Spectrum Labs - different slope than room/subwoofer frequency sweeps - Page 4 - AVS Forum | Home Theater Discussions And Reviews
Baselworld is only a few weeks away. Getting the latest news is easy, Click Here for info on how to join the Watchuseek.com newsletter list. Follow our team for updates featuring event coverage, new product unveilings, watch industry news & more!


Forum Jump: 
 
Thread Tools
Old 06-03-2015, 11:31 AM - Thread Starter
AVS Special Member
 
desertdome's Avatar
 
Join Date: Aug 2004
Location: Elkhorn, NE
Posts: 2,310
Mentioned: 50 Post(s)
Tagged: 0 Thread(s)
Quoted: 439 Post(s)
Liked: 488
Just to clarify, "10 dB per decade" is not the same as "10 dB per octave."
desertdome is offline  
Sponsored Links
Advertisement
 
Old 06-03-2015, 11:47 AM - Thread Starter
AVS Special Member
 
desertdome's Avatar
 
Join Date: Aug 2004
Location: Elkhorn, NE
Posts: 2,310
Mentioned: 50 Post(s)
Tagged: 0 Thread(s)
Quoted: 439 Post(s)
Liked: 488
Quote:
Originally Posted by Shreds View Post
SL shows disc levels accurately. RTA does not.
If by disc levels you mean energy at certain frequencies, then it looks like you are starting to get it.

To state the contrary, a frequency sweep in REW or an RTA of actual content shows the frequency response you will hear. SL does not.

Quote:
The guy behind the mixing desk takes into account how we hear and adjusts the mix accordingly. Why would you want to use an RTA to analyze disc content only to eliminate information?
Why is looking at data using two different methods eliminating data? Talk about attacking the straw man.

As I mentioned at data-bass, what started me on this whole thing was the fact that mastering engineers were using either an FFT with no tilt or an FFT tilted from - 3 dB to -4.5 dB. I wanted to know why. I found that they used the tilt to an FFT to represent how it would sound in room when measured with their RTAs.

1. In 2011 I noticed that SpecLab and other FFT analyzers were not consistent in how they represented the data. Some had a slope or tilt. I posted this at JRiver's forum on June 7, 2011, "What started me down this path is I'm trying to figure out why the movie frequency plots posted on AVS don't seem to match what I see in MC with spectrum analyzers.
2. I found out that some FFT analyzers have a 3 dB slope "baked" into them and other FFT's have the ability to adjust the slope. I also found discussion on gearsluz, cockos, krvaudio, and other forums regarding which slope to use when mixing the audio.
3. I corresponded with an audio mastering engineer and found that he mixes using the pink noise Accumulation Mode in SpectraFoo
Quote:
When the Accumulation Mode is set to Pink Noise is Flat, the Spectragraph shows the sum of the power of all of the spectral lines in the FFT that fit into each point in the displayed trace. This has the effect of displaying pink noise (or noise that has an equal amount of power in each octave) as a flat line. This mode is traditionally used when the frequency scale of the graph is logarithmic and corresponds to the type of display available on a traditional RTA.
4. I looked at the SpectraFoo manual and found graphs showing how the two methods of accumalation represented the data differently.
5. I read and tested more over the next few years and finally posted at AVS.
desertdome is offline  
Old 06-03-2015, 11:48 AM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 18,251
Mentioned: 71 Post(s)
Tagged: 0 Thread(s)
Quoted: 1237 Post(s)
Liked: 1557
"SL shows disc levels accurately."


data is presented in post 86 demonstrating that it does not. the amplitudes of all the signals are the same, yet the "level" in spectrum lab depends on duration. with an equal number of cycles in each tone, spectrum lab presents declining "levels" even when the amplitude of the signal is held constant.

Listen. It's All Good.
LTD02 is online now  
Old 06-03-2015, 12:25 PM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 353 Post(s)
Liked: 111
Quote:
Originally Posted by Shreds View Post
SL shows disc levels accurately. RTA does not. The whole thing with making a correction curve for the peak holds like from the Hulk Spectrum Labs - different slope than room/subwoofer frequency sweeps
is unnecessary. Spectrographs are the best tool we have for showing content levels. LTD, this post might help if you wanted to focus more on catching transients: http://data-bass.ipbhost.com/index.p...he-pot/?p=8145

Here is another test with 24 channels all playing a different tone, all set to the same level. SL accurately shows the levels while RTA does it's thing.




Is this some type of test?


The RTA tells you that the sound that you hear is biased to the higher end of your selected frequency range.


Instead of equal frequency spacing of your test tones, space the test tones in different octaves and see what you get for results in the RTA chart.
J_Palmer_Cass is offline  
Old 06-03-2015, 12:26 PM
Senior Member
 
Shreds's Avatar
 
Join Date: May 2014
Location: NC
Posts: 280
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 177 Post(s)
Liked: 73
Quote:
Originally Posted by desertdome View Post
Because Spectrum Labs has a different slope than other measurements you can't directly compare the charts to room measurements or subwoofer measurements such as those on data-bass.com.
  1. bass output peaks somewhere in the 40-50 Hz region and flattens or rolls off as you go down in frequency
  2. studios are viewing a different representation of the audio than what Spectrum Labs presents
Why would you want to compare SL of movie content with measurement sweeps when you can do a digital vs. mic'd comparison?



1. -might be true for some mixes but certainly not all
2. -Why haven't you shown us screen caps of the programs they use vs. SL with movie content to prove your points for your argument? Isn't your whole thing about using something other than SL because SL isn't accurate?
Shreds is offline  
Old 06-03-2015, 12:28 PM
Senior Member
 
Shreds's Avatar
 
Join Date: May 2014
Location: NC
Posts: 280
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 177 Post(s)
Liked: 73
Quote:
Originally Posted by LTD02 View Post
"SL shows disc levels accurately."

data is presented in post 86 demonstrating that it does not. the amplitudes of all the signals are the same, yet the "level" in spectrum lab depends on duration. with an equal number of cycles in each tone, spectrum lab presents declining "levels" even when the amplitude of the signal is held constant.
You're right, my bad. What program does show levels accurately?
Shreds is offline  
Old 06-03-2015, 12:31 PM
Senior Member
 
Shreds's Avatar
 
Join Date: May 2014
Location: NC
Posts: 280
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 177 Post(s)
Liked: 73
Quote:
Originally Posted by J_Palmer_Cass View Post
The RTA tells you that the sound that you hear is biased to the higher end of your selected frequency range.
Content is mixed by a human's ears so that sound bias is already cooked into the mix. When you apply the RTA to analyze it, it takes away from some of the content that is present.
Shreds is offline  
Old 06-03-2015, 02:22 PM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 353 Post(s)
Liked: 111
Quote:
Originally Posted by Shreds View Post
Content is mixed by a human's ears so that sound bias is already cooked into the mix.





True. No argument there by anyone who has posted on this thread.








Quote:
Originally Posted by Shreds View Post
When you apply the RTA to analyze it, it takes away from some of the content that is present.



Nonsense. If a sound track was theoretically mixed flat, then an RTA will display the FR a flat line. Nothing is added nor taken away from the mix.


The identical sound track will show a descending FR on an FFT display like SpectrumLab. The rate of descent is 3 dB per octave / 10 dB per decade. I show a target curve on all of my SpectrumLab spectrograms just to show the target for a theoretically flat FR. I never expect the sound track to actually be flat, but all sound tracks all follow the general 10 dB per decade slope.


What I look at is the deviation from the flat FR curve when I use an FFT display with a movie sound track.


If I wanted to, I could easily show how flat my speakers or amplifiers FR is with SpectrumLab.
J_Palmer_Cass is offline  
Old 06-03-2015, 04:35 PM
Senior Member
 
Shreds's Avatar
 
Join Date: May 2014
Location: NC
Posts: 280
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 177 Post(s)
Liked: 73
Quote:
Originally Posted by J_Palmer_Cass View Post
If I wanted to, I could easily show how flat my speakers or amplifiers FR is with SpectrumLab.
Not easier than comparing mic'd vs. digital. No wait, why don't you start a website using SL for measuring listening position and close mic response of different drivers and I can watch everyone not go there.
Shreds is offline  
Old 06-03-2015, 06:31 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 18,251
Mentioned: 71 Post(s)
Tagged: 0 Thread(s)
Quoted: 1237 Post(s)
Liked: 1557
Quote:
Originally Posted by Shreds View Post
You're right, my bad. What program does show levels accurately?

I don't know.


I too thought spectrum lab was showing amplitude as "levels" as that is pretty much how most folks have been interpreting the output for years. When I discovered that wasn't what I was looking at, I ended up in this thread.

Listen. It's All Good.
LTD02 is online now  
Old 06-04-2015, 03:49 AM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 353 Post(s)
Liked: 111
Quote:
Originally Posted by Shreds View Post
Not easier than comparing mic'd vs. digital. No wait, why don't you start a website using SL for measuring listening position and close mic response of different drivers and I can watch everyone not go there.




The mic'd vs. digital comparison does not tell you how flat a system is. Over at data-bass, you have about an 8 dB range for each color. In addition, only low bass frequencies are looked at.


Primarily, only frequencies below 30 Hz are of much interest to the boys over at data-bass. You have to own a pretty expensive measurement microphone to be accurate down into thru deep infra area.


Just as a note, I was doing that mic'd vs. electrical comparison before Bosso even knew how to use SpectrumLab. Basso uses that gimmick as a sales tool, so I take that into account.


There is more to audio than infrasonics.


One other comment. What makes you think that I have any interest in "measuring the response of different drivers" by use of any method? I have ZERO interest in doing that.
J_Palmer_Cass is offline  
Old 06-04-2015, 03:57 AM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 353 Post(s)
Liked: 111
Quote:
Originally Posted by LTD02 View Post
I don't know.


I too thought spectrum lab was showing amplitude as "levels" as that is pretty much how most folks have been interpreting the output for years. When I discovered that wasn't what I was looking at, I ended up in this thread.


Could you show me what you find to be not accurate in SpectrumLab? I am missing something here.
J_Palmer_Cass is offline  
Old 06-04-2015, 04:12 AM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 353 Post(s)
Liked: 111
Quote:
Originally Posted by desertdome View Post
Here is the 2 minute Fury clip with the FFT rotated to match the frequency reponse of the RTA. Looks like it matches pretty good.



Here is the original Spectrum and RTA taken in REW referenced at 1000 Hz.





Kind of interesting that the audio peaks from 200 to 2000 Hz is in a pretty tight range when viewing the data in the RTA form. The higher bass levels are about what one would expect.


It is still easy to see the same infrasonic roll off with either type of display.


When you used your "reference curve" to modify the slope of the SpectrumLab waterfalls, did the Peak and Long Term Average curves in the upper Spectrogram still work properly? As I recall, that is what I had problems with when I adjusted the waterfall 10 dB per DECADE.
J_Palmer_Cass is offline  
Old 06-04-2015, 08:05 AM
Senior Member
 
Shreds's Avatar
 
Join Date: May 2014
Location: NC
Posts: 280
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 177 Post(s)
Liked: 73
Quote:
Originally Posted by J_Palmer_Cass View Post
Nonsense. If a sound track was theoretically mixed flat, then an RTA will display the FR a flat line. Nothing is added nor taken away from the mix.
If a sound track is mixed with flat levels RTA apparently won't show a flat line.


Why would a track ever be mixed flat if we don't hear flat? Isn't the purpose of an RTA to show how we hear and show a flat line for a mix that is naturally compensated to our ears with a bumped low end by the mixer? Isn't showing a flat line for a bumped low end ignoring certain information that is on the disc?
Shreds is offline  
Old 06-04-2015, 08:15 AM
Senior Member
 
Shreds's Avatar
 
Join Date: May 2014
Location: NC
Posts: 280
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 177 Post(s)
Liked: 73
Quote:
Originally Posted by J_Palmer_Cass View Post
Primarily, only frequencies below 30 Hz are of much interest to the boys over at data-bass
That's why SL is just fine for most of us who use it for sub content. Subs take more energy to reproduce content as you go lower in frequency anyway right? SL seems to favor lower/longer waves for transient levels (but not necessarily either because longer tones will show levels accurately). So why not use it for sub content? Obviously nothing is perfect but SL is by far the best tool for the job. I don't see anyone going back to using RTA for sub content analysis.


Quote:
Originally Posted by J_Palmer_Cass View Post
Just as a note, I was doing that mic'd vs. electrical comparison before Bosso even knew how to use SpectrumLab.
And you don't know how to compare the digital with the mic'd? Maybe you can go over to DB and ask him how.
Shreds is offline  
Old 06-04-2015, 08:51 AM - Thread Starter
AVS Special Member
 
desertdome's Avatar
 
Join Date: Aug 2004
Location: Elkhorn, NE
Posts: 2,310
Mentioned: 50 Post(s)
Tagged: 0 Thread(s)
Quoted: 439 Post(s)
Liked: 488
@Shreds

Which of these two is correct?



Attached Thumbnails
Click image for larger version

Name:	EOT 1.PNG
Views:	101
Size:	40.0 KB
ID:	755169   Click image for larger version

Name:	EOT 2.PNG
Views:	101
Size:	48.9 KB
ID:	755177  
desertdome is offline  
Old 06-04-2015, 12:38 PM
AVS Addicted Member
 
LTD02's Avatar
 
Join Date: Oct 2003
Posts: 18,251
Mentioned: 71 Post(s)
Tagged: 0 Thread(s)
Quoted: 1237 Post(s)
Liked: 1557
Quote:
Originally Posted by J_Palmer_Cass View Post
Could you show me what you find to be not accurate in SpectrumLab? I am missing something here.

the second picture in post 86 has three sine waves with equal amplitude in the source, but the "level" presented in spectrum lab is different.


so when asking what seems like a really simple question, "what frequencies are in the source and what are their relative levels?" spectrum lab doesn't show that, yet that is the interpretation that most folks take away when they view the output.

Listen. It's All Good.
LTD02 is online now  
Old 06-04-2015, 12:39 PM
Senior Member
 
Shreds's Avatar
 
Join Date: May 2014
Location: NC
Posts: 280
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 177 Post(s)
Liked: 73
Dunno man, kind of hard to tell anything with no scale. What's your point?
Shreds is offline  
Old 06-04-2015, 01:16 PM - Thread Starter
AVS Special Member
 
desertdome's Avatar
 
Join Date: Aug 2004
Location: Elkhorn, NE
Posts: 2,310
Mentioned: 50 Post(s)
Tagged: 0 Thread(s)
Quoted: 439 Post(s)
Liked: 488
One image is using the voltage setting in Spectrum Lab and the other is using th dBFS setting. It illustrates that even the same program with with the same signal can display it in two different ways depending on what you want it to show. Just because the data looks different in each screen shot doesn't mean that one is wrong.



Quote:
Originally Posted by Shreds View Post
2. -Why haven't you shown us screen caps of the programs they use vs. SL with movie content to prove your points for your argument? Isn't your whole thing about using something other than SL because SL isn't accurate?
Spectrum Lab is accurate which I've explained many times in this thread. As illustrated above, different doesn't have anything to do with accuracy.

Let me ask you a similar question. Why would you use an oscilloscope if Spectrum Lab is accurate? Spectrum Lab is capable of showing the voltage requirements at each frequency. When the 10 Hz tone is playing in Edge of Tomorrow, what is the total output voltage of the signal? You can't tell because you can't tell from the Spectrum Lab graph which other signals are playing at the same time. You need to use an oscilloscope to measure the output voltage for the entire signal. Both tools are accurate, but are showing different info of the same signal.



Telling people their system needs to be capable of reproducing a 10 Hz signal at 937.7 millivolts is meaningless because people don't think of output as voltage. They think of output as SPL. The subwoofer measurements at data-bass are of max burst, basic response, long term output, etc. are showing SPL.

Spectrum Labs is showing maximum power at any given frequency, but is not showing how that energy relates to SPL. If someone wants to know the SPL requirements they need to measure using an RTA like the 1/48 RTA in REW. Nobody is doing that and showing graphs. It just so happens that an FFT, like Spectrum Lab, can have the response tilted by around 3 dB per octave and it will get close to representing the SPL. If people understand this, they can relate the Spectrum Lab peak hold and long term average graphs to actual SPL - when they are comparing to other graphs that use SPL. As far as comparing their own systems to the original signal, then the waterfall graph in SpecLab is best for that. Nobody is arguing that it isn't.

Perhaps the audio analysis software, Spectrafoo, can help clear this up. Spectrafoo is used by many mastering engineers and live sound mixers because it gives so much information about the sound, music, phase, etc. Spectrafoo has an FFT. It is described like this:

Quote:
SpectraFoo’s underlying FFT generates more data than can be displayed on screen at once. The way that this data is treated when the system computes the trace is called the accumulation mode. SpectraFoo provides two accumulation modes that you can choose using the Spectragraph Controls window:

1. When the Accumulation Mode is set to White Noise is Flat, the Spectragraph shows the maximum power of all of the spectral lines in the FFT that fit into each point in the displayed trace. This has the effect displaying white noise (or noise that has an equal amount of power at each frequency) as a flat line. This mode is traditionally used when the frequency scale of the graph is linear.

2. When the Accumulation Mode is set to Pink Noise is Flat, the Spectragraph shows the sum of the power of all of the spectral lines in the FFT that fit into each point in the displayed trace. This has the effect of displaying pink noise (or noise that has an equal amount of power in each octave) as a flat line. This mode is traditionally used when the frequency scale of the graph is logarithmic and corresponds to the type of display available on a traditional RTA.

The figures below show the results of the two weighting accumulation modes with the two different kinds of noise.


If you understand how an oscilloscope provides you with the sum of all voltages at all frequencies and Spectrum Lab shows you the maximum voltage at each frequency, then perhaps you can begin to understand how Spectrum Lab shows you the maximum power at each point in the FFT while an RTA shows you the sum of power at each point in the FFT. The sum of power is also know as sound power level. The sound power level without distance is the same as the sound pressure level (SPL).
Attached Thumbnails
Click image for larger version

Name:	Spectrum Lab Differences.PNG
Views:	105
Size:	57.0 KB
ID:	755617   Click image for larger version

Name:	Edge of Tomorrow Voltage smaller.png
Views:	104
Size:	39.8 KB
ID:	755625   Click image for larger version

Name:	Edge of Tomorrow Voltage.PNG
Views:	18
Size:	35.6 KB
ID:	755633   Click image for larger version

Name:	Spectrafoo.PNG
Views:	106
Size:	262.6 KB
ID:	755641  
desertdome is offline  
Old 06-05-2015, 04:56 AM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 353 Post(s)
Liked: 111
Quote:
Originally Posted by LTD02 View Post
the second picture in post 86 has three sine waves with equal amplitude in the source, but the "level" presented in spectrum lab is different.


so when asking what seems like a really simple question, "what frequencies are in the source and what are their relative levels?" spectrum lab doesn't show that, yet that is the interpretation that most folks take away when they view the output.


Actually SpectrumLab does show accurate voltage levels per bin no matter what the source is.


However, there is no way to easily translate the levels shown in any specific waterfall line into a meaningful SPL output level. As you have clearly shown, by eye a single frequency sine wave looks to be higher in Voltage Level than tone burst's Voltage Level.


While this is accurate, it is misleading if you are thinking about total SPL Levels. The single frequency sine wave displays as a narrow band of bins, and a tone burst shows up as a wide band of bins. In order to determine total signal level, you have to add up the levels of all of the bins per a well known formula. You can not do that by eye.


To observe total signal levels, the best that you can do with SpectrumLab is to activate the Amplitude graph option. The Amplitude graph shows total signal level as a percent of full scale. The graph displays in a linear scale rather than a log scale, so it is not that useful. A signal that is 6 dB down from 0 dB FS reads 50% on the linear scale if calibrated.


Anyhow, there is nothing wrong with doing a mic'd vs electrical comparison with SL subject to SL setting being used. As you have shown, it is difficult to translate those same waterfalls into SPL levels by "looking at all of da perdy colors".
J_Palmer_Cass is offline  
Old 06-05-2015, 05:21 AM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 353 Post(s)
Liked: 111
Quote:
Originally Posted by Shreds View Post

And you don't know how to compare the digital with the mic'd? Maybe you can go over to DB and ask him how.


Just wondering, do you have a learning disability? Why don't you say How! to Kemo Sabi over at data bass while you are communicating with the AVS exiles.


To simplify a lot of your silly talk, it would be nice if the SpectrumLab program had an RTA calculated bin option instead of the standard fixed FFT bin width standard.


I also think that is about time for you to put up or shut up. Produce a mic'd vs. electrical comparison of your own system with SpectrumLab either here at AVS or over at Data-Bass. Maybe you are afraid to show your results to the "full bandwidth" crew over at Data-Bass?
J_Palmer_Cass is offline  
Old 06-05-2015, 06:54 AM
AVS Special Member
 
popalock's Avatar
 
Join Date: Jul 2009
Location: Doha, Qatar
Posts: 4,583
Mentioned: 89 Post(s)
Tagged: 0 Thread(s)
Quoted: 996 Post(s)
Liked: 1364
Send a message via Skype™ to popalock
Quote:
Originally Posted by J_Palmer_Cass View Post
Just wondering, do you have a learning disability? Why don't you say How! to Kemo Sabi over at data bass while you are communicating with the AVS exiles.


To simplify a lot of your silly talk, it would be nice if the SpectrumLab program had an RTA calculated bin option instead of the standard fixed FFT bin width standard.


I also think that is about time for you to put up or shut up. Produce a mic'd vs. electrical comparison of your own system with SpectrumLab either here at AVS or over at Data-Bass. Maybe you are afraid to show your results to the "full bandwidth" crew over at Data-Bass?
Wellllll.... Not sure if it's a secret or anything but, I'm pretty sure Shreds comes from the loins of one of (if not the) most prolific full bandwidth advocates in the world today.

I'm not getting involved though, so you guys please continue. Wait, let me go get a beer...

...

...

...

Ok, go ahead.
popalock is offline  
Old 06-05-2015, 08:00 AM
Senior Member
 
Shreds's Avatar
 
Join Date: May 2014
Location: NC
Posts: 280
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 177 Post(s)
Liked: 73
Quote:
Originally Posted by J_Palmer_Cass View Post
I also think that is about time for you to put up or shut up. Produce a mic'd vs. electrical comparison of your own system with SpectrumLab either here at AVS or over at Data-Bass. Maybe you are afraid to show your results to the "full bandwidth" crew over at Data-Bass?


Put up or shut up eh? This from the guy who posts no data but is super critical of everyone who does. You bore me.
Shreds is offline  
Old 06-05-2015, 08:14 AM
Senior Member
 
Shreds's Avatar
 
Join Date: May 2014
Location: NC
Posts: 280
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 177 Post(s)
Liked: 73
Quote:
Originally Posted by desertdome View Post
Spectrum Lab is accurate which I've explained many times in this thread. As illustrated above, different doesn't have anything to do with accuracy.
I thought the whole thing with this thread was about SL not being accurate to SPL levels for short duration higher frequencies. I'm not seeing an alternative here to something that is accurate in this regard. If spectrafoo is the answer here because it does more things than SL, let's see some results of using it for some burst tests at different frequencies and see how it does.

Otherwise, matching SL's peak hold to an RTA is all but useless IMO. For transients, SL will favor the lower frequencies because they have more energy but some soundtracks hold out tones up in the higher registers and in that case it will show equal levels accurately so tilting it's response won't end up being accurate to levels either.

I appreciate you pointing out the voltage function on SL. I didn't know about that and it'll definitely be something fun to play with in the near future.
Shreds is offline  
Old 06-05-2015, 08:28 AM - Thread Starter
AVS Special Member
 
desertdome's Avatar
 
Join Date: Aug 2004
Location: Elkhorn, NE
Posts: 2,310
Mentioned: 50 Post(s)
Tagged: 0 Thread(s)
Quoted: 439 Post(s)
Liked: 488
Quote:
Originally Posted by J_Palmer_Cass View Post
it would be nice if the SpectrumLab program had an RTA calculated bin option instead of the standard fixed FFT bin width standard.
This sums up the purpose of this thread very well.
desertdome is offline  
Old 06-05-2015, 09:02 AM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 353 Post(s)
Liked: 111
Quote:
Originally Posted by desertdome View Post
This sums up the purpose of this thread very well.


The best we can do with SpectrumLab at this point in time is to place a "reference line" on the upper Spectrogram, or to add a reference offset to the SpectrumLab waterfall.




You can not talk to "the boys" over at the other site. To them full range means 0 to 120 Hz. They have no idea of how high a sample gunshot from say Open Range goes when you view a Spectrogram that is octave based rather than fixed bin based.




I am waiting for LTD to become totally exasperated and just give up over at the other site. I know better, so I just never joined "the secret society".
J_Palmer_Cass is offline  
Old 06-05-2015, 09:21 AM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 353 Post(s)
Liked: 111
Quote:
Originally Posted by Shreds View Post


Put up or shut up eh? This from the guy who posts no data but is super critical of everyone who does. You bore me.


Deception again? I posted data that you obviously chose to ignore or can not understand. Learning disability perhaps? Do you still think that we are talking about a 10 dB per octave roll off?


How about posting the mic'd vs. electrical comparison of your system instead of all of your "useless" data? "You bore me" is not a proper response from a "full range" advocate.


Bosso has my old comparison in his files. My old comparison is biased in that I used a digital RS SPL Meter as my microphone, and that adds in the SPL C Scale roll off to the Spectrogram. That was close enough for my needs.
J_Palmer_Cass is offline  
Old 06-05-2015, 09:45 AM - Thread Starter
AVS Special Member
 
desertdome's Avatar
 
Join Date: Aug 2004
Location: Elkhorn, NE
Posts: 2,310
Mentioned: 50 Post(s)
Tagged: 0 Thread(s)
Quoted: 439 Post(s)
Liked: 488
I should have added this to JPC's sentence:

"It would be nice if the SpectrumLab program had an RTA calculated bin option instead of the standard fixed FFT bin width standard, like both Omnimic and REW offer."

If I would have started with this, then maybe we wouldn't have gotten all the silly straw man arguments which have nothing to do with this sentence. (Wishful thinking )

Quote:
Originally Posted by Shreds View Post
I thought the whole thing with this thread was about SL not being accurate to SPL levels for short duration higher frequencies. I'm not seeing an alternative here to something that is accurate in this regard. If spectrafoo is the answer here because it does more things than SL, let's see some results of using it for some burst tests at different frequencies and see how it does.

Otherwise, matching SL's peak hold to an RTA is all but useless IMO. For transients, SL will favor the lower frequencies because they have more energy but some soundtracks hold out tones up in the higher registers and in that case it will show equal levels accurately so tilting it's response won't end up being accurate to levels either.
Now your talking about using burst tests? I was specifically told both here and at data-bass:
  • Fortunately, movie sound effects aren't logarithmic sine sweeps.
  • We are not analyzing measurement sine sweeps, pink noise, white noise or tone burst noise. We are analyzing audio files in various formats, showing a waterfall, peak hold trace, average trace and amplitude waveform.

So, I took actual movie content (starting in post 74) and showed exactly how there is a difference by doing the following:
  1. Showed that Spectrum Lab's FFT is the same as REW's FFT
  2. Showed that REW's FFT is different than REW's RTA
  3. Compared REW's FFT to REW's RTA for two movies to show everyone the difference
  4. Explained why Spectrum Lab's FFT and REW's FFT is different
  5. Explained that both Omnimic and REW allow two views of the data depending on if they are showing Maximum Power or the Sum of Power

Spectrum Lab's FFT and REW's FFT and RTA show the following:
  • Bass is mixed higher than other frequencies - I haven't argued against that
  • Low bass is present in many movies - I haven't argued against that
  • Bass has more energy and SPL than other frequencies - I haven't argued against that
  • The LFE channel is increasing bass levels starting at about 120 Hz - I haven't argued against that
  • Movies are mixed with a lot less content above 2 kHz - I haven't argued against that
desertdome is offline  
Old 06-05-2015, 10:16 AM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,585
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 353 Post(s)
Liked: 111
Note that Shreds is talking about SPL levels. It is improbable that anyone can calculate SPL levels by eye based solely on an FFT chart.
J_Palmer_Cass is offline  
Old 06-05-2015, 10:50 AM
Senior Member
 
Shreds's Avatar
 
Join Date: May 2014
Location: NC
Posts: 280
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 177 Post(s)
Liked: 73
Quote:
Originally Posted by desertdome View Post
Now your talking about using burst tests?
Yeah, for figuring out accurate SPL of different frequencies regardless of frequency duration. I thought that was what this whole thread was about.... -accurately assessing SPL of the frequencies on the spectrogram. You've met resistance with the RTA tilted curve thing because that would be no more accurate for SPL levels than using SL by itself. That and you claimed some horse crap about a ported sub with a limiter being able to accurately reproduce full bandwidth content.
Shreds is offline  
 
Thread Tools


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off