US2178217A - Production of sound motion pictures - Google Patents

Description

Oct. 31, 1939. l 1 0, BAKER 2,178,217

PRODUCTION OF SOUND MOTION PICTURES Filed Deo l0, 1937 4 Sheets-Sheet 1 F1a. 1. EG. 2.

PwcfmoM/JT/c P/vc//RoM/vr/c F'lLM F'VLM PICTURE g, .s100/vn P/CTURE g Sol/,VE

w/ 7' w/ 7W NEG/9W VE K Pos1 7'/ vE 'Ol//VD 71E/7c SOI/MD '177/9 C/f s s '0 'o H/GH CONTR/1.57' l] HIGH CONT/M57' Pos/ 7'/ VE Pos1 7'/ vg FILM 5 WLM U Pos/77ML' soz/,vn j 56'97" "E 777/76,( SOUND 'FH/70h bl D: Q

Q Q to 'o l sou/VD ,VEG/77V VE HELE/76E F'lL M F06/ 7'/ VE PWVTE 0H P/c v'l/RE RECORDED a,

NEG/7 7'/ VE 6oz/ND WMC/f 'SOUND RELEHSE Pos1 7'/ VE PICTURE mvp.

Jou/vn :inventor Lltorneg ct. 3l, 1939. l J., o, BAKER 2,178,217

PRODUCTION oF SOUND MOTION PICTURES Filed Dec. 1o, 1937 4 sheets-sheet 2 BEING/Ty l l l feu;4 EXPOSURE :Snventor Judd 0. Baer dttorneg O ct. 3L 1939. J. o. BAKER 2,178,217

PRODUCTION OF SOUND MOTION PICTURES Filed Dec. lO, 1957 4 Sheets-Sheet 3 L` W6 t` WS Fifa.

Snventor Judd Baler Oct. 31, 1939, J. o. BAKER PRODUCTION OF SOUND MOTION PICTURES Filed Dec.`

4 Sheets-Sheet e a 7|65 4 a 0.0 0 0. 0. 0A 0. OIO. 0 0 difwo 0 7 nv 1111s! [615114, a z Il \.4 0 0 0 0 0 vw. 9 RN .7 .w w a n Patented Oct. 31, 1939 PATENT OFFICE PRODUCTION OF sOUND MOTION PICTURES Judd 0. Baker, Camden, N. J., assigner to Radio Corporation of America, a. corporation of Delaware Application December 10, 1937, Serial No. 179,071

2 claims.

This invention relates to the production of Sound motion pictures, and has for its principal object the provision of an improved apparatus and method of `operation whereby a sound record track may be recorded on the same photographic film as a motion picture and may be subjected to the iilm processing technique most suitable for the picture without resulting in the reproduced sound distortion heretofore encountered.

It has long been known that emulsions with I ne grain and high contrast give superior results for variable width sound recording. The investigations made 'by Hoxie of the General Electric Company in 1921 led to the adoption of positive types of emulsion for this purpose. In single lm recording systems, such as news reels, where the picture and sound are recorded simultaneously on the same film, however, the sound must be subordinated to the picture. Thus the coarse grain of panchromatic emulsion together with the method of picture processing for an over-all gamma of unity results in a sound track inferior in quality to that obtained with fine grain recording emulsions. This, to a large extent, accounts for the poor quality of the sound reproduced in connection with news reels and other sound pictures recorded by the single iilm system.

When variable width sound track is recorded with ultra violet light on panchromatic lm, is processed in accordance with the usual commercial technique for motion pictures and is printed with ultra violet light on the usual positive film, the image denition on the panchromatic lm is inferior to that on the positive lm.` The ultra violet lter is used in recording to reduce the iniage spread and to improve the response at higher frequencies. This permits the recording of a higher track density but limits the maximum density obtainable due to the restriction of the recorder light to a narrow spectral band.

With White light recording on panchromatic nlm, the best results are obtained with equal negative and print densities of about 0.80. rlihese are the conditions which most nearly satisfy the requirements for minimum image distortion in release positives. Ultra violet light recording reduces the image spread in the negative thereby permitting a higher density in the recorded track.

The base or fog density of panchromatic lm is0.38 and, with a recorded track density of l or greater, the dierence in transmision through the opaque and clear portions of the track is considerably reduced. The printer light therefore penetrates the clear portion and produces on the print an exposure which, for perfect sound track, should be unexposed. This exposure or fogging of the clear part of the positive track results in reduced output and increased noise. In accordance with the present invention, this difliculty is largely avoided by the use of a high contrast emulsionlm such as Eastman No. 7100 in an intermediate step for printing the sound track from the panchromatic sound picture negative to the released sound picture positive.

If the sound is originally recorded on the panchromatic ilm in the form of a negative sound track, this new procedure involves the additional steps of printing the track on a high contrast positive nlm and of printing or rerecording the track to a negative from which it is printed to the positive sound picture release print. If the sound is originally recorded in the form of a positive track, only the rst of these additional steps is involved.

For the production of a positive form of reverse negative sound track in the original recording, use may be made of one of the various types of recorder disclosed in a copending application of Glenn L. Dimmick, Serial No. 168,173, led October 9, 1937, and assigned to the same assignee as the present application. As will be readily understood, this type of recording involves a reversal in the form of the recording beam. Otherwise expressed, this type of reversed negatinve recording involves the use of a light beam of a form complementary to the light beam used for recording the track in a negative form.

The invention will be better understood from the lollowing description when considered in connection with the accompanying drawings and its `scope is indicated by the appended claims.

Referring to the drawings:

Figure 1 is a block diagram illustrating one iorm of the invention,

Figure 2 is a similar diagram ilustrating a modied form of the invention,

Figure 3 ilustrates the characteristic curves of the photographic lms utilized in practicing the invention, and

Figures 4 to 7 are explanatory diagrams.

In accordance with the method of Fig. 1, the sound and picture are recorded on panchromatic lm and are processed in accordance with the usual motion picture technique, thus resulting in a sound picture negative having a fog density of about 0.38 and a gamma of approximately 0.55. In commercial practice, the gamma may range from 0.4 to 0.75. From this sound picture negative, the picture is printed to the release positive print and the sound is printed to a high contrast positive film such as Eastman No. 7100, thus resulting in a sound positive having a fog density of about 0.02 and a. gamma of approximately 4.0. From this high contrast positive, the sound track is printed or rerecorded to a negative from which it is printed to the same release print as the picture.

In the modied procedure of Fig. 2, the sound track is recorded together with the picture on panchromatic hlm in the form of a positive or reversed negative, is printed from the original panchromatic lm to the high contrast positive in the form of a negative track which is thereafter printed to the same release print as the picture. In either case, the procedure involved has the important advantage of greater volume range and lower ground noise in the sound reproduced from the release print.

How these advantages are realized will be more readily understood from the characteristic curves of Fig. 3 in which the various curves are indicated by self explanatory legends. It will be noted that the standard motion picture positive has a fog density of 0.05 and a gamma of approximately 2.2. For minimum image distortion, the' print track density must be equal to the negative track density. If A and B represent the print track density points of the characteristic curves of the two positive emulsions, then H and G indicate the fog densities in the clear part of the sound track. Thus, the fog densities for standard motion picture positive and for high contrast positive are respectively 0.25 and 0.05.

As previously indicated, fog increases the background noise and decreases the volume of the sound reproduced from the release print. The extent to which background noise increases and Volume decreases in the case of the two positive films under consideration may be determined by comparing these two tracks with a perfect track which, in the case of variable area recording, would be a track having its clear part altogether transparent and its dark part altogether opaque. Based on such a comparison, the relation indicated by the following tabulation is found to exist:

It Will be noted that the high contrast positive track has the advantage of much higher signal level, lower noise level and lower high frequency loss.

In establishing a common basis for comparing the characteristics of different photographic sound records, the expression densitometric level is hereinafter used to indicate the ability of a practical sound record to modulate the light beam as compared to the ability of an ideal or perfectA record tcmoduiate-the light beam.

The ideal or perfect variable density sound track would cover the full length of the scanning beam and vary the density between the limits of perfect transparency and perfect opacity. The ideal variable width sound track would have a maximum Width equal to the length of the scanning beam, the clear portion being perfectly transparent and the vdense portion being perfectly opaque and varying in Width from zero to the length of the scanning beam.

The most absolute basis for rating lm level is in terms of purely optical considerations without reference to any reproducing system. On this basis the level in decibels of an ideal film having 100% transmission at the peak of the Wave is used as a reference level andfilm recordings are rated as being so many decibels below the ideal film.

. .Since the photo-cell current is directly proportional to the amount of light received by the cell, a value of 50% transmission means that the current will be one-half as great with the lm in the light beam as it is with the lm removed from the beam. It is therefore feasible to measure transmission by measuring the photocell current obtained under the conditions of lm in the light beam and lm removed from the light beam, the value obtained in the former case being divided by that obtained in the latter case.

By employing a narrow light beam to scan the sound track and noting the resulting photo-cell current at many points along the wave, we may plot a curve similar to that shown in Fig. 4.

Such a curve is usually described by stating both the projected transmission, which is the average transmission representing the axis of the AC wave, and the percent modulation, which is the factor by which the photo-cell direct current representing the axis of the wave must be multiplied to obtain the peak value of the alternating current wave.

It is known from fundamental AC theory that the eiective (R. M. S.) value of a wave is related to the peak value in a definite ratio according to the wave form. Thus, for a sine wave the effective value is obtained by multiplying the peak value by 0.707 and for recorded frequencies which are sine waves, this value must be applied to the transmission and modulation values given above. This gives the following formula:

Ieff=0.707 TMI@ (l) in which Ieff is the effective current value of the signal, M is the percent modulation, T is the projected transmission, and 0.707 is the ratio of effective to peak value for a sine wave.

It is seen, by examining Fig. 4, that in order to have 100% transmission at the peak of the recorded wave of the ideal film, the axis of the wave must represent 50% projected transmission and the modulation must be 100%. Then by application of Formula 1 above and the denition of decibels, the following expression results:

Densitometric level 0.707 TML,

2o 10g lg=2o 10g TM (2) in which T and M represent the projected transmission (50%) and modulation (100%), respectively, of the ideal film.

The equations as developed for the projected *(Wsh" W2)Tc+ W2Td where (see Fig. 5)

WS=0.084-standard SMPE scanning slit length W1=greatest scanned width of olea-r portion W2=average width of dense portion W3=least width of dense portion,

and

Tc=transmission of clear portion of track Td=transmission of dense portion of track.

Multiplying Equations 3 and l4, reducing to least common denominator and substitution, Equation 2 becomes; densitometric level, (variable Width) For the purpose of developing the expression for densitometric level of a variable density sound track, consider the projected transmission and percent modulation in terms of the photo-cell currents.

Projected transmission and modulation Where Ia :average current with film in the light beam Im :maximum current with lm in light beam In :current with no lm in light beam v La Im' Ia Im-Ia TM- I., I., I,

The photocell currents are proportional to the transmission a-nd the width of the respective slits m=KTmWt, Ia=KTaWt and Io=KToWs Hence Then densitometric level (variable density) db.=zo1og 2(T,T)1og Figi (6) 'I'his expression of course can only be used when the recorded frequencies are pure sine waves.

The first term of the variable width Equation 5 considers the relation between the densities in the dark and clear portions of the sound track, while the second term considers the mechanics of the sound track and the scanning slit.

Rewriting the second term with Wt as a common multiplier db. =log (Te: Td)- 20 log and making an additional term log 20 log g Substitute (7) into (5) densitometric level db.=2o 10g (rpm-20 10g 2(W2-W3) is the amplitude of the recorded Wave and the ratio 20 log is the percent modulation.

then brne's simply the level of the wave traced by the galvanometer in decibels below 100% modulation which is convenient for measuring purposes when either the uni-lateral or bilateral types of track are used. The galvanometer can easily be set for 100% modulation by observation of the light vibrations on the slit. The galvanometer input can be measured with a decibel meter and as the input is reduced, the difference in decibel readings gives the value of the term 2(W2 Wa) 20 log Wt immediately.

The term Wa 20 10g wt Equation 5 reduced to its most simple and convenient form for routine measurements becomes densitometric level db.=20 10g (Te-Tf1) 0.9-i-K (9) where K is equal to the reduction in amplitude of the recorded Wave below 100%v modulation or full track width.

It can be seen from Equation 9 that the ideal film to be strived for would be one with 100% transmission in the clear portion and zero transmission through the dark portion. An ideal film of the variable width type with the present stand- -20 log =-20 10g 0.9 db.

ard dimensions would have a densitometric level A of D. L.: 0.9 db. f

The D. L. of the present ultra violet recordings with a track density of 1.3 and a fog density of .04 and 100% modulation would be Since densities are easier to measure than transmissions a conversion chart, Fig. 6, may be resorted to based upon the relation Fig. 7, which is a chart of 20 log (T2-T1) plotted against (T2-T1), makes it easy to determine the densitometric level of any lm.

For recordings of the variable density type, Equation 6 fory densitometric level is made up of two terms: densitometric level as in the case of the variable width, the rst term is the loss due to the diierence in average and maximum transmissions, while the second term is the loss due to the mechanics of the sound track and the scanning slit.

For the old standard of 100 mils for variable density recording Wt is equal to the scanned width and the second term drops out reducing the equation for densitometric level to densitometric level db.=2010g 2(Tm-Ta) (10) For a perfect lm of the variable density track with the old standard of width the maximum transmission must equal unity and the average transmission must be equal to 0.5. The densities for these values of transmission would be Dm: innity and Da=0.3. It is n-ot feasible to obtain an infinite density in practice but a density of 3.0 is a practical approach.

For the new standard (A. of M. P. A. & S.) of '76 mils for the variable density track width, the second term of Equation 6 becomes the same as that for the variable width and the expression for densitometric level becomes densitometric level db.- -20 log Z(Tm-Ta) 0.9 (11) It is not practical to measure the percent modulation of variable density recording as was done for the variable width type, hence it will be necessary to determine the densitometric level for variable density recording from the measurements of the appropriate densities.

The application of the equations just developed for the densitometric level would be as follows:

1. Determine the densities Dd and Dc, in the case of variable width; Dm and Da, in the case of variable density either from measurements or theoretical considerations.

2. Convert to equivalent transmission values by means of the chart of Fig. 3.

3. Determine thevalue of 20 log (Te-Te) in the case of variable width or 20 log 2(Tm-Ta) in the case of variable density from the chart in Fig. 4.

4. Substitute the values obtained in 3 into the following equations:

A. Variable width D. L. in db.='20 log (Tc-Td)-0.9|K

where K is the db. below modulation used in recording the test track. B. Variable density (a) D. L. in db.=20 log 2(Tm-Ta) for 100 mil recorded track, and (b) D. L. in db.=20 log 2(Tm-Te)- 0.9 for '76 mil recorded track.

Among the more important factors to be considered in investigating emulsions, developers and recording technique for sound recording on lm are volume range, frequency range and relative value of lm noise to signal level. When making comparisons of emulsions, developers or types of track, it is essential to have some standard to which all measurements or theoretical considerations can be referred. The ideal lm is logically the standard to use, since densitometric level is a measure of a practical lm in terms of the ideal lm. Its use, however, is limited to the comparison of volume range and relative output levels of signal and noise. 'Ihe modulated high frequency recording is stillnecessary, for the determination of proper negative and print densities for minimum distortion due to image spread. Likewise, it is necessary, at the present time, to make recordings of various frequencies in order to determine the frequency range. Densitometric level measurements on a reproducing equipment can only be made at the lower frequencies where the resolution of the lm is the greatest and the image spread is of least importance. Once the output readings of a fixed measuring system have been rated in densitometric level, all measurements made on this system can be easily and quickly converted to terms of densitometric level.

I claim as my invention:

l. The sound motion picture producing method which includes recording the sound and picture on a photographic record having a fog density of the range of 0.38 and a gamma of the range of 0.55, printing the sound track into a second record having high contrast positive emulsion, a fog density of the range oi 0.02 and a gamma of the range of 4, printing the sound from said high contrast record to a release print, and printing the picture from said photographic record to said release print.

2. The sound motion picture producing method which includes recording the sound and picture on panchromatic lm, printing the sound track onto a second record having high contrast positive emulsion, a fog density of the order of 0.02 and a gamma of the order of 4, printing the sound from said high contrast record to a release print, and printing the picture from said photographic record to said release print.

JUDD O. BAKER.

Images