That's just amazing. Last year I realized you can still download winamp (they ... freed it??) and it has that bizarrely good milkdrop 2 visualizer: https://www.youtube.com/watch?v=EO2YkCvOhNE - it's mindblowingly interesting.
I had a TON of classic Jazz 78s from my grandfather - unfortunately I couldnt find a 78 player at the time for the life of me - and I gave them to a friend who was a really accomplished Jazz drummer in SF - one of my biggest regrets - but they were so heavy and I didnt have space for them when I moved...
It's my understanding that it's more than just the speed. 78's will destroy the needle of a regular record player because they're made of very hard shellac, and not vinyl.
The trouble with digitizing analog mediums (especially vinyl) is how much your equipment can alter the sound. Given a different tonearm, cartridge, amp, etc, the same record can sound very different.
I’ve had some luck digitizing some of my collection with a DP-450 [0] and Ortofon Blue [1], but had higher end stuff end up sounding less interesting to my ears.
This is the trouble with audio in general: people not only prefer different sound profiles but also perceive sound differently, which makes it difficult to get rips (at least in my limited experience).
There is a lot of sibilance in the recordings I sampled from the OP link, whoever did the digitization would have benefitted from cleaner records, probably an anti-static brush, a fresh cartridge, and possibly some post-processing since they’re already digital. It’s better than nothing, of course, but it could be a lot better.
You might be interested to dig deeper around the archive and see what they're doing. Each record is transferred using four different styli, and then an engineer at GB picks which recording to use as the example - but all four are in the archive.
> This is the trouble with audio in general: people not only prefer different sound profiles but also perceive sound differently, which makes it difficult to get rips (at least in my limited experience).
The problem with worrying about this is that even the same person perceives sound differently situationally and is simply not repeatable. In most cases when someone hears a difference it is often just a change in their perception from one moment to the next.. and that’s even when something measurable has changed! The bass sounds better because of that new amp.. well no.. you just happened to focus on it more the second time around.
Longer term people’s physiology changes as well.
Only repeat, blinded A-B testing can clearly elicit an objective difference that is most likely not due to these perceptual inconsistencies.
What does this mean.. record it with decent equipment that captures as much useful raw information as possible (and yes 96khz is ridiculous as we’re not bats). People can EQ and mix to their preferences any given day.
It’s not clear there was something wrong with the cartridge. These are old 78 records.. they weren’t mastered with much to begin with and the useful fidelity in them is limited. If you want you can post-process them with whatever fancy shit, but I wouldn’t immediately assume the recordings weren’t the best given what they were working with for source material.
> The trouble with digitizing analog mediums (especially vinyl) is how much your equipment can alter the sound. Given a different tonearm, cartridge, amp, etc, the same record can sound very different.
Would a scan done with a laser turntable[1] be better? The wikipedia article mentions that it produces a much better capture than regular tonearms, although they are much more susceptible to dust.
I was wondering the same thing! For the dust problem, I’d like to imagine that the machines could be modified with an air pump to blow away anything ahead of the pickup. One probably wouldn’t want that extra noise in a listening room, but for archiving, it wouldn’t matter.
I guess when you can get a ‘good enough’ traditional turntable for a couple hundred bucks but an ELP laser turntable costs over $14,000[0] the economics just don’t make a lot of sense. I wonder how much the IRENE systems cost.
It's the equalization that had the kids thinking 78 was so obsolete & bad sounding compared to our 45's and the upcoming 33's back in the early 1960's.
Naturally it was all the fault of the RIAA.
By that time almost all new record players had the recent RIAA standard equalization built in, by compensation in the audio playback circuit based on the published RIAA engineering standard which had emerged.
Otherwise the 45's & 33's would have sounded nasty themselves since way less bass frequencies could be represented mechanically by the analog carving of the signal into the _microgrooves_ of the more-compact discs and not as many highs could be recovered at the slower playing speeds.
Compared to 78's which had plenty of space between the grooves since there was only one tune per side.
When they first tried to track smaller grooves more closely to get more tunes on a disc, the playback needle would jump out of the groove if more bass frequencies were not removed relative to 78's, before cutting the groove.
So almost all vintage record players which have speeds for 33, 45 & 78 [which is almost all of them] play everything through the RIAA circuit, so the 33 & 45 sound correctly equalized but the 78's are miserable because of the EQ even though they are turning at the right speed.
So 78's always sounded lousy since the beginning for me and everyone younger, even including a number of years older, and I was already a critical listener before The Beatles started making records.
78's had never been able to contain the full analog level of bass either, so their equalization had been engineered by each record company as best they could based on their advances in materials & electronics as the 78's were developed and had become popular. But there was no real standard and each record company made occasional improvements over the decades of 78 dominance.
To begin with a record album was a boxed set of these single-tune-per-side 78 records.
Then came the RIAA and an industry-agreed standard for equalization so popular music could continue to be distributed in audio form at prevailing prices on 45's & 33's, but at a cost of pressing closer to sheet music than 78 albums.
Plus there was going to be stereophonic sound with each channel on either side of the microgroove, fully compatible with monophonic by nature. Stereo 78's were never going be an actual thing, they had always been mono.
After RIAA, single 45's replaced the single 78, and 33's replaced the multi-78 albums with a single full-size disc.
Anyway by the time the 1990's came around I decided to do something about the EQ thing.
Firstly studying RIAA compensation using plenty of vintage 33 & 45 vinyl, as well as numerous other-peoples-circuits, then fully pursuing the NIH approach as if it had never been done before.
Numerous AM/FM/Turntable/8-Track combo players were being finally discarded, and they had come from various price-points in design & manufacturing, some very expensive consoles. A few times I was able to use a Dremel to simply saw out the audio portion of the PCB for testing, then later trim out only the RIAA sub-section for further testing in an otherwise all-reference rig.
Music CD's were already more popular than ever, so next had to design a flat analog circuit having less noise than a CD can reproduce. Could then fully confirm the best results would not be obtained using CD versions as references.
Then adding only passive components made my RIAA curve out of it.
Simply eating away at my excess gain & headroom willy-nilly compared to what I had seen and it came out great. Some EE's will probably relate to the math involved to select RIAA component values. The proof is in the testing. Much more testing than math & soldering.
Then came the 78's and no more math, just soldering & testing.
Sure had gotten a lot of pattern recognition done by then though, this was by design.
It was assumed the RIAA curve had been adapted from previous but disparate 78 EQ curves, so that was attempted to be reversed.
The best references were the ones by Elvis, Bing Crosby, Chuck Berry and a few other very popular artists who had simultaineously released new singles on 45 & 78 even after everyone knew 78's wouldn't be available for most releases after a number more years.
Some of the tunes could also be heard in vintage movies where they had sounded as intended, rather than ugly like 78's played on a late 1950's 60's or 70's turntable always did.
With numerous classical from US and Europe, some jazz, show tunes, pipe organ, and of course the 1812 Overture to go along with the pop music, another trend thought necessary was to make sure the Deccas came out good at the same time as the Capitols since they were often more numerous at the time.
During this reiterative process the best styluses could be selected.
When all was said & done what made people's jaw drop was The Tennessee Waltz by Les Paul & Mary Ford.
This was so popular a hit that the previous owner had obviously left it on their heavy-arm original 78 player for at least hundreds of plays, collecting little dust between plays but still the total amount of dust expected over that period of time. The music was about halfway scratched out of the groove and it was ugly :-(
Apparently they had never played the other side of the record though, it didn't have any words to the song anyway and you couldn't even dance to it at all. Record companies would never put a popular song (or even ones having popular potential) on both sides of the record, and Paul & Ford had already been on the air a lot. Nope, this was just Les Paul solo picking on some experimental electronic techniques at the time, using his own handmade equipment like nobody else had either.
This was also the harder phenolic pre-vinyl, vintage vinyl never sounded that good, even through the audiophile excesses of the '70's.
> 78's which had plenty of space between the grooves since there was only one tune per side.
Hence the appearance of the 12" 45 rpm disco single in the 1970's, which sounded a heluva lot better than the 45, especially in the base.
It was even better to score the "promotional copies" which had no bubbles in the plastic or other flaws. I don't know if the PCs were pressed on a separate line or were just hand -picked from the normal production run, but they sure sounded good.
The most high-fidelity record I know of is the direct-to-disc recording of Carlos Santana and Paco de Lucia in acoustic duets.
Unlike the others in the direct-to-disc series, there was supposedly no console between the artists and the input to the master record-cutting stylus, and no remixing after recording.
The difference was quite apparent, and I have always been convinced that the fewer electronic components between the artist and the listener would yield the best reproduction.
I've never even heard a simply amplified PA system sound as good as the acoustic instrument the mike is attached to. This is with high end pro equipment, not home audio stuff.
Literally the only advantage of a laser turntable is that they don't wear out the disc with repeated plays. They are worse in literally every other respect, especially sound quality, since they have no method of pushing dirt and dust out of the groove, so the record has to be perfectly clean, which is essentially impossible.
>They are worse in literally every other respect, especially sound quality, since they have no method of pushing dirt and dust out of the groove
How big of an issue is this? I imagine that the same peice of dust that can be pushed aside by the stylus can also be blown away with a blast of compressed air, so blowing the record prior to playing it should solve any dust-related issues. It might not be viable for everyday listening, but for archival purposes that might be worth it to get a better copy.
Laser players are pretty rare, but those who have them say it's a huge issue, and the noise from dust and dirt is very distracting. It's apparently very difficult to get the records to be sufficiently clean.
Old 78 records were produced with equipment from recording to cutting and pressing that is long since obsolete. Even if you overcome the challenges of dealing with laser equipment (dust, tracking).. what extra information are you really getting?. that it is useful signal more than even a dB or two above the noise I think is very wishful thinking.
Perhaps a laser could be used to first reliably replicate the original pattern of an LP into digital form, then one could feed that pattern to a real tonearm needle somehow, vibrating it suitably, in a nondestructive manner.
Like some kind of "fake vinyl", except the actual musical data could be downloadable.
This way one could stream/download vinyls while still playing the music with all the small imperfections which create warmth and the vinyl sound (=crackles of noise and static and wow and flutter and the harmonic distortions and so on).
I imagine someone with a signal processing background could design a suitable digital filter to emulate the effects of various tone arms. Designing the filter might require cutting a special tone record so that the frequency response can be measured.
Alternatively, the "hip" thing to do might be to train a neural network on a bunch of old record recordings, and then pipe whatever you want through the resulting model.
I wish I had not sold my 78's to Half Price Books a few years back. I had a couple hundred records, some of which had probably never been played. I found all of them in a house we bought right after getting married.
They were a mix of old Country music, Broadway musical shows, Jewish songs, and Big Band stuff. I always intended to digitize them before selling them but my phonograph only played 33 1/3 rpm and 45 rpm and had no facility for direct recording so everything had to be done through a connection to my pc.
I bought a USB interface to a phonograph and recorded about 40 of my albums and some that I bought at Half Price Books because I really wanted a particular song. After recording each album, I used Audacity to compensate for the speed difference and was able to get some great music.
It was a huge project though and I was working 7 days a week during this period so the spare time to complete it was just not available. That is why I decided to take them to Half Price Books. I knew that I would get almost nothing for them and that there was likely some very good music in the collection but I decided that there was a better chance for someone else to see them on the racks and tackle the job.
I love old music. It takes me back to my childhood in the same way that a smell can transport me to a specific point in time. I'm glad that someone has chosen to preserve this music in an accessible format and to make it available so others can hear those timeless compositions.
I knew going in that I wouldn't get much so I was not surprised to end up with a only few dollars which I spent on books for the kids. At the same time I also sold them a bunch of books so the records and books were co-mingled money. I know that HPB has always had a good selection of 78s on the shelves and that is why I took them. Hopefully they were able to send them to a store where someone could buy and enjoy them.
I also have a large collection of vinyl albums from the late 60's through the mid-80's that I tried to sell/give to a well-known local record shop in FtWorth. They wouldn't touch my albums or 45s and told me that most albums have no real value. This, in spite of the fact that their entire store was wall-to-wall used albums. They recently moved to a new building with a lot more floor space and their store has become a lot more like a HPB in that you can get memorabilia, posters, junk and crap along with your used albums. Pretty funny actually.
> I know that HPB has always had a good selection of 78s on the shelves and that is why I took them.
Oh, that's great. My local store doesn't have them.
> I also have a large collection of vinyl albums from the late 60's through the mid-80's that I tried to sell/give to a well-known local record shop in FtWorth. They wouldn't touch my albums or 45s and told me that most albums have no real value. This, in spite of the fact that their entire store was wall-to-wall used albums.
It's weird the way it works with vinyl: there's enough out there that it's still really, really cheap. But it's not necessarily easy to buy any given item.
I hope someone can put Internet Archive and Zero Freitas in touch with eachother. Zero is a businessman that has been buying up vinyl collections and archiving them in Brazil, amassing an enormous archive.
I try to understand how this is legal ?
I mean: what is the difference between Archive.org and any other torrent website ?
Aren't they both preserving the culture ?
Archive.org is doing 20M USD in revenue, lot of people are getting paid in the process, and suppliers as well, it's not like they are working for nothing.
They just don't have investors (but can have big salaries).
Genuinely wondering because I saw that you can download games there too (and some are still under commercial exploitation).
I think they're agreeing with you. There's no obvious reason this should be considered fair use.
Last time I checked, Archive.org also host ROMs of various copyrighted Nintendo games. Nintendo are famously litigious so I'm not sure how long that is likely to continue.
Thank you for sharing! If folks are interested in the code that powers the bot (it uses OpenCV to find the label, then draws frames and animates with ffmpeg) it's available here: https://github.com/thisisparker/78_sampler
I was thinking that this would've been really cool if they'd scanned the grooves with a laser to get the exact sound without the analog needle noise/imprecision. But I guess it's basically what I thought -- output from a record player in digital form. Which, I mean, is fine.
https://archive.org/details/78_rio-grande_leonard-warren-tom...