HD Isn’t For Movies…It Seems
After my morning run (a beautiful but hard 11 miles along the Southern California coast), I came to the studio to set up the room for a film mix that’s going to happen on Monday and Tuesday. Andy Hay, the go to guy in the world of independent film mixing and overall ace audio guy and I spent a couple of hours routing this and that, checking out the new near field monitors and making sure there was a hot microphone in the room for some lines that may have to be re-recorded.
And we got talking about film sound, monitor systems, the X Curve, DCP and…my favorite topic…high-resolution sound. I asked him if any films were being done in high-res audio. The answer is no, not that he’s aware of.
If you think going to the movies or listening to your favorite cinematic feature in your home theater delivers a great audio experience, you’re sadly mistaken. I thought that only music lovers are had fidelity problems…we’re getting great sound compared to film sound!
My mixing room is equipped with a THX certified set of JBL speakers…you know, the big ones with the horn drivers and a woofer for the low end. There is also a large subwoofer for the LFE (Low Frequency Enhancement) channel, my trusty TML Labs “Profunder” driven by 500 watts of Bryston power. Andy mixes “in the box”, which means that a surface controller makes changes to the Pro Tools sequences and plug ins. Mixers don’t want to mix dialog, music and effects with a mouse. Control surfaces use a simple Ethernet connection to allow the virtual faders on the computer monitor to be moved by the physical faders on the surface.
All audio for films is PCM at 48 kHz/24-bits. The actors are recorded by the production sound mixer at 48 kHz/24-bits, the foley and ADR (automatic dialog replacement) are recorded at 48 kHz/24-bits and the mix is output as part of a DCP (Digital Cinema Package) for final delivery with the visuals…at 48 kHz/24-bits PCM. Nothing high-resolution about any stage of the production. And it gets worse.
For “consistency”, there is an aggressive low pass filter applied to the full range audio before you get to hear the soundtrack. It’s called the “X-Curve”. Sounds pretty exotic, doesn’t it? How bad could a little high frequency rolloff be? After all, it’s only dialog, a little underscore and a few sound effects, right? It’s hugely bad…and I had no idea until today.
The X-Curve is based on the Film Academy curve. See the illustration below (which was borrowed and recreated from an interview with audio pioneer Loan Allen, on the editors guild website).
Figure 1 – The X-Curve and Academy Rolloff at the low and high end of the audio spectrum.
As you can see from the graph, the rolloff is very severe and basically removes any frequencies above 10 kHz. Along the way, the X-Curve attenuates more and more high frequency information…3 dB per octave up to 10 kHz and then 6 dB per octave. There’s no hope for high-fidelity in a movie house.
And do you think it’s any better at home? Nope…the studios that are preparing the 5.1 and stereo mixes for DVD or BD, are shackled with the same source material that was provided to the theater. Why bother thinking about better codecs and Dolby TrueHD for your movies when the content is already so compromised. I’ll have to do some checking with my engineer friends about the home theater world. But as far as I know, the same kind of rolloff (although a different curve) is applied to television programming.
It looks like it’s going to be a long struggle for film lovers looking for high-resolution audio on their soundtracks. There are a lot of theaters that would have to adopt a standard audio specification. There is the THX standard but this doesn’t apply to the non-THX theaters…too bad.
Will there come a day when 96 kHz/-24-bit surround mixes will happen in all movie theaters? Undoubtedly, yes. They’ve managed to adopt the DCP format as a digital standard for showing films digitally…let’s just hope the soundtracks are next.
How is the sound if you buy the after market CD or Blu-Ray audio of the music from your favorite films.
I have Gladiator, Chariots of Fire, Miami Vice, City of Angels etc. do they go past 10,000Hz?
It’s all been produced using 48 kHz/24-bit PCM, likely compressed and heavily processed and filtered to model the sound in a theater…so don’t get too excited. It is possible that the music soundtracks are better and will punch through 10 kHz but they are limited to 48 kHz.
We must be looking at different graphs. The red curve is ruler flat from 50Hz to 20kHz and ~2db down at 40kHz; I wouldn’t call that aggressive roll-off. Most speakers and headphones will roll-off more aggressively than that including the Oppo-1.
Whoops my bad, I read 1kHz as 10kHz. The curve’s 8db down at 10kHz and since the curve appears linear (on the log log scale) would extrapolate to ~10db down at 20kHz. Still how many headphones or speakers would have any meaningful output at 20kHz or above?
Actually re-reading it would be 6db per octave after 10khZ which would drop the output by ~15db (just by eyeing the curve).
That’s a lot especially compared to a good set of speakers.
There are actually a lot of new speaker designs and even headphones that deliver great performance to 20 kHz and some well beyond.
I suppose it means you can listen to it on rubbish systems, you sure ain’t gonna need extreme frequency coverage from your amp and speakers now, are you?
I have to say I am not a great film viewer and I have no desire to build a 5.1 system at home. We had some friends round yesterday for a meal and drinkypoos and their son was raving about the latest Marvel-type films just out and coming out soon and to be honest it left me cold. The kind of films I usually enjoy are just as good on an ordinary TV, for example: Bogart & Baccall’s ‘The Big Heat’ or ‘Dead Poet’s Society’ or Judi Dench’s ‘Iris’ or the David Lean films. Yup there are times when I have enjoyed films with lots of special effects, like: ‘2001 a Space Odyssey’ or the 1st 3 Star Wars movies or ‘Lord of the Rings’. But it seems that special affect are now used instead of a bit of drama.
Bill
Bill…I’m not with you on this one. I took my 65″ Panasonic TV home and hooked up my extra THX B&W speakers. Experiencing television or movies at home is a new experience. Even my wife, who was originally resistant to the “monster TV”, is on board.
It’s just the type of movie that people play on these home surround systems that doesn’t appeal to me. I am sure that a surround system made up of good ancillaries will sound OK, even with the X-filter in place. But the majority of systems I have seen are put together with pretty naff components.
It’s like when stereo replaced mono – you had to double everything up and so to keep the level the same you were doubling your outlay (well pretty much anyway). I can remember a guy I met at uni in the 1970s, who had a wonderful two-channel system. It was the nicest system I had heard up to then. Well we lost contact for a a couple of years but when we had hooked-up again he had ‘upgraded’ to JVC CD-4 quadrophonic and it sounded pretty ropey. This is an extreme case I know. But remember there are plenty of Japanese HiFi nuts who still only listen to mono!!!!!
The thing is I would rather spend more money on something to play great stereo sound rather than compromising on 5.1 (6.1 whatever), especially as I am not a big action movie fan.
As I said in my original post Bogart’s ‘The Big Sleep’ is one of my very favourite films and I don’t think it will be improved with a multi-speaker home cinema system. Bigger screen – yes, just to get a better view of Bogart & Bacall in action.
Bill
Mark –
this post is being discussed over at Head-fi. the guy that originally posted the link to your post did some digging and says that the graph you cite was referring to the state of motion picture audio in the 1970s. I haven’t gone to the editor’s guild website to confirm, but I thought I’d bring it to your attention in case someone is in error. here’s a link to his post
http://www.head-fi.org/t/685704/oppo-pm-1-a-new-planar-magnetic-headphone/1620#post_10475461
Thanks Chris…the date of the publication doesn’t change the fact that movie sound is still lacking and not high-resolution.
While I appreciate the goal of HD audio for movies, I can’t say I’ve ever heard a movie-goer complain about sound quality when leaving a reasonably configured and operating theater. And since movies are a business, you’re unlikely to get much traction calling for high-fidelity sound if changing production & distribution workflow won’t translate to increased ticket sales. And, in my opinion, movies have a more significant problem with visuals than sound. I recently saw CAPTAIN AMERICA: THE WINTER SOLDIER at my local multiplex that boasted Sony 4K projection, and the visuals were terrible. The curved screen (designed to reflect more light toward the audience) results in un-even focus; completely unacceptable in my POV. I don’t know how this technique has been embraced; it’s laughable. The Sony 4K projectors have terrible color reproduction, but thousands of theaters bought them because they were less expensive than the Barcos and Christies, and converting to digital cinema was a huge investment, so corners were cut. And there was a significant bright spot in the middle of the screen, particularly visible on fields of even color, like the Academy ratings cards. The corners were much darker than the center of this image that should be uniform. So, while I would love to hear real HD sound, I think the movie presentation business has a much bigger problem with low-tech, budget-minded projection systems that give us poor quality images.
Hey Andrew! I’m with you on this. During my discussion with Andy, my film mixer friend and knowing what the location recording people are doing…I merely wanted to point out that movies are not likely to move to high-resolution AND in fact, they are less than that right now. The picture issues…I almost always go to IMAX theaters.
Hi Mark,
while I am a fan of your opinions generally, I think you got this one wrong, especially the topic of the X Curve. The 24/48 lossless bit rate — okay, for the purpose of discussion on this blog we agree to define 24/96 lossless as the HD pass line, but the audible difference between 48k and 96k is either nil or very subtle.
Regarding X Curve, firstly, your post did not *clearly* mention that it has nothing to do with the recorded music, i.e. it is a room-tuning equalisation applied in the cinema. Buy a recorded movie and there will not be an X Curve embedded in it somewhere.
Secondly, the X Curve is NOT the EQ curve applied to the soundtrack during playback. It is the in-room TARGET. The actual EQ curve that is applied to achieve the target could be anything, depending on the speaker system in use and the acoustics of each individual cinema.
Thirdly, it is NOT an attempt to roll off the high frequencies, as if someone does not want us to hear the high frequencies (or 20-40 Hz low bass). Consider this: we have a perfect set of loudspeakers in a room, dead flat from 20 Hz to 20 kHz or even from 5 Hz to 100 kHz, and with no other sonic imperfections, and we play continuous pink noise into a room and measure it un-gated. Do we get a flat line of constant amplitude? Only outdoors at a great height — or in a perfect anechoic chamber. In a small listening room e.g. a typical home, we get a falling curve, depending on room absorption factors and other room factors, but generally smoothly falling from 20 Hz to about -10 dB at 10-20 kHz. And for an in-between room size, such as a large movie theatre, we get a curve that is initially flat and then starts to fall as frequencies rise from 1 – 2 kHz. That is, SOMETHING LIKE THE X CURVE!! You see? The X Curve is nothing more than an approximate compensation for the measurement method. It approximately represents no roll-off at all in the high frequencies.
Having said all that, modern frequency analysis technology would allow some more sophisticated methods than un-gated pink noise analysis, but it would have to be standardised. Yes, we can do better, over time.
In summary, I think you rather strongly exaggerated the severity of the issues affecting movie audio playback in cinemas. If I assume your post fully represents your opinions, I would have to conclude that your knowledge was mistaken on this topic. And the idea that the in-cinema sound quality is far short of an audiophile’s home system’s sound quality, overlooks the fact that 99% of audiophiles use no room EQ whatsoever, and the X Curve represents an admirable attempt to have the same room sonics in the mastering studio and in the cinema, so the audience hears what the mastering engineer hears. The X Curve is integral to the movie industry’s sonic *advantage* over the recorded music industry, in that they try to break the “Circle of Confusion” that Toole and Olive refer to, and that is as a major plague on recorded music.
Boo. Bad post. Way below your usual class-leading material.
cheers
Grant
Grant…I acknowledged that this is not my area of expertise. MY discussion with an expert and someone that mixes films here and around town everyday centered on whether the film world was (is) interested in high-resolution audio. It seems they aren’t, if you generally agree with my position on that standard. The X-Curve was created for the reasons you state (I read Tom Holman’s piece) but it does result in a different high end response than a mastering studio for music…like mine. If I play my B&W 801s and then switch to the THX JBL system WITH the X-Curve applied it is very different.
I understand why the film people are doing what they’re doing. We go to the movie theaters and expect the same experience no matter where we are. At home or in our cars, we can tweak as much as we want….but with sources that have more native fidelity than the movie soundtracks.
Thank you Mark for not over-reacting to my too-strongly-worded comments. I do apologise.
Regarding your comments above, I agree. Note that, although your THX JBL studio with X Curve does not sound like your 801 (straight) studio, once you EQ a soundtrack master so it sounds great (with no missing HF) on the JBL kit, it will sound similarly balanced in cinemas.
Grant…no need to apologize. I took your points and acknowledge that my expertise is lacking in this area. Onward and upward.
Thanks for this, Mark. I’ve always thought movie soundtracks sounded muffled and dull compared to CDs, and now I know why! Heck, this spec is inferior to FM radio.
In my opinion most movies sound duller because they’re not overly hyped with processing and aggressive radio-busting EQ like most non-classical CDs are, which puts them much closer to true reality in terms of sonic spectrum. While I’m not going to pretend that 48KHz 16 bit Dolby Digital playback is audiophile quality, I’d much rather listen to that any day (if watching movies were equal in value to music listening), given the dynamics and beauty of a good movie mix, than a loud brick of dead pop audio from an uncompressed CD or HDTracks.com. Movies have amazing music recordings, wide dynamics, and stunning sound staging. There is still even the rare movie where none of the channels actually reaches 0dB/FS. Find me any current pop CD that could compare. But I digress, this post isn’t about dynamics.
So bear in mind that Blumlein invented, demonstrated, and patented stereo film recording in 1931, but it took 45 years to get the industry to mainstream it. He did that with the stereo phonograph record also, and that took a quarter-century to get into consumer’s hands. Point: the industry is slow to accept certain types of change. Mark, keep preaching this but don’t expect much, even though it might be as easy for an engineer as pushing a button.