Why Does The Need for (Proper) Speaker Enclosures Continue to Elude?

[daniel_leavitt2000]

Aren't ESU sound files different for scales? Isn't this why I have to download the "N" scale version rather than "HO" or large scale? I'm pretty sure they are already tweaking frequencies.

As for enclosures.... There are plenty small speakers with an adhesive lip that would create an air tight seal with an enclosure. Since these speakers need a plastic bracket to isolate it from the chassis, there really isn't any excuse not to make a proper enclosure. Scale  Trains is a particularly egregious failure here.

One thing I have not yet seen, which would really ease assembly is the use of the fuel tank casting AS the  sound bracket and enclosure. The speaker would be sealed to the fuel tank which would include a cavity. This unit would then be press fit into the chassis with the speaker contacts pressing against contact plates in the chassis. Atlas and IM are 99% there. But they both use separate parts for the tank and baffle.

[peteski]

When my brother was studying for his electronics engineering degree, he learned of this little trick and passed the knowledge on to me. So, you'll have to learn about this on your own. I wish you luck--I suspect it's not the sort of thing you'll find in Wikipedia...

While I don't have EE college degree, I have been schooled (and worked) in the electronic field. I wish you has some more specifics than just a word-of-mouth.  I say that because while filters to select certain range of frequencies can be easily made using simple components, frequency doubling (while retaining the amplitude of the original waveform) is not that trivial. Nothing I would expect to see in a cheap transistor radio (where low cost and miniaturization are the main priorities).

[DKS]

[peteski]

LOL!
Not really disapproved, but a bit skeptical.
Speaking of TRW banners, I went to your banners page few weeks ago, and there are quite a few new ones I would enjoy seeing in the rotation.

[nickelplate759]

While I don't have EE college degree, I have been schooled (and worked) in the electronic field. I wish you has some more specifics than just a word-of-mouth.  I say that because while filters to select certain range of frequencies can be easily made using simple components, frequency doubling (while retaining the amplitude of the original waveform) is not that trivial. Nothing I would expect to see in a cheap transistor radio (where low cost and miniaturization are the main priorities).

It's not an electronics trick - it's a trick of psycho-acoustics.  When presented with sound at two distinct frequencies, your brain will perceive those two frequencies, the sum of the frequencies, and the difference.    So (oversimplyfing a bit, and assuming NOT equal temperament) - if you have A @440Hz, and E @ 660Hz, you will also perceive a lower A @220 Hz even if your speaker can't reproduce 220 Hz (and a higher C+ at 1100 Hz, but that doesn't matter here).   Composers used this trick years before electronics were invented.

[tehachapifan]

Aren't ESU sound files different for scales? Isn't this why I have to download the "N" scale version rather than "HO" or large scale? I'm pretty sure they are already tweaking frequencies....

I sure don't recall anything about scale when downloading ESU sound files. Have I missed something?

[DKS]

Speaking of TRW banners, I went to your banners page few weeks ago, and there are quite a few new ones I would enjoy seeing in the rotation.

PM @tom mann

[Steveruger45]

Aren't ESU sound files different for scales? Isn't this why I have to download the "N" scale version rather than "HO" or large scale? I'm pretty sure they are already tweaking frequencies.

As for enclosures.... There are plenty small speakers with an adhesive lip that would create an air tight seal with an enclosure. Since these speakers need a plastic bracket to isolate it from the chassis, there really isn't any excuse not to make a proper enclosure. Scale  Trains is a particularly egregious failure here.

One thing I have not yet seen, which would really ease assembly is the use of the fuel tank casting AS the  sound bracket and enclosure. The speaker would be sealed to the fuel tank which would include a cavity. This unit would then be press fit into the chassis with the speaker contacts pressing against contact plates in the chassis. Atlas and IM are 99% there. But they both use separate parts for the tank and baffle.

Here’s a quote from ESU site talking about the scale symbols on the sound files.

As a General Reference:
ANY LokSound Select Sound file can be written to ANY LokSound Select decoder. This means you can write a Micro(N Scale) 738xx file to an 8 pin or 21MTC Select, and alternately a 734xx(HO Scale) file to a Micro or Select Direct decoder. The ESU LokProgrammer software will recognize the difference and ask you to proceed. By Accepting you are giving the permission needed for the LokProgrammer to automatically make the conversion needed. However, these files cannot be altered for use with other ESU decoders. They are intended to be used with LokSound Select decoders only.


From above and the fact that different scales have different sound project numbers for the same loco identified It would appear that there is some difference relevant to scale but LP will make a conversion.

Not sure what is different in the projects but maybe to do with matching any hardware differences between the decoders for each scale.

[jagged ben]

With regards to EQing the sound so that more low frequencies can be heard, be mindful that just turning up the low frequencies is not a solution when the speaker just can’t produce the low frequencies. Unless there is a high pass cross over in use, the speaker is being fed the full range of sound at the same volume or what ever you’d like to call it. We’re not hearing it because as it’s been pointed out, the frequency response and there for the volume you can hear rolls off considerably after about 1000hz. If you increase the volume of those lower frequencies you risk damaging the speaker.

Some equalization to better match an enclosure is different but you can’t just turn up the low frequencies.

There is no replacement... for displacement!

 Craig.

It's not about turning up the low frequencies.  It's about turning down the high frequencies that are disproportionately loud because of both speaker response and scale phenomena. When a 4400HP locomotive passes by 1000ft away you don't hear the high frequencies from the prime mover.  But the people recording these sounds for decoders do the recording trackside (as they should, for other important reasons).   

The appropriate EQing solution also does vary by scale, not just because of speaker size but also because the typical real-world distance that the modeler stands from the model represents a greater scale distance at smaller scales.  Which means that hearing a tinny primer mover sound from a N scale loco 10ft away isn't unrealistic because speaker response is poor for low frequencies. It's because most of the frequencies you're hearing on a non-EQed sound effect aren't frequencies that would carry 1600ft at all in the real world.  But that's significantly less true for standing 480 scale ft from an O scale loco.

Granted until I actually have time to try it I am only speculating about how much improvement could be achieved.  But to repeat the original point, more effort needs to be put into the sound design and not just the equipment.  There is always a best sounding design (allowing for differences of opinion) for the given equipment.

Btw, this post applies primarily to the prime mover sound, not so much to others, which mostly just need to be turned down entirely, except for the horn.

[peteski]

Here’s a quote from ESU site talking about the scale symbols on the sound files.

As a General Reference:
ANY LokSound Select Sound file can be written to ANY LokSound Select decoder. This means you can write a Micro(N Scale) 738xx file to an 8 pin or 21MTC Select, and alternately a 734xx(HO Scale) file to a Micro or Select Direct decoder. The ESU LokProgrammer software will recognize the difference and ask you to proceed. By Accepting you are giving the permission needed for the LokProgrammer to automatically make the conversion needed. However, these files cannot be altered for use with other ESU decoders. They are intended to be used with LokSound Select decoders only.


From above and the fact that different scales have different sound project numbers for the same loco identified It would appear that there is some difference relevant to scale but LP will make a conversion.

Not sure what is different in the projects but maybe to do with matching any hardware differences between the decoders for each scale.

That sounds like the LokProgrammer is doing some conversion to the project, but I doubt it is modifying the actual sound of the recordings (like changing their equalization). It is probably just converting the sound from one storage format to another.   But it would be nice if we could get more details about the conversion.

[jagged ben]

That sounds like the LokProgrammer is doing some conversion to the project, but I doubt it is modifying the actual sound of the recordings (like changing their equalization). It is probably just converting the sound from one storage format to another.   But it would be nice if we could get more details about the conversion.

Had anyone asked them directly?

[CRL]

Why don’t you just have the low brass section play everything up an octave? 

[craigolio1]

It's not about turning up the low frequencies.  It's about turning down the high frequencies that are disproportionately loud...

If you turned down the high frequencies so they were proportionately quiet when compared to the low frequencies, then the over all volume would be very quiet. This would work great unless you wanted the over all volume louder because if you increase it then you’ve now ultimately made the same adjustment to the low frequencies that would have potentially damaged to speaker in the first place.

[jagged ben]

If you turned down the high frequencies so they were proportionately quiet when compared to the low frequencies, then the over all volume would be very quiet. This would work great unless you wanted the over all volume louder because if you increase it then you’ve now ultimately made the same adjustment to the low frequencies that would have potentially damaged to speaker in the first place.

Yes.  (Duh.) And in my opinion, other than the horn, all N scale sounds should be quieter from the factory.  So that would be one good thing, in my opinion.  But regardless of opinions about appropriate volume, sound quality doesn't necessarily have to be pitted against desired volume.  For example, I turn down the prime mover sound on all the sound locos I get.  But if the prime mover sound were better edited for quality as I suggested above, then maybe I wouldn't.

[jagged ben]

Why don’t you just have the low brass section play everything up an octave? 

With the opportunity to EQ the file, there's probably plenty of frequencies in that range to work with already.

The thing to do would be to find an amplifier that can drive an N scale speaker which can also be driven directly by a computer audio output.  Then edit the files to your liking in real time without transferring back and forth to the decoder.  With any decent PC based waveform editing software, one could do all the transformations one wanted to try.  It would take hours to do the setup and minutes to tweak the audio to perfection.  (Well, depending how perfectionist you are.)