anyone else noticed dolby's calibrated cat 69t's test film suffers from rather highish (4.0-4.6) failure rates? We're talking after pain stakingly aligning a dolby reader on a kinoton FP-30, and on a CP650 running i *think* it was 1.1.7 firmware.

I've seen prints running 20 mins after the alignments with error rates in the 1s and 2's.

Dolby's test films have been rather unreliable for some time and hardly the model of Digital Correctness.

If you get batches of Cat. 69T, you will find that the level does vary somewhat and it doesn't agree with the older B&W versions. The HF response on Cat. 69P is truely pathetic. If you tune your sound head with it, watch out on the slit loss...if you set Cat. 69P for flat, you will in fact have boosted the HF too much. If you still have any, check it against some B&W Cat. 69 (or even Ultra Stereo's TT-1).

We have taken to getting 1000' rolls and measuring them against reference films to derive an offset (in 1000' lengths, you sometimes even get a QC sheet to compare against).

In any event, yes 4s and 5s on Dolby Cat. 69T or P are more common than not. For what it's worth, I have found that the current Kinoton readers (with the new super sensitive lens...4/02 and later) to out focus all other readers I've tested, penthouse readers by Dolby included! I was quite surprised.

Steve

 |  IP: Logged

 |  IP: Logged

 |  IP: Logged

So who is correct? The SMPTE or Dolby? We decided that the SMPTE is the culprit because it is the buzz track that seems to vary, not the Cat97.

Also the latest version of Cat69P (which from memory is dated 99) is flatter at the top end than the '97 (or is it 94?) version. It is a shame that Dolby did not follow the long time practice for magnetic test tapes for studios whereby a test tape (film) is either absolutely flat, or if not, is supplied with a correction chart to tell you where and by how much it varies from flat. Otherwise how on earth do you know what is flat and what is not?

In our work developing reverse scan readers for another company we took the approach that the Dolby pink noise signal master was most likely flat when it was passed on to the printing lab, so any HF rolloff is due to limitations in the printing process and thus the latest version of Cat69 at any point in time should represent the same rolloff as feature films processed at that same point in time (albeit at that same lab). It is not a very satisfactory nor scientific basis for determining whether a reader is flat or not, but it is all we have.

------------------
Ray Derrick
President/Chief Engineer
Panalogic Corporation Sydney, Australia
Phone: 61 (0)2 9894 6655 Fax: 61 (0)2 9894 6935

Are you saying that I'm going to get wrong results by using the test films?

Why then should anybody use them?

Is Dolby aware of the problem? If so, why do they keep releasing bad test films?

I just paid a lot of money for new rolls from Dolby.

Oh...and to top everything off...my new tech from Miami says that somebody at Dolby told him (off the record) to set levels at 80dBC if we want to continue using 7 as the fader reference...or else play the movies at 6.

It's like nothing is standard anymore.

What test films are the folks at THX using? (I know, I know...they're a dying breed...)

 |  IP: Logged

Ken...your theory has some merit but it isn't the case. When on is setting up an A-Chain, they are going to focus and set slit loss on the same piece of test film. As such, any variences in the thickness of the film will be negated. Where the older style Kinoton head wins out is that once set, the focus will be spot on regardless of release print thickness. However, I've found that Kinoton's small drum (again, this is the style previous to their current reverse scan reader) is a nightmare to keep from having flutter...they having it spinning too fast; ANY problem with the bearings (drum or roller) shows up VERY fast. Use the Dolby Jiffy test film's channel sweeps...it will show up painfully well. On the current readers, I haven't had this problem.

Ray, I was leaving Cat. 97 out of it . The Current SMPTE Buzz track (P35-BT) is shot in a pin-registered camera. In fact, it is the same camera that shoots 35-PA/IQ. That camera is and target are consistantly checked for accuracy. Who checks Cat. 97? If it is the same people that check Cat. 69 then the answer is nobody. Therefore, I sided with the Buzz Track as being the one that tells the true tale. Have you found that the current style Buzz track (the one that has the name on each frame and has the "ticking" noise on the track to vary from batch to batch? Have you put it under a microscope to compare it to the standard to see if it agrees? How about Cat. 97, ever measured it?

How about Dolby Cat. 566...uniform illumination film, another film that will let you read the edges of the soundtrack. This is one of the most handy of test films and yet most people don't use it or have ever heard of it. It will let you set up your exciter lamps or LEDs...infact it will show up the problems of using non-adjustable LED mounts. Ever get a new soundhead (factory aligned) and found that when setting Dolby level, one was notably higher than the other? Run some Cat. 566 and the reason will be found. By properly setting the lateral position of the LED, you will then find the Dolby level is same on both channels. component engineering's readers beat everyone on that one.

Manny, as to your results...what I'm saying is that they are going to vary. As I said earlier, we have taken to buying the test films in larger rolls, measuring it and then making up offset files to align stuff. That way, from year to year, everything is consistant. As to the Cat. 69P, I believe you said you use CP-45s...don't worry your hosed anyway! The Cat. 222SR/A has an overly bright HF to start with, when in doubt on a reverse scan, leave the slit loss all of the way down, you would need very little boost anyway to achive "flat" response.

I would trust the Buzz track film for setting lateral alignment. With reverse scan readers, Cat. 97 is superfulous. At best, it can argue with Buzz track or show that your cell is not centered to the lens (which sometimes does happen but not too often). If you think about it, Buzz track locates the edges of the track, Cat. 97 locates the middle. You are still dealing with the same soundtrack area. Dolby was specifying to scan 102% of track (yes more than the whole thing) to avoid "clipping" the edges...except, it makes it more prone to noise. Thus, you can get a more accurate position by looking for the edges.

As to 80dB for setting level. The CP-650 people decided to "fix" a nagging problem that the reference level just wasn't right after all these decades...thus they fixed it on the CP-650. If you set a CP-650 to 85dBc, you will actually be 2-3dB louder (I forget what the exact number is but that is the ballpark). Now who asked for this? How many of you out there thought the old standard (using the Cat. 85 or the SG-1) thought "every time I set up a processor 'to spec.' people rush out saying it is just too quiet"? So, if you have a booth with mixed processors, you better set your CP-650s lower if you want the faders to match up in level. So go ahead, set up your CP-650 to 80, you still won't be running at fader 7 to please the bulk of the customers.

Steve

------------------
Ray Derrick
President/Chief Engineer
Panalogic Corporation Sydney, Australia
Phone: 61 (0)2 9894 6655 Fax: 61 (0)2 9894 6935

On your Buzz track film, was all of it the new style that varied from batch to batch? If it is, I'm most curious (I haven't found that...yet). I have found Cat. 97 to not always read the same though (even more curious since you say you have consistancy there!). Using Buzz track to center the film with respect to the cell, presuming that the Buzz track really does located the edges of the scanning area accurately, by definition, you are centered on the cell in a reverse scan situation. If Cat. 97 differs with Buzz, one or both of the test films is defective.

This is not the case for forward scan where the SLIT is fixed (on most system, Christie, for instance made it movable). In that case Buzz strictly sets the slit to track relationship, Cat. 97 set the Cell to film relationship. In revserse scan, the slit and cell are one assembly. Whereas the slit is affixed to the cell, you can't really move them independantly...the function of Cat. 97. For instance, on a AA-2, Cat 97 only merely confirmed that reverse scan has no cross talk.

Also with the overscan recommendation by Dolby, one could have a non-clipped Cat. 97 yet not be centered. Buzz track would be the more critical setting. If you get a system where the cell to lens relationship is off, Cat. 97 can be useful to recenter the cell. If the slit is off-center to the cell, you are hosed and at best can come up with a compromise.

Personally, I'd prefer if they went for only 100% scan and settle for the potential of a clipped signal at 100% modulation. I feel some of the noise assoicated with reverse scan is the junk at the edge of the applicated area being read by the overscan.

Steve

Whether P35-BT varies on a batch to batch basis in an on-going fashion I do not know. What I do know is that we possess a number of reels of this product purchased at different points in time, all of which bear the P35-BT lettering in the picture frame, and there are at least two widely differing lateral positions in this stock. Now it has long been known in the industry that there was a difference in lateral position between Cat97 and P35-BT, but as to whether P35-BT was originally incorrect has been corrected in more recent batches I don't have that answer, but I will try to find out for sure and let you know.

Also you are overlooking the fact that some reverse scan readers (the cinemeccanica comes to mind) allow you to change the focal length (that little grub screw on the side). Why they provided this adjustment is anyone's guess as it should never need to be touched, but it is there none the less, looks like a focus locking screw and is therefore a trap for young players. Changing the focal length alters the scanning width. It is also possible to alter the lateral position of the cell in most RV units.

More importantly, manufacturers of reverse scan systems need a reliable reference to work to when setting the internal parameters (such as focal length) of their products. Dolby recommends they use P35-BT Buzz track and Cat97. When you consider that these test films are to some degree inaccurate, or at least are in disagreement, it then falls on the manufacturer to determine which copies of these films are correct and which are not. As I said earlier, a test film should be precise, and if that is too difficult, then it should come with correction data.

BTW Dolby's recommended procedure for setting the scanning width of an RV unit is to alter the focal length on buzz track to the point where the cell is reading the whole of both buzz track square wave signals and then readjust the focal length to the point where the square wave signals are 20dB down.

------------------
Ray Derrick
President/Chief Engineer
Panalogic Corporation Sydney, Australia
Phone: 61 (0)2 9894 6655 Fax: 61 (0)2 9894 6935

It would seem that based on our various findings, our industry's test films are at best questionable. Your statements about Buzz track calls into question 35-PA as well since it is shot on the same camera.

What this industry needs is a varification procedure to ensure the test film is accurate.

Steve

Just one more interesting point on this subject. An LED contains 22 light emitting elements set in two rows of 11. For an LED which is correctly aligned laterally, only about half of the 22 light producing elements are actually directly behind the soundtrack and therefore seen by the solar cell. So there is some room for lateral error in the LED's position before left/right balance is affected.

Also Dolby recommend that the LED should be positioned precisely 1mm behind the film. If it is closer it will be too much in focus and the individual elements will be seen by the solar cell. This results in distortion because the light is not uniform across the soundtrack but has a dip between each light element. Similarly if the LED is mounted too far away it's image will be magnified and so a smaller number of light elements will be seen by the solar cell, and whilst they are out of focus, this greatly increases the chance of the solar cell seeing dips between the light elements.