#38
I am not interested in finding smallest differences audible, but listening to music. The last difference I heard was in cymbals' and piano's highest notes' tone and decay. And I knew what format/rate was playing
I am not interested in finding smallest differences audible, but listening to music. The last difference I heard was in cymbals' and piano's highest notes' tone and decay. And I knew what format/rate was playing
@newvirus2008 Obviously there are significant differences between compression formats and algorithms used in music playback. One is definitely not the same sounding as the other in terms of which part of the audio they discard or ignore. AC3 isn't quite so bad compared to typical mp3. DTS is in theory lossless (Meridian) and so are others for 2 channel ie. FLAC. They all prioritize some part of the audio bandwidth and dynamic range. I can hear the difference between most high bitrate mp3 and 44.1/16 PCM. It depends on the material and signal chain too.
Alot of later 70s and early 80s vinyl wasn't that great in terms of pressing quality. The pinnacle of analog playback was in the early 80s when tape recording technology was at its all time best. 30 IPS on a Studer A series 24 track was the preferred format. It was flat past 40khz with an 80 dB+ dynamic range. Even cassette tape did pretty well if you used good media with a decent 3 head closed loop dual capstan machine.
Alot of later 70s and early 80s vinyl wasn't that great in terms of pressing quality. The pinnacle of analog playback was in the early 80s when tape recording technology was at its all time best. 30 IPS on a Studer A series 24 track was the preferred format. It was flat past 40khz with an 80 dB+ dynamic range. Even cassette tape did pretty well if you used good media with a decent 3 head closed loop dual capstan machine.
@Michael C You won't get an argument back from me praising vinyl in terms of subjective sound quality.
I love analog audio and believe it has the perceived edge in SQ. However, if you look at hard data and specs, digital audio has a lower noise floor, lower distortion and consequently a greater dynamic range.
Comparing equal priced components, even a basic CD player will beat out an average TT. It ends up being a different story if you compare higher end equipment. The TT will surpass the CD player when the cost gets into the 4 digit range.
I love analog audio and believe it has the perceived edge in SQ. However, if you look at hard data and specs, digital audio has a lower noise floor, lower distortion and consequently a greater dynamic range.
Comparing equal priced components, even a basic CD player will beat out an average TT. It ends up being a different story if you compare higher end equipment. The TT will surpass the CD player when the cost gets into the 4 digit range.
You are kidding:
Type: 3-head, single compact cassette deck
Track System: 4-track, 2-channel stereo
Tape Speed: 4.75, 2.37 cm/s
Heads: 1 x record, 1 x playback, 1 x erase
Tape Type: type I, CrO2, Metal
Noise Reduction: B
Frequency Response: 20Hz to 22kHz (Metal tape)
Signal to Noise Ratio: 66dB (dolby B)
Wow and Flutter: 0.04%
Total Harmonic Distortion: 0.8%
...Year: 1979
Price: USD $1350 (1981)
IMHO: FR in those days was meant to be in between as +/-3 dB over the whole range. Today it marks the beginning and end of FR
Channel separation worse, dubbing between windings after storing some time.
Have fun, Toni
Last edited:
Its probably not quite that cut and dried. A well recorded CD through an SOA dac can do some things in terms of sound stage that isn't going to happen with the best vinyl. My vinyl setup is certainly worth more than $10k, probably closer to twice that (including DS Audio optical cartridge). OTOH, my less costly experimental discrete DSD dac with true-SOA clocks is like no dac most people have been exposed to so far. It can reproduce a holographic soundstage better than any vinyl system I have heard.
That said, a lot of CDs are not that well recorded, so a very exceptional dac is still of limited utility.
Also, all the great vinyl coming out of MoFi in recent years has been through a Merging Technologies HORUS system at DSD256. MoFi says it sounds better than piecing together a new master tape from multiple source tapes.
Last edited:
It's a poorly concealed secret that the trendy 15ips half track "master tape" copies have passed through an A/D/A stage on their journey to being re-virginized into analog. Manufacturers could not possibly admit these transgressions without danger of stoning, but it's obvious if we consider the value of the original tapes and their wear life. Your "master tape" was not a 1:1 real time transfer from the studio master, and your modern vinyl pressing's first engraving is very unlikely to derive from a direct transfer from playing the studio master tape. There are only a limited number of those playings that can be made, and the number is not thousands or more, and really isn't even hundreds.
If we admit to ourselves that any modern "analog" consumer production media has passed through an evil A/D/A conversion then we should be able to admit to ourselves that the earliest version, the A/D conversion, is the truest and best, and that the subsequent D/A "analog" versions are romantic variations. Nothing wrong with romantic variations; I love Brahms more than most folk would admit to. But they are not truer or better copies.
There's an interesting (almost)exception to the mastering issue: in the early 1990s Wilma Cozart Fine oversaw the transfer of the 1950s/very early 1960s Mercury recordings that she originally oversaw (credited to her husband Robert, but it takes two to tango). They used the original two and three track tapes, restored the original Ampex tape machines, etc. and tried their best to make as faithful versions as they could, mixed simply (as original) to Red Book. This origin would be "earlier" than a conventional vinyl's mastering cutting made from a specially EQ'd and level dynamically adjusted tape, which was not done for Mercury releases in the 1950s (!). There are even some SACDs that include the three mic original tracks separately, playable with a discrete center channel playback system. Rare nowdays unfortunately.
All good fortune,
Chris
If we admit to ourselves that any modern "analog" consumer production media has passed through an evil A/D/A conversion then we should be able to admit to ourselves that the earliest version, the A/D conversion, is the truest and best, and that the subsequent D/A "analog" versions are romantic variations. Nothing wrong with romantic variations; I love Brahms more than most folk would admit to. But they are not truer or better copies.
There's an interesting (almost)exception to the mastering issue: in the early 1990s Wilma Cozart Fine oversaw the transfer of the 1950s/very early 1960s Mercury recordings that she originally oversaw (credited to her husband Robert, but it takes two to tango). They used the original two and three track tapes, restored the original Ampex tape machines, etc. and tried their best to make as faithful versions as they could, mixed simply (as original) to Red Book. This origin would be "earlier" than a conventional vinyl's mastering cutting made from a specially EQ'd and level dynamically adjusted tape, which was not done for Mercury releases in the 1950s (!). There are even some SACDs that include the three mic original tracks separately, playable with a discrete center channel playback system. Rare nowdays unfortunately.
All good fortune,
Chris
Even direct-to-disk-lathe recordings, such as those of Lincoln Mayorga, have been made at different plants at different times. They don't all sound exactly the same. Possibly different copies of the master stamper, different vinyl quality, etc? Which is the truer version?
Not to mention there were different performances to make multiple master stampers, all marketed as the same album.
Not to mention there were different performances to make multiple master stampers, all marketed as the same album.
Last edited:
There's also the safety copies which were often done at 7.5 ips. Fleetwood Mac's Rumors album was mixed with the backup safety due to so many overdubs, killing the top end of the original analog multi track master. They had to manually sync the safety up with the master without the aid of smpte, so you can hear phasing and all sorts of other artifacts. So its a great album, but awful sound.
These folk might be the perfect source of parallel CD (or at least a bit perfect Red Book A/D) and vinyl versions of the same performance. Not going to change any minds, but within the means of a dedicated scientist. If I could still hear 6kHz I'd volunteer to try.
Haarlem has a namesake in the northern end of Manhatten, NYC, just west of Bronk's Farm, now called The Bronx, named when the place was called New Amsterdam. My Y chromosome was there nine generations ago, but got into some trouble with Peter Stuyvensant, details unknown. Goes without saying, we descendants didn't do as well as the Rockefellers. Arf!
Seasons greets to all, and be of good cheer,
Chris (Hoornebeek)
Haarlem has a namesake in the northern end of Manhatten, NYC, just west of Bronk's Farm, now called The Bronx, named when the place was called New Amsterdam. My Y chromosome was there nine generations ago, but got into some trouble with Peter Stuyvensant, details unknown. Goes without saying, we descendants didn't do as well as the Rockefellers. Arf!
Seasons greets to all, and be of good cheer,
Chris (Hoornebeek)
Last edited:
There's also the introduction of the digital multi track recorder. Most people don't know the Japanese had the tech to do the first commercial digital PCM recording in 1971, far earlier than the first multi track PCM machines showed up. They had the first equipment in the late 60s (ca. 1967) but it was Denon who commercialized it in 71.
I used to work with 3M and Sony PCM machines in the 80s. Quite a few recordings which were touted as fully digital (DDD) were often summed in analog due to insufficient headroom in the digital domain (DAD). If the album was released in analog, this was transfered with 1/2 track 30 ips used to cut a lacquer. If you listen to Dire Straits Brothers in Arms on vinyl, it was actually done from an analog master. Only the CD version was mixed down in full digital (DDD). The SPARSE code wasn't fully implemented until about late 1983, so you knew as a consumer which method of production a CD used. Even then it was still a little bit of an unknown which mixdown method was used.
The Sony PCM machines used a stationary head, so it could be spliced like analog tape. It was the most refined system IMO, but the Mitsubishi X80 and the 3M sounded better IMO. They had better low end and more linear treble. You can tell which machine some recordings used, as they sound quite specific in character. This translates into the analog versions as well, being some engineers would compensate for the converter linearity. The Sony 3324 DASH was the machine I liked most to work with. It was built like a tank and had excellent error correction. Didn't have the low end of the 3M, but sounded neutral in the mids. The converters were very fussy with any DC offset, which ate into the headroom if not removed before the mastering process. When playing a Columbia Jazz reissue (they often used 3324s), take a look at your woofers when playing it at a decent level. If the master compression VCA signal rode on top of the audio (as it frequently did, being removed before cutting laquer by the HP of the cutting head amps), you'd get significant VLF junk coming through on the CD. Some of the VLF was worse in amplitude on the CD than the arm resonance itself caused on the LP.
So my point is, even with digital recordings and remastering there are huge differences in SQ. Its not sufficient just to lump in the digital source as one type of SQ characteristic.
I used to work with 3M and Sony PCM machines in the 80s. Quite a few recordings which were touted as fully digital (DDD) were often summed in analog due to insufficient headroom in the digital domain (DAD). If the album was released in analog, this was transfered with 1/2 track 30 ips used to cut a lacquer. If you listen to Dire Straits Brothers in Arms on vinyl, it was actually done from an analog master. Only the CD version was mixed down in full digital (DDD). The SPARSE code wasn't fully implemented until about late 1983, so you knew as a consumer which method of production a CD used. Even then it was still a little bit of an unknown which mixdown method was used.
The Sony PCM machines used a stationary head, so it could be spliced like analog tape. It was the most refined system IMO, but the Mitsubishi X80 and the 3M sounded better IMO. They had better low end and more linear treble. You can tell which machine some recordings used, as they sound quite specific in character. This translates into the analog versions as well, being some engineers would compensate for the converter linearity. The Sony 3324 DASH was the machine I liked most to work with. It was built like a tank and had excellent error correction. Didn't have the low end of the 3M, but sounded neutral in the mids. The converters were very fussy with any DC offset, which ate into the headroom if not removed before the mastering process. When playing a Columbia Jazz reissue (they often used 3324s), take a look at your woofers when playing it at a decent level. If the master compression VCA signal rode on top of the audio (as it frequently did, being removed before cutting laquer by the HP of the cutting head amps), you'd get significant VLF junk coming through on the CD. Some of the VLF was worse in amplitude on the CD than the arm resonance itself caused on the LP.
So my point is, even with digital recordings and remastering there are huge differences in SQ. Its not sufficient just to lump in the digital source as one type of SQ characteristic.
Last edited:
The LP was recorded at 48kHz so you get more noise at that part of the spectrum. Not said it is music. This is comparing apples to tomatoes.
I know, for a fair comparison the record would have to be sampled with at least 96 kHz.
The same holds below 20 kHz: you can't see from the power spectrum what is music, what is distortion, what is normal noise and what are scratches. A moving spectrum plot like GoldWave produces - that gets refreshed several times per second while you listen - is much more useful in that sense, although distortion remains difficult to recognize.
The same holds below 20 kHz: you can't see from the power spectrum what is music, what is distortion, what is normal noise and what are scratches. A moving spectrum plot like GoldWave produces - that gets refreshed several times per second while you listen - is much more useful in that sense, although distortion remains difficult to recognize.
No it should be sampled at 44.1 so you have the same cutoff filters working.
The real test will be at line frequency. I remember from setting up my system to record lp's that 50 and 100Hz was the greatest issue. Routing cables differently made a huge (20 dB) difference. And we are looking at -50 dB for line noise. It really doesn't get better than that. Above 10kHz it is mostly surface nouse anyway with an lp.
The real test will be at line frequency. I remember from setting up my system to record lp's that 50 and 100Hz was the greatest issue. Routing cables differently made a huge (20 dB) difference. And we are looking at -50 dB for line noise. It really doesn't get better than that. Above 10kHz it is mostly surface nouse anyway with an lp.
Sorry, but that makes as much sense to me as stating that the CD first has to be transferred to vinyl record before doing the comparison.
One of the very few technical advantages of LP over CD is that it doesn't have a ridiculously steep roll-off above 20 kHz, but a more gradual roll-off. When you record it digitally at 44.1 kHz sample rate, you spoil that.
One of the very few technical advantages of LP over CD is that it doesn't have a ridiculously steep roll-off above 20 kHz, but a more gradual roll-off. When you record it digitally at 44.1 kHz sample rate, you spoil that.
Can't hear 6kHz played through speakers at 75dB SPL. Age, exposure to firearms (no ear-pro back during the draft), wind noise on motorcycles. Don't try this at home kids.
All good fortune, and ho, ho, ho,
Chris Kringle
I noticed that you brought up the "ridiculously steep filter" (digital antialiasing filter) earlier in this thread. They will be FIR types, not like analog filters, and so in the passband the effects are minimal to none.
I would really like to know why you find these so problematic?