Axiom Home Page Forums General Discussion The Water Cooler Today's Oddest Comment

In reply to:

---

no, it's not a presentation problem... they have a pretty good 27" tube TV and friends/family with box store RPs and Plasmas.

---

I really find that astounding.
Seeing the difference b/w a vhs tape and dvd video was rather startling even on our old 23" tv. The very day we rented that first dvd we never went back. I haven't used either of our vcrs in nearly 2 years now.

On another note, it appears we will be staying in the 'Peg and not moving to Calgary after all. More info on that later.

I think this subject came up in another thread a couple of months back. I seem to recall Bren saying something about what he had discerned with mp3s.
I have yet to try testing mp3s on my Axiom (or other) system. I have only used some small monitors in my office and with the electrical noise that emanates from my tower, along with the occasional noise outside my window, even cds can sound a bit 'off'.

I've tried with ABX programs to ensure good results. On heavy metal, 128 is somewhat noticeable(darker highs). I used that since there's lots of high frequency energy in heavy metal drumming. At 160k, barely noticeable. At 192k.. I'm not even sure. I doubt there was.

I was using LAME 3.90.3 with --alt-presets to test. I forget which headphones, but I'd like to repeat the test now that I have ones which are much better. Once I get this otitis media cleared up, I'd like to do another test. The --alt-preset standard VBR mode was completely transparent on everything, so I don't worry about the audio quality of well encoded MP3s. Good enough for me, although I like to have an original copy.

I know that with Ogg Vorbis, there was little to no distinction at -q4(128k VBR). With most songs there is no difference. The problem there was that the highs were a bit raised overall, barely noticeable though. -q5(160k VBR) was completely transparent. This was with an older encoder, however, so by now it's probably much more transparent at lower bitrates. I use -q6(192k VBR) with a newer encoder to ensure that there aren't any audible artifacts with my portable player.

Ogg Vorbis performs much better on test clips, too. For naysayers who say 128k is CD quality on everything, I usually point them to this clip. If with decent headphones or good speakers you can't hear any artifacts at 128k with that clip in comparison to the original, you're probably deaf. It's a worst case scenario clip, yet, as with the rest of the similar clips, MP3 usually performs horribly in comparison to newer codecs.

careful now dont be bashing high end pc speakers I have a set of Logitech Z560's and the sub on that pounds harder than the one in my HT
Then again you cant really compare pc speakers to HT hehe

Recently I've done extinsive testing with MP3s.
iTunes MP3s at 192kbps and even higher sound like crap compared to a cd.
LAME does a much nicer job, although I can still hear a small difference at 192kbps.

Yeah, but AACs sound pretty good. If you're going to encode MP3s, iTunes wouldn't be my first choice, but I'm happy with AACs. Of course, I'm not really all that discerning...

Yea, I agree. The iTunes AAC do sound better than the their MP3s, but in my opinion still not as good as the 192kbps LAME MP3s.
I've also found that it heavily depends on what type of music your converting.
Still experimenting though.

Jason, you might find this encyclopedia article which covers MP3 history,encoders,listening tests,links etc. in a fairly unbiased way to be interesting.

JohnK
Thank you very much for that article. Gives me more to "chew" on.

Whoa, now that's timing... my GF just bought an MP3 player (Creative MuVo?) this weekend... the ripper that comes with it rips MP3s (or WMAs) at about 8x-10x speed... got bro and I into a conversation about audio compression (for those who don't follow my posts - bro and I have differing opinions on audio, I'm a technician, he's a musician) and we did some moderately scientific testing recently on MP3 vs Redbook.

Same selections played back on three different sets of transducers - my Axioms (2xM3s, VP150, 3xQS4s w/STF-2), my Edirol monitors (MA-20Ds) and his Sennheiser cans (sorry, I don't know the model) from 2 different sources - MP3s played off a CD-R in my DVD player through the TOSlink (optical) connection, Redbooks played through my H/K carousel (FL8550) over the S/PDIF (coaxial) connection. MP3s were encoded at 128Kbps with Fraunhofers CODEC & through whatever Creative's software uses and at 256Kbps through Fraunhofers CODEC.

Method: listener sat in the sweet spot eyes closed, luckily both sources were pretty close in gain so we didn't need to pad either source to match levels. Tester would randomly mute the receiver (for Axioms), turn down the Edirols (run through the stereo outs on the receiver) or the headphone level, cue up one of the tracks and play about 5 seconds of a random selection, mute, and restart.

-The Fraunhofer codec at 256K did very well against Redbook (subjectively) but was still picked out of the lineup more often than mathematically "null".
-Fraunhofer at 128K was easy to determine but subjectively didn't sound that bad until you heard it "against" redbook or the higher bitrate.
-The Creative encoder might as well have been AM radio.

Also, the Axioms seemed to be the easiest to make the determination on (though I did have to deal with bro's constant b****ing that they were too bright - sorry bro, next time bring over your 1-way paper cone 80s era towers!) with (to my surprise) the Edirols being second and the Sennheisers third... I would have thought having the music pumped directly against my eardrums would have been better than this pair of $250 powered nearfields.

Also, I had a much easier time of it than bro did (he's got hearing damage from his years on stage) and once you put the high and low bitrate/quality tracks onto the MP3 player and played them through a pair of consumer headphones (whatever Sonys GF has - not the included Creative earbuds - I can't wear them!) most of the extra quality was lost. By the time you take your heart beating in your ears into consideration while jogging (her) or cycling (me) it's probably a wash.

Again, not exactly as scientific as it could be, but more than anything, it was a way of getting a "blinder" listening test between the compression methods. Of course, since he's a Mac guy, he wants to add in AAC and a bunch of other variables if we do it again.

Bren R.