher bump from ~18-30hz vs. if timing ref. one is selected. The rest of the graph appears the same, just at lower SPL.

Appreciate the answer.
AustinJerry's Avatar AustinJerry 03:55 PM 08-20-2013

If you are sending only the left or right signal to the subs (instead of the combined signal as per my example), then check in the AVR to see if the trim level for one channel is higher.  this would cause a difference in the sub measured output, depending on which channel you configured.


jamsys's Avatar jamsys 03:56 PM 08-20-2013
Quote:
Originally Posted by AustinJerry View Post

I don't think I have seen this message.  Can you post a screen shot?  Please also provide the name of the mic you are using, and are you using Java or ASIO?

Sorry for the long delay, but here are the screen shots of the two errors. Also plots of the best I was able to get. It seems with Audessey on and the AntiMode at Lift 25 was the best. I am still not happy... Thoughts?

Scott



AustinJerry's Avatar AustinJerry 04:33 PM 08-20-2013

The lighter green measurement is not great.  It shows a 55dB drop (95dB@30Hz to 40dB@70Hz).

 

The darker line shows a somewhat better response, although there is still a 20-25dB variation.

 

The level is also quite low.  It looks like the output from REW is only in the 40-60dB range, when it should be at least 80dB.  I personally use 90dB to give the measurements a 5odB headroom above the 40dB noise floor. 

 

Regarding the error message--unfortunately, I cannot read the screen shots because they are too small.  If you are using a Windows system, please use the "Snipping Tool", which can be found in the Start Menu under Accessories. 

 

What make/model mic are you using, and did you calibrate it?


Roger Dressler's Avatar Roger Dressler 04:44 PM 08-20-2013
Quote:
Originally Posted by jamsys View Post

Sorry for the long delay, but here are the screen shots of the two errors. Also plots of the best I was able to get. It seems with Audessey on and the AntiMode at Lift 25 was the best. I am still not happy... Thoughts?
Regarding the response plots, did you apply the ~10 dB gain change for clarity, or is there such a gain difference in the room between Antimode on/off?

The notches appearing at ~70 and 95 Hz in the Antimode plot suggest to me that the latency of the antimode is not being compensated. The 8033 manual does not say if the Bypass mode still goes through the DSP or not (my guess is not -- that it is hardwire bypass), and it does not say what the latency is. I wonder if you added several feet of "distance" to the subwoofer output (10-20), would it kill the notches.

Was Audyssey run with Antimode active or bypass?
ggsantafe's Avatar ggsantafe 05:27 PM 08-20-2013
Quote:
Originally Posted by pdxrealtor View Post

Question-

When running sub graphs only there is a difference, aside from output, between timing reference set to

'HDMI Audio Out' 1 and 2.

Selecting 2 ups the output, that's obvious, but does REW intend for you to set timing reference to 1 or 2 when testing subs only?

I'm sure it wouldn't matter I have two subs since they are a mono out from my receiver, but none the less FYI I do have 2 subs and use 2 sub outs on my receiver.
Quote:
Originally Posted by AustinJerry View Post


When testing subs only, set the REW output to HDMI Out 1, and set HDMI Out 2 on the timing reference channel.  This sends a mono signal to both the left and right inputs.  On the AVR, select either Stereo or PLII Cinema (doesn't matter which).  Turn off the mains and run your frequency sweep, which will measure the subs you have connected, assuming of course that your mains are set to "small" and you are using bass management.  If you are using both sub out connections from the AVR, you can measure sub1, sub2 or sub1+sub2 by selectively powering off the subs.

Jerry - just to confirm my understanding - if your laptop/netbook provides 8 channel capability, you can select HDMI Out 4 to measure SUB only, and HDMI Out 4 on the timing reference channel for a dual sub configuration.
jamsys's Avatar jamsys 05:29 PM 08-20-2013
Quote:
Originally Posted by AustinJerry View Post

The lighter green measurement is not great.  It shows a 55dB drop ([URL=mailto:95dB@30Hz]95dB@30Hz[/URL] to [URL=mailto:40dB@70Hz]40dB@70Hz[/URL]).

The darker line shows a somewhat better response, although there is still a 20-25dB variation.

The level is also quite low.  It looks like the output from REW is only in the 40-60dB range, when it should be at least 80dB.  I personally use 90dB to give the measurements a 5odB headroom above the 40dB noise floor. 

Regarding the error message--unfortunately, I cannot read the screen shots because they are too small.  If you are using a Windows system, please use the "Snipping Tool", which can be found in the Start Menu under Accessories. 

What make/model mic are you using, and did you calibrate it?

The first is the error message

Excessive variation in measurement
The sound card measurement varies by 91db between 20hz and 20khz, that is much higher than it should be
The measurement may not be valid and should not be used to make a calibration file





The second is the error message when doing a sweep
Clipping was detected in the measured input, the frequency response may be distorted as a result
Try repeating at a lower volume setting on your processor or try reducing the sweep level


I like to be in the 80db range too, but it clips immediately and REW freaks out.

Here is a bunch of the different runs. The top is with Audessey on and bottom with Audessey off, both with various settings on the AntiMode. Mains are set at 60hz.


I beleive the mic is a AudioTek USB Mic from RadioShack.... Yes I know garbage, but I didnt want to blow $$ on a good one till I saw where I was at. My cross roads is to borrow the Audessey Pro Kit and spend $160 on a lisc or beg Mark Seaton to come back out.... (or the mini DSP)

Thx in advance!
Tags: Dayton , Dayton Audio , Room Equilizer Wizard Rew
First ... 146  147  148 149  150  ... Last

Up
Mobile  Desktop