|
|
 
|
|
Author
|
Topic: Cyan Tracks
|
|
|
John Hawkinson
Film God
Posts: 2273
From: Cambridge, MA, USA
Registered: Feb 2002
|
posted 07-22-2003 05:27 PM
My, this thread is getting long. Way back on page 4, I asked "what's the quantitative difference between red and infrared on reverse scan readers? And on forward scan, too?" I did quite a bit of digging today, and I'm not pleased with the results.
By far the best source of information seems to be the SMPTE Journal paper from Goldberg, Allen, Ricotta, & Sehlin, Red Readers, Silver, High-Magenta, and Cyan Analog Sound Tracks-Review and Status. That's available on the dyetracks.org web site as well as dolby.com.
Curiously, most of the data is available in pictorial form only, and not everything is so clear. Measurement of soundtrack reproduction is done in 3 ways: amplitude, crossmodulation cancellation, and signal-to-noise ratio (all measured in dB).
Data from Goldberg, et al
A conventional tungsten exciter lamp will see an 11dB loss in amplitude going from a silver redeveloped soundtrack to a cyan soundtrack.
A red LED reader will see 2dB less gain on a silver soundtrack than a tungsten exciter lamp.
A red LED reader will see 20dB worse crossmodulation cancellation on a silver soundtrack, versus a tungsten excitier lamp.
A red LED reader will see 8dB (on average) more noise (not SNR) on a silver soundtrack, versus a tungsten exciter.
A red LED reader will see 3dB worse crossmodulation cancellation on a hi-magenta print than a tungsten exciter lamp.
Data from Kodak
Kodak's Lab Guide claims a red LED reader will see 2dB worse SNR on a silver soundtrack than a tungsten exciter.
Data from Dolby's patent
Dolby's patent No. 5,587,749 ("Apparatus and method for reading motion picture film photographic dye soundtracks") claims a cyan-dye track with a red LED reader will see 3dB poorer "audio noise performance" than a conventional tungsten exciter.
No reverse scan data?
It's very unclear to me whether these numbers include the gains due to reverse scan or not. I really could not decide reading the various documentation. I also couldn't find any good information on the quantitative benefits that reverse scan provides. I looked for patents, and did some web searches, but came up empty-handed. Does someone have this data?
About water
I believe the issue about water has been misunderstood by some people. It is not about simply use of water. It is that the water used to wash the silver redeveloper off is contaminated with silver, and thus cannot be discharged into the environment directly, and costly cleanup measures must be taken. Even then, the water still contains environmentally harmful silver.
Trailers
Hey, Brad, I think that's a good idea. Would you please email the dye track folks and let us know what they say?
--jhawk
| IP: Logged
|
|
|
|
|
|
|
|
Stephen Furley
Film God
Posts: 3059
From: Coulsdon, Croydon, England
Registered: May 2002
|
posted 07-22-2003 06:02 PM
Steve wrote
quote:
The time scale of prints have never been considered that of a few months to a few years despite your factual statement about how long the typical print lasts. Most would consider that a print that is designed with such a short life-span would be considered defective.
Agreed. I have, or have had in the past, 16mm diacetate and 35mm nitrate prints going back to the '30s. I have seen colour prints, mainly Technicolor, a few in various odd processes, from the same era still in good condition. I have seen, at the National Film and Television Archive, orginal film going back about a hundred years, still in good condition. If I had a print which faded very quickly then yes, I would think it was defective. I don't believe that a print with a cyan track made on modern stock and proerly processed, would have 'such a short life-span', I believe that, unless stored under poor conditions, it would have a life probably extending into decades without serious fading. If you store the print in hot, humid conditions, then it will fade faster, and you will also get other problems.
I am not saying that a print would only last a couple of years, but, in the case of normal release prints it only *needs* to last that long. If it lasts longer, as I believe it would, then thats fine. If a track was used which would last only a few weeks, then that would not be satisfactory.
The words 'archive' and 'archival' in relation to film tend to get given two meanings:
Various original and intermediate picture and sound elements intended for permanent storage.
and
Projection prints retained indefinately after the general release of a film.
The former are not going to use cyan tracks, so it is not an issue.
The latter are, in my view, the only place where the life of a cyan print is an issue. How long these prints should need to last is open to debate. Certainly years, probably a few decades, probably not longer than that. I would suggest that rather than trying to make a print last forever that it would be better to concentrate on keeping the true archival preservation evements, as above, in good condition so that new prints can be producd if required. Having an 'archive' print that doesn't fade won't help you in that print gets damaged in a projector. Some degree of expendability has to be expected of prints. Given that the number of these prints is only going to be very small, compared to the total number of prints made, then, if the life of a cyan track is not considered adequate, perhaps there is a case for a few special prints being made for this purpose with silver tracks. If these prints went straight to the vaults, rather than finding their way there after a normal cinema run, they would be in better condition. Of course, This option would be more expensive, a few 'special' prints will always be so, and I believe that a different track negative would also be required.
For the vast majority of 'normal' release prints, the cyan track will last far longer than it needs to.
The paper on which my newspaper is printed will yellow and become brittle if a fairly short time, but it will last far longer than the day or two that it needs to. If I need to keep a story long term, then I will make a photocopy on good quality paper, and I would hape that both the text and pictures are kept in archival storage somewhere on some suitable media.
A cyan track will last longer than it needs to on a normal release print. There may be other disadvantages to it, but that is a different matter.
quote:
As to fading, certainly today's stocks are superior to old, that I have never disputed. As to comparing an audio track fade to a picture fade, one CAN NOT equate the two. If the cyan density changes in the picture, while one can chart the change the human brain is pretty good about mentally shifting things so they appear correct. An audio reproducer would not have such benefit. While you can see a print that may look off color, pretty soon, most people stop noticing that and end up watching the film. To a lesser degree elivated constant noise will also be filtered by one's brain.
Agree about the overall colour balance, I think you do still notice that the shadows are not 'black'. Glad you brought up the brain. We see with our eyes and brain, and hear with our ears and brain, and it's difficult to prodict how the brain will respond. Some people notice these hings much more than others. I find a certain amount of hiss, as on old magnetic recordings much less disturbing than mains hum (it's not just that my response at the top end is dropping off with age, I've always found it that way) I can hear the hiss if I listen for it, but I find it less disturbing. Some people react in exactly the opposite way.
quote:
If the gain is not stable over time, you will have other issues. Dolby-SR (the only proposed system for cyan at the moment) is a level dependant compander (as is Dolby-A). If the levels and dynamics are not matched to the encoder, the playback will mistrack...simply turning up the gain will affect overall playback.
The Dolby issue only crossed my mind as I was writing the previous post. I don't know a great deal about the technicalities of Dolby, I understand the general principals, and I can set up an A chain on a CP, but it would take somebody more knowledgable than I to say how much fading of a track would
be needed to cause an unacceptable, and uncorrectable change in the sound as heard in the auditorium. I'm not sure how you would define 'unacceptable', would it be acceptable if one person with very keen hearing could just detect a very slight change in the sound, but the rest of the audience could not? How about if it could be detected by instrumentation, but not by any of the audience?
I still think that a cyan track will last for longer than it needs to on a normal release print.
quote:
When looking at older prints that have had considerable dye fade, you have to look at two things, you have the fact the application is on there AND the magenta layer is still there. Application plus magenta = "high-magenta." Glory be, that works for all kinds of readers, albeit a bit noisier than proper cyan + magenta and application. With cyan only, you are hedging your bets on a single dye layer that must precisely interact with the narrow wavelength of the light source.
What you seem to be saying here is that, in absolute terms, a cyan only track will be less stable over time than a silver plus dye one. I would agree with you, if you were to measure the optical density of the track, at verious intervals, with a sensitive transmission densitometer, at the appropriate wavelenghts, I am sure that you would observe a greater change in a cyan track than in a conventional one, over an equal time period. It comes down to how great that change will be, and over how long a time period, and whether such a change is acceptable for the purpose of a release print; the only place where its use is proposed. I think it will be, you think otherwise, this seems to be the point on which we disagree. We seem to be pretty much in agreement otherwise.
I would rather see a conventional track used; I still don't know whether cyan is going to happen or not, I still haven't seen a single foot of it (anyone care to lend/donate me some to test?) If it does happen then I see the cost and compatibility issues to the theatres, who will face greater costs, for smaller benefits, than other parts of the iduustry, as being more of an issue than the life of a track which I expect to last much longer than it needs to, even if not quite as long as a conventional track.
quote:
Your comments about SDDS are interesting...of the various problems associated with SDDS, one of them is actual track damage (tracks that used to play but now don't). If the cyan layer is so well protected (2nd layer and covered with the geletin) then why are those tracks being damaged? Now I will grant that the SDDS track are will come into contact with just about any film guidance system, roller or what have you and the optical track should not (though double flange pad rollers do)the scratching effects of the cyan layer are real. Furthermore, we are using collimated light...any scuff on the outer layer will create a distinct shadow that will be accurately read by a reverse scan reader. This is the chief issue I have with both IR and visible LED based readers. As to other SDDS issues, I believe density at the labs have been one of the major ones. Certainly there are other issues with the SDDS tracks.
If I have understood you correctly, you are saying that mechanical wear and damage to the track areas of the print is the main cause of tracks which originally played later faling to do so, and that poor control of exposure and/or processing of the track at the lab is sometimes resulting in a track which is less than ideal in the first place. Is that correct? The lab issues wouldn't cause a track to deteriorate over time, but might result in a marginal track, which was more likely to become unreadable after minor damage, so that would make sense. I have no experience of running SDDS, it seems to be by far the least common of the three surviving digital systems here, I've only ever heard it once, ond that was because I went to the Odeon Marble Arch, before it was devided up, to see 'Mary Shelley's Frankenstein' just to hear the system. I wasn't very impressed by what I heard, but, of course, I cannot really judge a system by hearing one print of one production, in one theatre.
I do see prints with SDDS tracks from time to time, and they do seem to suffer more from both dirt and wear in the SDDS track areas than elsewhere; this would seem to be the likely cause of prints failing to play over time. Of course, if the dye in dhe track was fading, this could have the same effect, but I do not observe any obvious fading of the track if I compare a print a year or so old with one of a much more recent release, I don't think I've ever seen a brand new SDDS print. I have compared the clip of film with an SDDS track which comes with Dion Hanson's 'Fantastic Formats' booklet (about four years old now, I think, so the clip must be older than that) with a track on a film a rew weeks old. I haven't measured the density, but again, by visual inspection they seem to be closely matched, so fading of the track doesn't seem to be a problem, at least over this time span. Obviously, I don't have any fifty year old SDDS tracks to try!
That was a much longer reply than I intended!
| IP: Logged
|
|
|
|
|
|
John Hawkinson
Film God
Posts: 2273
From: Cambridge, MA, USA
Registered: Feb 2002
|
posted 07-23-2003 04:22 AM
No, Frank, it is not an error. This 20dB crossmod cancellation is the reason we have the interim step of hi-magenta instead of going directly to cyan-only.
I cannot speak to exactly what crossmod cancellation is or its relationship with reverse scan's features (better high frequency reproduction?), but I hope someone else can.
Incidently, can someone then confirm for us that the US is at 100% hi-magenta? For the big labs (Deluxe & Technicolor), the small labs (Duart, lablink, Tape House Digital, etc.), and the custom houses (Filmack, Wilming-Reams, Tim Reed, etc.)?
--jhawk
| IP: Logged
|
|
Stephen Furley
Film God
Posts: 3059
From: Coulsdon, Croydon, England
Registered: May 2002
|
posted 07-23-2003 06:52 AM
Steve wrote:
quote:
An excellent reply indeed. We seem to be in agreement on most issues with the longevity of the cyan track of chief dispute.
The chief difference in our arguements is that you feel or believe the cyan tracks will have enough life in them while I claim (see many previous posts) that the cyan-people have offered no long-term testing to back up your feelings or belief. Remember, they are proposing a new and universal soundtrack to the industry and ramming it down everyone's throats...it is up to them to prove its durability...and I don't mean with a limited test run. They are already starting in the hole on performance. Their dubious claims about the cost of applicated soundtracks have not and will not show up on the exhibitor's bottom line. If the labs were to send out a red reader to anyone that asked for it for the first year of release, then the issue of $$$ would be moot. Instead they have chosen for the exhibitor to pay for their downgrade.
I reserve judgement of the subjective quality of the sound from these tracks until I have been able to hear several different ones, on several different readers; since there is no sign of me being able to find any film in the forseeable future, that may still be some time away.
There's still one point in your concern for the stability of these tracks that I'm not clear about. Do you doubt only that these tracks will last as long as the print is capable of doing, about which I agree that there must be some degree of doubt, or are you unsure that it will even last as long as a normal general release print has to until it is scrapped without noticable change in the sound, which I think is most unlikely.
From the point of view of the exhibitor there are, at least in some cases serious issues. Take the case that Scott was talking about; is somebody going to pay the cost of some custon engineering to convert those Western bases, or even to buy new projectors.
The theatres are being told to convert their machines, and have been told for years that these tracks are coming, but few people have even seen any film. You cannot blamepeople for being reluctant to spend money converting machines when there is no indication when, of even if, these tracks are actually going to arrive.
Finally there is the point that the labs are going to be making substantial financial savings on every foot of film they print, at almost no cost to themselves, while the exhibitors are going to be facing the greatest costs, for, at best, minor advantages. I repeat my call, if these tracks aregoing to be introduced, for the labs to pass on some of the savings, via the distributors, to the exhibitors.
I don't dispute the advantages of these tracks as John has given in his advocation of them, but there are also disadvantages, and I don't see why those who stand to see few of the advantages, but most of the disadvantages, should be paying most of the cost.
| IP: Logged
|
|
|
|
Stephen Furley
Film God
Posts: 3059
From: Coulsdon, Croydon, England
Registered: May 2002
|
posted 07-23-2003 08:22 AM
Ken wrote:
quote:
People seem to forget that some old movies are still here with us today not because a studio carefully preserved a film, but rather because some film collector preserved an ex-cinema print! Sometimes studios 'lost' negatives, master postives, track negatives, etc, or had fires that destroyed precious originals. The film collector to the rescue! Soundtrack stability is very important on a print.
Of course don't forget that studios don't or won't reprint some films and revival showings can't get anything but a beatup, faded print to show. Again another case where dye stability is important.
Films have indeed be restored from a number of unlikely materials. Silent films copied from old Pathescope 9.5mm prints and long discarded episodes of Dr. Who where chroma from old Betamax home recordings was combined with luma from 16mm black and white kinescope recordings made for export come to mind.
If you were to be considering restoring a film from a faded release print you would have to be getting prtty desparate. I feel that this is less likely to happen in the future than it may have been in the past, for two reasons.
Firstly, most old films in the past were largely seen as being of little value, they were discarded to save vault space, or neglected, and deteriorated beyond restoration. Today there is a large market for old films on broadcast television, cable, video etc., therefore they are more likely to be seen as a potential source of future profit, and therefore looked after.
Secondly, if you were proposing to restore a film you now have many more potential sources of the track available to you, and in some cases far more copies of each. Let's say you go to the vaults, and all of the archival elements, optical and magnetic are missing or decayed. In the past the best you could hope for was probably to find an old print somewhere, probably worn, and possibly with frames missing. Today you have other options, various masters will have been made for video transfers, if you were lucky you might be able to get hold of a multi channel PCM or Dolby E master, if not you might have to settle for a consumer Dolby Digital or DTS master which would have to be decoded and then re-encoded in the required format, with some loss of quality. At worst you go out and get a NTSC DVD and take the track from that, there will have been far more of those made then there ever were film copies, and it is highly likely that at least one can be found which is still playable. Other possibilities would be to find a set of DTS CDs, or even to reproduce the digital tracks from a print.
An old print with a cyan track really would be the last possible option, and I think you would be able to find something better for a film produced today.
| IP: Logged
|
|
John Pytlak
Film God
Posts: 9987
From: Rochester, NY 14650-1922
Registered: Jan 2000
|
posted 07-23-2003 08:34 AM
Kodak also has an on-line tutorial about cross-modulation testing to assure optimum printing of analog soundtracks with minimum distortion:
http://www.kodak.com/US/en/motion/support/h44/
The original concept was published in the SMPE Journal:
J. O. Baker and D. H. Robinson, "Modulated High-Frequency Recording as a Means of Determining Conditions for Optimal Processing," Jour. SMPE, 30:3-17, January 1938
And is specified by SMPTE Recommended Practice RP 104-1999, "Cross Modulation Tests for Variable-Area Photographic Sound Tracks"
The optimum density for traditional (cyan + magenta + silver) tracks was different for white light (tungsten) readers and red LED readers, meaning that a print optimized for one, would have some distortion on the other. Cross-modulation distortion may cause voice passages to contain an excessive amount of sibilant distortion. That is, many of the sibilant speech sounds (s, sh, ch, z, etc.) and some of the voiced and voiceless stop sounds (B, P, and T) will be distorted and spitty.
"High Magenta" (magenta + silver) tracks have an image spread characteristic that reduces this difference, making "high magenta" tracks more compatible with today's mix of old tungsten readers and the newer red LED readers found in most theatres.
| IP: Logged
|
|
|
|
|
|
|
Powered by Infopop Corporation
UBB.classicTM
6.3.1.2
The Film-Tech Forums are designed for various members related to the cinema industry to express their opinions, viewpoints and testimonials on various products, services and events based upon speculation, personal knowledge and factual information through use, therefore all views represented here allow no liability upon the publishers of this web site and the owners of said views assume no liability for any ill will resulting from these postings. The posts made here are for educational as well as entertainment purposes and as such anyone viewing this portion of the website must accept these views as statements of the author of that opinion
and agrees to release the authors from any and all liability.
|
Home
![[Wink]](wink.gif)
Printer-friendly view of this topic