PC Sound and Speakers: A Subjective LookWritten on January 15, 2009 by
To address these issues, two people listened to all of the testing. Our resident sound tech, Dan Hermanson, jumped at the chance to be involved. He works as a sound technician at the state university where he is also a student, running sound for professional and amateur concerts and other events. If there are differences in sound quality, he will hear them. Though he's very familiar with the differences that speaker quality will make, he was curious to look at PC sound cards, as he'd never had the opportunity to work with them before.
I (Melissa) was the non-audiophile. I rarely even listen to music, and happily ignore it when someone else turns it on. I appreciate good sound, but I also appreciate money in my bank account - I was curious to see what was worth it, and what was unnecessary for someone with very low demands.
With these backgrounds in mind, we set out to find what matters, and how much.
We tested each sound card with the highest quality set of speakers, and each set of speakers with the highest quality sound card. While more combinations were possible, we chose to limit the test to keep it simple. As testing went on, we did try a few additional combinations, which will be discussed in detail later.
For each setup, we listened to a battery of tests. We started with pure tones, from 30 hertz (the low end of human hearing - this tone is felt as much as heard) to an ear-splitting 15,000 hertz (a range of hearing that people often lose as they age). After that, we moved into more subjective analysis - MP3 playback in various bitrates as well as lossless audio (Windows Media Audio Lossless). While it wasn't directly related to the rest of our testing, we wanted to see if we could tell a difference between MP3s and lossless audio. We used three songs from very different styles to give an idea of all-around performance.
All tests were performed in the quietest area with the best acoustics that we could find. Dan would have liked to use a professional sound stage, but since they're a little hard to come by, we had to make do with a section of the offices. The background noise was similar to what you'd find in a suburban house.
We tested the sound cards first, and went from the bottom up. They were all tested using the best speakers we had (the Logitech Z-5500) to minimize the impact of any flaws in the speakers. The onboard audio did very well with pure tone replication. The only problem was an overtone in the 30hz test, we could hear a higher note along with the sound that was supposed to be there. Everything else seemed clear. The MP3 listening was unremarkable - there wasn't anything particularly horrendous about it, although Dan said it wasn't particularly good, either.
During pure tone replication, the overtone at 30hz disappeared, although an undertone appeared in the 15,000hz test. The MP3 testing was the first major surprise we encountered - the Xonar DX really didn't sound better than the onboard solution. I could tell that there was a difference between the two, but couldn't describe what it was. Dan was able to put words to the difference - the Xonar DX offered more depth in the vocals than the onboard sound, but lost detail.
|The Xonar DX also had some equalization problems. The mids and high mids (vocals and guitars) were too loud, and the bass was too quiet. I didn't pick up on the problem right away, but once Dan pointed it out, it was very obvious. The equalization can easily be adjusted in software, so it's not an incurable problem, but adjusting the EQ of a brand new sound card shouldn't be necessary.|
The D2X suffered from the same undertone at 15,000hz as the DX. At 1000hz, which is within the range of the human voice, it produced two tones instead of one, like it was playing a chord. This was very concerning, but as far as Dan could tell the problem never appeared outside of pure tone replication.
|In music playback, the D2X produced significantly better sound than the DX or onboard sound. The equalization problems in the DX had been cleared up, and there was far better definition than the other solutions had shown. It was so clear that at one point I was able to pick out flaws in the original recording. To be fair, I'm not sure if I would have noticed the difference if the cards weren't being tested back to back. But after listening to the D2X, the DX and onboard chips sounded cheap and 'fake'.|
|Creative Elite Pro|
The Elite Pro is $100 more expensive than the D2X, and includes a remote and an I/O console that offers connections for most types of audio gear. After working with professional sound boards, Dan was inclined to turn his nose up at the Elite Pro's console, but acknowledged that it would be useful for a home enthusiast. The remote is also useful for some people, though I'd probably just lose it.
|During the tone replication, the Elite Pro had a number of difficulties, including a return of the overtone at low frequencies that we'd heard with the onboard sound, and undertones in the high frequencies. At 1000hz, the right speaker was approximately twice as loud as the left - we weren't able to fix it, but it also didn't show up anywhere else in our testing. |
In music playback, the Elite Pro offered similar sound quality to the D2X. It was noticeably better than the lower quality solutions, and we had to switch back and forth between it and the D2X before Dan could pick out differences. His opinion was that the Elite Pro seemed to put a faint sustain over everything, which smoothed out the voices nicely, but blurred the background instruments together. It wasn't as crisp as the D2X, but it certainly wasn't bad. It appears that the extra hundred dollars in the price tag is for the accessories, not for better sound quality.
As expected, our old, cheap speakers were exactly that. They struggled to produce 50 hertz, a note well within the range of human hearing. MP3 playback was similarly horrible. Even while using the ASUS D2X, the lyrics were hard to make out, electric guitars dissolved into fuzz, and all the real nuances of the music were overwhelmed by a flood of noise. It was, as expected, gross. Just for fun, we switched back to onboard sound from the D2X, and couldn't even tell that anything had changed.
|The point here isn't to rag on Creative (although that can be fun). The Creative SPS15 set consists of two soup can sized speakers, and no subwoofer. They might have cost $20 when new, and are now several years old. We knew they'd be bad when we chose them. The point is that many people use speakers like this for day to day use, and just think that's how computers sound. Speakers are easily overlooked in the upgrade shuffle, as most people would rather put money into a monitor or video card than a new speaker set. This makes good sense, but cheap speakers really are just that. You don't have to be an audiophile to tell the difference.|
The Logitech X-230s, our midrange speaker option, has two desktop speakers and a small subwoofer. Thanks to the subwoofer, 50 hertz was clear and crisp, and the system even made a passable attempt at 30 hertz. The MP3 listening tests sounded remarkably good, and we decided that these speakers are very easily worth their $40 price tag. Out of curiosity, we tried these speakers with both onboard sound and the Xonar DX , and found that there was a noticeable difference - not a huge one, but definitely there. These speakers are good enough that the input quality makes a difference. For their low price tag, we were very impressed.
Even though we'd used these speakers for the sound card testing earlier, it was interesting to come back to them after listening to less expensive speaker sets. The Z-5500 offers five speakers for surround sound, and a large subwoofer. The audio files we used for our listening tests were not in surround sound though, so just the front speakers were used. This kept the comparisons even between speaker sets.
The speakers are wired together using very cheap boom-box style clips with unshielded wires, which Dan was disappointed to see and described as an 'interference nightmare waiting to happen'. I asked if he'd tell people to avoid these speakers, but he said no. If he were going to spend several hundred dollars on a speaker set, he'd insist on shielded cabling and quarter inch jacks, "but that's because I'm used to professional stuff". It's disappointing to see cheap wiring on such a high-end speaker set, but it's not enough of a concern to warn people away from this set, either.
|During pure tone replication, the subwoofer did a beautiful job of producing the 30 hertz tone. The main speakers tried to produce the tone as well, though, and failed miserably. Ideally, low tones should be played through the subwoofer only, and the main speakers should drop out once the tones go below what they can effectively produce. These speakers didn't drop out, which Dan said was a problem with the crossover. It wasn't terrible to listen to, but it's a sign that the speakers weren't put together as well as they might have been. |
In music playback, the Z-5500 speakers produced excellent sound, and were noticeably better than the X-230 speakers. Dan described the difference as 'more precision in the mids'; I described it as 'nice speakers'. The large subwoofer added a beautiful amount of kick to the bass, and the midrange speakers performed admirably. The Z-5500 speakers performed very well, as we'd expected.
At the beginning of this review, we'd asked if sound cards are worth the extra money, as compared to onboard solutions. The answer was sometimes. The high end sound cards easily outperformed the onboard sound, even to my untrained ears. The less expensive card was different than the onboard sound, but not necessarily better. It probably comes down to what sound codec is built into the motherboard. We did our testing with an eVGA 680i board, which had a high quality Realtek codec. Most of Puget's current mid- and high- end motherboards use similar codecs, so I'd expect similar quality out of them. There is common wisdom among computer builders that onboard sound is terrible and that sound cards are a necessity, but it doesn't appear to hold true any more. If you have a good quality motherboard, a low or mid range sound card probably won't give you better sound. If you have a cheap motherboard with lousy onboard sound, you'd be better off putting the $100 into getting a better motherboard, which will give you both decent sound and improved reliability.
To my surprise, the high end sound cards were very much worth their price tag. I hadn't expected to be able to tell a difference, but it was there. It was like pulling cotton balls out of your ears - I could hear the music with the less expensive sound card, but the high quality ones made it much more vibrant and alive. Dan and I were both quite impressed. If you have decent quality speakers and an extra two hundred dollars, a good sound card will add a lot to your listening experience.
Dan was surprised by the number of problems we had with pure tone replication. Every setup we tried had some sort of over- or undertone, and a few had worse problems than that. We're not sure if it was the result of our test environment, bad parts, or if computer sound equipment is possibly just like that. Fortunately it didn't seem to impact MP3 playback, but it is a little worrying.
For MP3 vs. lossless testing, Dan ripped the same song to MP3 format at 128kb/s, 196kb/s, 224kb/s and 355kb/s, and in Windows Media Audio Lossless format. Using the highest quality sound equipment, there was a difference between the WMAL format and the 128kb/s MP3... but just barely. Neither of us could tell a difference between the higher quality MP3s and the WMAL format. Our conclusion was that if you're using very high quality audio gear or
consider yourself an audiophile, using a lossless format is worth the hassle. If you're not a
purist, just use a midrange or better MP3 quality, and don't worry about it.
The toughest question to answer was where to put your money: speakers or sound cards.
We had both expected speakers to be more important, hands down. But the Xonar D2X sound card really impressed us, as did the midsized speakers. Given a choice between the D2X and the largest speaker set, Dan said he'd pick the D2X. He emphasized that it's personal preference, though, and he wouldn't fault anyone for choosing the speakers. My conclusion was that it depends on the application - people who really sit down and listen to music would be best served by the sound card, while gamers should get the biggest subwoofer they can afford. For a casual listener like me, either option is probably overkill. Good onboard sound and a $40 speaker set are plenty for light use.
With computer audio gear, you get what you pay for. However, the law of diminishing returns applies very strongly. Higher quality gear does sound better, even to a non-audiophile, but the difference is relatively small compared to, say, differences between video cards, or a faster CPU. If you have the money and audio matters a lot to you, go ahead and get a really nice sound card and speakers. Otherwise, there's no need to sink a lot of money into it. (Just avoid the super-cheap stuff.)
In the end it'll always be a matter of personal preference. Listen to as many different speakers and sound cards as you can, visit your local hardware shop, ask your friends about what they have. If you're considering buying a Puget computer, talk to our sales reps. They'll be happy to help you customize a computer and sound equipment to match your needs, whether you need an amateur recording studio or just a couple of speakers on a desk. We have experience with both ends of the spectrum, and enjoy tailoring computers to match the needs of the people who use them.