The Audiophile Thread

[Firth]

Here's a visual representation of what I'm talking about. Here's a clip from Kaipa's latest, recorded from the vinyl:

Attachment 14178

Here's the same clip ripped from the CD:

Attachment 14179

It's easy to see how a stylus trying to follow the lower waveform carved into a groove would be problematic. It'd be like trying to play a saw blade.

Looks like the scale on the CD version is different so that the waveform is cut off. CD versions can be brickwalled however I have read where Steve Wilson and another well known engineer will convert High res digital to analog and then use a tube stage to perform analog compression similar to a cutting lathe for a record.

[Firth]

This article is about the CD or HDCD mastering of CSN’s first album. There are several masters, but the Joe Gastwirt version wins here. It’s very interesting and Steve Hoffman has a version which is compared. There is a lot of other info about making the album.
https://audiophilestyle.com/ca/the-b...ed-debut-r826/

That article compares an HDCD version of CSN (Grundman) to a Joe Gastwirt version. The HDCD version sounded great when I heard it relative to the older version I have. The comment section gets into dynamic range controversies.

My HDCD versions of Joni Mitchell’s CDs were mastered by Joe Gastwirt and they sound great. I read an article which I can’t find where he would use a tube stage as a processor for Analog compression.

[progmatist]

Kapia brickwalled the hell out of the CD version. Many have complained it sounds way too loud and overly compressed.

[Firth]

Kapia brickwalled the hell out of the CD version. Many have complained it sounds way too loud and overly compressed.

Sorry to hear that. I’m surprised that vinyl would make a difference unless the mastering was completely different for vinyl.

[progmatist]

^^ I would absolutely say the mastering is different. Just listening to vinyl compared to the CD is night and day in terms of loudness, noticeable to just about anyone. It's not at all subtle like the differences between CD and Hi-Res, which will escape most people's attention.

[Firth]

^^ I would absolutely say the mastering is different. Just listening to vinyl compared to the CD is night and day in terms of loudness, noticeable to just about anyone. It's not at all subtle like the differences between CD and Hi-Res, which will escape most people's attention.

I think is an overreaching generalization to assign bad mastering to all CDs. It’s also possible that vinyl recordings come from the same digital master as a CD.

I would be interested in knowing whether that piece or a song from that piece was on ITunes and mastered for ITunes. Apple requires 24 bit 96khz files for input to their processing that performs data compression. Apples processing insures little saturation in the stages of processing. It is possible that the ITunes version sounds better than the CD.

[progmatist]

^^ The AAC+ format Apples uses has the distinct advantage of being able to encode directly from Hi-Res. MP3 must first be converted to a 44.1K or 48K wav or aiff, then to MP3.

I don't recall ever saying the "mastering" in general on CDs was bad. Not everyone brickwalls the hell out of their music. But when when excessive brickwalling does happen, it isn't transferable to the vinyl format.

[Firth]

^^ The AAC+ format Apples uses has the distinct advantage of being able to encode directly from Hi-Res. MP3 must first be converted to a 44.1K or 48K wav or aiff, then to MP3.

I don't recall ever saying the "mastering" in general on CDs was bad. Not everyone brickwalls the hell out of their music. But when when excessive brickwalling does happen, it isn't transferable to the vinyl format.

If the mastering brickwalls the digital data and that data drives a cutting lathe and then a needle, it can’t sound good.

[Zeuhlmate]

CD has the potential of being more dynamic, have a tighter bottom, and less surface noise.
Vinyl has a bigger frequency range, and thus a potential of more musical info sent to your ears.

But it all depends of the production how it sounds. Equipment, raw material, mix, mastering, skill and ears, pysical production (pressing). You can ruin both media - as we all know.
I generally prefer CD's because they are practical, but I hate the covers.

And besides, our HiFi systems are different and sounds different. Taste.

[Firth]

CD has the potential of being more dynamic, have a tighter bottom, and less surface noise.
Vinyl has a bigger frequency range, and thus a potential of more musical info sent to your ears.

But it all depends of the production how it sounds. Equipment, raw material, mix, mastering, skill and ears, pysical production (pressing). You can ruin both media - as we all know.
I generally prefer CD's because they are practical, but I hate the covers.

And besides, our HiFi systems are different and sounds different. Taste.

Absolutely vinyl does not have more frequency range than CDs done right. Record playing is an extremely expensive thing to get right, and almost impossible to get extreme low end correct without environmental interaction if speakers are involved. I have Telarc recordings on vinyl of Also Sprach Zarathustra a needle will barely stay on.

[Zeuhlmate]

Absolutely vinyl does not have more frequency range than CDs done right. Record playing is an extremely expensive thing to get right, and almost impossible to get extreme low end correct without environmental interaction if speakers are involved. I have Telarc recordings on vinyl of Also Sprach Zarathustra a needle will barely stay on.

Correct - its the other end that vinyl has potential to cover better, it doesn't stop at 20 khz as the CD does.

[Firth]

Correct - its the other end that vinyl pas potential to cover better, it doesn't stop at 20 khz as the CD does.

Doesn’t matter because you can’t hear it. But, the real reason why systems need more bandwidth beyond the range of hearing is the practical aspect of making sound linear in the band we can hear. That’s why digital production and even my receiver has 32 bit 192 KHz processing. The accurate digital filters that downsamples to the CD rate, does so with extreme precision.

[progmatist]

If the mastering brickwalls the digital data and that data drives a cutting lathe and then a needle, it can’t sound good.

Brickwalling would be audio compression, independent of any recording medium. It's the most extreme form of limiting, which itself is the most extreme form of audio compression, with an infinite compression ratio. Typical per track compression uses a ratio of between 1.2:1 and 5:1. Live sound can be brickwaled, without any recording whatsoever.

Correct - its the other end that vinyl has potential to cover better, it doesn't stop at 20 khz as the CD does.

Speaking of that: Magnetic tape moving at 30 Inches Per Second is capable of ultrasonic frequencies, much like Ultra Hi-Res digital. The fact both fast moving tape and Ultra Hi-Res digital have an ultra-wide frequency range is neither here nor there. I don't think anyone in their right mind would vehemently deny tape moving at 30IPS sounds better than 15IPS, which in turn sounds better than 7.5IPS.

[Zeuhlmate]

DAT tapes wasn't limited to 20 khz as I recall it.

[Firth]

Brickwalling would be audio compression, independent of any recording medium. It's the most extreme form of limiting, which itself is the most extreme form of audio compression, with an infinite compression ratio. Typical per track compression uses a ratio of between 1.2:1 and 5:1. Live sound can be brickwaled, without any recording whatsoever.



Speaking of that: Magnetic tape moving at 30 Inches Per Second is capable of ultrasonic frequencies, much like Ultra Hi-Res digital. The fact both fast moving tape and Ultra Hi-Res digital have an ultra-wide frequency range is neither here nor there. I don't think anyone in their right mind would vehemently deny tape moving at 30IPS sounds better than 15IPS, which in turn sounds better than 7.5IPS.

The specification of DSD/SACD was designed to match the best tape machines, 100 KHz bandwidth and 100 dB dynamic range. However, DSD’s impulse response is more accurate than tape or any other digital format. DSD requires the absolute minimum digital or analog filtering. However in reality DSD is converted to PCM or Analog for mixing boards, and converted to the desired format. Mixing in DSD is very exotic.

Audio compression (analog or digital) can be brick walling but not necessarily. A true analog compressor such as DBX, could theoretically achieve expansion with a net linear response. Digital compression could approach analog compression however IMO, it would need to have much finer quantization than 24 bits would afford. Perhaps 64 bit floating point computation.

[hFx]

Speaking of that: Magnetic tape moving at 30 Inches Per Second is capable of ultrasonic frequencies, much like Ultra Hi-Res digital. The fact both fast moving tape and Ultra Hi-Res digital have an ultra-wide frequency range is neither here nor there. I don't think anyone in their right mind would vehemently deny tape moving at 30IPS sounds better than 15IPS, which in turn sounds better than 7.5IPS.

Well it's not that simple in the world of analog...

https://www.tangible-technology.com/media/media_2.html

[Firth]

Well it's not that simple in the world of analog...

https://www.tangible-technology.com/media/media_2.html

And that article is 23 years old

[BobM]

I've been doing a good bit of mixing and recording the last few months, all digitally of course. It is absolutely critical to put a limiter on the master track to help the signal pull back down to the zero level if it breaks through. Having said that, you do need a compressor to get the signal as close to that zero mark as possible on loud passages to give you the dynamic. It's a real balancing act and not at all easy to do. As a result my digital mixes don't quite sound as loud as commercial releases because, well, I don't like over compression, it flattens things out dynamically.

With tape you can press quite a bit past that zero mark because the tap saturation creates its own compression. I think you tend to get a better dynamic with tape. Old vinyl records only had tape to record on and I do believe that old vinyl tends to sound better than digitally remixed digital on those songs, probably because the remixes use more digital compression and, though it sounds consistently louder, it also flattens the macro and micro dynamics. Ask yourself why most binaural recordings (only using 2 microphones to record as opposed to close miking everything) sound so much more natural and have clearly defined dynamics than modern recordings. Sure, their overall level is lower, but when it gets loud it gets loud.

[Firth]

I've been doing a good bit of mixing and recording the last few months, all digitally of course. It is absolutely critical to put a limiter on the master track to help the signal pull back down to the zero level if it breaks through. Having said that, you do need a compressor to get the signal as close to that zero mark as possible on loud passages to give you the dynamic. It's a real balancing act and not at all easy to do. As a result my digital mixes don't quite sound as loud as commercial releases because, well, I don't like over compression, it flattens things out dynamically.

With tape you can press quite a bit past that zero mark because the tap saturation creates its own compression. I think you tend to get a better dynamic with tape. Old vinyl records only had tape to record on and I do believe that old vinyl tends to sound better than digitally remixed digital on those songs, probably because the remixes use more digital compression and, though it sounds consistently louder, it also flattens the macro and micro dynamics. Ask yourself why most binaural recordings (only using 2 microphones to record as opposed to close miking everything) sound so much more natural and have clearly defined dynamics than modern recordings. Sure, their overall level is lower, but when it gets loud it gets loud.

I read where Joe Gastwirt would use a tube stage for compressing analog before digital mastering. Intuitively I don’t understand how digital limiting or compression unless done on the very extreme peaks would not have a negative impact on the sound.

[progmatist]

I've been doing a good bit of mixing and recording the last few months, all digitally of course. It is absolutely critical to put a limiter on the master track to help the signal pull back down to the zero level if it breaks through. Having said that, you do need a compressor to get the signal as close to that zero mark as possible on loud passages to give you the dynamic. It's a real balancing act and not at all easy to do. As a result my digital mixes don't quite sound as loud as commercial releases because, well, I don't like over compression, it flattens things out dynamically.

With tape you can press quite a bit past that zero mark because the tap saturation creates its own compression. I think you tend to get a better dynamic with tape. Old vinyl records only had tape to record on and I do believe that old vinyl tends to sound better than digitally remixed digital on those songs, probably because the remixes use more digital compression and, though it sounds consistently louder, it also flattens the macro and micro dynamics. Ask yourself why most binaural recordings (only using 2 microphones to record as opposed to close miking everything) sound so much more natural and have clearly defined dynamics than modern recordings. Sure, their overall level is lower, but when it gets loud it gets loud.

One reason binaural recordings sound so good is every instrument has to be positioned just right, relative to the mics. This skill was developed over decades of positioning musicians around an acoustic horn, which cut directly into the wax master, then a single mic in the early electric recordings. Close micing everyone and artificially positioning them via the mixing board does leave much to be desired.

[hFx]

if you check the frequency response of a normal analogue master recorder, it is pretty linear up to about -20 dBVU and gets increasingly nonlinear above, dropping fast in the higher frequencies. at 0dBVU the response is quite off and above that, even more. Used that way the analogue tape machine is a sound processor. It does sound nice with the soft compression, added distortion and treble roll-off, but is has less and less to do with the actual signal it tries to record. Digital on the other hand gain fidelity by increasing the number of available bits until it hits the 0dBFS roof where the distortion gets really nasty. Anyway, to utilize any increased frequency range of analogue tape, stay way down in recording level (and some good NR to keep the tape noise floor down).

[Firth]

I’m not sure why all recording is not done with a DSD recorder at the 5.6 MHz rate, and having all channels archived at that rate. If you want analog, then use a DSD to analogue converter.

[progmatist]

^^ Because DSD equipment is substantially more expensive than standard PCM equipment. At the dawn of CD technology in the early 80s, there were 20bit, 96KHz recorders. But they too were bleeding edge, super expensive, and out of reach for most.

[Firth]

^^ Because DSD equipment is substantially more expensive than standard PCM equipment. At the dawn of CD technology in the early 80s, there were 20bit, 96KHz recorders. But they too were bleeding edge, super expensive, and out of reach for most.

Well yeh, but the reason that ADC became cheap was the invention of the Delta-Sigma converter which is essentially a DSD converter. Even now, when you get a 24 bit 96 KHz output from a ADC, it started as a single bit output stream integrated and filtered to the PCM format. There is no tech reason the single bit decisions can not be stored. The bias against DSD is rediculous, the bit stream bandwidth of 24/96 PCM stereo is more than 4.8 Mbps. A DSD stream on a SCD is 2.8 Mbps. DSD is the most efficient, least expensive and accurate way to get HD audio.
On the other end, all good DACs are Sigma-Delta which Convert PCM or more directly DSD, to a bit stream which requires very benign analog filtering at the output.

[progmatist]

^^ It's not just the conversion hardware. The computing power required to capture and process DSD is far more massive than the most commonly used 24/44.1 PCM. The same applies to anything above 48KHz PCM by the way. Let's not forget: it isn't even possible to earn a living as a recording artist anymore.