Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Subsonic Filter
#11
(15-Mar-2015, 16:36)Confused Wrote: I did consider posting this in the "effects of SAM" thread, but though on balance it deserved a new thread.

Since getting SAM for the KEF Blades, have noticed the very occasional track that sounds a little strange with SAM, or displays some kind of "bass artefact", for want of a better term.  These are very low frequency "puffs" of other not very musical low bass effects.

I had a theory that these seemingly negative effects of SAM could be eliminated by adjusting the speaker positioning.  For  while, if I ever listened to a track that I felt was producing some sort of bass artefact, or if the bass was simply extremely overblown in some way, I placed the subject track into a playlist.  The idea being that when I had a few of these tracks compiled, I could experiment with speaker positioning, to see what effect it would have on these particular tracks.

Today I spent some time trying to tidy up the cabling of my system, and to facilitate this I had moved the speakers.  With this complete, I positioned the speakers a full 10" (250mm) further away from the wall, an the listened through the playlist of tracks with rouge bass.  The result?  Well moving the speakers away from the wall did change the overall balance toward the "lean" end of the scale.  Plus, almost as proof of concept, the majority of the "rouge" tracks either no longer displayed any of these negative bass artefacts, or at least, the effects were reduced to the point that they were no longer of any real concern.  After much experimentation, I found that actually a 3" (75mm) shift was enough to tame the bass issues, whilst giving a much better overall balance to the sound than was apparent from the 10" (250mm) shift.  I will be leaving the speakers in this position for a while to see how everything sounds after some more prolonged listening.

But there was one track.....   Tricky's "Hell is Around the Corner" (Maxinquaye).  (CD rip, AIR Ethernet)  In addition to sounding a little odd in the very lowest bass with SAM enabled, parts of the track with SAM on produce very large, slow, cone movements.  This is the kind of cone movement you can sometime see using vinyl due to environmental feedback if a subsonic filter is not used.  It is hard to judge, but I would say the cone movement is below the 20hz / audible range.

So some questions.  Firstly, it would be interesting if other users with SAM enabled systems could try this track.  Is it just the SAM / Blade combo that is upset by this track, does it impact some other speakers, large or small?  I am curious!

As a second question.  I am aware that it is possible to enable a "Subsonic filter" in the general settings section of the configurator.  I am thinking of trying this, to see if it tames the unwanted (and slightly alarming) low frequency cone movements that SAM + tacks like "Hell is Around the Corner" can produce.  However, I do wonder if there are any negative to using the subsonic filter.  After all, it does not make sense to compromise 99% of ones listening, just to prevent a minor issue with one or two specific tracks.

My last question, I am sure these cone movements are somewhere in the 10 to 20hz range.  I always thought redbook CD was limited to 20hz?

Hi, i have also the same problem with those rough base artefacts produced by SAM.
(Devialet 120 + MartinLogan ESL)
Since i have enabled the subsonic filter it is less, but to completely prevent it i have also some highpass filters in EquelizerAPO (running von Windows from where i play my music)
Reply
#12
I'm not certain that the problem is a SAM problem and, re-reading the earlier posts in this thread including mine, I'm wondering whether it is actually a room problem or a problem related to the interaction of SAM and the room.

Let's look at the issue:

- it's a bass frequency problem;

- SAM affects bass frequencies below 150 Hz;

- rooms produce bass frequency response issues due to the effect of modal frequencies.

SAM does 3 things. First, at a setting of 0% it corrects the speaker's phase response at frequencies below 150 Hz and that will produce stronger bass at some frequencies. Second, it allows you to extend the speaker's bass response by raising the SAM setting above 0% and doing that will progressively boost low bass response. Third, it has a protection feature "built in" which cuts the boost if you raise the volume to high levels in order to protect the woofers.

What do room modes do to bass frequencies? Depending on your location in the room (where you put your listening chair), the level of bass frequencies related to the room dimensions will be boosted or reduced to some degree.

Let's say your problem occurs at x Hx. SAM may be correcting the phase of that frequency to a degree that it makes that frequency louder at the listening position and you may be increasing the boost at that frequency by using a SAM setting higher than 0%. How do you work out whether SAM is contributing to the problem? Turn SAM off. If the problem is less noticeable then SAM is contributing to the problem.

How do we work out whether the room's modal frequencies are a factor? Move backwards and forwards from the listening position along a line equidistant from the 2 speakers. Is the problem better or worse in some locations than it is at the listening position (ear height can make a difference so try to keep your ears at the same height as they are at the listening position). If the problem reduces or disappears at a location which is different to your listening position then the room is having an effect.

What to do?

If it's only a SAM problem then turn the SAM setting down or turn SAM off. If it's only a room problem then either move the listening position to a location where the problem is minimised while still keeping a sound balance you like and/or move the speakers a little closer or further away from the wall behind them. If both SAM and the room are contributing, then try a combination of those suggestions.

If the subsonic filter is helping then that suggests that the problem is occurring at a frequency below 20 Hz because subsonic filters apply a steep roll off below 20 Hz. If you need to run additional highness filters in order to prevent it then that suggests the problem is above 20 Hz. You didn't say what the highness filter frequencies were but if they're above 20 Hz then the problem is above 20 Hz and you probably don't need to engage the amp's subsonic filter because the highness filter should be sufficient.

If the problem frequency is above 20 Hz then there is another possible cause which is speaker related. If the speaker is ported then what you may be hearing is a "chuffing" sound as air is expelled from the speaker's port at or around the frequency the port is tuned to. If that's the case then you could try sealing the speaker ports if the speaker comes with foam plugs designed for that purpose and see if that solves the problem. You'll probably want to adjust your SAM setting if that suggestion works.

Regardless of whether you can avoid the problem by using the amp's subsonic filter or a high pass filter on your computer, I would try to avoid using that solution because filters are going to affect all frequencies below the set frequency and that means they're going to affect other bass frequencies which aren't causing the problem so I'd first try to solve the problem without using a high pass filter of any kind.

I hope there's something in the above suggestions which helps.
Roon Nucleus+, Devilalet Expert 140 Pro CI, Focal Sopra 2, PS Audio P12, Keces P8 LPS, Uptone Audio EtherREGEN with optical fibre link to my router, Shunyata Alpha NR and Sigma NR power cables, Shunyata Sigma ethernet cables, Shunyata Alpha V2 speaker cables, Grand Prix Audio Monaco rack, RealTRAPS acoustic treatment.

Brisbane, Qld, Australia
Reply
#13
Besides the 3 things mentioned in the previous post, SAM does two more things, one desired from the design (which is positive or negative, depending on many factors) and one undesirable (which is negative).

The desired design is that it progressively reduces frequencies below 30 Hz, which Devialet considered insignificant for music. If I remember correctly, frequencies below 20 Hz are strongly reduced, and below 15 Hz are cut completely. You can test this with audio files with pure and fixed sinusoidal frequencies.

The undesirable by design is that for pure and fixed sinusoidal frequencies in the range of 20-35 Hz, the SAM software "enters into resonance" (I haven't found a more suitable term).
Even at low volumes and at SAM % far from 100%, the bass amplification gets out of control, it is done in an exaggerated way and the protection of the speaker from the SAM is automatically triggered. This happens significantly and visibly only at frequencies of 20-35 Hz (it does not happen either below or above this range).
The SAM % at which this happens depends on the volume and the type of speaker. For example, in my system it happens at SAM 33% in one set of speakers and at SAM 66% in another set of speakers (this at the same volume).
I asked Devialet about this bug and the answer was that SAM was designed for music, not for sinusoids with fixed frequencies. Obviously, I immediately asked myself the question "And if a song contains 1-2 seconds of such a sound?..."

There is something else very important about SAM: if you watch the YouTube video about how the SAM profile is generated, you will easily notice that there are many things that can go wrong. One of the most important is the speaker used by Devialet for measurements: it may not have been burnin (sometimes they use speakers they find at dealers).
There are also the manual decisions / settings made for generating the SAM profile.
For example, I suspected such a situation in the case of a set of speakers of mine and Devialet provided me with an "adjusted" SAM profile (different from the one in the configurator).

In theory, SAM is something very good and intelligent. However, in practice the result depends on luck and many other variables.

After all my experience and all the tests I done, I think the most recommended option is to call Devialet to your home and measure YOUR speakers (Devialet offers this service), AFTER you have burnin the speakers.
Reply
#14
@daniel.avasilichioaei

Your post is the first time I've heard mention of SAM introducing a progressive reduction of frequencies below 30 Hz. I haven't conducted any testing of my own but I don't think that can be quite accurate.

I have no doubt that SAM does roll off at the lower end of its range for a given speaker in order to protect the speaker driver but if one looks at Devialet's quoted claims for the bass extension SAM can deliver for different speakers one finds that for many "small" speakers it does not extend bass to 30 Hz while for many "larger" speakers the response is extended to below 20 Hz. Looking at the claims for a variety of speakers some years ago I came to the conclusion that it extended bass response by something between half an octave and an octave and that the amount of extension claimed increased as the speaker manufacturer's claimed bass response got lower. That's kind of in line with a comment I remember reading in Floyd Toole's book "Sound Reproduction" which said something like phase correction in the bass resulted in an extension of around half an octave.

Speakers roll off naturally in the bass and SAM can't avoid that, in fact trying to extend the bass response of every speaker to, say, 20 Hz would result in damage to a lot of woofers. There's a limit to how far you can push a driver without causing damage and some speaker drivers are more "robust" than others.

So I think that you're right is stating that SAM progressively reduces frequencies at the low end but I think that reduction actually occurs at the low end of the profile for each speaker rather than from a set frequency such as 30 Hz given that Devialet's claims for the extension provided for some "smaller" speakers does not extend as low as 30 Hz and logic suggests that SAM has to be rolling off its action progressively at the low end of it's claimed extension capability for those speakers.

I would love a fuller explanation from Devialet about just what SAM is doing. I suspect at 0% it is simply correcting phase with a "roll off" of that correction at the upper end of its range so that there's a smooth transition back to the woofer's natural behaviour and a "roll off" at the lower end also in order to deliver a natural roll off at the bottom end. I would love to know what's going on in between as one increases the SAM setting. Is the change between 0% and 100% achieved with simply a graduated increase in the amount of phase correction applied, is the phase correction profile fully implemented at 0% with an increasing amount of bass boost applied as you increase the setting, do they use a combination of those 2 strategies, or is there some other "secret sauce" involved in what SAM does?

I hadn't thought of the possibility that SAM responds differently with pure sine waves than it does with music but having read your comment I do think that is plausible but I think that Devialet is probably being honest when they told you that this was because SAM was designed for music rather than for pure sine waves though I suspect that it also is a factor in how the driver protection function is implemented.

I've used SAM with 2 different speakers since getting my Devialet, originally with Dynaudio Contour 1.3 SEs and then with my current Focal Sopra 2s. I ended up settling on a setting of around 50% with the Contours but I've ended up running SAM at 0% with the Sopra 2s. Dynaudio claims bass extension to 35 Hz for the Contours and Focal claim extension to 33 Hz for the Sopra 2s, a relatively small difference in extension but my experience is that increasing the SAM setting above 0% delivers very different results in my room with those 2 speakers and I put that down to differences in speaker placement plus the fact that the woofers of the Sopra 2s, a floor stander, are much closer to the floor than the woofers of the stand mounted Contours which means they load the room differently to the Contours and get more room reinforcement. I think SAM is beneficial but I also think that there can be a lot of difference in the SAM profiles for different speakers, thaty different speakers can load the room at bass frequencies in very different ways, and that speaker placement also has a significant effect on bass response in the room. Taken together, those factors mean that one needs to spend a fair amount of time playing with speaker placement and the SAM setting if one is to get the best out of SAM because whatever SAM is doing, it is operating in a frequency range where the room is also a big player in the bass response we hear and interactions between what SAM is doing and what the room is doing are probably a factor in any situation where someone is reporting problems with SAM. Those interactions are also probably a major factor in why different people report using very different SAM settings with the same speakers much less with different speakers. They're probably also the reason why Devialet use a laser device to measure the woofer's response when creating profiles rather than microphone measurements because the physical behaviour of the woofer, which is what the laser device is measuring, is probably going to be far more consistent from room to room than microphone measurements will be.
Roon Nucleus+, Devilalet Expert 140 Pro CI, Focal Sopra 2, PS Audio P12, Keces P8 LPS, Uptone Audio EtherREGEN with optical fibre link to my router, Shunyata Alpha NR and Sigma NR power cables, Shunyata Sigma ethernet cables, Shunyata Alpha V2 speaker cables, Grand Prix Audio Monaco rack, RealTRAPS acoustic treatment.

Brisbane, Qld, Australia
Reply
#15
I received the information about the cut off from Devialet, along with a graphic that shows exactly how the cut is done. I will look for it, to see if I still have it. Anyway, the information matches what I tested with pure sinusoids. Obviously, the result also depends on SAM % (but not at very low frequencies - these are always cut).
Reply
#16
I'm not suggesting that very low frequencies are cut. I'm just suggesting that the frequency at which they are cut and the slope of the cut depends on the SAM profile for the particular speaker, not at the same frequency and slope for all speakers.
Roon Nucleus+, Devilalet Expert 140 Pro CI, Focal Sopra 2, PS Audio P12, Keces P8 LPS, Uptone Audio EtherREGEN with optical fibre link to my router, Shunyata Alpha NR and Sigma NR power cables, Shunyata Sigma ethernet cables, Shunyata Alpha V2 speaker cables, Grand Prix Audio Monaco rack, RealTRAPS acoustic treatment.

Brisbane, Qld, Australia
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)