Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › From CD to SACD how much of difference
New Posts  All Forums:Forum Nav:

From CD to SACD how much of difference - Page 9

post #121 of 161
Quote:
Originally Posted by PhilS View Post
Would it also be true in your opinion that there is no difference in audible sound between one redbood CD player and another for two channel playback? Between one MP3 player and another? Between an MP3 player and a CD player?
I compare equipment that I own to make sure it matches the output of my audio workstation. Of my six iPods, there's no noticeable difference between lineouts. The two iPod models that I have compared to my CD player are identical using lossless files. My old Sony CD player and my old JVC DVD player sounded exactly like my Philips SACD player, and the Philips sounded just like a friend's Yamaha CD player that I compared it to. My turntables and cartridges all sound different from each other, however. So do my R2R and cassette tape decks. My car stereo sounds way off, but that's probably because of the speakers. Every set of speakers I've ever heard sounds different.

It's entirely possible that a Zune sounds different than an iPod. I've never heard a Zune. It's also possible that there are CD players that don't perform up to spec. But I have yet to find a player myself that sounds different using line out. Headphone output quality on iPods do vary from generation to generation. And line out levels vary on everything I own. I have to calibrate that before I can compare anything.

If a CD player or iPod sounded noticeably different, I'd return it because it wouldn't be able to follow the equalization curve I use for all my sources. I need a consistent baseline frequency response.

See ya
Steve
post #122 of 161
Quote:
Originally Posted by bigshot View Post
I compare equipment that I own to make sure it matches the output of my audio workstation. Of my six iPods, there's no noticeable difference between lineouts. The two iPod models that I have compared to my CD player are identical using lossless files. My old Sony CD player and my old JVC DVD player sounded exactly like my Philips SACD player, and the Philips sounded just like a friend's Yamaha CD player that I compared it to. My turntables and cartridges all sound different from each other, however. So do my R2R and cassette tape decks. My car stereo sounds way off, but that's probably because of the speakers. Every set of speakers I've ever heard sounds different.

It's entirely possible that a Zune sounds different than an iPod. I've never heard a Zune. It's also possible that there are CD players that don't perform up to spec. But I have yet to find a player myself that sounds different using line out. Headphone output quality on iPods do vary from generation to generation. And line out levels vary on everything I own. I have to calibrate that before I can compare anything.

If a CD player or iPod sounded noticeably different, I'd return it because it wouldn't be able to follow the equalization curve I use for all my sources. I need a consistent baseline frequency response.

See ya
Steve
Isn't this music by numbers? Does the source have to fit within certain parameters designated by a formula before it sounds good?

I have a Cyrus Redbook player which I consider to be a mid-level source. I have also had the good fortune to listen to a Cary CD306 for a hour. The Cary was light years ahead of the Cyrus with regard to SQ. The Cyrus is quite good, but the Cary is magic.

I thought music was about feel, enjoyment and all the other cliche's you can think of, NOT science and numbers. Part of the fun is the variance between different bits of gear.

I am not the most eloquent of people, but hope I have got my thoughts across.

cheers
Simon
post #123 of 161
Quote:
Originally Posted by bigshot View Post
I compare equipment that I own to make sure it matches the output of my audio workstation. Of my six iPods, there's no noticeable difference between lineouts. The two iPod models that I have compared to my CD player are identical using lossless files. My old Sony CD player and my old JVC DVD player sounded exactly like my Philips SACD player, and the Philips sounded just like a friend's Yamaha CD player that I compared it to. My turntables and cartridges all sound different from each other, however. So do my R2R and cassette tape decks. My car stereo sounds way off, but that's probably because of the speakers. Every set of speakers I've ever heard sounds different.

It's entirely possible that a Zune sounds different than an iPod. I've never heard a Zune. It's also possible that there are CD players that don't perform up to spec. But I have yet to find a player myself that sounds different using line out. Headphone output quality on iPods do vary from generation to generation. And line out levels vary on everything I own. I have to calibrate that before I can compare anything.

If a CD player or iPod sounded noticeably different, I'd return it because it wouldn't be able to follow the equalization curve I use for all my sources. I need a consistent baseline frequency response.
Thanks. I think that helps in evaluating your comments in response to the original question.
post #124 of 161
Quote:
Originally Posted by The Pieman View Post
Isn't this music by numbers? Does the source have to fit within certain parameters designated by a formula before it sounds good?
I've carefully equalized my system to correct for the imbalances in the speakers and the acoustics of the room. If my CD player sounded different than my DAP, and both of them sounded different than my DVD, I'd need a separate EQ setting for each component. That would be WAY too much work.

I don't doubt that there are components designed to provide colored sound. But I'm not interested in them. I buy components that have a flat response, so they all match to my system properly. That way, *everything* sounds like "magic".

As you say, music is all about feel and enjoyment. Music is a creative artform. But sound reproduction isn't a creative thing. It's all about *fidelity*.

See ya
Steve
post #125 of 161
I downloaded the 5 sample FLAC files from HighDefTapeTransfers.com which are recorded in 24/96. Every one of them (although none belong to any genre I regularly listen to) sound better than any of the 1000 CDs I have in my collection. They all have more depth and width, more accurate timbre and tone, greater imaging and separation, and a more neutrally balanced frequency response. When you compare audio natively sampled at 96 khz, you can hear that it just sounds cleaner and less fatiguing to the ears. It is not as choppy as 44.1. But it should be easier on the ears; it is sampled twice as many times per second.
post #126 of 161
Quote:
Originally Posted by IPodPJ View Post
I downloaded the 5 sample FLAC files from HighDefTapeTransfers.com which are recorded in 24/96. Every one of them (although none belong to any genre I regularly listen to) sound better than any of the 1000 CDs I have in my collection.
That isn't a valid comparison. There are a dozen reasons other than bitrate that the 24/96 tracks sound better to you. Compare apples to apples. Get a good audio editing program and knock it down to 16/44.1 with the proper dither. Compare the 24/96 to that. I bet you find that the difference is MUCH harder to detect when you're comparing the exact same recording.

See ya
Steve
post #127 of 161
I did that Steve. I resampled it to 44.1 to see what the difference was and it sounded choppy like any other 44.1 kHz recording. Don't get me wrong, I have some amazing CDs that sound incredible. I would never have thought 44.1 sounded "choppy" until I heard a good 24/96 audio file. A 24-bit file can store more information and at 96 kHz it is sampled at more than twice the rate of redbook CD. I even tried taking that same 24/96 file, and then after downsampling it to 16/44.1, resampled it back to 24/96 and that super fidelity wasn't there.
Most CDs out there are mixed and mastered so many times, none of them sound accurate to what would have been heard in the studio, and that's typically because they layer all the tracks on top of eachother. But a good live stereo recording will sound better than any album thats mixed from multiple tracks since all the instruments are recorded simultaneously and with only 2 channels. And typically, the only genres of music you find this with are classical and jazz.
post #128 of 161
I am DL some Stravinsky and Brahms right now from HighDefTapeTransfers.com I will see if these tape transfers are any good
post #129 of 161
Quote:
Originally Posted by IPodPJ View Post
I did that Steve. I resampled it to 44.1 to see what the difference was and it sounded choppy like any other 44.1 kHz recording. Don't get me wrong, I have some amazing CDs that sound incredible. I would never have thought 44.1 sounded "choppy" until I heard a good 24/96 audio file. A 24-bit file can store more information and at 96 kHz it is sampled at more than twice the rate of redbook CD. I even tried taking that same 24/96 file, and then after downsampling it to 16/44.1, resampled it back to 24/96 and that super fidelity wasn't there.
Most CDs out there are mixed and mastered so many times, none of them sound accurate to what would have been heard in the studio, and that's typically because they layer all the tracks on top of eachother. But a good live stereo recording will sound better than any album thats mixed from multiple tracks since all the instruments are recorded simultaneously and with only 2 channels. And typically, the only genres of music you find this with are classical and jazz.

Probably a bad downsample. 96khz doesn't downsample neatly into 44.1khz, and who knows what was going on with the 24-bit. It could have been -24db on the highest peaks of the 24 bit source, and putting it at 16-bit made it extremely low resolution with a poor resample.

"Choppy" is NOT a term I've ever heard used for 16/44.1khz. At least not a proper 16/44.1 source.
post #130 of 161
Definitive XRCD vs 24bit/96Khz Review of Erick Friedman Violin Show Piece

I have the XRCD version and I think its one of those most well mastered/recorded CDs I have. It's got the XRCD2, and the 20bitk2 processing.

I downloaded the 24/96 tape transfer version from HighDefTapeTransfers

WOW what a difference 24/96 tape transfer makes; the high def tape transfer made the XRCD version sounds like mp3!!!

Soundstage is greatly improved in the 24/96 version. The soundstage instead of stopping at a imaginary 'brick wall', now extends far far far beyond. All the instrument seems to have the real reverberation, and air around them (like in a real concert hall!!).

The sound of the high quality tape transfer just sounds more relaxed, more analog, much more smooth. Erick's violin took on a new life. Not just 'hi-hi' sounding in the XRCD version (a little grainy I might add), but 'real life' sounding. Smooth yet real, detailed without feeling edgy. The whole performance took on an ease I have only experienced with real instruments.

One noticeable area the XRCD substantially differs from the tape transfer is a mid-bass hump that's present in the XRCD version and not in the tape transfer version.

I think this is the best example of high def vs CD (in this case XRCD). If there is this much difference between a 24/96 tape transfer and XRCD (presumably much better mastered than regular CD), I am 1000% sold on the 24/96 format. THERE IS NO CONTEST.

I used SD TP to cue up both versions of the same track and just compared them back to back to back to back, the difference is obvious and amazing.

Update: I cued up the XRCD in the Rega Saturn and A/B against 24/96 playing back from TP. No surprise there, the TP with 24/96 tape transfer made Saturn sounds like a cheap machine. The difference is much great than the difference between the Linn CD12 and Saturn (yes I compared them with the same XRCD (my default stereo shop demo CD. yes I would also say TP with 24/96 would put Linn CD12 playing XRCD to shame too.
post #131 of 161
Quote:
Originally Posted by bigshot View Post
That isn't true. Higher sampling rate increases resolution in all frequencies, but higher bitrate just increases the bandwidth, which translates into dynamic range and frequency extension. .
Not quite according to Dan Lavry. The higher sampling rate increases the bandwith from 20k to say 48K or 96K, the waveform if bandlimited can be perfectly reconstructed by 2B fs where B is the bandwidth, in that respect higher sampling (fs) does not change resolution.

Bit-depth simply increases the dynamic range.

Dan Lavry did a very interesting white paper on why higher sampling rates are problematical and often redundant. Then he goes and makes 96K DACs anyway, go figure.
post #132 of 161
Quote:
Originally Posted by IPodPJ View Post
I did that Steve. I resampled it to 44.1 to see what the difference was and it sounded choppy like any other 44.1 kHz recording.
What kind of dither did you apply? That's very important. If it sounded choppy, there is definitely something wrong with the way you resampled it. You probably resampled from 96 to 44.1 without dithering.

See ya
Steve
post #133 of 161
Quote:
Originally Posted by hciman77 View Post
Not quite according to Dan Lavry. The higher sampling rate increases the bandwith from 20k to say 48K or 96K, the waveform if bandlimited can be perfectly reconstructed by 2B fs where B is the bandwidth, in that respect higher sampling (fs) does not change resolution.
I'm afraid you lost me there. Are you saying that the higher sampling rate provides frequency extension and the bit rate accounts for the increase in dynamic range?

As I understand it, high bitrate sound does have additional resolution across all frequencies, but only at extremely low volume levels. It has something to do with the increase in dynamic range. I know that our ears are much less sensitive to bandpass filters at low volume levels. That's the principle that dynamic noise reduction operates on.

Sorry
Steve
post #134 of 161
Quote:
Originally Posted by hciman77 View Post
Not quite according to Dan Lavry. The higher sampling rate increases the bandwith from 20k to say 48K or 96K, the waveform if bandlimited can be perfectly reconstructed by 2B fs where B is the bandwidth, in that respect higher sampling (fs) does not change resolution.

Bit-depth simply increases the dynamic range.

Dan Lavry did a very interesting white paper on why higher sampling rates are problematical and often redundant. Then he goes and makes 96K DACs anyway, go figure.
Quote:
Originally Posted by bigshot View Post
I'm afraid you lost me there. Are you saying that the higher sampling rate provides frequency extension and the bit rate accounts for the increase in dynamic range?

As I understand it, high bitrate sound does have additional resolution across all frequencies, but only at extremely low volume levels. It has something to do with the increase in dynamic range. I know that our ears are much less sensitive to bandpass filters at low volume levels. That's the principle that dynamic noise reduction operates on.

Sorry
Steve
The higher sampling rate increases the bandwidth. The Nyquist-Shannon theorem says that to get a perfect mathematical reconstruction of a waveform you must limit the bandwidth to fs/2 i.e if you want to be able to render a 22,050 tone you must sample at a frequency of 44,100 (fs). If you up the fs from 44,100 to 96,000 you up the available bandwidth that can be dealt with from 22,050 up to 48,000. Intuitively you would think that upping the sampling rate would improve the resolution as yuo have more samples, however this means that the Shannon-Nyquist theorem would have to be wrong ans that fs is insufficient (N-S works in an ideal world and we can argue the toss about it but for now assume it is valid) , so to say you have to have 96k sampling means you assume that Nyquest-Shannon is faulty.

Bit-depth increases the dynamic range simply because it gives you more descrete levels of voltage difference, so instead of 65536 for 16 bits you might have 16777216 for 24 bits. Now we get into questions of how well humans can distinguish the difference.

So for instance is it possible for a human to tell the difference between 65535 and 65536 ? , my guess would be no as it would be well under the 0.2db that gets bandied about ( it is a percentage difference of 0.00152%) , I think technically the difference is 0.01db as far as I am aware nobody has put the JND for human hearing at anywhere near that level, lets try another one , can a human tell the difference between 10 and 11, probably yes. Now the difference between 10 and 11 on a 16 bits system is 1 , the difference between 10 and 11 on a 24 bit system is er 1.
post #135 of 161
Redbook digital most likely sounds chopped (or whatever) because people actually do not know how digital works. They have seen too many pictures of a stairstepped waveform - and then gone on to make a conclusion that it would look and sound smoother if the sampling was made at higher frequency.

My estimate is that 99% of these reported night-and-day differences would simply vanish in a properly conducted blind test. But we're audiophiles and we don't wanna know...
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Dedicated Source Components
Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › From CD to SACD how much of difference