Head-Fi.org › Forums › Equipment Forums › Sound Science › Test transient response
New Posts  All Forums:Forum Nav:

Test transient response

post #1 of 28
Thread Starter 

Can anyone point me in the direction of any music/tracks that can help test the transient response of a speaker/headphones? And possibly also what I should look out for when listening?

 

Thanks

post #2 of 28

There's no difference between "transient" response and regular high frequency response. So if bright sounding music sounds crisp and clear, you're good to go.

 

--Ethan

post #3 of 28
Thread Starter 

Thanks
 

post #4 of 28
Quote:
Originally Posted by EthanWiner View Post

There's no difference between "transient" response and regular high frequency response. So if bright sounding music sounds crisp and clear, you're good to go.

 

--Ethan

 

So you wouldn't  count a single cycle of a 50Hz sine wave as a transient??? Not even if it was near the maximum amplitude the system could handle? If this is standard audio terminology, then it is very silly, but I really doubt that it is - I can't find anything to agree with your definition, and plenty that strongly implies that is wrong in that people talk specifically about high frequency transient response - not to mention bass transient response, and stuff like the need for careful design of bass speaker cabinets so they can handle transients properly!

post #5 of 28

From wikipedia:

 

 

Quote:

 

 

http://en.wikipedia.org/wiki/Transient_(acoustics)

 

In acoustics and audio, a transient is a high amplitude, short-duration sound at the beginning of a waveform that occurs in phenomena such as musical sounds, noises or speech.[1][2] It can sometimes contain a high degree of non-periodic components and a highermagnitude of high frequencies than the harmonic content of that sound[citation needed]. Transients do not necessarily directly depend on the frequency of the tone they initiate.[citation needed]

Transients are more difficult to encode with many audio compression algorithms, causing pre-echo[citation needed].

 

 

"Duration" in this context is simply relative to wavelength. Both cymbals (high freq) and drum shots (low) are excellent examples of transients.

 

Otoh, a sustained whistling sound maybe high frequency but will NOT be transient.

post #6 of 28

Chesky has some good transient test tracks (mostly drum strikes) in his reference CDs.

post #7 of 28

I think his point is, if a system can accurately reproduce a cycle that occupies 1/20,000th of a second, it isn't going to have any trouble with a drum hit that measures in the hundredths of a second.

 

In most cases, bad transients are a dynamic issue more than a time one. Big sloppy transducers can smear over them. Not really an issue in headphones, I would think. I haven't even found it to be an issue in speakers unless they're seriously underpowered or damaged.

post #8 of 28
Quote:
Originally Posted by scuttle View Post

 

So you wouldn't  count a single cycle of a 50Hz sine wave as a transient??? Not even if it was near the maximum amplitude the system could handle? If this is standard audio terminology, then it is very silly, but I really doubt that it is - I can't find anything to agree with your definition, and plenty that strongly implies that is wrong in that people talk specifically about high frequency transient response - not to mention bass transient response, and stuff like the need for careful design of bass speaker cabinets so they can handle transients properly!

Ethan is correct.  You might be confused by considering a transient in the time domain only because it's of short duration, but it can also be considered in the frequency domain.  It's the same data, just a different view.  Any system capable of full bandwidth audio is quite capable of passing any transient in the audio band.  Transients may have many different spectral distributions.  But in audio systems, passing low frequencies isn't usually much of a challenge, where passing highs might be.  So, Ethan is correct in his statement, especially regarding transients with frequency distribution in the upper octaves.  

 

The issue with speakers handling a bass transient is mostly one of damping, which also can show up as a response issue too, if you know where to look in the waterfall.  Well damped smooth bass response in a speaker is often incorrectly called "fast bass", which is an oxymoron.  There's nothing fast about bass, not even a so-called transient. 

 

You seem to also be trying to define a transient by its amplitude, which has nothing to do with the definition.  A transient can be of any amplitude. 

 

Sorry if you feel audio terminology is silly, it was invented by some rather non-silly blokes a long time ago.  

post #9 of 28
Quote:
Originally Posted by jhamdg View Post

Chesky has some good transient test tracks (mostly drum strikes) in his reference CDs.

 

The ultimate transient here:

 

http://www.audiocheck.net/testtones_impulse.php

 

;-)

post #10 of 28
Quote:
Originally Posted by bigshot View Post

I think his point is, if a system can accurately reproduce a cycle that occupies 1/20,000th of a second, it isn't going to have any trouble with a drum hit that measures in the hundredths of a second.

 

I don't think that was his point - it certainly wasn't what he wrote! - and if it was, it was utterly wrong: the ability to deal with high frequency transients is no guarantee at all of the ability to deal with low frequency ones. Why should it be? It's quite possible to build a membrane that vibrates well at 20,000 Hz and poorly at 50Hz - indeed, if systems didn't have different responses at different frequencies then neither musical instruments or transducers would have frequency limits or varying response curves.

 

(In fact, it's pretty damn common for headphones to have widely differing bass transient response: if I play something that's pure treble on my HD25s and 668Bs they're not easy to tell apart, but the difference when bass transients are involved is super obvious.)


Edited by scuttle - 2/19/13 at 5:55am
post #11 of 28
Quote:
Originally Posted by jhamdg View Post

Chesky has some good transient test tracks (mostly drum strikes) in his reference CDs.

 

Drums will be bass transients; cymbals will be treble transient. Mid-range transients could come with from staccato piano or acoustic guitar - flamenco might be a good bet.

post #12 of 28
Quote:
Originally Posted by scuttle View Post

 

I don't think that was his point - it certainly wasn't what he wrote! - and if it was, it was utterly wrong: the ability to deal with high frequency transients is no guarantee at all of the ability to deal with low frequency ones. Why should it be? It's quite possible to build a membrane that vibrates well at 20,000 Hz and poorly at 50Hz - indeed, if systems didn't have different responses at different frequencies then neither musical instruments or transducers would have frequency limits or varying response curves.

 

(In fact, it's pretty damn common for headphones to have widely differing bass transient response: if I play something that's pure treble on my HD25s and 668Bs they're not easy to tell apart, but the difference when bass transients are involved is super obvious.)

Sorry to disagree, but he was utterly correct.  Again, you're confusing time and frequency domain representations.  Your reply to jhamdg illustrates the misunderstanding.  A drum hit contains a wide spectral distribution, not just bass.  The actual hit of the beater to the drum head is actually made up of some fairly high frequencies as well as low, which if you remove them, you end up with a very dull thud, not a drum hit. Perhaps that's what you calling a "bass transient", in which case we're (once again) arguing language usage.  Perhaps a re-reading of your Wiki excerpt might help, "short-duration sound at the beginning of a waveform", and, "It can sometimes contain a high degree of non-periodic components and a higher magnitude of high frequencies than the harmonic content of that sound."  

 

All of the other transient sounds you mention are also distributed more broadly, especially acoustic guitar where plucked strings attacks are quite high.  

 

It may also be prudent to consider the source of the information you claim is "utterly wrong".  It might be that a post comes from someone new to audio without much experience, or possibly he's a person with an entire career in the industry, a manufacturer of audio products that depend on an thorough understanding of the science, and a writer of authoritative works on the subject.

 

Might want to pick your battles.

 

Or, even better, don't battle at all.  

 

Thanks for your consideration.  

post #13 of 28
Quote:
Originally Posted by jaddie View Post

Sorry to disagree, but he was utterly correct.  Again, you're confusing time and frequency domain representations.  Your reply to jhamdg illustrates the misunderstanding.  A drum hit contains a wide spectral distribution, not just bass. 

 

 

EVERYTHING that isn't computer generated contains "a wide spectrum". (For a certain value of wide, anyway.) Saying that because a drum hit contains components that are not bass it is not a bass transient is like saying that because an elephant has fleas it isn't a mammal - it goes beyond mere lack of intelligence into bizarro land. To see how stupid this idea is, imagine a very sharp pure bass wave - say a 50Hz sine wave; now impose higher frequencies upon it - has doing so made it any less taxing for a speaker membrane that is bad at low frequency reproduction? Regardless of who is right or wrong, your logic stinks like said elephant after a hard day pulling logs and no bath.

 


It may also be prudent to consider the source of the information you claim is "utterly wrong".  It might be that a post comes from someone new to audio without much experience, or possibly he's a person with an entire career in the industry, a manufacturer of audio products that depend on an thorough understanding of the science, and a writer of authoritative works on the subject.

 

Congratulations on a new logical fallacy: the idea that everyone should imagine that anyone they speak to is an expert!

 

And even if the guy is an "expert", it was a horribly useless answer - because it would mislead the reader into thinking that any transducer that reproduces treble well will produce good bass.. well, what are we going to call drum hits if we do not call them bass transients? This is certainly not the case! Once again, many headphones do well with treble and are poor at bass transients. If this is precluded by some bizarre audio industry definition of transient response, then a reasonable "expert" reply has to make this clear.


Might want to pick your battles.

 

Or, even better, don't battle at all.

 

Yes: I get that you are carrying cross-thread butt-hurt. However, as personally as you seem to take everything, this is not a battle for me (and I had been trying to avoid making you look silly - just because you're incapable of logic doesn't mean you're a bad person afterall!)

 

Anyway, you might want to read http://www.hydrogenaudio.org/forums/index.php?showtopic=72512&st=25

 

- Someone there makes a similar claim about HF and transients, and AB Krueger (a professional engineer and the inventor of ABX testing, so a pretty reasonable authority) debunks this - its's true that good HF is needed for SOME waveforms, but not all, and HF response will tell you nothing about a transducer's ability to reproduce bass transients without phases errors, as the ear is unlikely to detect them above 1000Hz.

 

I suspect that you are confused because HF is needed to make a nice 50Hz sq wave. However, it is not needed to make 50Hz sine wave - and, again, if a half cycle of a 50Hz sq wave at maximum volume isn't a bass transient, then I would like to know what it is? 


Edited by scuttle - 2/19/13 at 9:41am
post #14 of 28
Quote:
Originally Posted by scuttle View Post

 

EVERYTHING that isn't computer generated contains "a wide spectrum". (For a certain value of wide, anyway.) Saying that because a drum hit contains components that are not bass it is not a bass transient is like saying that because an elephant has fleas it isn't a mammal - it goes beyond mere lack of intelligence into bizarro land. To see how stupid this idea is, imagine a very sharp pure bass wave - say a 50Hz sine wave; now impose higher frequencies upon it - has doing so made it any less taxing for a speaker membrane that is bad at low frequency reproduction? Regardless of who is right or wrong, your logic stinks like said elephant after a hard day pulling logs and no bath.

Lets see, we have "lack of intelligence", "stupid", "logic stinks like said elephant"...  you aren't mad, are you?

 

I'm taking exception with "bass transient", because there's more to a drum hit than bass, and the transient part of the hit especially isn't just bass.  I also take exception to the separation of transient and frequency response.  There can be no separation, you can't have good transient response without good frequency response.

 

Quote:

Originally Posted by scuttle View Post

Congratulations on a new logical fallacy: the idea that everyone should imagine that anyone they speak to is an expert!

 

And even if the guy is an "expert", it was a horribly useless answer - because it would mislead the reader into thinking that any transducer that reproduces treble well will produce good bass.. well, what are we going to call drum hits if we do not call them bass transients? This is certainly not the case! Once again, many headphones do well with treble and are poor at bass transients. If this is precluded by some bizarre audio industry definition of transient response, then a reasonable "expert" reply has to make this clear.

 

Do yourself a favor.  Google Ethan Weiner.

 

Quote:

Originally Posted by scuttle View Post

Yes: I get that you are carrying cross-thread butt-hurt. However, as personally as you seem to take everything, this is not a battle for me (and I had been trying to avoid making you look silly - just because you're incapable of logic doesn't mean you're a bad person afterall!)

 

I'm doing what now? 

 

I don't take anything personally, I just think your post is mean-spirited, self righteous, contains little sense of humility, and is full of name calling.  If you only didi it with me, I might take it personally, but I can see that I'm not the target, it's just a general thing.  

Quote:
Originally Posted by scuttle View Post
Anyway, you might want to read http://www.hydrogenaudio.org/forums/index.php?showtopic=72512&st=25

 

- Someone there makes a similar claim about HF and transients, and AB Krueger (a professional engineer and the inventor of ABX testing, so a pretty reasonable authority) debunks this - its's true that good HF is needed for SOME waveforms, but not all, and HF response will tell you nothing about a transducer's ability to reproduce bass transients without phases errors, as the ear is unlikely to detect them above 1000Hz.

 

I suspect that you are confused because HF is needed to make a nice 50Hz sq wave. However, it is not needed to make 50Hz sine wave - and, again, if a half cycle of a 50Hz sq wave at maximum volume isn't a bass transient, then I would like to know what it is? 

 I read the link, and I'm very familiar with Arny, been reading his papers since the early 1980s.  I don't see anything in his post that I would disagree with.  

 

What instrument produces a half cycle of a 50Hz square wave?  Just kidding, I see where you're going.  

 

So, here's what I did.  I digitally generated a signal like you mentioned...50Hz square, half cycle, then ran a spectrum analysis on it.  Here it is:

 

 

Then, just for grins, I generated a 50Hz sine wave, and took just the leading half cycle, and ran spectrum analysis on it.  Here it is:

 

 

The waveforms were generated in Audition, 48KHz 24bit, saved as .wav, then opened and analyzed with Audacity's spectrum analysis.  

 

Comments?

post #15 of 28

Just for more grins, I just tried this: 50Hz sine, 1 cycle fade in, 3 cycle fade out, thus removing the "transient" nature.  Here's the spectrum:

 

 

 

One note: because of the longer duration of the signal greater FFT resolution was possible.

 

I'll hold my own comments for now.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Test transient response