Tuesday, August 25, 2015

Two Too Noisy

Unfortunately my newest DVR, the Magnavox MDR-557 doesn't have a quieter chassis than the older MDR-537 I just started using recently.  It seems slightly noisier, and more importantly, slightly more irritatingly noisier.  While the MDR-537 does produce a slightly tonal operating noise, a midrange whine, the 557 produces a tonal noise that warbles slightly, making it significantly more irritating despite being barely louder.  This was with just the HDD operating, as it would often be overnight while I'm sleeping mostly.  There is also a random clicking as the HDD stores the previous 6 hours of video.  I'd like to be able to turn that on and off, but it does not appear to be a user-controllable feature.

It's very hard to measure most chassis noise.  It's clearly audible in my bedroom at night, but isn't above the self-noise of the SPL meter built into my Android phone, as I quickly determined.  Even with purpose made SPL meters, in my previous experience, chassis noise is still hard to measure.  The noise itself may be only 15dBA or so but still stand out irritatingly in a quiet room because of the tonality.

WRT measuring, or at least more objective comparison, the best thing to do is make low noise recordings.  I have a studio quality 1" condenser microphone which has 5dBA self noise.  It would make a fine recording, which could be measured afterwards, or analyzed with a post processing RTA such as a program I once wrote called GFFT.  I had been thinking about measuring chassis noise for many years when I bought my R0DE NT2000, so I knew to look for a microphone with low self noise, and that is good for general recording purposes also.  About the only lower noise microphones are from Neumann.  Cheap SPL meters have microphones 1/4" or smaller and these have many times more self noise.  Small microphones are good for high frequency response but bad for low self noise.

You may think it's curious to be concerned about chassis noise.  But it's a far more objectively important thing to be concerned about than whether source material is 16 bit or 24 bit.  The noise level in a 16 bit recording is 96dB.  At a recurring peak level of about 85dB which is a typical "loud" listening level (louder than most non-audiophiles listen at) that puts the noise floor -11dBA in the environment.  Meanwhile, if you have a disc player producing noise at 20dBA, it is 31dB, or 35 times louder.  And 20dBA would be a relatively average one, I think…the Magnavox could be higher (of course this depends a lot on how measured), my guesstimate is about 25dBA (537) with just HDD operating, as it always done when the unit is on.

Another thing I think is almost always overlooked is the fact that most equipment has internal resonances that react to ambient noise.  I'm less concerned about the effect this has on the signal path through transistorized electronics (which is minimal) than on direct radiation of the resonances into the acoustic environment.  Even with tube electronics, except for phono preamplifiers, the effect on the signal path is probably the smaller part of the story.  Direct resonances are one thing good about having "overbuilt" equipment in heavier gauge metal boxes than is absolutely necessary.  Of course the Magnavox players were built according to a different set of criteria, namely to be as low priced as possible.  This means they use thin gauge metal which is not helpful to keeping the noise level down, in fact it may increase the noise level as compared with having no metal casing at all.

My comparison so far is further corrupted by the fact that I have the 537 sitting on top of the 557.  Even when only operating one unit at a time, the other unit serves as a resonating chamber for the operating one.  Further, the position is slightly different in the cabinet, and might matter which unit sits directly on top of the shelf.  Obviously a fair measurement would only measure one unit at a time, sitting directly on the shelf and with nothing on top of it.

While some people might worry about the heat of having one unit on top of the other, that hasn't been a problem for me.  I've never noticed the case of either unit to get above ambient temperature, they have no vent holes on top, and are constantly fan cooled.  Also my house is continuously climate controlled to have ambient temperature no higher than 79F.

I plan to do the fair comparison, each unit on the shelf by itself.  Then I will compare the better unit operating by itself with having the other unit above or below it.  If it does really help to have only one unit sitting on the shelf, I will put the noisier unit somewhere else.  I will operate it less either way.  I will only operate both units at the same time if I need to, but mostly just the quieter one.  And I will use acoustical foam wedges and other measures external to the chassis to reduce the noise as much as I can easily do that way.  I have already carved a large piece of 6" thick Sonex foam that fills up the shelf space above the top unit, and blocks some or most of the fan noise coming from the back.

That's just the short run.  In the long run, I intend to make the 537 first, and later the 557, into the quietest unit I can.  I think the hard drive is the largest source of noise in the chassis, and the hard drive can be replaced, I hope, with a quieter one or even an SSD.  The fan can be externalized into a larger fan that runs slower.  I have done that sort of business with computer fans before, and it is not a walk in the park.  It took years to perfect the semi-external fan system I ultimately used with my Amiga 2000.  It featured a circuit to break the fan-sensor line in the computer when/if for any reason the external fan wasn't running.  It never failed in the two or so years I used it, in fact the nature of such "ultimate" tweaks is that shortly after you have finally reached the summit of a new invention, it becomes absolutely necessary to switch to a different unit, such as my subsequent Amiga 3000, for which all the old interventions are useless.

Then for all it's unsolved (though heavily tweaked, even fan replaced) noisiness, my Amiga 3000 is now unused, though it could be fired up again if I ever excavate the junk in the computer room back where I can set up the video distribution amplifier and hook it to the new home networking panels, bringing Amiga Vision back to life.

But for the last two years, I've been busy with other matters, now an all new bedroom video server, which suddenly became necessary because the last one, a Sony DVP-995C 400 disc carousel from 2006, quit working last month and nobody makes those for reasonable prices anymore.  Besides, a system based on hard drive like the Magnavox is far neater to operate.  You can switch between titles quickly and it even shows previews beforehand, all without doing any metadata editing.  But a constant whine is a big drawback, and the Sony was quieter, especially when there was no disc playing.

No comments:

Post a Comment