Slashdot Mirror
Building a "Reference" Home Theater
An anonymous reader writes "FiringSquad has recently written a 14-page article on building a 'reference' home theater. They go through step-by-step and define all of the issues you need to think about when going with a new home theater setup. Exceptionally detailed but also easy to read."
I want to drag this out as long as possible. Bring me my protractor.
...I spent years building my own home theaters in each house I owned and lived in. Since I performed all the labor myself, they actually did add value upon resale.
One area that, in my experience, offers the most bang-for-the-buck is a two-part issue: room dimensions and sound-proofing, i.e. room treatment.
There are a LOT of expensive and probably useless room treatments. For me, the ultimate sound didn't come from watts or speaker power-handling but in properly sound-proofing the room against external noise. The lower the noise floor, the clearer the sound. This is key to having a good movie experience, I'd say, because you don't need it loud to be dynamic.
Room dimensions can be just as important, as certain rooms (square is the worst) have standing waves that emphasize or de-emphasize certain frequencies at certain locations. My ultimate theater had an odd shape (slightly angular walls and ceiling) but the sound was amazing. We also covered the walls in fairly cheap acoustic foam of varying lengths, and covered the foam with nice acoustically transparent cloth (red) so you couldn't see the varying foam squares. It was a slightly dead room, but it really had punch for the action films my friends liked.
Lastly, the proper bass crossover combined with the proper bass drivers is the final key for those who want action-style entertainment. I am a HUGE fan of Bag End from Barrington, Illinois. They make an ELF crossover and driver system which is just a miracle in a box. It is the flatest, most dynamic bass system imaginable, and the crossover was wonderful since it didn't overemphasize higher frequency bass to muddle male vocals or punchy sound effects.
One sidenote: I almost never focused on surround sound. Honestly, I was more happy with pseudo-surround out of low power, but dynamic speakers, than I was at have 16.5 channel surround sound. When I removed my rear channels, my visitors were always blown away by the clarity and depth of the properly positioned, amplified and mounted front 2 to 3 speakers I had installed. My current home theatre only has 3 speakers, and we're extremely happy with the install, which I did for a fraction of the overblown sound system my neighbor has. Even better, we're only driving the efficient and dynamic speakers with HEAVY 30 watt amps each, but since our noise floor is so low and our room is so quiet and dead, the sound is gorgeous, even for music.
I'm done with my theatre days, as the money is best spent elsewhere, and the upgrade bug is finally over.
You can build your own homemade projectors for a fraction of the price of a commercial one. Check out http://www.lumenlab.com/. Their forums are excellent.
Some of the advice for the video display was ok, but the rest of the article was very poor. One of the main determinations of audio performance is room acoustics - yet this is not covered at all. Monster cables are generic stuff that is horribly overpriced with outrageous markups. Never buy that stuff. Power filters are a total wast of money for 95% of people, and can often hurt more than they help. The speaker selection (Polk) is sort of a mass-market default - there is much better to be had out there in other brands, especially from the Canadian companies like PSB and Paradigm. Polk is by no means a brand that you would expect in a reference home theater.
It appears to me that this article was written with a lot of feedback from a big box store like Best Buy because the brands they recommend are typically what these stores carry, and in particular they push Monster stuff and power filters hard because of the huge markups.
No, no, no, no, no, no, no, no, no.
80-conducter IDE cables still only have 40 pins. The other 40 pins are insulation and grounding, used to space and isolate the active conductors so there isn't any crosstalk. There's not just a performance difference if you plug a modern disk drive into the controller using a 40-conductor cable, there's inevitable data corruption. This has NOTHING to do with expensive speaker cable, or the Monster myth. It's one of the stupidest justifications I've ever seen.
AND THEN they go on about PCM jitter with a straight face. Holy god, people still believe in this?
Even Jesus hates listening to Creed.
The way you do it is you take a CD with a couple of songs you typically listen to (also maybe a sample from a movie soundtrack). Important is that it should be something you're familiar with and preferably uses natural instruments (not synthesizers).
Then, compare speakers in your budget two at a time, make notes on which sounds closest to what the recording should sound like. When doing this make sure of two things:
1) the receiver base and treble settings are set flat
2) No other speakers (or subs are on).
Listen for how clear each instrument (and voice) comes out on each speaker.
When you get your two main speakers you need to get a matching center, you can switch between dolby pro-logic and stereo to compare the sound of your mains vs your central, or better, if they'll let you, hook up one the center channel in place of one of the main speakers and use the L/R fade on the reciever to compare the sound between the two (your music should be pretty even on both channels for this to work well). If you got bookshelf speakers you may want to really consider getting a 3rd identical speaker as your main (you probably will have to buy a 2nd pair since they won't sell them solo but in some cases it's worth it and you have a spare if something goes wrong in the future).
Don't assume the center channel provided by the manufacturer is the best matching center for your mains either. Get the best matching speaker period. When sounds pan across the front you don't want the quality of the sound to change too.
Rear speakers should be reasonably close matches to your fronts. When they're in your room they'll sound different anyway (since the quality of sound we hear is different when the source is in front of us vs the rear). You can also get dedicated surrounds (with drivers facing in different directions) vs traditional speakers for the rear. Always audition the rear speakers as you did your mains. Make sure they're good quality as well.
Last but not least, higher model number, more expensive *does not* mean better sounding. Louder.. maybe. But there are alot of considerations that go into balancing the sound from a speaker and some manufacturers get it right with their mid range models but lose it as they stray from that design.
Oh and if you're gonna listen to a Bose offering (typically not the best value for the price), really listen to the alternatives, you'll be surprised.
Georgia Tech, the leader in Chia(tm) technology.
Now is a great time to build an HTPC for watching high definition movies. LG just released their dual-format hddvd/blu-ray reader drive for under $300, and nvidia/ati have low-cost HDCP enabled video cards which can offload the decoding work from the CPU. This allows you to make a cheap dual format HDM player which is essentially future proof to future changes in the various standards.
The LinuxMCE project looks like it would be better than the wimpy/uptight WindowsMCE for running a home theater in a feature-stuffed home media network, including content, telephony, automation, alarms, remote monitoring, and all kinds of bundled features of disparate apps for "the Home of the Future". But it also looks like it's got plenty of holes in support and reliability. It could use a lot of attention from developers and users feeding back improvements.
FWIW, if the project porting X and codecs to the PS3 had more developers, the PS3 would be an excellent home media terminal running LinuxMCE without whatever Sony's planning to saddle it with.
--
make install -not war
There's some misinformation here, all right, but it is you who are the source. You're confusing things across domains.
The digital signal is coming in at a rate much, much higher than the reproduced signal. For instance, if we were dealing with 8-bit audio at 20 KHz, then for every analog sample there are 8 digital bits of a bitstream that have to be recovered. In an uncompressible signal, this requires 8 bits to come in at a rate of 20 KHz. Time must be allowed for this, even in the case where the signal is (at the moment) compressible. So the bitrate has to be eight times higher than 20 KHz. But - Nyquest explanation goes here - the signal can't actually be 20 KHz, it has to be 44 KHz so we can avoid aliasing by filtering. Yet the frequency response of the system is 22 KHz (less, actually, but theoretically, it could be 22 KHz.) Now, if one of those bits, arriving at a bitrate of 8 * 44 KHz, or 352 KHz, jitters (because of the clock) and "arrives late" to quote you, how "late" can it be without shoving the next bit out of the sample recovery slot? The answer is less than 1/2 bit time at 352 KHz - which is 704 KHZ; Now, ask yourself, using that same knowledge you gleaned in basic engineering class: How much of a 704 KHz phase jitter do you think will get through a steep filter at 22 KHz, if indeed the byte being recovered jitters at that rate? The answer of course, is none of it. The rate of change that a 22 KHz filter allows is so slow that signals at hundreds of KHz don't have any effect at all on the signal integration; they're simply too fast and too transient to slew the summing process.
But wait - we're not talking about an 8 bit signal, are we? No. We're talking about a signal that is many bits deeper, and coming a whole lot faster - but still sees a brick wall filter way, way, down low, often still at 20 KHz. So the jitter is even smaller, and the filter is even stronger with respect to it.
Digital domain signals certainly have potential problems, but THD due to jitter isn't one of them in this particular application (hifi and home theater.)
I've fallen off your lawn, and I can't get up.
Actually what you said is "you're not going to get harmonic distortion" which is the opposite of reality. Here's the peer-reviewed watershed work on jitter and harmonic distortion, published in the Journal of the AES. Educate yourself. http://www.essex.ac.uk/ESE/research/audio_lab/malcolmspubdocs/C44%20DAC%20Transisiton%20distortion,%20jitter,%20slew%20rate.pdf
Aside from the fact that the barrage of ads on that site nearly melted my laptop, a few observations:
The article starts off with a disclaimer in the first page. When an opinion piece has to state, "this doesn't represent an endorsement from FiringSquad, but rather our research," the opposite is generally true.
He decides the lack of 1080p content is a myth. His argument? You can convert 1080i back to 1080p. While this is fundamentally true, so is the statement that there's little 1080p content available. Converting it doesn't mean that the content was natively available in 1080p. I know, seems trivial... but this is a precursor to an argumentation style he uses that gets tiresome and downright disingenuous as the article goes on.
He has a table of data for maximum viewing distances to appreciate the full benefit of 1080p quality. He doesn't bother to cite any sources or what methodology was used to arrive at the data... but nevermind, it does get worse.
He's right about LCoS being superior, but for the wrong reason. It isn't the control circuit but rather the fact that there's a liquid crystal display element for each pixel and, in principle, support for resolutions well above 1920 x 1080. Also the fact that baseline LCD displays and DLP do not display a red, green and blue pixel simultaneously for each pixel of resolution on the final image.
Then he uses THX and SMPTE theatrical standards and applies them to a home auditorium, to support his argument that the opposite is true? What ever gave him the idea that these standards were ever applicable to small home auditoriums in the first place? Theatrical exhibition is a different deal entirely, whether digital or optical... but I'm guessing that the THX/SMPTE specs he's quoting were for 35mm which has much higher effective resolution than 2k/4k digital theatrical projection systems.
He confuses the term motion blur with the issue of print clarity. Motion blur is a side effect of optics whereby an object in motion is blurred by way of the aperture and shutter timing of the exposure. This is actually a good thing because in a motion picture format, i.e. a series of still images, it assists the brain in perceiving fluid motion from a series of still images. Motion blur is NOT correlated with effective clarity but exposure length. Therefore it's ridiculous to say that 35mm is equivalent to 720p. In fact, 35mm, depending on the film stock used and the style of cinematography (e.g. sharp, grainy, diffuse glow, etc.) used, motion picture can render images whose digital equivalent would extend up to 6000 pixels of horizontal resolution... three times that of HD 1080p.
While it's true that theatrical Dolby Digital is 320 Kbps and DVD Dolby Digital is typically 448 Kbps, he makes no mention of additional parameters in Dolby Digital home encoding (e.g. dialogue normalization, Dynamic Range Compression, etc.), he doesn't discuss theatrical DTS (an ADPCM-style format with a 1.5Mbps bitstream), nor does he observe that Dolby Digital at 448-640 Kbps is acoustically transparent relative to an uncompressed source. He also confuses the term "lossless" with "uncompressed"... Lossless refers to compression formats, but "high resolution audio" like that on a DVD-Audio recording is typically 20 or 24-bit Linear PCM, which is an uncompressed format. Calling it lossless is superfluous.
He doesn't mention that in addition to the majority of sound being in the front channels, 5.1 is actually 5.1, whereas 7.1 is not. At best, 7.1 is actually 6.1 with the two rear surround channels paired in mono as in DTS-ES. At least, 7.1 is actually Dolby Digital 5.1 discrete with the two paired rear surround channels still mono, but also stereo matrixed into the left and right surround channels the same way that Dolby Surround analog carries the rear surround stereo matrixed into the front left and right channels.
I mostly ignored his commments on power filtering and cables because that subject has been beaten to death already. H
S/PDIF encodes the audio using PCM frames, which are electrically encoded using Biphase Mark Code
A S/PDIF representation of a "zero-crossing" would be a series of PCM frames where the binary value in a channel goes from >0 to 100 or so "electrical" zero-crossings. This is by design. Even a completely silent audio signal (all numerical 0) is encoded by a repetitive, square-wavish mark/space sequence on the channel.
There is ZERO harmonic correlation between the encoded S/PDIF signal and the logical analog signal. This is by design... to ensure a noise-resistant channel by which clocking and data frames can be transmitted with resiliency.
Moreover, most output DACs don't actually use the incoming clock signal for playback. They use an internal oscillator. The input clock is merely used to delineate frames and get data into intermediate buffers.
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
The party of stupid and the party of evil get together and do something both stupid and evil, then call it bipartisan.