Jump to content

Bass test signals


Kvalsvoll

Recommended Posts

(Although this is bass content in the form of videos and signals, I think it is more suited here in the gear section)

 

I will create a small selection of bass test signals - sweeps, impulses, LOUD sine waves.

 

Some of these I will put up on youtube - easy access for a lot of people - unfortunately the quality limitations makes it necessary to download files for anything performance.oriented.

I will post links to those files when they are ready.

 

Here is the first one I made - a simple sine sweep from 5Hz - 100Hz:

https://www.youtube.com/watch?v=Sxs7GVbmWqs

 

Not very useful for all the experts with REW, but still fun and easily accessible for a larger audience.

 

Description:

 

How low can you hear?
Watch the sound frequency increase and observe when you can hear the very deepest bass sound.

If you hear clearly audible sound below around 16Hz, this is distortion - what you hear is sound with higher frequency, created in your speakers trying to reproduce the low frequencies you can not hear, or due to the AAC data compression.

Sound down to very low frequencies can be observed - we feel the pressure changes, and we feel the movement from floor and seating. We can even feel the sound waves as wind, if the sound intensity is very high.

A high performance sound system will be able to reproduce this strange feeling of airy waves and a sense of large scale. Depending on the sound system, room acoustics and floor foundation you will typically be able to sense low bass from around 10-14Hz.

On headphones it will be difficult to hear anything below around 20Hz.

Warning:
Do not turn up the volume too loud!
The sound is recorded at 0dBFS* - even though you may not be able to hear it in the beginning, it will get very loud and may destroy your speakers if turned up too loud.

 

On a calibrated (85dB for -20dBFS) system the sound will produce 111dB SPL at 0dB master volume setting, if the frequency response of the system is flat. Most systems have a tilt with some gain at lower frequencies, so add maybe around 6dB, or in the 3-10dB range, and we will typically measure around 117dB in the room, which is quite loud.

Start with a lower master volume setting, -20dB will give around 97dB and this is sufficient to hear sound down to well below 20Hz.

Unfortunately the AAC data compression at 192K bitrate creates very audible distortion and artifacts.
At master volume level around -20dB this distortion is barely audible, at -10dB the distortion is easily heard and quite annoying, espescially when I know how it is supposed to sound.
This AAC data compression is a technical limitation on YouTube.
We can only hope that YouTube will improve and soon allow for better audio resolution.
With ordinary music signals these faults may go unnoticed, because the distortion will be masked by higher frequency sound.


*: Actually, the level is -0.2dB, necessary due to limitations in the AAC encoder.

Link to comment
Share on other sites

Made a distortion-free (no crappy aac) movie in .mkv format with full 7.1 lossless flac soundtracks and low frequency sweep encoded at 0dBFS in LCR and LFE:

http://www.kvalsvoll.com/Articles/TestSignals/ulf_5_100hz_71_lcr_lfe.mkv

 

Any distortion you hear now is from your playback chain - processor internal clipping, pre out clipping, subwoofer amp, speakers.

 

Sweep is encoded to -0.1dBFS in channels LCR and LFE.

This adds up to 121dB (LFE alone is 115dB).

 

Adding a typical +6dB hot bass we expect to measure 127dB rms or 130dB peak on the spl logger for 0dB master volume. 

Very few systems will be able to reproduce this clean.

Start from -30dB and work your way up.

Also consider the heat build-up with this 30 sec long sweep, very possible to burn some drivers.

 

This can be used to find the limits on your system, with no need to measure anything.

When it starts to sound bad, that is your limit.

 

This can also detect problems with internal clipping of the bass-managed subwoofer signal in the processor.

If the signal sounds distorted at very low volume, the fault is in the processor.

Link to comment
Share on other sites

Indeed.  And unfortunately, I'm not sure you'll always be able to tell before it's too late when you're dumping too much power into your coils.

 

They usually start to smell really bad..

 

This is problem mainly when testing subwoofers, because you can pump out several 100W - even KW - and it still is not so loud it kills your ears.

If a driver is rated for 800W it does not mean it will take a continuous 800W for several minutes, the voice coil is far too small to be able to dissipate the heat.

Link to comment
Share on other sites

People, heed the warnings above, or you will be opening your wallet for more drivers.

 

I still have a  7.1 Test disc on the back-burner somewhere (man does time fly), looks like Kvalsvoll is gonna do it for real.

 

JSS

 

I already have many test files for 7.1 hosted on the web, but they were never announced and they are audio only.

 

If you have some suggestions for test files, I can make them.

A video is better because it is possible to add real-time information on display for what is going on.

 

REW and measurement software is great, but usually limited to 2 channels.

Test files with 7.1 signals will often be a much quicker approach to see what is going on and check things.

Link to comment
Share on other sites

This is great. I've been pushing for a new omnimic disc for some time now. The fact that there is still only 5.1 is silly. at least 7.1 and a pipe dream for atmos channels as well...It doesn't seem to me like it would be all that hard. maybe even a good pink noise track for each channel would be nice.

Link to comment
Share on other sites

I don't have an encoder for Atmos, unfortunately.

Would be nice to have 7.1+Atmos test signals with pink noise to verify level and frequency response.

 

The TestSignals-folder on my web site has 7.1 test signals, but they are audio only:

http://www.kvalsvoll.com/Articles/TestSignals/

 

The level for lfe is off in those test signals, I think, it will not match level correctly.

 

Checking level and response is quick and easy with those signals.

You set up your mic and REW in RTA mode, then play the test signal for the speakers you want to check.

 

Some of you may not have a computer/htpc set up as source in your system.

You need that to be able to play the files.

This is easy to do now today, just install Kodi on a laptop and connect with hdmi to your avr/processor.

Make sure you go into system settings - audio and enable the proper settings, and also verify that your processor/avr was detected.

 

Pink noise is good enough to verify frequency response, it will match up quite well with a sine wave sweep measurement especially at higher frequencies, where the sound will be dominated by the direct sound.

Link to comment
Share on other sites

They usually start to smell really bad..

 

This is problem mainly when testing subwoofers, because you can pump out several 100W - even KW - and it still is not so loud it kills your ears.

If a driver is rated for 800W it does not mean it will take a continuous 800W for several minutes, the voice coil is far too small to be able to dissipate the heat.

 

I just figured that once you start smelling something bad, damage is already done.  That means it's hot enough for something (adhesive most likely) to be decomposing and vaporizing.  Then there's the issue of how long it takes for that yuck smell to escape a sealed box and travel to your nose.  I think a safer approach would be to look for signs of power compression in the output.  Once the coil starts to heat up, it's probably not going to be able to take much more power before it gets too hot.  Of course, this requires accurate measurements in conjunction with the test signals.  An even more elaborate test scheme would be to measure impedance in real-time and look for the rise in resistance, but for speakers with passive crossovers, the passive crossover might overload before the driver does in some cases.

Link to comment
Share on other sites

I just figured that once you start smelling something bad, damage is already done.  That means it's hot enough for something (adhesive most likely) to be decomposing and vaporizing.  Then there's the issue of how long it takes for that yuck smell to escape a sealed box and travel to your nose.  I think a safer approach would be to look for signs of power compression in the output.

..

 

Absolutely.

 

If power specs could be trusted, it would be possible to use the numbers for a good estimate, but since even the pro drivers are specified either as AES or some "program material" definition, it is not so easy when you play continuous sine waves. 

 

I noticed that the drivers in the horns started to smell a little after some rounds of testing, and I don't think that is a problem, but if it starts to smell very bad - noticeable at a long distance, that should be taken as a sign that something is really running far too hot.

 

Those drivers definitely held up to the abuse, but those are pro drivers, they have very high thermal power handling and motor assembly designed for convection cooling.

Link to comment
Share on other sites

You have to be absolutely careful with high duty cycle test signals. Coil smell is an immediate stop point for me because damage is either already occurring or not far behind. It is easier to smell in a vented cab or tapped horn. Once you start to smell it from a sealed or FLH type cab it usually has caused some damage IMO.

Link to comment
Share on other sites

Fresh drivers can give off a slight smell of "hot". the voice coil can take quite high temperatures without damage, and this will cause evaporation of gases from glue and surrounding materials.

 

If it start to smell like something is melting or burning, it is definitely getting too hot and something is damaged.

Link to comment
Share on other sites

You have to be absolutely careful with high duty cycle test signals. Coil smell is an immediate stop point for me because damage is either already occurring or not far behind. It is easier to smell in a vented cab or tapped horn. Once you start to smell it from a sealed or FLH type cab it usually has caused some damage IMO.

 

 

So when your room smells like a tire factory you mean to tell me I have to turn it down some? 

Link to comment
Share on other sites

So when your room smells like a tire factory you mean to tell me I have to turn it down some? 

 

Hell nah. :huh: Just break in. Gets the dirt off the coils and gets the party buzzed too. Ain't a real party till Puff the dragon shows up to pack a few subs in his magic hookah.

Link to comment
Share on other sites

Now it gets more interesting.

 

Impulses with different base frequency.

100hz - 50hz - 30hz - 20hz - 16hz - 12hz.

They look like this:

 

post-181-0-32182600-1456522553_thumb.png

 

Example spectrum of the 20hz impulse:

 

post-181-0-54693900-1456522572_thumb.png

 

Spectrum has a peak around the base frequency, but there is content spread around a wide frequency range.

The different impulses will sound different, and they have different tactile feel.

 

The 12hz and the 16hz is quite violent, shakes everything.

The 100hz sounds more like a short bass impulse, and hits in the chest.

 

2ch sound only version:

http://www.kvalsvoll.com/Articles/TestSignals/sponk%20100hz__12hz%202ch.flac

 

7.1 with impulses enoded on all channels:

http://www.kvalsvoll.com/Articles/TestSignals/sponk%20100hz__12hz%207_1.flac

 

The 2ch version is quite safe, but still, beware of the volume, it may not sound loud at all, then suddenly you have bottomed out a sub driver and perhaps damaged something.

 

The 7.1 IS NOT SAFE.

If you do not have limiters properly set up you will bottom out drivers and cause serious damage. 

 

Peak level is -2.1dB, I may update the files later on with levels set close to 0dB. 

Still, the level is more than sufficient to find the limits of most systems.

 

Purpose of this is to find and check limits of the system, and to find analyze tactile feel at different frequencies.

 

I found that these test signals reveal a lot of rattle and noise in the room and surrounding rooms, it may not sound loud at all, because there is no higher frequency content, and the duration is short.

It is quite clear that distortion from the subwoofer system is not an issue before you exceed capacity limits, all the rattles an noise form vibrating objects drowns out any distortion from the subwoofers.

 

I also found the limit of the current system in the demo room, it does not play the 7.1 12hz impulses at 0dB MV clean.

Link to comment
Share on other sites

Generated in Audacity using the Risset Drum generator.

Unfortunately it turns out is does not work well for this purpose, so I ended up making one 50hz impulse and then shifting that one in frequency to create the others. 

 

The signals could also be generated by filtering a step impulse or unity/dirac.

Or from a sine wave, cutting and filtering and fading.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...