Topic: 24/96 of a 1974 master analog?

[Belexes]

Someone on Dime asked if the seeder would post a 24/96 of a 1974 master he has.  That's fairly pointless is it not seeing as the master tape was not recorded in 24/96?

[ghellquist]

It depends. A good analog tape can have both a large dynamic and frequency range which can benefit from a bit more than the standard 44.1/16. Depending on the operator doing the transfer I believe the 24 bits is recommended.

It also depends on the AD converter as some of them does not quite deliver what you expect and can sound decidedly different on different sampling rates.

Gunnar

[boyacrobat]

get over 16/44
served us well BUT SERVES US NO MORE.

all analogue masters need to move from 16/44

g

[sunjan]

Someone on Dime asked if the seeder would post a 24/96 of a 1974 master he has.  That's fairly pointless is it not seeing as the master tape was not recorded in 24/96?

I'm with Gunnar and Moke on this.
Besides, the comparison you make between analog and digital is moot, since analog tapes are not recorded in bits... 

[Kevin Straker]

I'm fuzzy on how upsampling can insert meaningful information that was not there to begin with? I'm sure someone will enlighten me.

[aegert]

I'm fuzzy on how upsampling can insert meaningful information that was not there to begin with? I'm sure someone will enlighten me.

It is not an up-sample

You are sampling for the first time.. analogue does not have a Bit depth or sample rate.... :-)

A

[Brian]

as others have said it's not pointless at all.  as a matter of fact i'd transfer it at 24/192.  it's an analog source...might as well get as many samples per second as you can out of it! even if it is just a cassette

remember....when something is recorded on an analog medium, what's on the media IS the source.  when digital, it's just a "representation" of the source.

[Digital Quality]

What is the dynamic range of the cassette? I was thinking it's around 50dB but that's pretty much just a guess. I think the frequency response for tape is pretty high but the range is pretty low. If the range of magnetic tape is really around 50dB, then 24 bits would really be overkill IMO but you would still want a high sampling rate since the frequency response is so high.

[Roving Sign]

I'm fuzzy on how upsampling can insert meaningful information that was not there to begin with? I'm sure someone will enlighten me.

Its about capturing the sound wave in a more naturally rounded form, rather than the stepped staircase, chopped up effect, that you get with 16bit audio.
Of course you're not adding anything to what wasn't there to begin with. You're trying to keep analog more like analog in the digital domain.

Isnt that more about sampling rate than bit depth...? (stepped staircase)

[rastasean]

So if this fellow does want to convert this over to digital, what would be the best way to do this?

[Kevin Straker]

I'm fuzzy on how upsampling can insert meaningful information that was not there to begin with? I'm sure someone will enlighten me.

It is not an up-sample

You are sampling for the first time.. analogue does not have a Bit depth or sample rate.... :-)

A

Thanks, i should have realized that. The analog source will be just like being at a show.

[BlingFree]

So if this fellow does want to convert this over to digital, what would be the best way to do this?

Heres what Menke is doing..

MOTB Release: 0062 24/96
Release Date: 2008-03-07
Band: Grateful Dead
Date: 1979-11-25 (Sunday)
Venue: Pauley Pavilion - University of California
Location: Los Angeles, CA
Analog Audience Source: Master Cassettes [FOB, 60' from stage] (MAC)
Medium Stock Brands: MAC = Maxell UD XL II
Analog Lineage: 2 x Nakamichi CM-700 => Sony TC-D5 >> MAC
Analog Sound Preservation: MAC >> Nakamichi DR-8 => Korg MR-1000 >> DSF [1-bit 5.6448 MHz Stereo] >> Korg MR-1000 => Korg AudioGate >> WAV [24/96]
Taped By: Bob Menke
Transfer By: Bob Menke
Mastering By: Derek McCabe

[DSatz]

Imagine that you have a perfect recording medium that adds no noise whatsoever to any signal that you choose to record. Then imagine that you use this perfect medium to make a copy of an existing recording that has its noise floor at, say, 54 dB below full scale (the 0 level of the recording). Quick: What would be the noise level of the copy that you just made? That's right, it would be 54 dB below full scale.

The same is true if the recording medium is not perfect, but its noise floor is distinctly lower at all frequencies than the noise floor of the recording that you're copying. If the copy is being made on a system with a 95 dB noise floor and the original recording is as above, there will be no difference in the noise floor of the result--you might as well have made the copy with a perfect recording medium.

People, p-p-p-p-p-please let's get this straight: Word length in digital audio--the number of bits per sample--affects ONLY the dynamic range of the recording. You don't get any other improvements in sound quality whatsoever. More bits per sample won't give you "rounder" or less "stairstepped" signals, lower distortion, better "resolution," better sound quality at low levels (apart from the effect that noise always has, just like in analog media). More bits per sample simply makes the noise floor lower. So if you're recording a signal that has a lot more noise than the medium you're recording it onto, there's simply nothing to be gain by increasing the number of bits per sample far past the dynamic range of your source material.

It's 6.02 dB per bit, so you can easily do the math. If this is a very carefully made Dolby "B" cassette recording on pure metal tape (which was newly available in around 1974), 11 bits would be enough to render it. Of course you want the noise floor of the digital transfer to be several dB below the noise floor of the cassette all across the audio frequency range, so by all means go for 12 bits. But don't expect any sonic improvement beyond about 12 bits because by then the noise floor of the digital channel is much lower than the noise floor of the signal you're recording.

I'm sorry but all the references in this thread to "stairsteps" and such are simply wrong and the people who are saying them need to find out how digital audio recording actually works and stop giving advice until they have so learned. There, I've said it--I truly don't mean to antagonize any individuals in this very nice place, but some things are a matter of opinion while others are a matter of knowing how something actually works or doesn't work. Which you only prove that you don't know, if you speak of stairsteps. That metaphor outlived its usefulness in the 1980s. Die, metaphor, die.

--best regards

[sunjan]

It's 6.02 dB per bit, so you can easily do the math. If this is a very carefully made Dolby "B" cassette recording on pure metal tape (which was newly available in around 1974), 11 bits would be enough to render it. Of course you want the noise floor of the digital transfer to be several dB below the noise floor of the cassette all across the audio frequency range, so by all means go for 12 bits. But don't expect any sonic improvement beyond about 12 bits because by then the noise floor of the digital channel is much lower than the noise floor of the signal you're recording.

Thanks for clearing this up, DSatz!
But all that being said, wouldn't you agree that it's still better to do the first A>D transfer at a higher bit rate, allowing more editing in post? And downsample to 12 bits (or 16 bit, whatever) for the final output???
In practice, how would you perform this operation?

/J

[aegert]

Moke +T

I am with you on that... I think that sample rate plays a very big part in sound as well. I really don't want to get in an argument so I won't be detailed about it.. Forum threads usually degrade over this discussion...

From a qualitative perspective 24 bit sounds better... and it affords better editing results.

Higher sample rates make things sound better to me as well.

I think there is a lot lost in the quantitative discussion .. Yes gear designed for certain rates and bits will sound better at them and yes filters for one sound better than at others.. Yes the math plays out differently but I am constantly amazed at the difference I hear...

Now is that a Placebo effect? maybe... But if you believe that saltwater cures cancer and your belief proves to convince your third eye of that truth and your cancer goes away because you believed and healed yourself.... Would you take the placebo or say the math shows something different? LOL

I do feel that the higher and faster you go from an archive perspective gives you the better chance in the near and distant future to revisit the piece and help it out with newer technology.

You can't put Bits or sample in after that fact you will always still have your effective resolution so I do everything as best as I can and damn the torpedos!

Love A

[Belexes]

If this is a very carefully made Dolby "B" cassette recording on pure metal tape (which was newly available in around 1974), 11 bits would be enough to render it.

Sadly, it was a Scotch normal bias cassette.    So even less bits to render that.

Thanks for weighing in on this DSatz.  Something didn't seem right to me about a 24/96 of a very old analog master.

[Brian]

Imagine that you have a perfect recording medium that adds no noise whatsoever to any signal that you choose to record. Then imagine that you use this perfect medium to make a copy of an existing recording that has its noise floor at, say, 54 dB below full scale (the 0 level of the recording). Quick: What would be the noise level of the copy that you just made? That's right, it would be 54 dB below full scale.

The same is true if the recording medium is not perfect, but its noise floor is distinctly lower at all frequencies than the noise floor of the recording that you're copying. If the copy is being made on a system with a 95 dB noise floor and the original recording is as above, there will be no difference in the noise floor of the result--you might as well have made the copy with a perfect recording medium.

People, p-p-p-p-p-please let's get this straight: Word length in digital audio--the number of bits per sample--affects ONLY the dynamic range of the recording. You don't get any other improvements in sound quality whatsoever. More bits per sample won't give you "rounder" or less "stairstepped" signals, lower distortion, better "resolution," better sound quality at low levels (apart from the effect that noise always has, just like in analog media). More bits per sample simply makes the noise floor lower. So if you're recording a signal that has a lot more noise than the medium you're recording it onto, there's simply nothing to be gain by increasing the number of bits per sample far past the dynamic range of your source material.

It's 6.02 dB per bit, so you can easily do the math. If this is a very carefully made Dolby "B" cassette recording on pure metal tape (which was newly available in around 1974), 11 bits would be enough to render it. Of course you want the noise floor of the digital transfer to be several dB below the noise floor of the cassette all across the audio frequency range, so by all means go for 12 bits. But don't expect any sonic improvement beyond about 12 bits because by then the noise floor of the digital channel is much lower than the noise floor of the signal you're recording.

I'm sorry but all the references in this thread to "stairsteps" and such are simply wrong and the people who are saying them need to find out how digital audio recording actually works and stop giving advice until they have so learned. There, I've said it--I truly don't mean to antagonize any individuals in this very nice place, but some things are a matter of opinion while others are a matter of knowing how something actually works or doesn't work. Which you only prove that you don't know, if you speak of stairsteps. That metaphor outlived its usefulness in the 1980s. Die, metaphor, die.

--best regards

so you have nothing against a sampling rate of 192kHz?  i suggested recording at 24bit because storage really isn't an issue these days and most of the gear taper's use have ADC's that perform better at 24bit. but as was pointed out earlier before your post, 16/192 would be sufficient for this application.

plus....if any major digital editing is done(eq, NR, etc etc etc) dithering is generally a great idea to remove any unwanted digital artifacts introduced in the editing process.  that's also why a high sample rate would be useful so any kind of artifacts will be well up into the inaudible supersonic range. now whether there is a advantage to dithering to 16 from 24 over just dithering 16 to 16, that is where my memory becomes foggy.  care to share your thoughts? (without the condescending tone this time? ) 

thanks for your continued input. i always like reading your posts

[jmz93]

Analog is a continuous waveform, limited only by the recording equipment and the storage medium used to capture it. Digital is a
sliced up representation of that waveform. So, I think you would want to capture as much of that analog signal as possible in digital form. 
Also, any processing done after the fact, to the digital version, will benefit greatly from it being in 24 bits.

[rastasean]

So we have the audiophiles who want all their 1974 tapes converted over into digital format and then we have people who still prefer the old analog format that records provide. Pretty funny how that can work and both are considered audiophiles.
Do any of you ever jam out to a 1960s record of the Mama's and Papa's?

[Brian]

So we have the audiophiles who want all their 1974 tapes converted over into digital format and then we have people who still prefer the old analog format that records provide. Pretty funny how that can work and both are considered audiophiles.
Do any of you ever jam out to a 1960s record of the Mama's and Papa's?

in terms of *recording* mediums....

the difference between a cassette and a 2" reel is quite staggering.  i don't think you'll ever hear "audiophiles" claim greatness to cassettes as a recording medium

comparing vinyl and cassette as playback mediums......yeah.....vinyl still wins out by a long shot.

people want to transfer their cassette masters to digital for longer storage.  those tapes don't sound the way they did when recorded onto forever.  once transferred digital...."quality" at the time of the transfer is more or less maintained. maintaining digital playback gear and media is much easier than maintaining proper cassette storage and keeping a cassette deck tuned up to spec.

[Digital Quality]

I thought about this some more last night, there are still some pros to sampling at 24 bits even though the range is easily covered in 16. I know - I'm a flip flopper    There will be less sampling error, so less noise injected for quantization. Also, the playback chain of the 24 bit system probably is nicer sounding - a little easier on the ears for sure and possibly louder. Plus they are your masters, who cares if it's over the top, I say go for it!

Which SACD disc is it that is supposedly just a high res copy of the 16 bit version? I forgot the title but I remember I have it and I like the sound of it better than the CD. edit - unless it was Dark Side of the Moon, that SACD is weak!

[bgalizio]

Which SACD disc is it that is supposedly just a high res copy of the 16 bit version? I forgot the title but I remember I have it and I like the sound of it better than the CD. edit - unless it was Dark Side of the Moon, that SACD is weak!

Was it a Norah Jones album?

[Digital Quality]

Ya that's it http://stereophile.com/thefifthelement/1104fifth/ I still like it a lot but I guess I should try both layers and see.

[Roving Sign]

I've done a bit of this transfer business - so I want to throw this notion out there...

Is there such a thing as "gain scaling"? - (thats my word...I think Im really referring to an effect that normalization causes)

Preface: (my cassettes dont seem to have enough level to hit anything near a full scale digital peak..not sure of thats just my collection or just the way it should be...)

When I make a transfer - I have two choices...


Option A

Set the A/D to unity (no gain added)...play tape...record...and normalize the final result...

OR...

Option B

I can add gain in the analog domain - and get the levels up to full scale digital...

Strangely enough - if I use option A - I almost always think the copy sounds BETTER than the master...(and better than Option B)

Option B tends to get mushy sounding...as if adding gain in the analog domain changes something in an undesirable fashion...(now obviously I've introduced other gear here...)

Now - "sounds better" can mean a lot of things - I tend to think "Option A" makes my downstream gear work better - more gain, less noise...and really more detail (or the perception of such) - not sure the recording has changed much, other than having a more gain...?

If you look at the wave form as just a shape...and then scale that shape up to a bigger scale(ie normalize)...so - its the same shape  - just bigger...thats what option A "sounds like" to me...(sorry if thats a weak analogy)

So - yeah - Im confused with what I hear - and I havent even gotten into the 24 bit question...sorry for the tangent - but I was thinking there was another aspect beyond the 24 bit debate...

What is the favored approach? - add analog gain to full scale digital (no matter the bit dept/sample rate) - or take it unity from the tape and normalize to full scale?

[aegert]

I try to transfer tapes by getting as full a wav as possible that means main peaks at -12 to -6 transients to -3 to -2... Then Let post handle the gain issues. Much as you do but unity is a problem as what the tapes are recorded at usually goes over 0db..



A

[DSatz]

Brian, I'm quite sure that 192 kHz is past the point where one could be getting any audible benefit from the high sampling rate; there comes a point at which the drastically shortened sampling intervals actually hurt conversion accuracy, so that you end up "knowing more and more about less and less."

To my way of thinking the point of diminishing returns might be in the neighborhood of 64 kHz, though of course there's some opinionatin' that goes into choosing any one number like that. But except for special purposes (e.g. Jamie Howarth's system for filtering flutter out of old analog recordings by tracking the residue of the bias frequency), I can't imagine any benefit from sampling rates higher than, say, 96 kHz--and even that seems like overkill, really. The meaningful issue is the audio quality within the human hearing range, not how high an inaudible frequency we can record.

--best regards

[DSatz]

moke, I wanted to say a special word to you, because I thoroughly agree that our ears must be the final arbiters. But here's the thing: Say you have a recording device called "X" which I don't have, and the "X" recorder has a switch for various sampling rates. On your "X" one sampling rate might sound better than another one, and since I don't have an "X" I have no reason to doubt that you hear what you hear.

All I'm saying is that your experience with the equipment you have used up to now, plus the beliefs which you seem to have about how digital audio recording works, might lead you to make general conclusions about sampling rate which are not, in fact, generally correct. But they could very well be correct for the equipment you have, or have heard being used.

So when someone asks what sampling rate and word length they should use for transferring a 35-year-old audio cassette (and by the way I have an entire wall of shelves full of such recordings, since I was a very active taper back then, and I used cassettes when I didn't bring my stereo Nagra with me to certain places, including hundreds of recordings I made in Europe between 1972 and 1978), I'm answering from the standpoint of what the better equipment that's commonly available today will do in general--not your "X" recorder which I don't know anything about. In general, 24-bit recording really makes no sense for sources that have such a narrow dynamic range to begin with.

You see, in addition to the documentary and "on the go" recording that I did with cassettes, I also recorded classical concerts (everything from baroque chamber music to full orchestras and percussion ensembles) and in the analog era, I often used 15 ips Dolby "A" or Telefunken's even more powerful c4 noise reduction system--and 16-bit PCM outdoes both of those by a considerable margin. So you just can't tell me that it takes 16 bits to copy a cassette without adding noise, let alone 24 bits; both the math and my years of professional experience in that area are very clear about that.

Again, your equipment might have different sound quality when you run it at 24 bits as opposed to 16 bits; I'm not there in the room with you so I don't know. But properly designed, properly functioning digital recording equipment doesn't--it only has a lower noise floor in the 24-bit mode.

I just wanted to make it very clear that this isn't personal--or actually, of course it's personal because music is personal and recording is personal to us--but I mean, I'm not trying to say that I'm better than you! Just that there aren't stairsteps in digital audio recordings (at least not in the sound that comes out when you play them back), so you don't need to cover them up by throwing more bits at them than you really need for the dynamic range you require.

--best regards

[aegert]

Moke I agree with you full heartedly

I have Nak dr-1's 2pc... A dragon

V3's, Benchmark adc-1, and a mytek 192

I have a 722, hd-p2 and a korg-mr1000

I listen through a benchmark dac-1 and for a time a lynx studio twoB

At every turn in any combo I experience the same things

Higher sample and BIT win out every time either on reference speakers or in the ultrasone Pro750 cans

Funny that the cassettes sound even better ANALOG is king... It plays out in my mind in only one way.. The closer you get to the source in representation the more it will sound like the source...
LOL

A

[Petrus]

You have my deepest symphaty DSAtz. Against ignorance even the Gods themselves fight agains in vain.

Way I see it: 16/44 has in many tests proven to be undistinguishable from live analog feed, or from so called hi-resolution formats. Remember the recent AES SACD versus CD test for example. Old cassettes have bad noise levels, bad freqency range, were recorded with so-so microphones. Now some people think that to copy these ancient tapes they need resolutions even modern mic/preamp/ADC systems can not fully ulilize for live recording. Are the tapes really BETTER than live sound today? Something awfull must have happend to the air!

By the way, I have the 16/44 versus 24/96 test file availabe at http://hosted.filefront.com/Jullepoika/

With that you can test if your ears and systems can make any difference between a hi-resolution format and plain old CD quality. Download also the explanation TXT file.

[digifish_music]

You have my deepest symphaty DSAtz. Against ignorance even the Gods themselves fight agains in vain.

Way I see it: 16/44 has in many tests proven to be undistinguishable from live analog feed, or from so called hi-resolution formats. Remember the recent AES SACD versus CD test for example. Old cassettes have bad noise levels, bad freqency range, were recorded with so-so microphones. Now some people think that to copy these ancient tapes they need resolutions even modern mic/preamp/ADC systems can not fully ulilize for live recording. Are the tapes really BETTER than live sound today? Something awfull must have happend to the air!

By the way, I have the 16/44 versus 24/96 test file availabe at http://hosted.filefront.com/Jullepoika/

With that you can test if your ears and systems can make any difference between a hi-resolution format and plain old CD quality. Download also the explanation TXT file.

I have been closely following the 16 bit 44.1 kHz vs all other formats for many years. I came across a paper by David Griesinger (Lexicon DSP engineer) and this lead me to his site, very interesting (although cluttered) ...

http://world.std.com/~griesngr/

There's a lot of text...but 1/2 way down...

"And now for something completely different... Being currently over 60, and having in my youth studied information theory, I have a low tolerance for claims that "high definition" recording is anything but a marketing gimmick. I keep, like the Great Randi, trying to find a way to prove it. Well, I got the idea that maybe some of the presumably positive results on the audibility of frequencies above 18000Hz were due to intermodulation distortion, that would covert energy in the ultrasonic range into sonic frequencies. So I started measuring loudspeakers for distortion of different types - and looking at the HF content of current disks. The result is the paper below, which is a HOOT! Anytime you want a good laugh, take a read.

Slides from the AES convention in Banff on intermodulation distortion in loudspeakers and its relationship to "high definition" audio.

http://world.std.com/~griesngr/intermod.ppt

Conclusions -

1. Adding ultrasonics to a recording technique does NOT improve time resolution of typical signals – either for imaging or precision of tempo. The presumption that it does is based on a misunderstanding of both information theory and human physiology.

2. Karou and Shogo have shown that ultrasonic harmonics of a 2kHz signal are NOT audible in the absence of external (non-human) intermodulation distortion. (This BTW: means they can't be heard in the real world and that filtering them from the recording is a good thing as they can only do harm).

3. Their experiments put a limit on the possibility that a physiological non-linearity can make ultrasonic harmonics perceptible. They find that such a non-linearity does not exist at ultrasonic sound pressure levels below 80dB.

4. All commercial recordings tested by the author as of 6/1/03 contained either no ultrasonic information, or ultrasonic harmonics at levels more than 40dB below the fundamentals.

5. Our experiments suggest that the most important source of audible intermodulation for ultrasonics is the electronics, not in the transducers.
Some consumer grade equipment makes a tacit admission of the inaudibility of frequencies above 22kHz by simply not reproducing them. Yet the advertising for these products claims the benefits of “higher resolution.”

6. Even assuming ultrasonics are audible, loudspeaker directivity creates an unusually tiny sweet spot, both horizontally and vertically.

7. A/B blind listening remains the gold standard for audio comparison.

digifish

[Digital Quality]

Forgive me for reviving a dead thread. Did anybody else catch the interview of the guys from http://www.plangentprocesses.com/ on the GD channel last Sunday?? In 50 words or less; they are restoring audio from tape to correct for stretching, as well as limited wow and flutter capacities of analog tape recorders. One part of making an analog tape recording includes AC bias (typically about 120kHz). The machine runs a bias signal to the tape right before it lays down the signal you want to record. The bias signal helps "loosen up" the tape "magnets" so you get a better SNR with the desired signal. Well, this bias signal was generated with a very accurate oscillator, much more accurate than the wow and flutter specs of the recorder, and the plangent processes folks do the job of recovering the bias signal. Then they run the desired signal and recovered bias through a digital process that speeds and slows the music down based on errors in the bias frequency. In drastic cases where the tape has really stretched, this corrects warbling and pitch shift so they can fix up old movies, but even when the source is still relatively clean it can make a big difference in terms of depth and soundstage.

That really got me thinking about this discussion. These guys are a one-off boutique service right now but who knows what will happen in the future. It's entirely possible that this sort of thing will show up in the high end tools some day. If you archive at a high enough frequency and resolution _maybe_ you will be able to recover the bias later but if you stick to what is needed just for the audio portion, you'll be hosed. Anyway, I didn't want to fire up the whole debate again but it was such an interesting interview. I didn't check out the NPR interview on their web page yet but I bet that's pretty cool too.

[DSatz]

Steam Powered, in my reply about five messages up from here, I mentioned Jamie Howarth--Plangent is his project. He's doing amazing work and as I said, that's the only kind of reason I can think of to preserve inaudible high frequencies in a recording. Otherwise, they're nothing but trouble, and the only real question is how to get rid of them with the least harm to the audible signals.

--best regards