I'm on a bit of a mission at the moment to try to work out what the hell to do about DRACULA, which is (much like HALLOWEEN and THE EXORCIST) a historically significant favourite of mine which has been treated frustratingly poorly over the years.
This more or less involves me trying to source every single known unique release and examine the viability of each one as a source.
I currently have:
1992 US LD from Warner
1993 JP LD from Warner
2000 GB VHS from Warner (which I own but don't have in my home because of COVID, frustratingly)
2013 GB BD from Lionsgate (2007 BFI master with "theatrical"* and extended cut with some JP footage re-integrated)
2017 DE BD from Anolis (2007 BFI master with "theatrical"* and further extended cut with more JP footage)
2018 US BD from Warner (2007 BFI master of "theatrical"* cut with the colour now corrected, but with severe black crush from what looks like an RGB range error)
2019 DE BD from Studio Hamburg (2002 Warner master of US theatrical cut with "HORROR OF DRACULA" title and 1.78:1 framing instead of 1.66:1)
I'd like to get (but probably won't):
2016 DE BD from Studio Hamburg (stupidly expensive)
2020 ES BD from ... well, it's hard to say, but I think it's Resen, which would make it most likely an unlicenced bootleg, which I don't want to financially support by buying
* The actual UK theatrical cut had censorship to a scene involving a vampire being, er, "staked", but this is present in every version I own (possibly excepting the UK VHS, but given it's a Warner release from 2000, I expect it to be in there).
The problem is primarily that overall, the most watchable (and likely most true to source) release to date is the 2018 Warner Archive release... but they have clipped the s**t out of the blacks to such an extreme extent that I'm pretty convinced it must be a technical error where they've mixed up their RGB ranges (unless I'm getting this confused, I take it what they must've done is erroneously assume that a full range RGB production master was limited range, clipping off anything below 16 and over 235). The sound, at least at a glance, appears to be head and shoulders over just about anything else including both LaserDiscs. But you'll notice that I've not checked any DVDs yet, so it's theoretically possible (however unlikely) there may be a better source in DVD format.
Here's the first of at least two short videos on the subject, comparing the two different extended cuts (from 2012 and 2017) to the 2007 BFI "remaster" which introduced that now rather infamous blanket blue-green tint @"The Aluminum Falcon" did a pretty good job of masking in his restoration a while back.
To very briefly illustrate what this looks like, here's the same frame from a few versions:
US Prime Video stream / 2019 DE Studio Hamburg BD, 2002 Warner master from newly-struck IP:
I believe this is pretty much exactly what it's *supposed* to look like, aspect ratio (and US title card) aside.
2013 GB Lionsgate BD, 2007 BFI master from OCN provided by Warner:
Clearly all kinds of messed up.
@"The Aluminum Falcon"'s restoration of the BFI master:
Better, but still extremely pushed toward the cold end of the spectrum... only so much anybody can do with a source that heavily screwed. It's still the best version of the extended cut available atm, imo.
1993 JP Warner LD, unspecified early 90s Warner master (evidently a release print):
Very very close to the 2002 master. Again, I'm convinced this is how it's meant to look.
A second video is upcoming showing some of the differences between the 2002 Warner master, the 2007 BFI master, and the aforementioned problematic 2018 Warner version of the BFI master (which at least fixes the colours but obliterates shadow and highlight detail).
The *easiest* solution would probably be to take the best available encode of the 2002 Warner master, possibly bringing the levels down a bit and the saturation up (it's quite bright / washed out compared to the LD), then crop down the UK title card to match and cut that in, using the audio from the 2018 Warner Archive BD. And that may very well be what I do to start with. But I'd ideally prefer to preserve it in 1.66:1, which makes the best source the Warner Archive release, but that has the insanely clipped levels, making it significantly less ideal... I don't know yet.
(This post was last modified: 2021-01-23, 04:40 AM by pipefan413.)
Hell of a collection there. Not sure I can offer much advice but I'm very happy you are taking up this project. Although that movie is beloved, it never seems to get all the way over the finish line. Every version has at least one thing wrong.
(Also pretty great comparison video)
(2021-01-23, 05:16 AM)PDB Wrote: Although that movie is beloved, it never seems to get all the way over the finish line. Every version has at least one thing wrong.
I already see one other possible "issue" in the 2018 Warner Archive version that nobody seems to have mentioned yet: they regraded the (original UK) titles to look more like their master of the US title. And in theory, this is sound logic, because the US master was scanned off an unfaded interpositive check print in 2002 for the original DVD. However, it was a *newly struck* interpositive rather than a contemporary colour graded source, and other sources indicate that the BFI grading is actually more correct here! The early 90s LaserDisc releases of what appears to be a US 35 mm release print are warm/coppery (US title), as is the 2007 BFI master (UK title) and all others derived from it, apart from the Warner Archive one. It's possible that whatever print was scanned for the early 90s LaserDiscs was maybe just a little faded and that's what gave it that coppery look, but I'm not sure about that...
The UK title was only restored in 2007. Supposedly it went something like this:
1. BFI head of "picture quality" goes hunting for a UK print, finds a 16 mm one in the old BFI lending library
2. He sends a frame grab of this to Warner in LA who start hunting for it (having previously no idea where they might have an element with the UK title)
3. Warner is finally able to locate a source and scans it for the BFI to restore
Now in fairness, the BFI title text (as in the actual text itself) looks like crap, it's stupidly dull: almost brown rather than big bold red. Aside from the screwed up levels, Warner's version looks better to my eyes in this respect. But I'm inclined to agree with the BFI grading generally because it looks almost exactly like the LaserDisc scan of the US version of the title.
A lot of 1990s and 2000s era HD transfers were subject to magenta/copper tinge (to combat the supposed blue push on CRTs) and EE to "enhance details" so they still show up on low res DVD masters. Also, it was speculated Warner retimed the BFI HD master to match a Technicolor IB print as close as possible. How accurate the Technicolor IB (not IP) is is anyones guess, but it looks like Warner could only do so much on a flawed master with baked in blue-teal.
(This post was last modified: 2021-01-23, 06:42 AM by SpaceBlackKnight.)
(2021-01-23, 06:40 AM)SpaceBlackKnight Wrote: A lot of 1990s and 2000s era HD transfers were subject to magenta/copper tinge (to combat the supposed blue push on CRTs) and EE to "enhance details" so they still show up on low res DVD masters.
Yep, I know. What were you mentioning this in relation to, sorry? I reckon the only version of the 2002 transfer that appears to have either problem to any sort of noticeable degree is the 2019 encode of the 2002 Warner master from Anolis which definitely has a wee bit of magenta push going on, but it also has some (at least vaguely restrained) DNR that isn't on the streaming version. The streaming version definitely has less of a magenta bias, though not necessarily *no* magenta bias (can see it in the background of that first screenshot), and has some visible flecks of minor debris etc. that aren't there any more on the 2019 Studio Hamburg version. But the older transfer is still an older transfer and shows its age in other ways: in some ways it has *more* detail because of the lack of cleanup, but in others it has less (understandable since it's a 2K scan of an IP made for the standard def era). The shot if the tomb at the beginning with the blood splat looks a lot softer on the streaming master of the 2002 transfer, though that could in some part just be a streaming bitrate thing (it's around the 10 Mbps mark but the Studio Hamburg encode is maybe about double that, I can't remember off top of head what the exact bitrate of that BD is).
(2021-01-23, 06:40 AM)SpaceBlackKnight Wrote: Also, it was speculated Warner retimed the BFI HD master to match a Technicolor IB print as close as possible. How accurate the Technicolor IB (not IP) is is anyones guess, but it looks like Warner could only do so much on a flawed master with baked in blue-teal.
Certainly could be the case. To be clear, I am very aware of the difference between an IP (interpositive) and IB (imbibition) print and wasn't confusing the two. I mentioned an IP in connection with the Warner *master* from 2002, not the 2018 Warner release of the BFI masfer in 2018. The BFI stated that they used this 2002 IP as reference in 2007 but like... where? That dark blue crap looks like no print I've ever seen...
Anyway I'm not sure what you mean about only being able to do so much with the tint, have you seen the 2018 one? It honestly looks pretty amazing colour-wise except that it has levels badly clipped, which isn't the BFI's fault as I don't see that it'd be anything to do with the tint. Honestly, I kinda feel like Warner might have gone back a step to the pre-colour-grading master from 2007, which would both be the sensible thing to do and quite likely possible for them because they held all the cards in 2007 and every move the BFI made was at Warner's behest. Warner gave the BFI access to the negative materials and the "final mix" sound master but all work on the physical elements had to be done at Warner's premises, so presumably they'd have the OCN scans and possibly the cleaned up DI from the BFI work before grading was done. It certainly could just be that they only have the BFI graded master like everyone else but if anybody other than the BFI/Hammer has a pre-grading version of that master, it'd be Warner. I don't know how the hell they could possibly have done this good a job otherwise. Which makes the black crush and drastically inflated contrast all the more infuriating...
EDIT: I should maybe mention this too. The BFI picture quality guy claims that he didn't really try to grade it to look like any particular stock or print type (Eastman, Tech, whatever) in 2007, he just graded it based on "what he saw in the negative" (but mentions the "check print", meaning, I believe, Warner's IP from 2002). It's also from him that I know the 2002 DVD transfer was made from said newly struck IP. Honestly, a lot of the film looks pretty good to me in his grading, but the scenes where he's cranked the hell out of that awful cool bias seem to me like he just took massive liberties with it (it looks very much like a "creative license" grade, e.g. with dark blue tinting on night scenes and scenes that somebody wanted to evoke night even if they don't necessarily take place at night... though other indoor daytime scenes also have an extraordinary amount of green, so it isn't just that).
(This post was last modified: 2021-01-23, 01:23 PM by pipefan413.)
I'm wondering if this might be a another good example of something that should prompt me to finally get round to testing out AutoOverlay: what if, say, I were to overlay the somewhat cropped 1.78:1 2002 Warner master onto the generally more opened up 1.66:1 2007 BFI master, or vice versa...?
Here's what the Universal logo looks like cropped and upscaled to match the aspect ratio of the Warner master.
Source 1.66:1 version from 2013 encode of BFI restoration, with the black pillarboxing cropped off accurate to the individual pixel (60 off each side):
Upscaled to match the exact framing of the streaming version of the Warner master:
I did the same thing with the German encode of the same source, and the original UK title sequence (for both the UK source and the German encode of the same master), and the same again for the Universal logo that was at the end of the film on the BFI master (again, UK and German versions both done to give me options). I stitched those onto the Amazon streaming version with the start and end trimmed off. And I'd already resynced the US Warner Archive Collection audio, so that's done too.
Result: 2002 Warner DVD master in 1080p, with original Universal logos and UK "Dracula" title sequence, and the 2018 US Warner Archive Collection audio!
I think I'm going to want to improve this further but I'm thrilled that it's now at least watchable. I think the next step is probably exercising my languishing colour grading muscles to remove what's left of that magenta bias and perhaps slightly darken the image, seeing as I think the old master is probably a little bit too bright compared to all the newer versions (and indeed I do think it probably looks better a little darker, just not with all the shadow detail sliced off in the process). I might also investigate the viability of using AutoOverlay so as to not waste the additional picture area afforded by the 1.66:1 BFI master, though I don't know how well that'd work here.
I've started the colour fiddling by just doing *really* basic adjustments to the opening titles from the UK master (in AviSynth, but I'm going to export and have a more sophisticated mess around in Resolve) because they look really dull on the source but going OTT with it results in the text looking pretty good but the blood drip looking stupid. WIP.
The screenshots look different (more exaggerated) than the screenshots in AvsPmod, which bothers me. But I'll figure that out later, I need sleep.
In terms of the actual grading of this sequence, the massive elephant in the room is that the Warner releases all grade it really cool and make the stone look grey, but in the UK BFI master it's extremely brown/reddish. And y'know what? I'm actually feeling like the BFI got this one right. Look at the LaserDisc, which is from a print:
(This post was last modified: 2021-01-24, 07:12 AM by pipefan413.)
I really do wonder where the hell this dark blue/green thing ever came from.
According to the BFI, Warner's more consistent (and less alarming) grading was built on top of an interpositive struck from the negatives in 2002. Now, being a *new* interpositive, who knows if they timed it right compared to contemporary IPs of 1958, but it would almost certainly be at least a *bit* different.
The trouble is that the BFI grading is based on using *that same IP* as reference and yet looks wildly different from the conclusions Warner has come to over the years: none of their releases (LDs, DVDs, BD, streaming) have that heavy blue/green push. Paraphrasing but the BFI guy said more or less "I didn't want to give it too much of a signature that would make it seem like we were trying to make it look specifically like an imbibition print or a Kodak print or anything else. I just wanted to bring out what I saw on the negative and on that check print" (meaning the IP from 2002).
The trouble is, the BFI isn't going to come out and say "I basically did whatever the hell I felt like so I gave it a really strong blue tint toward the end to make it feel more dark and oppressive". Conversely, it's possible that Warner may have "corrected out" things that were present on the source but seemed to them to be excessively pushed toward a particular colour.
Take the title sequence: there are probably two main possibilities that occur to me. In any case, either Warner got it wrong from 2002 onwards (since they released the LaserDiscs in the 90s and those are like the BFI grade) or Warner got it wrong on LaserDisc and the BFI got it wrong in 2007, but they'll both have based their decisions on *something*. How do we figure out which is right?
Well, we know that the thing that prompted Warner to finally locate the original UK title in the first place was the BFI finding a 16 mm lending library print and sending them a picture of a frame from it, meaning the BFI definitely had reference in the form of that 16 mm print. They also had Warner's 2002 IP scan. We also know that the early 90s LaserDisc matches (roughly) the BFI grading, though it's even more pushed toward that coppery brown seen on the BFI title sequence... and that LD is quite possibly from a release print, since there's a crapload of dirt/damage on it. Perhaps Warner saw that brownish colour on their new 2002 IP but felt that the brown stone was wrong somehow, and corrected it out. Perhaps it just wasn't present on the IP, being that it would have been timed somewhat differently from however a lab would've done it in 1958, but the BFI fell back to using their 16 mm UK print as a reference, and that did have the brown colour. Or maybe all the print references with that brown colouring are simply faded to brown through age, and never would have looked like that originally, which would likely make the Warner grade with grey stone more correct. It's worth mentioning as well that we *also* have the very very worn out and severely damaged last 4 Japanese reels, and the end sequence on those absolutely does *not* have that oppressive blue wash that's all over the BFI master. Hell, Molinare (the UK restoration house that Hammer/BFI brought in to restore and integrate footage from the Japanese reels) specifically claimed to have graded the Japanese print to match the BFI master, but that's very decidedly not what it looks like in the end result: watching the 2012 extended Hammer cut, every time it cuts to Japanese frames during that end sequence, the blue disappears and the colour looks more brown (which, again, might just be faded dyes for all I know). To be fair, they got a closer match on the Horny Mina shot earlier in the film, but the disintegration scene footage looks nothing like the BFI grading.
I'm going to look through the rest of the LD to see if the brown is really noticeable elsewhere in the film. If so, it might just come from fading dyes. Certainly, the Japanese reels look pretty brownish throughout, if I recall correctly... I'll check those again too.
Great thread that really needs to exist for this film which still has problems on video. The WAC Blu was indeed the 2007 BFI master but retimed to try and remove their appalling revisionism to the color and time against IB prints as stated by WAC. In fact I think they nailed the color largely for what their source was but at the cost of the black levels-but if you look at the UK Lionsgate BD of the 2007 master it already has the shot black levels with no depth so they're apparently baked in to the master.
A print would have the greater range and darker look of most scanned 35-there's a few shots of a collector's print in the very long Bluray.com thread for the film. We've gone over this there back and forth and it's still an issue. The new German disc apparently does have a higher gen version of the DVD master on it which is cool as still for a number of reasons it looks quite nice and balanced.
I've always wanted to get the US LD but it's probably not necessary. I have no idea if the Japanese one is any different.
My theory is WB made new IPs for their three main Hammer titles around the same time in the late 90's and because LD was ending these showed up as follows: Curse of Frankenstein on a late release LD and all three (CoF, HoD, The Mummy) on the snapper case DVDs. These have been the primary source for the further releases of all three films-CoF was in the worst shape and only recently have digital tools allowed for proper restoration, The Mummy looks stupendous, but sadly the BFI in their 2007 "restoration" decided to completely alter the look of the film and admitted to doing so.
Why it was decided we had to stay with their crap master and not simply revisit the Warner IP and vault materials for a new disc is beyond me. I would hope after CoF coming out so wonderfully that they could revisit it in the same manner.
The DVD master is overcropped and yes a tad reddish in places but it still looks great and has the IMO superior title. The BFI restored the UK title by using a still frame and that's why the UK title looks crummy on all releases....because it's a dressed up frame grab...
Anolis did a much better job in handling the extended scenes being better integrated in the film. The 2012 Hammer master shoehorns it it a bit forcefully. The problem is they tried to mesh the theatrical and extended together when they are edited and scored differently. It's the same problem one faces when comparing foreign Hammer versions with differences-you cannot simply smash them together and expect it to work.
The ultimate defense that the BFI color is revisionist garbage in addition to the admittance of revisionism is not only do all other Jack Asher shot Hammer films look nothing like this but the Japanese reels even without the red fade look glorious and exactly as they should with no cold blue!
Damn Fool Idealistic Crusader
The brownish look of the LD you indicate is in my WB 1.33 LD of The Mummy. I think it's because they were straight print transfers and certain elements get highlighted by NTSC where others don't. I've seen this happen on a number of print transfers on LD in this era.
Damn Fool Idealistic Crusader
(2021-01-24, 09:26 PM)captainsolo Wrote: Great thread that really needs to exist for this film which still has problems on video. The WAC Blu was indeed the 2007 BFI master but retimed to try and remove their appalling revisionism to the color and time against IB prints as stated by WAC. In fact I think they nailed the color largely for what their source was but at the cost of the black levels-but if you look at the UK Lionsgate BD of the 2007 master it already has the shot black levels with no depth so they're apparently baked in to the master.
I dunno about that, the levels are definitely more screwed on the WAC version. If you put them side by side there are things you can see in shadows on the BFI one that you absolutely cannot on the WAC one. That said, I suspect there is a little more shadow detail on the older Warner master.
(2021-01-24, 09:26 PM)captainsolo Wrote: A print would have the greater range and darker look of most scanned 35-there's a few shots of a collector's print in the very long Bluray.com thread for the film. We've gone over this there back and forth and it's still an issue. The new German disc apparently does have a higher gen version of the DVD master on it which is cool as still for a number of reasons it looks quite nice and balanced.
Aye, as I mentioned, the 2019 Studio Hamburg disc is indeed the Warner master. However, it's had digital cleanup done and the magenta push is pretty evident. Comparatively, the streaming version is actually better in a lot of ways: less magenta push, less/no cleanup (there are a lot of wee flecks of dust etc. that make that pretty clear).
(2021-01-24, 09:26 PM)captainsolo Wrote: I've always wanted to get the US LD but it's probably not necessary. I have no idea if the Japanese one is any different.
I have both. Honestly? I wouldn't bother. But I'm glad I did, otherwise I would have wondered forever if the audio was any good and what the colour looked like. Now I know.
(2021-01-24, 09:26 PM)captainsolo Wrote: My theory is WB made new IPs for their three main Hammer titles around the same time in the late 90's and because LD was ending these showed up as follows: Curse of Frankenstein on a late release LD and all three (CoF, HoD, The Mummy) on the snapper case DVDs. These have been the primary source for the further releases of all three films-CoF was in the worst shape and only recently have digital tools allowed for proper restoration, The Mummy looks stupendous, but sadly the BFI in their 2007 "restoration" decided to completely alter the look of the film and admitted to doing so.
Why it was decided we had to stay with their crap master and not simply revisit the Warner IP and vault materials for a new disc is beyond me. I would hope after CoF coming out so wonderfully that they could revisit it in the same manner.
Do we actually know for certain that Warner *didn't* go back to a pre-graded version of whatever work BFI did? I mean, rather than spend the money and time redoing the entire scan etc., I just wonder if they might have at least gone back to an earlier version that didn't have the hugely distracting dark blue wash over half of the film. Because as you say, their colour regrade of it is pretty amazing in general, apart from the levels thing. The one thing that makes me wonder about the accuracy of Warner's grading is the colour of the stone in the opening title sequence: they make it firmly grey, but on their own LD and on the BFI version (which they've clearly based on *something*) it's definitely red/brown.
But yes, I too wish they'd just started again, given the state of what the BFI did to it. It's appalling.
(2021-01-24, 09:26 PM)captainsolo Wrote: The DVD master is overcropped and yes a tad reddish in places but it still looks great and has the IMO superior title. The BFI restored the UK title by using a still frame and that's why the UK title looks crummy on all releases....because it's a dressed up frame grab...
Are you certain that's 100% accurate? It certainly doesn't appear to be what the BFI claimed it was. Their story is that they found a 16 mm print with the title on it and took a frame grab of that, and sent that to Warner in LA, *who then found an original element* based on having the frame grab as reference. So, not that they edited said frame grab into the film. But perhaps they're lying / misrepresenting / misremembering, or they just made an incorrect assumption about what Warner actually did.
Ben Thompson, BFI National Archive Wrote:They didn't know where it was and it was really not until I found a 16-mil print in the BFI Films Collection, which is actually like a lending library of old, and an 18-minute 16-mil unfaded print had reference, or it had the UK main title in it. So I took a frame grab and sent that out to Bill Rush at Warner Brothers in LA, and they located it.
(2021-01-24, 09:26 PM)captainsolo Wrote: Anolis did a much better job in handling the extended scenes being better integrated in the film. The 2012 Hammer master shoehorns it it a bit forcefully. The problem is they tried to mesh the theatrical and extended together when they are edited and scored differently. It's the same problem one faces when comparing foreign Hammer versions with differences-you cannot simply smash them together and expect it to work.
I actually disagree with this quite strongly: the Anolis version inherits all the sync issues of the Hammer extended cut, and adds more. Hammer at least tried to avoid using the audio from the JP reel here, because it doesn't match the rest, and I respect that; the trouble is that in doing so, they caused noticeable desync at one point. In contrast, Anolis clearly just used the Japanese audio for the bit they added in, and then clumsily faded it out and faded the Hammer audio back in again right in the middle of the following shot. Did you watch the comparison video? I went over that in detail.
(2021-01-24, 09:26 PM)captainsolo Wrote: The ultimate defense that the BFI color is revisionist garbage in addition to the admittance of revisionism is not only do all other Jack Asher shot Hammer films look nothing like this but the Japanese reels even without the red fade look glorious and exactly as they should with no cold blue!
Yep, exactly.
(2021-01-24, 09:28 PM)captainsolo Wrote: The brownish look of the LD you indicate is in my WB 1.33 LD of The Mummy. I think it's because they were straight print transfers and certain elements get highlighted by NTSC where others don't. I've seen this happen on a number of print transfers on LD in this era.
That might be all it is, but if so, how come the BFI one is much the same? I mean, I get that they clearly made a hash of the grading of the whole film, but even so: it seems awfully coincidental that they happen to have come to the same conclusion of the LD release, y'know? It could very well be some kind of fading from the print they scanned it from, and the BFI's reference material had the same problem so they've reproduced it, I don't know. They definitely had access to a 16 mm print with the UK title on, which was supposedly unfaded, in which case that brownish-reddish colour may actually be correct.
|