kl Posted November 14, 2013 Report Share Posted November 14, 2013 I'm trying to feed my slate with off speed TC for a music video playback (via Comteks). The display though shows/updates to my numbers very spottily - and the reception LED mostly stays on red no matter what levels I'm sending. I have the slate on READ (dip 4 + 5 are on) and also have dips 1,2,3 as off, off, on - according to the manual. Anybody know of ways to trick the slate into behaving....? Thank you, Karl Quote Link to comment Share on other sites More sharing options...
kl Posted November 16, 2013 Author Report Share Posted November 16, 2013 Same issue with a Denecke ts-1 slate. It's a little less jerky, but jerky still. what's the issue here? Quote Link to comment Share on other sites More sharing options...
Tom Visser Posted November 16, 2013 Report Share Posted November 16, 2013 The mein issue is that your not sending timecode. Quote Link to comment Share on other sites More sharing options...
kl Posted November 16, 2013 Author Report Share Posted November 16, 2013 I'm sending plenty in standard speed and it works nicely in standard speed too - my issue happens when I'm sending off-speed for 40fps slow motion....! In other words sped-up audio + TC! Quote Link to comment Share on other sites More sharing options...
Tom Visser Posted November 16, 2013 Report Share Posted November 16, 2013 Sorry I was not in a place where I could be more verbose. I'm certainly not the authority on SMPTE timecode, but the standard rates as I understand it are 23.976 (aka 23.98, aka 24p), 24, 25, 29.97, 29.97DF, 30, and 30DF. Anything that is not one of the above is simply not going to be readable by any of the products we use that are typically jammed by incoming timecode, but for the most part free-run in a generator or generator / reader mode. The slates that you are working with are really designed to keep accurate timecode in one of these standard rates and utilize some error correction / free-wheel ability to read errors in incoming timecode signal and output something that is a stable clock. When you are feeding it a non-standard vari-speed-like signal, your confusing the internal generator as it tries to read the incoming code, and in the case of the Denecke, does so somewhat successfully, but interprets the mis-speed as an error and the jitter you see is a result of that. It's like sending 37.8KHz digital audio to a DAC that is expecting 32, 44.1, 48, or 96KHz signal. There is no reason why I can't interpret the data as a 37.8KHz stream, works from a theoretic technical perspective, but will only be interpreted by the DAC as an erroneous incoming signal. You might have better luck using a device that has no built in generator functions and free-run timecode, such as the Ambient timecode app on an iPhone or iPad, or a MOTU audio interface that has some LTC timecode reading abilities on an audio input. The slates are not misbehaving, they are doing exactly as they are expected to do in that situation. Quote Link to comment Share on other sites More sharing options...
John Blankenship Posted November 16, 2013 Report Share Posted November 16, 2013 What Tom said (sorry, Tom) doesn't really apply here. In READ mode the slate should be taking in whatever the code is and displaying it. A Denecke TS-1 will display pretty much anything you throw at it. Some versions (such as the early TS-3 model) wouldn't handle some of the higher speeds, but the newer circuitry will. (Earlier versions can be updated.) There are several things at play in order to read good time code. One is how clean the signal is. Time code is basically square waves. With rounding of the waves (caused by high frequency rolloff, etc.), there are some devices that can't handle it. Also, the level of the signal can be a factor. Try different levels into the transmitter. You don't want to overmodulate as that will severely damage the square waves. Also try different levels out of the receiver into the slate. I've had quite good success with Comtek 72s and Denecke slates but have little experience with Ambient slates. I've used their Lock-Its, but not their slates. Quote Link to comment Share on other sites More sharing options...
kl Posted November 16, 2013 Author Report Share Posted November 16, 2013 Thanks John. Yes both slates are in read mode. I've played around with levels and also used different speed conversions - still jerky rolling of the numbers. It kinda works - not smoothly though. Quote Link to comment Share on other sites More sharing options...
John Blankenship Posted November 16, 2013 Report Share Posted November 16, 2013 What is your time code source? Have you tried a direct feed to see if the issue is with the Comteks? Quote Link to comment Share on other sites More sharing options...
kl Posted November 16, 2013 Author Report Share Posted November 16, 2013 Yes, tried that too Quote Link to comment Share on other sites More sharing options...
kl Posted November 16, 2013 Author Report Share Posted November 16, 2013 It's puzzling as it works perfectly on 24 - only thing I can think is that the pitch correction does something to the tc signal. Quote Link to comment Share on other sites More sharing options...
John Blankenship Posted November 16, 2013 Report Share Posted November 16, 2013 It's puzzling as it works perfectly on 24 - only thing I can think is that the pitch correction does something to the tc signal. So, are you saying it did not work with a direct connection? If that's the case, it helps us to rule out the Comtek link. So, how did you do the pitch correction? It sounds like it may be the culprit. Quote Link to comment Share on other sites More sharing options...
kl Posted November 16, 2013 Author Report Share Posted November 16, 2013 One version done in protools - using 'speed': speeding up to 1.6683 and pitch correction down at the same time. The other version, same same but with built in tools of 'audacity' - 'change tempo'. Thank you for hanging in there! Quote Link to comment Share on other sites More sharing options...
Tom Visser Posted November 17, 2013 Report Share Posted November 17, 2013 If you look at the weave forms, are they still square waves? LTC timecode happens to be transmitted over audio wires at audio levels, but is in fact a data stream. Quote Link to comment Share on other sites More sharing options...
kl Posted November 17, 2013 Author Report Share Posted November 17, 2013 Tom, did you ever prepare TC for slow motion playback? If yes, did it work? What tools did you use? thanks Quote Link to comment Share on other sites More sharing options...
Tom Visser Posted November 17, 2013 Report Share Posted November 17, 2013 Everything I've done has been half speed or double speed, and whenever i do playback, bring a full Pro Tools system for playback flexibility. I've never had to worry about SMPTE with respect to real song time as everything is already inherently in time. I've just restripe the resultant playback track with real speed timecode which is double or half the song speed. What if I did need actual song time slated? I would probably find something capable of displaying MTC timecode or if I could just use an attached display, throw "Big Time" up on the display expanded to full size and have cameras point at that (unprofessional but would work), or fly in a small wired monitor as a slate - maybe even velcroed to an actual slate or something. Going back to your specific problem, have you tried it without any pitch correction at all? LTC is something like an 80bit word represented by square waves. Pitch correction in the traditional audio sense would be keeping the rate of the number of zero crossings the same for any given audio signal (same frequency). If you apply pitch correction to a data square wave set, I'm not sure how that would affect the signal, but in many cases it would simply stretch a "1" or "0" position as the software would detect an f of zero if it were just looking at a small portion of the signal... or in some instances, if not round off square edges of the wave or even worse, insert waveforms where none should exist or possibly deconstruct the square waveforms altogether and reconstruct as sinusoidal waveforms. If you just let the signal go through withou pitch correction, it should be the original intact data stream except at a faster or slower timing. I don't think that LTC timecode is all too particular about the carrier's clock frequency, so as long as the change is not too drastic, the LTC reader's logic too selective about incoming signal speed, it might work, assuming you have your slate set to reader mode only (3,4 DIP switches set in as you mentioned). Quote Link to comment Share on other sites More sharing options...
Tom Visser Posted November 17, 2013 Report Share Posted November 17, 2013 Sorry about the horrible spelling - I hate the iPhone keypad, will go back and correct later. Quote Link to comment Share on other sites More sharing options...
kl Posted November 17, 2013 Author Report Share Posted November 17, 2013 Thank you Tom, I did not try without pitch correction! Maybe that's the light at the end of the tunnel I'm looking for Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.