Jump to content

Bouke

Members
  • Content count

    21
  • Joined

  • Last visited

Everything posted by Bouke

  1. It should be easy... If it's about getting TC input as well, I would think any recording software will do. Just record all your tracks, and record incoming TC on one of the tracks as well. Post processing is needed, but that I can provide. This thing accepts Wave as input, and can output BWF with correct timestamps (subframe accurate!) https://www.videotoolshed.com/product/ltc-convert-auxtc/ Now, this was intended to aid simple post, but I could alter it for you guys so you can also input track names / other metadata. If you need LTC out, that will be a different story. PC's have bad clocks, so LTC out WILL drift like hell. You'll be better off with an external TC generator / lockit / tentacle / yadda as a master clock. Bouke
  2. Delivering files from multiple recorders

    Hi guys, I'm highly interested in the why of ALE use. (Maybe in the old days it served a purpose, but nowadays...) One big problem I've encountered with Avid is the UMID memory. It is hard to test multiple outputs if the UMID does not change, as Avid (thus ProTools as well I assume) will take metadata from it's database rather than the file. That's not bad for normal production, but a PITA when testing. (If you output with Master selected, import that and do the same test again without, you may NOT see what's actually there.) So, next step is to generate a new UMID. If someone has specs on what kind of UMID is used for BWF, please tell me! Meanwhile, version 0.9.1 is out that at least upnumbers the UMID by 1 point from the source UMID (And there are a few other improvements.) Get it here: https://www.videotoolshed.com/product/bwfmerge/ And there are a few screenshots explaining what the 'master' function does. (in short, trimming / padding all extra input to one recorders clip in/out.) Bouke
  3. Delivering files from multiple recorders

    Vale, That is not difficult, but I fail to see the point. Could you elaborate on why you want this? Avid imports a lot of metadata already. What does it not do then so you would need the ALE route? Meanwhile, I have a working version for both Win and Mac. https://www.videotoolshed.com/product/bwfmerge/ Bouke @all, It took some time, as I wanted it to be good, and sample accurate. (And i've managed to do so!) I have a working version. It accepts all kind of input files (mono, poly, even sr / bd can differ), and lets you merge them into poly files. Metadata is retained for all inputs (well, channel name. Comment is taken from the first / most important one.) You can also set the output patching if you want to re-order the tracks. You can choose if you want one recorder to be 'master', and have other recordings trimmed to that in/out, or have (a gazillion) different files, where as long as there is overlap between any files all will become one large one. Multiple stop/starts of a rogue recorder is supported as well! Get the demo, toy with it, feed me with anything you like to see! (It's working, but still I like to add a lot more, but I need input!) Last but not least, should I make a new thread about this? Bouke
  4. Delivering files from multiple recorders

    Guys, I just made something that can do exactly this. (It works under lab circumstances here.) Input mono or poly files, output one big poly. It does not matter if the files start at the same time or not. Track names / other metadata is retained, in both the BEXT and iXML chunk. @pindrop, if you can elaborate on what you wolud like to see, I can probably build it in. You can reply here (so others might add,) or drop me a line directly. I hope to have a beta version on my site tomorrow. Bouke
  5. Delivering files from multiple recorders

    Well, I would think to just strip the beginning of the files, since they start at about the same time. (Less work :-) Then, altering the track names in the BEXT and Ixml is doable as well, but the venim is in the tail: What to do with the UUID's? What output filename? And, while I'm at it, what else could be handy that WaveAgent does not do? (I could create sound reports in other formats than PDF, or make a customized PDF with your logo, etc..)
  6. Delivering files from multiple recorders

    Hi Contantin, I'm afraid it is a bit more difficult. If two recorders are used that start at 'about' the same time, it's not always that they know about each others existence, is it? So, the metadata of the files from recorder 1 contains other info than the metadata of recorder 2 So those have to be merged. Then, that metadata lives in both the Ixml as well as in the BWF chunk. (Where the timestamp is the least of my worries.) And of course, chances are big that the poly goes over the 2 gig limit, so that will be fun as well. That's why I would like to see some real world files :-) Bouke
  7. Delivering files from multiple recorders

    Well, BWF isn't that hard to alter. If someone can send me some testfiles, and you all write specs on what a new app. should be able to do, I can make something. https://videotoolshed.wetransfer.com/ Making multiple files match in duration / start is not that difficult. Renaming is also not the issue. What I need to know is what to expect, and what metadata lives where, has to be (auto) altered, and perhaps other stuff I don't get. (Mind you, I'm an editor and coder, not a sound guy.) But I don't get the Poly part when you're talking about 10+ tracks. But this is an old discussion that does not have to be repeated. (Note, in most lingo a Track can be mono or poly, having 1 or more Channels. (especially for AV files this is important, as an AV file can have one or more video tracks, one or more data tracks, and one or more sound tracks, where each sound track can have x channels....) Bouke
  8. Hi guys, Strange thing happens. I have a client who claims a 2 channel (poly) Wave file from me comes in in FCP 7 as dual mono, while a ProTools file (not sure if it's a wave or something else) comes in as stereo. (Just the pan settings.) I forgot close to everything about FCP 7... Anyone knows how this could happen? I never heard of metadata inside a Wave / BWF file that contains panning info... thx, Bouke
  9. Hi Alex, Long time no see! I've got no clue how / why this happens. It's a single 2 track file (stereo in this case), that seems to become mono panned. It should not be an issue as the editor 'should' have been paying attention, but it was a commercial that went out mono... (Also strange that the broadcasters QC missed that one...) Bouke
  10. Suggestions - kitchen table unscripted discussion

    Hmm, I have had very good results with a BLM in a similar situation (On table, but with clear directions not to tab pens, and there were no drinks / cuttery). Note that it does not have to be on the table, but can be on 'any' surface. So, mounted on a piece of chipboard and hanging over the table would be a simple / cheap way to avoid boom shadows. hth, Bouke
  11. Timecode solution

    Can someone enlighten me why / when / what situations going over midnight will cause a problem? If all is well, the user bits contain the date, so there should be no issues. (Yes, that's theory. And as we all know, in theory there is no difference between theory and practice. In practice however...) Bouke
  12. What about simple math? First, I did NOT watch the OP's video. I refuse to have my time abused this way, simple as the topic name sorta kinda says it all. I'm not going to watch a long video to get the info I could read in 30 seconds if it was in text. First. You need clients. Get them. Spend money on that. No clients, no income. Next, there is a difference between renting and owning, but if write off goes by year and you have little jobs, renting could be cheaper than owning. (A rental kit works at least 100 days a year, yours might be working less.) So start with spending money on getting clients, and decide what you WANT, not what you can afford. Do some simple math. Say you want 20K worth of gear. If you borrow the money and pay off in 2 years, you pay some 2K of interest (assuming you get a loan for 10%). Not sure about the UK, but over here interest is deductable, so you end up paying less than that. Thats 22.000 / 24 months = 917 per month. (Tax returns not included.) Thus, if you have jobs that you can put 200 a day for gear, you need just 4 days of work a month to get this cost covered. The math changes with different payback time / interest rates, but it's not rocket science. Again, get cliens, buy whatever is needed for the job and the clients are willing to pay for. Although I know that clients in your bizz refuse to pay for the good gear, again do math. See if you can bill enough to get by. That's the only thing that's important. Running a business is way more that owning gear. (If you're a sound guy, does it matter if you bill 100 an hour for you and bring your equipment for free?) Bouke (who's borrowed way over 100K to buy equipment, but I'm in post / developing.)
  13. BWF files are the same for 24 and 25, as the BWF timestamps converted to TC have no non-drop issue. (23.976 has this 'problem' ) Now, in Avid, you might need to override the 'baked in' speed, so TC matches. But Avid syncs sample accurate based on the SAM (if you turn that on), so it should be no issue. What CAN be an issue, if you have fed TC to the cam from the bwf recorder, I have no clue what happens then... I would say, test a couple of clips, if they sync with a 'little' offset, it's an easy fix if the offset is always the same. If the cam did not lock at all, or the offset differs between clips, it's manual or Plural Eyes.... But again, since it's 24 vs 25, I cannot imagine it's a progressive drift. (That can be compensated as well with a little math.) If that happens more often? Yes! One of my clients had his Zaxcom setup for 25, the cam was 25, but LTC on a soundchannel was 24 He used a TX/RC Zaxcom system, where the RX was set to 23.976, so the LTC (on a soundchannel, NOT tc input) was in 23.976. (In this case / piece of hardware, the RX regenerates, it does NOT relay incoming!) This was an easy fix, as there was no slaving, it can be done with simple math. Another example, today I talked to a sound guy who had a gig in the US, and the DOP refused to tell him the framerate they were shooting (Don't ask why, I have no clue.). He being European, unknowing, set up for 30 fps where they were in fact shooting 23.976. This will probably a 2K screwup from the DP, unless there is a very tech savvy editor, who will curse anyways...) hth, Bouke
  14. Pluraleyes Good or..not?

    Now i'm pissed. How long does it take for Plural Eyes to sync a day of ENG of material? (say, 100 clips) With what succes rate? Last time i've tried, it took me a couple of hours to have only 1/3 done. LTC will do this within a minute, and will give a very clear warning if it worked or not. (And if if did not, YOU are to blame, but there are still backup plans that will save the day.) You're not in post,are you? Did you ever do this yourself? Please enlighten me if your experiences (hands on) are different! Bouke
  15. Batch Combining sequential BWAV takes

    Hi Nick, As I needed a better way to cat files anyways, I put in some effort and wrote the stuff, also made a GUI to do just (and alone) this. So, IOW, I just made the app. you're looking for :-) Input is mono or poly BWF files, output is one big poly (can be 1 up to XXX tracks) BWF, where the timestamp is from the first file, and silence added to make all others fall in place. (I still think this is a stupid way of dealing with the problem, but hey, who am I to judge...) Sample rate and bit depth is taken from the source. (I could customize that....). A bit of math is done and it will warn you if you go over the 4 gig file size limit. Drop me a line if you want to toy with it. Bouke
  16. Batch Combining sequential BWAV takes

    Well, since my name came up... It's certainly not hard to cat the files, but aren't you afraid of the 4 (Or sometimes even just 2) gig file size limit? If so, an option would be to cat to a QT audio only file that has the same audio data but in a different container. I don't know PT, but in Avid, you can AMA link them (Thus, no import time), then do an 'autosequence' on the clips, where you end up with a timeline with all the clips at the correct position. From there it's just an export to Wave (Avid will generate a BWF with the correct timestamp.), or QT. That takes close to no time. Isn't there a similar function in PT? If it's to spot, you can have a look at my Transcriber app. (Runs on both Mac and Windows), that can load an entire folder with wave files and play it as if it was one big file. And if you do load a long file with silence, it also has silence detection so it auto jumps over those parts. (That was added for a poor fellow who has to transcribe air traffic communication, and had to sit out very long times of silence...) But, if I have to write something custom, that will set you back 200 bucks.
  17. Videotoolshed

    Hi Phil, Do note, my first name is Bouke, my last name is Váhl, but that seem even more difficult to pronounce :-)
  18. Pluraleyes Good or..not?

    Mea Culpa. I'll start another thread soon, for the benefit of all :-) Bouke
  19. Videotoolshed

    Philip, As always, you can mail me with your issues / wishes / ideas / bug reports, even if you're not a registered customer! Bouke
  20. Pluraleyes Good or..not?

    @ Constantin. Yeah, sorry I’ve hijacked the thread, but to my understanding there is consensus that TC is to be preferred over Plural Eyes. So, should we create a new thread? @ Callou & Daniel Not sure if you fully understood me. The idea is to connect the LTC to an audio input of the cam, if the cam does not have a proper TC input. You all know that LTC is in fact an audio signal? (Although LOUD, you probably have to attenuate.) This LTC can come from the Tentacle hardware, or wired/wireless from your TC out on your recorder. This way the LTC on the cam matches your BWF. Now say the cam’s internal TC at rec start is 01:00:00:00, and your recorder starts at 01:00:01:00, there is a one second offset. My software knows this offset, and can set a new timestamp of 01:00:00:00 on your BWF files. This is altering just a few bytes, and post can now sync just as they always do when both video TC and BWF timestamp match. Yes, the DSLR’s have run rec TC that drifts a lot, but that isn’t a problem for syncing. (The lack of gunlock on long takes is an other issue, so in all cases shoot short takes, or be happy with the 4 gig file limit on a lot of DSLRs that will generate multiple clips anyways.) @ Philip (@ Callou) Bleeding of TC into the other channel should not really be a problem. First of all, the other cam sound track is probably just a scratch track, and it ‘should’ not bleed if you don’t put the LTC too loud. -25 dB is loud enough, even on a 16 bit recording. Then, you do NOT need to fill the entire take. A few seconds in the beginning of the video recording is good enough. I have a client (Sound guy) who made a simple switch to choose between LTC and scratch audio. When he knows the cam is running, he simply presses it for a couple of seconds to send TC, and after that the signal becomes scratch audio. My software analyses the first XX seconds (you can tell it how much) of the video and will find it. Dropping out is not really an issue, as it only needs a few frames of decodable LTC. And those don’t have to be adjacent. To all: By all means, use your BWF recoder as master TC if you shoot on a DSLR. (Well, IMHO have the audio recoder be master always, no matter the cam, unless you are in a studio with a master TC gen.) My software does NOT target a specific workflow, but it assumes you have LTC on a sound track. That can be on the camera, or even on the sound recorder (Yes, there are people working with Marantz / Zoom etc. without proper LTC out, using Jumpstart or alike on a stupid phone / old Ipod gives them a master clock accurate enough.) My software allows to output dailies in original codec or Prores, it can even do audio patching, it is smart enough to know when there are multiple sound takes for one video take and so on. Besides creating new video with the desired audio channels, it can also (as I started with) change the BWF timestamp to match the video, or in case the video is in QuickTime (.mov) format, set the TC of the video to match the BWF. (And this is just a tip of the iceberg :-) Does this makes sense to you all? Bouke
  21. Pluraleyes Good or..not?

    Ok, let me jump in here. First of all, hello. My name is Bouke, and I'm a software developer (as well as editor and multicam shooter.) I'm around for quite some time, and some of you already know me. Since the topic is now more or less about Tentacle sync, let me throw in some of my experience, as well as some shameless self promotion. What I'm missing here is the more logical way to sync up. Close to any cam has internal TC. (Ok, the Canon 5D MkII has not, but it has THM files that give subsecond accurate TC that I can use.) So, if you have LTC on an audio channel, and the cam has normal TC, why go trough the extra data to create new clips? The more logical way is to set a new timestamp on your BWF files that match the video TC. (And yes, that goes up to a few samples accurate.) This also means that on takes where there is no TC on the sound, the offset will be the same and there will STILL be a match found, and it will STILL sync! This is lighting fast, and will be exactly the same for post production as you have actually slaved the cam to the TC. My LTCconvert app. can do so. https://www.videotoolshed.com/product/ltc-convert-auxtc/ And it does a gazillion things more not found in the Tentacle software.... Let me know if you want me to elaborate on this, or have specific questions. (Either here, or drop me a line.) Bouke
×