Why upload bootlegs in high-resolution?

[ILuvHoney]

There's one thing here that I have wondered a bit about. I see that many of the bootlegs posted here are posted in high-resolution files (e.g. 24/96). I wonder whether there is any point at all with doing this, when taking the source material into account?

Of course there are those who claim that high-resolution is unnecessary anyway, as the limits of human hearing doesn't exceed CD resolution. This point, although interesting, is not the one I am trying to make here. There are of course many people who disagree with those who claim that high-resolution is pointless, who swear that they can tell the difference and argue that the filtering process affects frequencies within the audible range as well. I¨ve also heard that some studies indicate that people get more concentrated about the music and that people enjoy it more when listening to music in resolution up to 96 KHz, even if people can't hear those high frequencies directly.

Therefore, I am not going to argue that high-resolution itself is pointless in all cases; if you transfered the master tape of Dark Side of the Moon to digital, for instance, it would probably make sense to transfer in high resolution. Yet, most of the stuff we have here is sourced from cassette tapes (in the case of audience recordings often recorded on portables) and FM broadcasts etc. Is there really much information in the signal over 22 KHz, and at least over 24 KHz, on any bootleg?

As far as I know, there is no way a cassette tape can have a frequency response of 48 KHz (which makes me see 96 KHz sampling rate useless), and actually very few cases in which it will achieve a frequency response over 22 KHz. According to Hifiengine, the frequency response of the Nakamichi Dragon (by many considered the best cassette deck ever) reaches 22 KHz using metal tape - virtually identical to CD resolution. A metal tape recorded on a Nak Dragon is pretty much what we can call ideal conditions for cassette recording, reaching the absolute maximum achievable with the medium, yet even under such ideal conditions, the tape will not be able to store more information than a digital sample rate of 44 KHz. To be fair, much vintage equipment was conservatively rated in terms of specs, so it's possible that the Dragon would be able to record slightly above 22 KHz on metal tape. Yet, I think it's fair to say that at least if you increase sampling rate to 48 KHz (frequency response up to 24 KHz), that will be enough to capture the whole signal captured by a cassette tape recorded even under such ideal conditions. Anything above that will be noise and hiss only. The question then is; why would anyone use a sampling rate of 96 KHz (frequency response up to 48 KHz) in order to transfer a recording from a format which even under ideal conditions can only store half of that?

And my argument does not end there. I am pretty sure that nobody ever brought a high-end cassette deck to a concert in order to record it; it would have been extremely unpractical, it would need an 230V power supply during the whole gig; in order to do this, the bootlegger would have to bring a 230V battery (or a lower voltage battery and a transformer). I am pretty sure that this never ever happened; the equipment would almost certainly have been confiscated upon entry, in addition to the inconvinience bringing this heavy equipment into a concert alone would entail. Audience recordings were pretty much always recorded on portable cassette recorders. Even high-end walkmans often had a frequency response of only approximately 15 KHz, meaning that 44 KHz sampling rate should be more than sufficient to capture it all. The Nak 550 - a high-end recorder known for its use by prominent bootlegger "Mike the Mic" among others - has a frequency response of 17 KHz, still well below CD resolution. As far as I know, there were never any portable cassette recorders able to record over 20 KHz, making the use of higher sample rates for audience recordings, as far as I can see, pretty pointless.

In addition, most of the recordings available on R&D and the Y! archive are not masters but 1st, 2nd, 3rd generation or worse. Copying of cassette tapes entails generation loss, which affects frequency response. If the original master cassette had a frequency response up to 17 KHz, for instance, it's likely that the third generation tape has no data at all except noise over 10 KHz. Furthermore, cassette tapes tend to lose audio quality with age if not stored well; in addition they get worn during playback, and it's likely that many of the tapes transfered had even been recorded over several times. Usually, this loss of quality with age and wear leads to a loss of the high frequencies, decreasing frequency response further, and the tapes transfered here on Y! were most often 20-50 years old upon transfer, meaning that they likely had lost some of the high frequencies that were originally recorded on them.

In addition to the quality of tapes and tape recorders, the mics might also be an issue. As 20 KHz is the generally accepted limit of human hearing, even most live mics usually are not able to capture frequencies much in excess of this. The mics that were plugged to the aforementioned tape recorders were likely usually smaller mics who were more convenient to bring in, and I assume most of these mics had a lower frequency response considerably below that. This means that even if some portable cassette recorders might approach a 20 KHz frequency response under ideal conditions, what actually was recorded was likely lower.

To put it shortly: I am pretty sure that pretty much every audience recording on cassette has a lower frequency response than CD resolution. I therefore find it highly unlikely that any of the high-resolution transfers of audience recordings on Y! or R&D contain any more information than a CD-resolution transfer would, apart from probably a little bit more noise.

When it comes to soundboards, these are most often sourced from radio taping. FM broadcasts, even if they might sound very well, have a frequency response significantly below CD resolution. FM broadcasting contains a 19 KHz pilot tone, and the highest frequencies of what is broadcasted is filtered out in order to not disturb the pilot signal. Usually, this limit is around 15 KHz. This means, as far as I can see, that on any recording from FM radio, the use of high-resolution files is absolutely pointless, as the actual frequency response of any FM recording will be far below CD resolution anyway. If any difference is heard between a 44 and a 96 KHz sampling rate of an FM transfer, the difference is noise and nothing else. In addition, most of these FM broadcasts were probably recorded on cassette, which itself might have reduced the frequency response even further, and usually with an MPX filter enabled, which would eliminate all the frequencies in excess of 15 KHz (including pilot tone) anyway, and probably even lower if the filter was not working perfectly. The question again is: why use high resolution for a transfer of an FM broadcast which 100% certainly contains no music above 15 KHz, which is far below CD resolution anyway?

Vinyl, on the other hand, theoretically has a higher frequency response than CD; from what I've heard, it can at least theoretically capture frequencies up to around 50 Khz (equivalent to a 100 KHz sampling rate); hence a 96 KHz sampling rate might theoretically make sense for transfering vinyl. However, that is pretty much theoretical; even officially produced studio LPs rarely have any frequencies even approaching that. In addition, very few cartridges are able to reproduce any sound even near 48 KHz, as are most amps; so even if there's information that high stored on the record itself, it's unlikely that you can reproduce it as every aspect of your system would need to be able to reproduce them. As far as I can see, if your amp or cartridge has a frequency response up to 22 KHz, using a sample rate over 44 KHz wouldn't make sense at all. When it comes to bootleg LPs, I assume that the source material they were pressed from were usually audience recordings done on cassette, or cassette recordings of radio broadcasts. This means that they have the same limitations as I noted above with recording equipment, cassettes and FM radio. Hence, even in the case of bootleg LPs, I can't see that 96 KHz transfer makes any sense at all. Probably 48 KHz in some cases, but in most cases I assume that redbook would be sufficient to capture all the information on a bootleg LP.

Note that I am talking about sample rates only here; bitrates are a different thing which I have little understanding of, so maybe using 24 instead of 16 might make sense, but I can't see that using high sample rates really makes sense for bootlegs due to the lower frequency response of the source material.

I don't mean to put anyone down for using high resolution for transfering their boots, but I am interested in why you are doing this, because at least from my theoretical perspective, I can't see that it makes any sense. If you have good reasons for thinking otherwise, I would be very glad to hear those thoughts.

[Skinny12]

I don’t know about the equipment used. But there one very strong argument for high resolutions: If the transferred audio would or should be remastered in any way one should do it at the highest possible resolution not to loose any audio down the way. So even though the information might not be on the original recording and in the end your ears might not respond to it, i might be careless not to use every possible resolution on your transfer. At least when the sources are analog.

And do not forget: It’s also about archiving. If you got this one transfer, you never know about the possibilities coming in a few yrs. Would be a bummer to sacrifice any opportunities due to a low res transfer.

Best
Nils

[MrFender]

(Skinny beat me to some of my arguments)

There are endless "discussions" (to put it lightly) about this topic over at TapersSection.com. Many who have argued your same point and only record maybe at 24/44 or 24/48 these days. I'm not saying you are wrong one bit but my argument is as simple as "Why not?" (Disclaimer: I only know enough recording tech to get myself in trouble)

An argument used to be made on storage or processor limitations to deal with the larger files but those days are now past.

We now have recording decks that will do 24/192 so they would be even farther past the hearing thresholds of humans. What's the point? Could the audio engineers (and tapers) back in the 1960s & 1970s have conceived of such resolution when they were recording back then? Maybe but probably not. They were going for the best and highest resolution that was available to them which is what I do. Or try to do.

I say record as high as you can in order to capture every scrap of sound/information you can and maybe someday in the future, as technology progresses, those generations still listening to music will thank us for it as there was something extra in that 24/96 recording. When they have to "bake" that DVD, Blu-Ray, SD card or hard drive, maybe they'll be glad we did. The cost of doing so isn't much more than only recording at 16/44.

Not very technical but it's just how I feel.

[Chorale]

I would generally agree, but...

There are certainly a couple of worthwhile advantages of 48 kHz over 44.1 kHz. A lot of low-cost digital audio devices run natively at 48 kHz and have to do sample rate conversion to play 44.1 kHz material, so using 48 kHz avoids a resampling step on playback for many users. And in practice you don't want to be working with frequencies very close to the Nyquist limit, because the ADC needs to have an analogue filter in front of it to remove frequencies that can't be represented (plus some digital operations you might want to do while editing will also involve filtering), and the behaviour of filters around the cutoff point will always be a bit of a compromise by nature - so 48 kHz means that you can represent more of the typical human hearing range without getting close to the Nyquist limit.

The same two arguments work for 96 kHz to some extent. 96 kHz to 48 kHz downconversion is just decimation (LPF then take every other sample), so it's easy to implement efficiently without introducing artefacts. And 96 kHz is certainly overkill for the frequencies we can hear in Nyquist terms - but those frequencies are then very definitely in the region where filters will be well-behaved! It's also worth bearing in mind that we're always distributing losslessly-compressed audio here anyway, so increasing the sample rate doesn't actually increase the file size as much as you might think (the residual left after FLAC's prediction ends up being spread out over more samples, but still contains the same amount of information).

Regarding 24-bit samples vs. 16-bit samples, the difference is in terms of dynamic range rather than frequency response - you can represent a wider range of volume levels accurately. 16 bits is generally plenty for material that's been mastered, but having a few dB of extra headroom is very useful when you're dealing with raw recordings that may need some level adjustment later without losing precision. The sample depth does make more of a difference to the file size with lossless encoding since it's capturing more random noise (i.e. there's more information to represent in the residual).

[MrFender]

For sure some of it is about ego or *ahem* "manhood" size. It's like someone that blows $100k on a stereo system that is not about actually listening to the music, it's the "I have so-and-so speakers or amps so I am better than you". Which ones sells better, the puny CD quality music at 24/48 or the High Resolution/Blu-Ray/4k level of 24/96? (I have Endless River in both! Envy me!)

[Skinny12]

Funnily enough the (L)PCM thingy is almost exactly what i studied back in the late 90s when the first mobile phones came up. It was called “Kommunikationsforschung“ and our professor explained it quite nicely. „Never give up information you have, ever. Only once you do not need it anymore or you are limited somehow, break it down. Resolution is king.

Best Nils

[Son of Nothing]

Vinyl, on the other hand, theoretically has a higher frequency response than CD; from what I've heard, it can at least theoretically capture frequencies up to around 50 Khz (equivalent to a 100 KHz sampling rate);

Interesting claim about vinyl. I've heard the opposite, from a musician who had experience with the compromises made when cutting his digitally recorded master onto vinyl. IIRC his calculation was that a vinyl album only had about 12-bits worth of dynamic range due to limitations of the groove size. Not sure about the frequency response. 12 inch 45 rpm has better dynamic range because you're not trying to cram 20+ minutes of signal onto a side.

[ILuvHoney]

Dynamic range and frequency response is not the same thing. Digital has a superior dynamic range to vinyl, but in practice it does not mean that much because vinyl's dynamic range is sufficient to capture all the dynamics of the vast majority of albums anyway. Furthermore, since the digital era, music producers have unfortunately not utilized digital's superior dynamic range, but have instead used digital compression to decrease the dynamic range of recordings in order to increase overall volume so that they should sound better on low-end equipment, so the dynamic range of releases from the digital era are actually overall much lower than those from the analog era. This phenomenon is known as the loudness wars. Actually, some albums from the 1990s onwards actually have higher dynamic range on vinyl than on CD, because the volume limitations on vinyl make it more or less meaningless to squash it too much. Yet, it's true that digital has the potential to reproduce a much higher dynamic range than vinyl, but this is not as significant when the vast majority of music recordings - new or old - have too low dynamic range for this to really be a big issue.

The frequency response of vinyl is at least theoretically much higher than CD, and indeed higher than on any other mainstream consumer audio format. But while vinyl can theoretically reproduce frequencies up to 50 KHz under ideal conditions, this does not mean that it is able to accurately reproduce everything that is in an entire signal up to that point, it just means that it might be able to store some amount of signal up to that point. Because vinyl is an analog format, it does not have a straight cutoff point where everything is filtered out above that point like digital has, it's more a long range where it gradually fades out up to the point where there is no longer any signal at all. And of course, the actual frequency response also depends on the frequency response on the equipment (mics, amps, tape machine, mixing board etc) used to perform, record mix and master the music on the record, the quality of the pressing equipment used to press the record, length of the record etc. For instance, if the equipment in just one of those stages could only reproduce frequencies up to 20 KHz, the record would not contain any music above 20 KHz either. Furthermore, even if the record itself contains very high frequencies, they cannot be reproduced unless your equipment (cartridge, preamp, amp, speakers) can also reproduce it, so it depends on a lot of factors whether this potentially superior dynamic range is actually utilized, but a well recorded and mastered record played back on a high-end system should at least be able to reproduce frequencies somewhat higher than CD's limit.

As anything above 20 KHz is considered above the limits of human hearing anyway (and CD's frequency response reaches 22 KHz), it might not be that much of an issue; however, many argue that even though the cutoff is above the audible range, the hard cutoff filtering of digital sources affect the audible frequencies, and it's also argued that some supersonic frequencies can be felt if not heard, thus making the smoother, more gradual fading of very high frequencies on analog vinyl more pleasant to listen to than digital's hard cutoff, and this is one of the arguments that is frequently used in the digital vs analog/vinyl vs CD debate.

[YASHA]

https://ravinganddrooling.com/forum/...-sound-quality

[daemonspudguy]

There's another, really big argument as to why digitizing bootlegs in high definition is especially a good idea, and one I'm surprised no one has brought up until now: analog and digital audio storage are nothing alike, and things like resolution don't easily map 1-to-1. It's probably better to digitize at a resolution higher than the analog source is, to account for any differences that might arise from the fundamentally different ways analog and digital storage works.

[Nodak]

I was of the opinion that crappy bootlegs are even more in need of high resolution captures than a good sounding bootleg.

I cant remember which recording it was or even when it was but there was a thread on Y that showed how much more detail one could hear in a 24 bit capture than a 16 bit one. The difference was apparent. The way it was explained was that all the noise in the recording was fighting for the same space as the music and only so much will fit. It might sound absurd but I absolutely remember the difference. I wish I could remember which thread it was.

Now If you don't have a great stereo you're probably not going to hear that difference, but I do and have sat people down for A/B test of 16 vs 24 bit recordings. The difference doesn't show up everywhere but with recording with a lot of dynamic range the SMALL difference is there.

[YASHA]

Yes, there was such a topic with links to audio tests, etc. If I find it preserved in the archive, I will put a link here.

[ILuvHoney]

Maybe indeed 24 bit vs 16 bit might make a difference, but with a recording with no signal above 15 KHz, I can't see any reason why there should be any point in using a sample rate higher than 44KHz. The only difference higher sample rates makes is higher frequency response. And I can't see any reason that a high sample rate can make a difference if the peak frequency of the recording is far below the limits of CD resolution (that is, 22KHz, or a sample rate of 44 KHz). On the vast majority of bootlegs, there is no sound above 22KHz anyway, as I made clear in my opening post, so using a higher sample rate seems like overkill as far as I can see, as the frequencies filtered out using CD standard 44 KHz sample rate contain no sound anyway, except probably a bit of noise that we don't want anyway.

[vince666]

OMG! Reading this sort of brings me back like to 15 yrs ago at Y!

I can understand people complaining about HiRes back then, when HD storage space was more of a problem than nowadays and DSL connections were slower.... but now?
Why complaining, if capturing analog sources at the lowest decent digital quality (which is 24bit 96khz) became the standard?
Why complaining if most of us struggle to get the most from our captures and try to offer the same HiRes to everybody?

I would understand people complaining when we'd give them lower resolution, not higher resolution.

Just for the record... thinking about the numbers (i.e. frequency responses and such) is pointless... it's all about accuracy and accuracy at CD quality is poor, especially if the source sound isn't a professional studio recording.
The more we have "fragile" sound quality in the original recording, the more Hi-Res is a good choice to minimize further losses.

Moreover, if you go and process the captured sound, Hi Res source material is a godsend, compared to CD or DAT quality.

I'd say that, unfortunately, we still have some good sources which were never captured in Hi Res... that's the true problem.

So, please, do yourself a favour and be HAPPY that Hi Res became the standard just since many years ago, at least here in our PF community.

[ILuvHoney]

I had to stop reading even at your first sentence. Let me ask, who is complaining here? It's not me at least, it's rather you who are complaining because I am even asking this question. I am just asking a question to find information, yet certain people decide to attack me rather than give serious answers!
Because that's what you're doing here. You do not even make an attempt to give an answer to my question, which is asked as a serious question, and with all respect to all those who are choosing to upload in high-res.
You're not at all even trying to adress my main point here, namely that most of the material on old bootlegs (whether audience recording done with portable cassette recorders, or FM radio recordings) have a frequency response lower than 22 KHz. What I am asking respectfully is simply what the point is with transfering recordings with far below 22KHz frequency response to much higher sample rates is.
And no, bringing in frequency response is not "pointless" - sample rates do not affect anything else than frequency response, actually, it's irrelevant to anything else. The frequency response is equal to half the sample rate. So a 96KHz file has a max frequency response of 48 KHz. What I am asking here is simply what the point is with uploading files in a sample rate far exceeding the frequency response of the original recording. And, to repeat it once again and more explicitly: there is only one person in this thread who has complained about anything, and that is you.

[vince666]

a couple examples...
the vinyl rips of Pictures of Pink Floyd, Big Pink and Waters Gate i shared a couple years ago were captured at 24bit 192khz...
And, while working on them (i.e. declicking or the crazy speed correction at AHM on Pictures of PF) it was just easy to notice how the sound got less affected by the restoration tools at 192khz... it was even better than working on it at "only" 96khz.
Btw, after finishing my work, I downsampled to 96khz which was just sufficient for listening purposes while getting the impression of listening directly to analog.
Maybe I also shared the 16/44 versions of those... don't remember exactly about those 16/44 versions because I don't even keep them into any serious consideration... but, hey, I am sure somebody downloaded the 16/44 while ignoring the 24/96...

from next time and on, I will totally avoid making any 44/16 versions of any Hi Res transfers I'll be working on... this way, who thinks the resolution is too high then he will avoid downloading it at all... LOL!

PS : vinyl, R2R or cassettes sources.... they all benefit from Hi Res transfers, just because Hi Res sounds better and more natural/lifelike as a digital format in itself... also because certain typical drawbacks/artifacts of digital sound will be kept well above the useful sound and the "hearable frequency range".
Also, not all soundcards/converters sound the same... many do make weird things when pushed close to their limits in both audio frequency and level.... hey, did you ever attach an analog oscilloscope at the output of a consumer soundcard which cannot do 96khz while playing test tone sinewaves at more than 3-4 kiloherts? (not to mention if you go over 10khz).
Please, try it and then tell me what you see there in the oscilloscope screen! certainly NOT a clean sinewave... now, go figure what happens with music, which is a complex signal with more harmonics and much higher frequencies than a 4 khz sinewave... distortion and aliasing would be the least you'd get... and that's only the playback side of most digital systems out there..


Being someone who made Hi-Res transfers (etc.) for 20+ years, I'd say I do know what I am talking about... don't I?

[vince666]

ILuvHoney : no worries, I also didn't bother to read more than the first couple lines of your opening message, and also your latest one.... but, the way you started this thread sounded negative to me (unless, maybe, I go and read your extremely long opening post but without losing myself into it? )

anyways, i've actually explained a few things... especially some which come from my own direct experiences (read the point about the oscilloscope and the sinewave tones, for example) and which you won't easily find on the web...
but it would need from you to just waste some more time to read, of course.

[vince666]

anyways, it's night here...

tomorrow, i will explain you several more things about why digital sound, especially at normal resolution, simply sucks.

it's a promise.

[Skinny12]

Maybe just maybe we close that thread soon. All is said. Again and again and again. There is nothing to gain and learn anymore. And to the creator of this thread: please stop testing this community. You got many great answers but you seem just not to accept them.

[vince666]

that said, he might be right about that he was not complaining (will know it for sure after reading the long opening post with care, though, but all of his points seem to go against HiRes).

for the moment, another (short) answer about why Hi Res is better...

Hi Res is a better choice both for the software (the audio data) and for the hardware (A/D and D/A converters), no question asked.