Laurel and Hardy had a different scheme: they'd leave up all the sets, and later re-take all the scenes that had made it into the final cut, phonetically, in up to 4 additional languages. All of the other actors were replaced by native speakers, but audiences everywhere wanted to see, and hear, the real Laurel and Hardy.
Randy Skretvodt's "Laurel & Hardy: The Magic Behind The Movies" quotes Stan Laurel: ...after shooting the entire picture in English, editing it and previewing it, they would bring in "French, German, Italian and Spanish interpreters. And they translated our dialogue into each language...we'd set up the camera, and we'd do the French version of the first scene...and each scene we did four times before we moved the camera for the different change."
> but audiences everywhere wanted to see, and hear, the real Laurel and Hardy.
This was probably due to the fact that until some time earlier people had not actually ever heard _anyone_ speaking, these were amongst the first movies which had audio, they just had not thought of dubbing.
But they started applying normal dubbing soon after: according to wikipedia, the "self-dubbed" movies were produced between 1929 and 1931 but then stopped.
A fun side effect of this is that the new dubbers kept the odd english accent! So if you get a copy of later movies where Hardy is dubbed by Alberto Sordi (one of the greatest actors of italian cinema) you can hear him make this funny fake accent, and mispronouncing words (e.g. very famously "stupìdo" rather than "stùpido").
In the 70s Monty Python picked up this tradition by shooting two episodes of 'Monty Python's Fliegender Zirkus' for German TV.
The first special was written in English, then translated to German, and all Python members learned their German lines by heart.
They didn't have time for the second special, so that one was dubbed.
It goes the other way with Kraftwerk, that released every single album both in German and English, except for Tour de France which is appropriately available in German or French
The 1999 soundtrack for Tarzan by Phil Collins was also recorded by Collins himself in English, Italian, German, Spanish and French: https://www.youtube.com/watch?v=m03Wcfx8dPE
I have often wondered recently why this isn't done more? Just re-shoot the same show/movie with Spanish French Chinese Japanesse Arabic etc.. speaking casts. make the smae show for diffenrt markets with different actors use the same sets the directors already have the scenes set up hell you could record camera positions and the camera could move exactly the same in different versions.
> I have often wondered recently why this isn't done more?
Because you have to pay N casts, shoot N × as much film, do N times as much editing, and use the sets of N times as long. The only thing this saves on is constructing the sets.
> hell you could record camera positions and the camera could move exactly the same in different versions.
Which probably is of less value that you think unless you do the same thing with the cast, who usually aren't as mechanically programmable.
You know, that might be about the first legitimately good use of deep fakes.
Personally I've tried to watch programs dubbed into English before and found it very distracting, to the point I'd almost rather watch it in the original language subtitled (though that takes considerably more concentration).
I'm not sure how much of an issue this is to people who don't have English as a first language, where they have presumably encountered dubbed English films since childhood.
I'm a native French speaker, watched dubbed English movies since childhood but started hating it and being disturbed by dubs when I was a teen (didn't really notice or care about it when I was a child). Lips not moving in sync with the audio break immersion for me. Plus, more often than not, voice actors for dubs are just not very good and that destroys the actor's performance.
So, I transitioned to only watching subtitled movies around 16 once I could just buy dvds and have access to the original version.
Growing up in the Netherlands (everything subtitled) we had access to German television (everything dubbed) and we always frowned about how everything sounded the same. I'm still convinced that the same three voice actors were doing all the characters in all the properties I ever saw on german television. When you grow up with subtitles, it's second nature, and dubbing is annoying.
I don't think that lip movements are the thing holding dubbing back. Most dubbed voices have a cheesy Days of Our Lives vibe. Its quite easy to tell a dubbed TV show without ever looking at the screen.
You're downvoted at this point but I think this is definitely going to happen in the future, even if it's just rewiring mouth movements to match the new audio.
Probably because it would still be significantly cheaper to just sub/dub the english version. And finding quality casts for all the languages would be a lot harder too.
I'd much rather watch a foreign production with top quality actors subbed in english than an assembly line english version with sub par acting.
(Although I can see it being alright for stuff you already don't expect to be quality, like soaps)
But in this case all the actors speak English, you don't need a whole new cast. Plus they put on that Norwegian accent that makes it great for English audiences.
It's a great show that manages to be both about Vikings and modern Norway at the same time.
Hollywood does this with productions from other countries, for example "City of Angels" [1] for "Der Himmel über Berlin" [2],
or "Vanilla Sky" for "Abre los ojos" [4]. And not only do the use American casts, they re-shoot the whole movie.
Sergio Leone's spaghetti westerns had a mixture of languages spoken during filming. In the English releases it certainly looks as though some of the minor characters are dubbed from another language, but my multilingual lipreading skills are not up to determining whether that other language is Italian or Spanish or even something else.
I have to imagine that it would be tough to cast actors in multiple languages that all fit the same vibe as the original. My understanding is that the casting process is relatively difficult and takes a lot of effort; why do it multiple times and risk putting out a mediocre imitation?
Kids TV is really good at this, until quite recently I didn't even know that Paw Patrol was originally a US show which then gets dubbed by a UK cast alongside the more expected regions with entirely different languages.
>Then in the early 1990s it transpired there was a copy in Cuba, a fact confirmed by the Cinemateca de Cuba in Havana. Still, it took four meetings – the UCLA Film and Television Archive had to fly out to take part in negotiations – to arrange a temporary loan. As of 2015, the film sits in the US Library of Congress
So, they made a deal and Library of Congress broke it? Or is it still on loan or is it a copy?
I was wondering the same thing when I read that. According to https://www.miamiherald.com/entertainment/movies-news-review... (2015), the loan was made by 1995, and it does not mention anything about it not being returned, so it seems somewhat unlikely that the US Library of Congress version (which was added in 2015) is the original.
Cuba interestingly enough already has a policy of not accepting money from the US for stolen property. The US writes a check each month for the rent of Guantanamo Bay that the Cuban government then doesn't cash in protest.
This sort of diplomatic “loan” with extensive negotiations is typically done with the expectation from both sides that, in practice, it will never be returned. I expect (and hope) that Cubans got something in return.
The article makes it sound like this was some skunkworks project the crew did behind the producer's back, but as far as I know alternative language productions using the same set during off hours was standard practice before dubbing became widespread.
Yes, it's strange. The article completely acknowledges this was a normal practice, but the title makes it sound sneaky. I guess it's just to make it more clickbaity.
Is this https://lccn.loc.gov/2003636872? It’s in Spanish and directed by George Melford, but it’s on VHS and it’s dated 1992. Plus it has English subtitles.
It would be fun to see this, but I get the impression that it would involve a visit to the Library of Congress. Or maybe Cuba?
Is there more information on the Library web site that I’m missing?
It's an extra on the Dracula Complete Legacy collection; I don't think it's rare or anything, usually it gets bundled with Universal's collections of Dracula films i think.
Not rare, and outside of Lugosi's performance, is considered the better film. Think about it, the Spanish language crew could watch the English dailies and improve on their work.
It looks like any registered reader at LOC can request it for on-site use, but that seems like the only way to see it. Anyone can become a reader but you need to complete registration in-person at the building.
Aside, @personlurking[0] on HN (you won't see this but thanks) mentioned FlixOlé[1] the other day and I've been hooked on Spanish cinema for a few days now. $4/mo (14 day trial) and there are no subtitles, so for the HN crowd it's only good for Spanish learners forced to finally wean off subtitles.
While Netflix has stellar Spanish shows/movies from El silencio del pantano (movie) to El desorden que dejas (show), it's a fun break from Netflix with a large catalog of eras (1950-2000) that Netflix doesn't have at all.
I wouldn't normally watch a movie from 1957, but there's something easier about doing that when it's a culture you weren't born into. I think this is the spice of learning another language, especially for us native English speakers since we have no real impetus beyond leisure to engulf ourselves in other languages/cultures. IME this is something people take for granted who learn English as a 2nd language and wonder why English speakers are monolingual.
This comment doesn't fit here (just triggered by Spanish and Dracula and a couple statements in TFA) but it's something I was thinking about over the last couple days that I wanted to jettison into the ether.
This reminds me of putting the LimeWire stub installer into production the first day of the COO/CTO's vacation.
We were the most popular Java desktop app at the time, circa 2005, P2P file sharing. At the time, we had approximately 60 million client GUIDs connecting to the network per month (or was that quarterly?, in any case, tons of users). It was a small company, 5 developers, 13 employees total.
One of the largest costs was bandwidth for the free version of the software. We already had a mechanism for using the P2P network to save bandwidth on version upgrades of the free version. The hello handshake between peers included the serial number for a signed (1024-bit DSA) XML message containing the SHA-1 hash of the latest free installer. If the serial numbers didn't match, the out-of-date peer would download the latest XML message from the other. Within 60 seconds of a new release coming out, over 99% of the online nodes would have the SHA-1 hash of the latest free installer and start downloading the latest free version installer from each other.
Edit: Originally, the XML update messages just had a URL for the installer. We accidentally DDoS'd ourselves the first time we used it, with millions of users flooding in within minutes of the latest release, so we added a randomized delay over 24 hours for actually notifying the user of a new version. After we added the P2P download of the upgraded installer, we still redirected the user to our website for the upsell opportunity. I don't recall how we got the already-downloaded installer to launch after visiting our website.
So, one of the other developers and I wrote a proof-of-concept minimal C++ installer. Its only GUI was a progress bar. It had just the DSA verification code from Crypto++, a minimal parser for just the subset of XML we were using, and enough of the P2P protocol to bootstrap into the network and download the latest free installer. It would check the SHA-1 of the downloaded installer and then launch the downloaded installer. It was less than 1/100th the size of the actual application installer, and had the advantage that it would always install the absolute latest release version of the free version of the application.
Also, the monthly profit sharing payment was roughly the same as my salary. For more senior developers, it exceeded their salary, so we were all highly incentivized to cut costs.
The COO/CTO shot down our proof of concept with just vague worries that we might have broken something, that maybe it only worked from the office and our apartments. He had no concrete criticisms.
So, the next time the COO/CTO went on vacation, we went into the webserver and replaced the full sized free version installer with the tiny stub installer. The first time the free version started up after install and successfully bootstrapped onto the P2P network, it would make a tiny HTTP request to our webserver, so we could monitor for problems.
We watched carefully the metrics for successful first-time P2P bootstraps, and they were fine. We were all very happy. About an hour later, the COO/CTO called into the office from the beach on a Monday morning. He said we broke something. We asked what we broke, and he said he wasn't sure, but he was checking total badwidth usage, and it dropped off a cliff about an hour earlier. We told him to calm down, than free version successful connection rates were unchanged... but that stub installer he had nixed was saving bandwidth.
He wasn't happy, but agreed to let us to continue trialing the stub installer while he was on vacation. We were the only two C++ developers at LW, and we both left for Google within a year. The stub installer stayed in production for a year or so after we left, but eventually they got rid of it due to maintenance concerns.
Edit 2: "The Bulgarian Nuclear Option": I noticed a flaw in that we used a Java long for the XML message serial number, not an arbitrary precision integer. When the recording industry started threatening a lawsuit and the owner asked us to start researching options for a censored version of LW, one of the developers claimed he took my observed flaw and signed a message with serial number Long.MAX_VALUE to lock everything at the current version. He claimed he had this final version message on his home machine and had a chatbot mechanism he could activate from his cell phone to be able to lock out the owner from being able to auto-update people to a censored version. The developer in question was from Bulgaria, and hence, this was "The Bulgarian Nuclear Option". News got back to the owner, and I heard that the developer in question came back white as a sheet from being called into the owner's office. The Bulgarian wasn't the other C++ dev.
> So you just put it into production without permission? Makes sense that you two went on to Google.
Well, we went to Google because the RIAA filed a lawsuit and it was clear the owner didn't have a great plan of what to do, vacillating between shutting down the company and fighting the lawsuit. Another year or two later, a third of the five developers joined us at Google and the other two were at hedge funds.
It was a 5-developer startup. Everything was played fast-and-loose. It was a P2P file sharing startup, post-Napster, clearly a risk-taking environment. It was a minor offense in that corporate culture, and it made a 90% drop in one of our biggest expenses. It was a 13-person company pulling in over 1 million USD / month in revenue. It was clearly a smart risk to take, we were monitoring it, and it would have taken literally seconds to undo it.
Also, the Bulgarian Nuclear Option guy (not the other C++ dev) didn't get fired. You could plot a technical mutiny to thwart a corporate legal settlement, and get off with a stern warning. That's the kind of company it was.
Randy Skretvodt's "Laurel & Hardy: The Magic Behind The Movies" quotes Stan Laurel: ...after shooting the entire picture in English, editing it and previewing it, they would bring in "French, German, Italian and Spanish interpreters. And they translated our dialogue into each language...we'd set up the camera, and we'd do the French version of the first scene...and each scene we did four times before we moved the camera for the different change."