The CD is just the plastic disc which holds the bits... but you know that.
You might as well use all the dynamic range if it's there so yes, recording closer to the maximum level is sensible. What do you mean by compression artefacts?
LOL. I guess tastes are different. I find Grado HPs so sheiky I fear they'll rip my ears off.
I don't find them reveling, I find them way too hot in the top end and that can sound awful on some recordings. Or to my ears, most recordings. They do seem to be popular, so ears and tastes vary.
Getting levels up to 15 db higher on a CD than with analog formats is the 'cheap' way to improve its SQ, at the cost of brutalizing the high level transients with overuse of dynamic compression. I believe this is one factor that has contributed to modern recordings being so limited in dynamics with many spending much of their time occupying a 1 or 2 db dynamic range.
Last edited:
I read and hear this a lot, but I don't find that well designed tube circuits have a sound. I don't hear the coloration. Solid-State, on the other hand, almost always has a grit and grain that I simply don't hear in natural sounds. To my ear, it's the transistors that have an unnatural coloration, not the tubes.
Not all transistor circuits, of course, but most.
I'm not sure I understand the point. CD or memory stick or hard drive I regard the 16 bits as simply a range with the possibility for dither at the bottom end and a hard limit at the top end - so don't record so loud you exceed the range. Compression? You shouldn't need to compress if you don't want to because the dynamic range is so wide (unlike, say, vinyl). However, you may have a specific application in mind, like sounding louder for radio or for people to listen to on headphones while cycling in busy streets - sounds terrible but that's not the fault of CD or 16 bit audio.
Last edited:
how about QA manager who has no idea what electronics is about, in ems companybut yeah, got masters degree in god-knows-what
i love self-promoted professionals with diploma printed on toilet paper with potato stamp on it
priceless
A-flippin-men brother!
i have no problem with someone like that who is humble! who says "look guys i dont know, but you guys do! what can i do or get you to help?"
but dosent happen often!
Some very good discussion going on at the moment - very pleasing to see ...
I would be concerned with any A/B'ng where you go rapidly from one state to the other. One thing that whacked me hard against the side to the head, over 25 years ago, is that digital sound will relatively slowly, over many minutes or more, evolve in quality, either better or worse depending upon everything. Whenever I alter anything, unless it profoundly sounds worse, I'll let it "play with my mind" for 24 hours or so, so that I can make a judgement call that I feel confident about.
To put this into context, back in those good ol' days, all the equipment ran 24/7 - including the Perreaux with its 60degC heatsinks, so I was going through some juice - and the big Yamaha CD beast took 3 days to fully come on song. It really p!ssed me off when I had to shut it down for an extended time, because I knew I had wait 3 days again to get back to a known good state ...
Close to half of CD's dynamic range suffers from quantization distortion in excess of 1% - figures that would rule out a good quality amplifier or most quality speakers if they possessed them across the entire bottom half of their useable dynamic range. One cannot assume that 'dither' will have been properly applied, because in most cases it hasn't. The 'cheap' approach to minimize this effect is to chop off peak dynamics to push the music to higher levels dynamically on the CD.
Another factor that markedly deteriorates many moderate rate digital recordings is jitter, either during the record or playback process. 1ns of peak to peak jitter will deteriorate red book CD audio to about 11 bit precision at 20khz and such levels are common in dedicated CD quality equipment. And then envelope ripple increases linearly with frequency, as I have stated before, reaching about 5% at 16 khz with CD. These plus the brick wall pre-ringing artifacts would seem to be sufficient reasons for CD's woeful high frequency performance. I recall reading an early 1950's amplifier review where the amplifier in question was slammed for having something resembling this between 20-25 khz due to some magnetic nonlinearity.
Another factor that markedly deteriorates many moderate rate digital recordings is jitter, either during the record or playback process. 1ns of peak to peak jitter will deteriorate red book CD audio to about 11 bit precision at 20khz and such levels are common in dedicated CD quality equipment. And then envelope ripple increases linearly with frequency, as I have stated before, reaching about 5% at 16 khz with CD. These plus the brick wall pre-ringing artifacts would seem to be sufficient reasons for CD's woeful high frequency performance. I recall reading an early 1950's amplifier review where the amplifier in question was slammed for having something resembling this between 20-25 khz due to some magnetic nonlinearity.
Last edited:
Yeah, it works so much better if you can wait a day between selections. It's so well known to improve acuity. 😀
I gather you haven't actually looked at this experimentally. Well, at least you've traveled from 3% to 1% in just one afternoon. At this rate, you'll hit the real numbers in the next 12 hours or so.
This is right on the money. When CD sound comes together properly, you feel like you need unlimited ability to pump out SPLs, because no matter how loud the sound objectively is, it's never too "loud" - if that makes sense!! 😉
And, this is the nightmare world of what we're dealing with here. People clamour for meaningful measurements of what's going on, and I've always felt instinctively that it was going to be damn hard to do - this confirms that difficulty, even when one has lots of the right measuring tools ...
Emotiva ERC-2 CD player Measurements | Stereophile.com
Simaudio Moon Evolution 650D CD player Measurements | Stereophile.com
First two examples I came up with; I can list a few dozen more if you like. Perhaps you can point me to these intrinsic artifacts of the CD medium that you bring up?
Simaudio Moon Evolution 650D CD player Measurements | Stereophile.com
First two examples I came up with; I can list a few dozen more if you like. Perhaps you can point me to these intrinsic artifacts of the CD medium that you bring up?
Sorry, it just doesn't work that way: just consider a situation where, say, you're trying to improve a glue which takes 24 hours to dry each time, and you only have one sample of the glue to play with at a time. You need to evolve a technique which as accurately as possible determines whether you're better off or not after that period: this is why I use really "nasty" recordings to test where I've got to at any point -- if I want to test a bridge I drive 100 fully loaded semitrailers on to it at one time, then I know it's strong enough ...
Humans aren't glue. I know it's painful to actually do an experiment, but set up a pair of files, where one has a 0.3dB level difference with the other, or a similar amount of EQ. Try doing an ABX using (for example) foobar, with switching between selections essentially instantaneous. You'll pick up a difference if your hearing is sharp. You should be able to score 10/10.
Now listen to the randomly chosen selections a day apart, without being able to switch instantaneously. Heck, try them an hour apart. Your score will be random. This is all really well known stuff in sensory science, but that comes from people who actually do controlled experiments as opposed to waving their hands.
Now listen to the randomly chosen selections a day apart, without being able to switch instantaneously. Heck, try them an hour apart. Your score will be random. This is all really well known stuff in sensory science, but that comes from people who actually do controlled experiments as opposed to waving their hands.
Could be smooth, very smooth. I have a nicely working '86 Yamaha "battleship", and it's mellowed with age, I guess: switched on from not being touched in months, it makes LP sound coarse and edgy ... 😀, 😀
It's apparent that quantization distortion of a signal with a peak amplitude 100 times larger (40db) than the quantization step will be close to 1%. I've bought plenty of CD's in the past, but basically have given up on them. I was hoping that SACD or DVDA would become more popular, but the public mindset has been headed for the shallow end of the audio pool.
Low level LP distortion is in the hundredths or thousandths of a percent.
Low level LP distortion is in the hundredths or thousandths of a percent.
Last edited:
You're still missing the point. Being able to pick up level differences is working at one level of acuity, "quality" or distortion detection is in another area of hearing, so to speak.
To me, when listening I can turn the volume up and down, dramatically, and subjectively it doesn't change - I'm not registering the SPL as being meaningful, because that is not what I listening to. My brain is actually tuning into the story behind the story, listening for the signature indicators of the various distortion artifacts.
On that aspect, when a system has these elements of distortion in the sound then volume changes do register very strongly; perversely, the better the sound the less impact a major change of volume makes, because the ear/brain adapts to the variation, automatically, unconsciously. This is why one can get into clipping very easily, one doesn't perceive the sound as being "loud", subjectively.
- Status
- Not open for further replies.