BSC discussion - split from Usher two way thread

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
sreten said:
Hi,

What can be suggested to be a timing issue, i.e. it sounds like it has
more timing, doesn't mean it has more timing. Naim amplifiers sound like they have really good "grip" of the speakers but they don't "grip"
a speaker better than other amplifiers - they are built with internal
details that cause this apparent phenomena / type of sound.

:)/sreten.


Ok... I am on board. So, I think I have made it relatively clear what I mean by "timing" - well as clear as such a subjective thing can be. I am absolutly sure that whatever I am refering to is something very real about a hi-fi component (and not just speakers) so what is it? If the speaker is not ACTUALLY messing with any cues in the time domain then what is it that I am identifying as sloppy timing? Or good timing in that case? It certainly SOUNDS like multiple sounds that should be made at the same time are being made slightly apart from each other, but I am sure other things could be responsible. Perhapse some sort of linear distortion? Well whatever it is that I am identifying as "poor timing" I am still fairly convinced that it relates somehow to BSC. Does anybody else know what I am talking about (because apparently I dont)? :xeye:

Thanks,
-Chad
 
Hi Chad,

I have a few thoughts about why BSC may appear to affect the "PRAT" of a speaker.

My first few DIY speakers had no BSC. My comparision speakers at the time were commercial monitors with flat frequency response. Listening to my new speakers I could hear oodles more detail. I thought that drums had a sort of "crack" that made be blink, guitar strings had a "click" when strummed. Perhaps I could say that all these sounds put together eccentuated the rythm of music. Does this kind of describe what you would call pace and timing?

Anyway, to cut a long story short, I simply think that unbalanced frequency response can create what you are describing. This can be a design decision, it doesn't necessary mean it's a bad speaker. Perhaps a little rise in response can sound good.

I always use BSC now. I think it does reduce this "PRAT" effect, but is of course more natural. I've also used active and passive line level BSC. The result is the same as with speaker level BSC.

I also think that speakers without BSC encourage louder listening volumes, which exaggerates this issue when trying to compare speakers with even frequency response.

Has anyone seen measurements of Naim speakers? I haven't, but I have heard them on many occasions.

Andrew.
 
Totally agreed with Andrew.

Non flat frequency response can add some excitment to certain types of music but at the same time can make some other music worse. At the end of the day, if we are looking for accurate reproduction of music and to make a pair of speakers suitable for all types of music, we should be looking for flat response.

However, frequency response is one thing, power response is another. If a BSC introduces power loss at high frequencies due to close distance to the front wall so that the reflections come mixed with the direct sound, I would probably compromise it a bit by not introducing the full BSC so that power response is more even.

Regards,
Bill
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I have been through this issue and I have to say 3 things:

1. Elevated mids and/or 3rd harmonic cause false speed up.

2. In speakers going down to the 30s use 3dB of bsc, in speakers going down to the 50s use 4 dB of bsc.

3. Foam and cloth suspension cones time better.
 
I suggest care when calculating BSC, and ideally using a program such as Svante's Edge. It turns out that tall cabinets can shift the baffle step down by close to an octave if the box is much higher than wide.

For example, a 13" wide cabinet should have a step 3 dB point around 330 Hertz, but for a 72" high baffle the actual step is closer to 160 Hz. You'd end up with a nasty lump at 220 Hz if you use 330 Hz BSC for that box.
 
Hi,

So one person thinks BSC causes timing errors by causing it to lag,
and another thinks no BSC causes timing errors by false speed-up.

As an engineer is fairly obvious where this is going, e.g.
Too bright a tweeter - oodles of apparent detail. Can I fix it ? of
course I can. But someone wants to keep the detail and lose the
brightness - I cannot do that - you cannot have it bothways.

I do find it amusing serious equipment has no tone controls, but
that equipment judgements are often based on tonal balance.

Not that I like standard tone controls - they don't work well.
I do like passive line level tone tweaking - costs peanuts.

:)/sreten.
 
So sreten, you are a bass player AND and an engineer? How could you possibly be worng !!! ;) I am an engineer and the smartest man alive, so I am probably right too :D .

Anyways, joking aside, I will try to post some graphs tonight, but I am beginning to wonder if an oversized series inductor can possibly correct for baffle step in an even and ballanced way. What I am having trouble with is making the slope caused by the inductor go all the way down to the start of the baffle step. The transfer function for the crossover shows that the inductor is effecting the 200-800 region, but when you plot the driver with and without the inductor it doens't seem to start effecting the FR noticably untill more like 600-800, giving the upper bass a bloated sound. So I can get a fairly flat looking FR, but it doesn't sound like it does when I use an L/R circuit. The other thing that I have noticed is that when I use the L/R I still need a very large inductor to hit my target slope (2200 LR4). Like I said, I will post some graphs tonight, but any ideas you guys have before I go through all the work to put these ideas together graphicaly would be great.

Thanks,
-Chad
 
So one person thinks BSC causes timing errors by causing it to lag,

If something improves (i.e. linearises) the amplitude response then it is most likely also improving the phase response and therefore timing (I am almost sure that I am with sreten on this one).

There are many people who are NOT used to accurate bass reproduction. If they hear the decay of the bass-drum resonance they say the speaker under test is slow - while the speakers they are used to listen to simply "peculate" the low-end content.

So I belong to the camp who would call the lack of BSC "false speedup" - using sreten's terminology.

It is physically impossible that the resonance of a snare starts at the very same instant as the initial "crack" BTW.

Regards

Charles (another engineer who is playing the bass)
 
If something improves (i.e. linearises) the amplitude response then it is most likely also improving the phase response and therefore timing (I am almost sure that I am with sreten on this one).

A few years ago, I measured my Lowther ML TL speakers with and without the BSC and the results were as you describe. Without the BSC at the point where the SPL level started to rise a gentle phase shift occured, with the BSC the SPL and the phase curves were smoothed out.

Having an overly bright speaker can at first sound more detailed and a listener will be drawn to its perceived speed and bright sound. It will sound dynamic and live while a compensated speaker will sound somewhat veiled. However, I found that this only lasts a while and you quickly become fatigued without a BSC circuit. In my opinion the best way to determine the values for a passive BSC is by measurement or calculation to get a preliminary design and then final adjustment by listening. For my room and my system, I tend to end up with 3 to 4 dB of correction for my full range speakers.
 
Oops, this was a fast growing thread... :)

My opinion on the baffle step is that it is no different* from other things that affect the frequency response. If the driver has a little peak somewhere, the system would benefit from a compensation. If there is a room gain present, it is a good idea to compensate for that, at least partly. And if there is a baffle step, it should be compensated for.

Adding these compensations results in a system that has better frequency response and better transient response, when the system is seen as a whole.

* The baffle step adds a tiny bit of directivity that other effects might not do, but that is beside the point of this.

Edit: It is not nessecarily so that BSC is visible by looking at the crossover of a loudspeaker. In some cases the voice coil inductance and the response of the drivers can be used to acheive a flat response without the typical RL compensation circuit.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.