This one is special for several reasons: while it’s not the earliest disc I’ve featured here (that goes to Imdad Khan’s 1904 sitar solo), it’s among the earliest Algerian records ever made. It is difficult to pinpoint a date when it was recorded, but I’ve been advised that it was no later than October of 1907. The precise identity of the artist, Zerbib, is still unknown to us, but he was almost certainly a Jewish singer, and perhaps a member of the Zerbib family of Constantine, Algeria. He sings Andalusian classical music here on this 11 1/2″ disc, reminding me of the style performed later by artists such as the ones I featured on Opika Pende, Sassi Lebrati and Mahieddine Bachtarzi.
This information, while scant, doesn’t come out of thin air. Nor is it often found in books – or if it is, it’s in wildly disparate sources that need to be triangulated. It also helps to know kindred spirits. Our friend Thomas Henry, aka Ceints de Bakélite, along with our local comrade Chris Silver of the Jewish Maghrib website, are also keen collectors and researchers of early North African music. Thomas recently delivered a presentation about early recordings in the Middle East and North Africa at the Médiathèque Musicale de Paris last Thursday, and this made me think that it was time to try and transfer this baby. I consulted them both for additional information, as well as engineer Chris Zwarg, and they were able to help solidify some facts.
This is the very first “vertical cut” recording that’s appeared on Excavated Shellac. Those take some special care. I know I have a number of hardcore 78 collector maniac readers, but I also know that the majority are not entrenched in that insular camp. So, I’ll briefly explain this process, hopefully without too much needless pedantry. Bear with me…
Most 78s, and most records as we know them today are “lateral cut” – meaning the sound vibrations of a recording are cut by a stylus into the sides of the groove. In the early years of recording, cylinders and discs by a few record companies were made using a vertical process, also known as “hill and dale.” With the vertical process, the cutting stylus etched the sound recording vibrations into the bottom of the groove. The most notable companies who practiced this are now legendary names: Edison in the United States and Pathé Frères in France, who issued today’s entry. To play back these records on contemporary equipment, you either have to manually switch the pin connections to your turntable cartridge, or use one of the couple of preamplifiers that have this function built into them.
The early records that the frères produced seem designed to be contrarian. Apart from their non-standard record sizes (a group that ranged from 8″ to a mammoth 20″), Pathé’s records were “center-start” until the mid-19-teens, meaning you started the record at the center and it played outward toward the edge. Until the mid-1910s, they did not use a paper label, and instead engraved the artist, title, and disc information right where a label might be, filling the engraved letters and numbers with a light pigment as you can see in the top scan. Pathé’s vertical cutting method caused discs to sometimes have a deep or a shallow groove, which today can cause tracking problems with some contemporary turntables, sending lighter tonearms skating across records. Consumers at the time had to use a special Pathé-branded sapphire ball-shaped stylus with their records, but that ball-shaped needle was not usable with any other 78s that were vertically cut besides Pathé’s! Maybe most importantly, Pathé used a reproduction method that was different than other companies, and it often lessened the sound quality of their releases. Most companies were recording acoustically onto beeswax masters, which were electroplated to create metal masters or “shells.” Pathé had to be different. Until 1929, the company would make initial recordings on blank wax cylinders, and then mechanically (or “pantographically”) copy the grooves of the cylinder as it played, onto a master disc as it spun. If on this or any Pathé record you hear some repetitive scraping noise, it’s very likely the sound of that master cylinder spinning! And let’s not even begin to discuss speed, which for Pathé can be anywhere between 85 and 100 rpms. It seems a large percentage of this oversized disc wasn’t even used, as the tracks are a little over 2 minutes long and play at about 90 rpm. This would indicate that it was probably originally released as a cylinder. Who here has seen an Algerian cylinder recording? They have near vanished.
Pathé’s history as it pertains to global music is still being discussed and discovered. Those discs appear so infrequently, it’s difficult to complete the picture. What we do know is that Pathé, like the other major labels of the day, recorded global folk music quite extensively. During their “etched” era, well over 100 years ago, they recorded as far away as Tatarstan, India, and Thailand. While they may not have been the very first to record in Algeria (the Gramophone Company first made recordings there in 1906, Odeon (Disque Yafil) sometime in 1907), Pathé was on the forefront to capture local sounds.
To hear more of what are likely the only examples in existence, please browse the “Archives de la Parole” collection in Gallica, the Bibliothèque nationale de France’s digital library; here’s one example to whet your appetite.
With thanks to Thomas Henry, Chris Silver, and Christian Zwarg.
Notes
Label: Pathé
Matrix Number: 10144
Transfer/Serial Number: 21422 BC
It looks better in an original sleeve:
Fascinating Jon, thanks to all involved in this one! I never heard about Pathé’s technology before. Amazing that (despite the cylinder speed whooshes) the music was so faithfully and convincingly transferred from cylinder to disc. But why not spell Zeghbib as written? Or have I missed something?
Thanks, Tony! According to my resident experts, the “gh” is really a French transliteration of “r” – so “Zerbib” is proper…
I had a vague suspicion that there was such an explanation:)
Another ‘How does he do it?’ moment. 110 years old (at least) and it sounds fantastic. And it looks great. I love ‘designed to be contrarian’, too!
Thanks, Ray!
Great + rare (!) stuff + one of the clearest explanations of the Pathé process I’ve read.. Many thanks again!
Thanks, John! I could have gotten more technical, but I feared I’d lose everyone (that I hadn’t lost already…).
Thanks for posting this, truly wonderful music. And the history of the recording process is fascinating, even to a non-techhead. . . .
Great post. Interestingly the two titles on the record both contain mistakes. The ‘erid’ in English should be ‘trid’ (or terid, or similar, given transliteration is an inexact science), which is clearly written in the Arabic (تريد) and clearly audible both in the first line of the spoken introduction and the first line of the song proper, and is in any case the only thing that makes sense (it’s the second person singular form of ‘to want’; the full title means ‘O, you who wants to kill/fight me’). I wonder how that mistake came about…
In the Arabic title meanwhile, the q of qitali (= ‘ektali’ in the transliteration) is missing a dot, so it actually says ‘fitali’–but that sort of ‘typo’, if you’ll excuse the anachronism, is common so it’s easier to see how it came out.
On the question of Zerbib/Zeghbib, the comment that ‘the “gh” is really a French transliteration of “r”’ sounds the wrong way round to me–I believe French transliterations of Arabic often have ‘gh’ transliterated as ‘r’, reflecting the proximity in pronunciation–so I’d be really curious to see it written in Arabic. Both names (زربيب وزغبيب) seem to exist in north Africa but Zeghbib (زغبيب) gets a lot more hits on google, so it may well be that the transliteration on this record is actually correct, ie the gh actually corresponds to (غ) (rather than there being any interference from the french method where ‘r’ is used indiscriminately to represent two separate Arabic letters).
Ahh, trivia!
Thank you, klh – I will alert my friends about this and will make any amendments needed. Not trivial at all – this is precisely what I hope for with each post 🙂
Incidentally, my colleague Chris Silver (Jewish Morocco) who told me about the “gh” and “r” situation, clarified:
“North African Jews took to spelling the common Jewish last name of Zerbib as Zerbib. The actual letter is indeed an Arabic “ghayn” which means it should be spelled with a gh — in theory. Like Baghdad. But that’s not how Zerbib would have spelled his name. He would have spelled it how you wrote it. […] Zerbib would have spelled his name in Hebrew characters.”
Cool, I thought it was probably that way round. Mystery solved!