Let's get honest with SPL

[killdozzer]

In some forms of music, there can be significant differences in level with time, as we all know. If we have a speaker that works well at 86 dB, but falls apart after about 94-95 dB, it will do fine provided the level doesn't exceed the 94 dB limit. For something like compressed pop where we are listening at around 80 dB all is well. The speaker sounds great and we are happy. If we are listening to classical music with a number of quiet passages, we are likely to turn it up enough for them to be audible. But when there is something like percussion or crescendo of the orchestra the levels may go up much higher. Say 95-100 dB. Here we are beyond the limits where the speaker can produce undistorted sound, and it will misbehave in some way. Usually a well-designed woofer can cope by just getting to its excursion limits and limiting any further travel in a controlled fashion and it will sound muddy or maybe have some crackle, but tweeters can produce a lot of nasty high frequency harmonics when they are overdriven. This means that instead of the desired awe of orchestral power, we will get lots of unpleasant distortion and noises during those passages as we have exceeded the limits of the speaker. In many ways, its like clipping since these effects seldom come in gradually, but fairly rapidly as we get to the limits of the transducers in my experience. Ideally you want enough reserve to not go too far into the large signal portion of the transducer's response as that is where things start getting nonlinear.

Yes, that I know. I still don't see why would it be more important in determining what is the preferred upper limit of SPL that members here enjoy listening at? You turn it up to whatever tickles your fancy and just write it down. Also, that's why I proposed a well produced song. Hey nineteen shouldn't be compressed. At least I haven't come by one that is.

There is no sense in determining the upper volume limit by quiet passages in classical music nor by the short loud burst. I hoped we've ruled those out by saying it's the loudest volume you stick to during at least one album. So, no short quiet passages or short crescendos.

If you rarely listen to pop or such, maybe you could use another music program. But my guess is that you would listen to what ever SPL still comes out clean at the loudest part.

Maybe the loudest part of Carl Orff's "O Fortuna" is long enough and loud enough for you to measure. Just don't turn on your SPL meter during the quiet part because the average will get screwed. Wait for the loudest part, turn it up to your genuine liking and take down the number.

 

[dasdoing]

Just don't turn on your SPL meter during the quiet part because the average will get screwed.

the thing is, the average is not what we should look at in this context. the speaker has to be able to reproduce the loudest part.
on a sidenote: since the big majority here agrees that EQing the room response is necesary, there is actualy the added factor of necessity of headroom.

 

[killdozzer]

the thing is, the average is not what we should look at in this context. the speaker has to be able to reproduce the loudest part.
on a sidenote: since the big majority here agrees that EQing the room response is necesary, there is actualy the added factor of necessity of headroom.

I don't see what's so hard to understand here. We look at the average because music has dips and peaks. I can't say I listen to 81dB because Hey Nineteen goes up to 81dB in some microseconds. I also can't say it's the bottom 71dB. If your speaker distorts at 96 and you turn it down, then 96 is not the volume you listen at because, as it says, you turn it down. Perhaps some don't care if it distorts, this is completely subjective, maybe they will leave it at 96dB and distorted and listen to it like that. You listen how you listen and at your chosen level and conditions. I'm only interested in what reaches you at your seat. If your gear limits you, then it's anyone's guess. We won't know until you get new equipment.

Loudest part says nothing about your preferred SPL because it's often brief so everyone just goes through the loudest part. I assume you don't jump at the remote and turn it down because of a few loud seconds.

Another thing is, if your SPL is limited by your equipment as in "I'd like to go louder but it distorts/clips" then you simply don't have the gear that plays at your preferred level but at the limit of distortion. Which is a number often given in measurements so I wasn't that curious about it. My speakers also distort at one point and it is too low of a point in Amir's opinion, but luckily, it's way above what I need and like. If your speakers distort before you reach the level that you like, then we still don't know what that level is, right? However, one other way you could determine the SPL you like would be to use program material with less dynamic range so that it doesn't go into clipping/distortion before you reach it. This is strictly to determine the loudness you like. You don't have to carry on listening to it.

And, I point out again, that's why I picked (a rather boring IMHO) song that doesn't have the "wildest" dynamic range and is very clean with mostly pleasant sounds. You won't be turning it down because of some Overdrive pedal effect.

 

[Cars-N-Cans]

Yes, that I know. I still don't see why would it be more important in determining what is the preferred upper limit of SPL that members here enjoy listening at? You turn it up to whatever tickles your fancy and just write it down. Also, that's why I proposed a well produced song. Hey nineteen shouldn't be compressed. At least I haven't come by one that is.

There is no sense in determining the upper volume limit by quiet passages in classical music nor by the short loud burst. I hoped we've ruled those out by saying it's the loudest volume you stick to during at least one album. So, no short quiet passages or short crescendos.

If you rarely listen to pop or such, maybe you could use another music program. But my guess is that you would listen to what ever SPL still comes out clean at the loudest part.

Maybe the loudest part of Carl Orff's "O Fortuna" is long enough and loud enough for you to measure. Just don't turn on your SPL meter during the quiet part because the average will get screwed. Wait for the loudest part, turn it up to your genuine liking and take down the number.

I think two things are being conflated here. Your original comment was in regards to a speaker not being happy at 96 dB in Amir's testing, and that was the relevant part in my comments regarding the transient capabilities. The 96 dB limit is an absolute one in that once the SPL level, regardless of time scale, gets there the speaker may start to exhibit nonlinearities. Whether or not they will be an issue depends on the nature of the issue, and how long it persists for. We should not conflate this with comfortable listening level since that is the average level over some measurement interval without regard to how high the SPL gets at any particular time. If we listen at 80 dB, and the music spectrally contains passages that are 5 dB beyond that, then the speaker needs to, in theory, be able to achieve at least 85 dB of clean playback. If we have a speaker that does ok at 86 dB, but has internal limiters or performance limitations that cause it to fall apart at 96 dB, then we are only just meeting the requirements. There is little headroom to spare. On the other hand, if it can still hold it together at 96 dB and still be at least somewhat useable, then we are golden. And that is at least in part relevant to the thread in that just because we only listen at 80 dB doesn't necessarily render the output at 96 dB to be a gross excess of capabilities that we don't need. Actually trying to use that same speaker at an average of 96 dB would be an obvious disaster since we would be overdriving it much of the time. Its sort of analogous to maximum RMS output from an amp. Beyond that we get clipping and distortion. Ideally we want to reasonably accommodate everything that comes through the speakers during normal use. I would say for most people the 86 dB one is what matters, but the 96 dB is also relevant as it shows what happens when we exceed those limits. Id be at least slightly concerned if all the distortion just went off-scale in the 96 dB plot. That means the speaker will suddenly fall apart if I turn it up more than normal or there is a sudden loud passage of music. Edit: I don't want to get too far off-topic, but I do have a similar speaker to one tested here at ASR that both have the same tweeter that just makes it under Amir's "red line" in the 96 dB distortion plot. On occasion, the tweeter will sometimes produce audible distortion, indicating that, at least over a short time scale, its limits are being exceeded and likely going beyond 96 dB for short periods of time. This, however, is quite reasonable as 99% of the time at around an average of 80-85 dB, they produce clean audio. If these just absolutely cratered and turned into distortion factories by the time they got to 96 dB I would probably end up tossing them and get something else. There would be audible distortion any time I try to get to spirited, but comfortable, listening levels.

 

[dasdoing]

if distorsion in trasients is audible is something that should be blind tested. I personaly never heard realistic transients before I got speakers that are capable of 120dB (not saying that 120 is necessary).
however, when it comes to bass, we often get close to 0dBFS without it being a transient.