Conformed original BWAVs are out of sync by up to a frame compared to the AAF

[pindrop]

Philosophical Opinion on Responsibility:

 

An NLE or any other software that does a conform or otherwise places or shifts things on a timeline needs to shift them according to the finest expected resolution of its content. If your software operates on the time relationship between picture and audio, that "finest resolution" is the sample. It is therefore the responsibility of the NLE to operate with sample accurate movements on a sample accurate timeline. (Not saying this is easy for Avid, just saying it's their responsibility).

 

While it may be a good, smart business move for a company like SD to compensate for another company's shirking of its responsibilities, it is certainly not the responsibility of recorder manufacturers to record on frame boundaries.

 

Or perhaps more succinctly, "You move it, you buy it". 

Yes good points, to the extent that it could obviously be argued that SD has made the issue worse by unilaterally doing what it has, and could be some kind of appeasement.

[studiomprd]

"  SD has made the issue worse by unilaterally doing what it has, "

good grief!

[Constantin]

Yes good points, to the extent that it could obviously be argued that SD has made the issue worse by unilaterally doing what it has, and could be some kind of appeasement.

How exactly have SD made it worse? Because they have made sure that files recorded on their systems will always be oerfectly in sync and compatible with all editing systems?

[John Blankenship]

Apparently they have "made it worse," as in, "How dare Sound Devices design their equipment to avoid some of the post pitfalls of editing systems!"

[pindrop]

Apparently they have "made it worse," as in, "How dare Sound Devices design their equipment to avoid some of the post pitfalls of editing systems!"

No not at all, but as in, a big company releases something with an obvious glaring fault and a small company comes along and accommodates the fault for them and by doing so relieves some of the pressure for the big company to correct the fault, and could lead the big company to think, in a bullying kind of way, that all small company file recording hardware will be forced to do the same as a result of post production declaring a preference for the machine that means they have less work to do, whereas post production could be directing their dissatisfaction at the source of the problem, which might make the big company give fixing it a higher priority........

[Constantin]

Sound Devices are doing what they should do, by helping sound recordists in the field avoid issues with post (although I have never heard of a case where post actually made an issue about this for the recordists). It would be unrealistic of them to think that Avid would change their system, since they haven't done so in the past.

It'd be the same blaming Schoeps for building the SuperCMIT and arguing that this might further decrease sensibility to sound problems.

[pindrop]

Sound Devices are doing what they should do, by helping sound recordists in the field avoid issues with post (although I have never heard of a case where post actually made an issue about this for the recordists). It would be unrealistic of them to think that Avid would change their system, since they haven't done so in the past.

It'd be the same blaming Schoeps for building the SuperCMIT and arguing that this might further decrease sensibility to sound problems

The analogy is not at all the same, sound problems are an objective fact, not a manufacturer's decision.

Avid are moving sound files along the time line to suit their inadequate software.

I would say it's not unrealistic to expect Avid to do something about it, just on the basis that they haven't in the past. If there is a groundswell of requests to do something about it, Avid might just concede and give it higher priority, or alternatively if Zaxcom concedes to accommodate Avid, the cause might be lost and we'll all be doomed.

[Matt Mayer]

Accommodating a glaring fault?  I am fairly certain when the 744T was in development, no one had any idea this problem even existed.

 

Very early on we decided it was a good idea to start files on the 00 frame, for cross-jam and frame rate conversion reasons.  Also, in studying the most common playback equipment of the time (primarily the Fostex DV-40), all of their files started on frame boundaries, so it made sense to conform to that paradigm.

[Philip Perkins]

I would rather have the recordings stop and start predictably, with a view towards how NLEs actually work with them in the post syncing process (no subframe movement fo clips).

 

philp

[macrecorder]

yep, good decision by SD. Other manufacturers will just have to catch up. It makes sense for more reasons than just Avid compatibility. With a little pre roll, which most machines have anyway, can it be that difficult to roll back ? I'd much rather know that my files are compatible and accurate than harass Avid in a probably futile way. Just do it, other manufacturers. If editors are noticing it, and moaning about it, which they are, then it's probably time to update, and not waste time casting blame on other systems, even if you have a point. Sometimes you just have to bow to existing practice, rather than wish for an ideal solution.

[Wandering Ear]

I think if the desired solution is to get manufacturers to start recording on the frame edge, the best way to ensure that is to lobby to change the BWAV standard to include a frame edge start requirement.  Currently there is none.

I can see the merits in pushing for frame edge recording starts, but I also still think an NLE shouldn't be moving things out of sync, and I can think of many workflows where you may want to work with audio that isn't necessarily recorded on a frame edge.

[studiomprd]

" the best way to ensure that is to lobby to change the BWAV standard to include a frame edge start requirement. "

let us know how that goes...

[takev]

It is going to be difficult to add this to the BWF standard, I believe that the specification for how audio recorders should interpret a start of frame or even a start of a second will be 2 times larger than the BWF standard itself 

 

The biggest problem is that the timestamp in a BWF is recorded as "number of samples since midnight" and that the frame rate is not recorded.

 

It is pretty easy to know where to start a second with 24, 25 or 30 fps, however it is going to be a lot more interesting when recording at 29.98 and don't mention on what to do on drop-frame-rates?

[studiomprd]

another example of something that at first, sounds simple, but isn't !

[Wandering Ear]

It is going to be difficult to add this to the BWF standard, I believe that the specification for how audio recorders should interpret a start of frame or even a start of a second will be 2 times larger than the BWF standard itself

Ha!

I see this issue as being a workflow issue that is more than just this recorder manufacturer or that nle, and standards like the bwav specifications are designed to provide the basis for implementation throughout the workflow. It may not (probably isn't) practical to update the bwav spec, but it's worth considering. There are lots of issues, this one included, that are caused by differing implementation of bwavs and their metadata.

[Jeff Wexler]

"There are lots of issues, this one included, that are caused by differing implementation of bwavs and their metadata."

 

What are all these additional "issues" you are referring to? How have we gotten away with thousands of Sound Rolls filled with BWFs all these years?

[John Blankenship]

***THIS JUST IN***

All major motion picture studios are recalling the hundreds, if not thousands, of films whose production sound was recorded on any device that didn't commence recording on a frame boundary.  When reached for comment, studio heads agreed, "We may not have noticed it, but now that it's been drawn to our attention, we're unable to collectively sleep until this catastrophic situation is rectified!  We're not just going to repair them, we're going to re-shoot every one of these films!"

 

In a related story, all movie theaters worldwide are converting to having only one row of seats since it's been discovered that for every (approximately) twelve rows further back one sits in a theater, the sound is delayed by an additional frame.  "Scandalous that they should keep this critical information from us," lamented one theater owner.

 

Meanwhile President Obama has created a new government agency tasked in part with determining if the slower speed of sound relative to that of light, is another by-product of global warming.  The agency will also be charged with commissioning in-depth studies on whether it is more feasible to speed up the propagation of sound waves, or to curtail the speed of light.

 

Stay tuned for film at eleven and non-stop coverage on CNN with intensive interviews conducted with every person who has ever even seen a motion picture.

[Wandering Ear]

"There are lots of issues, this one included, that are caused by differing implementation of bwavs and their metadata."

What are all these additional "issues" you are referring to? How have we gotten away with thousands of Sound Rolls filled with BWFs all these years?

None of them are such a problem that can't be readily worked around. I didn't think my post was worded as to be so alarmist, i didn't mean it that way.

Issues like fcp only reading some of the bwav metadata on ingest, and deleting all of it upon export. This imo wastes a lot more time in post if a reconform is needed than the avid issue discussed here. Despite that tons of people use fcp successfully.

Just because we suck it up and fix these workflow hickups as they arrise doesn't mean we shouldn't also be looking at our workflows as a whole and considering all the factors that go into it.

[studiomprd]

" this catastrophic situation is rectified "

OMG...

I've been watching movies with sound that is not frame edge accurate...

[Jeff Wexler]

"Just because we suck it up and fix these workflow hickups as they arrise doesn't mean we shouldn't also be looking at our workflows as a whole and considering all the factors that go into it."

 

I have been very much involved in "looking at our workflows" since it was really only a very small number of us early adopters to file based production recording that had to deal with these issues. The post-productions systems in place (not just Avid) at the time (1997-98) were never faced with the "problem" of utilizing sound files directly (since no production sound mixer, prior to the Zaxcom Deva I) was delivering files. One has to remember that the word "workflow" was not even in the common daily vocabulary for production sound mixers, sound transfer operators, etc., there was no need for discussion --- we turned in the tapes (1/4" or DAT) and post-production took over using whatever workflow they had devised. It was hardly a standardized procedure, I might add, because each production handled things in post (after "ingesting" our production sound) the way they wanted to. The move to file based production recording was one of the few technological changes that was motivated and promoted by PRODUCTION sound people. Early on we discovered that although there was a Broadcast Wave File standard, adding support for metadata, timecode and so forth, few if ANY of  the systems in post were aware of this standard. So, it really didn't matter if a file started on 00 or had proper metadata, the post people didn't care --- their whole workflow, once we started delivering files, was a "workaround" and many of the post people were basically just pissed off at this small group of sound mixers pushing for this new way of recording. Fortunately, there were quite a few enlightened post people who could see the future and realized that the advantages to be had with file based production recording were potentially very beneficial to the whole process --- even if it meant re-thinking their particular systems and workflow. 

 

All I'm really saying is that the big players in post, the recipients of our production sound tracks, whether it is Avid/ProTools, FinalCut Pro, etc., will always be much slower to adhere to standards which they never supported in any way in the past. That doesn't mean we shouldn't try and keep the pressure on them, big as they are and slow to change. Had we not been able to prevail with post facilities in the early days, we might not be turning in sound files today and we wouldn't even be discussing which recorder from what company starts the file on the 00 frame.

[macrecorder]

Regarding NTSC frame rates and what to do, ask SD, they seem to have it figured. A mountain is arising out of a molehill here. It's a simple enough convenience which ensures NLE's keep the audio in sync. Just because people didn't realise where these issues arose before doesn't mean that they don't exist or are irrelevant. There is always room for improvement, why shouldn't we want that? One recorder already does it successfully and is appreciated for that by some editors, which proves the point. No audio will be harmed in the process. Keep calm and carry on.

[Jeff Wexler]

"One recorder already does it successfully and is appreciated for that by some editors, which proves the point."

 

It's great that "some editors" appreciate that one recorder accommodates their faulty editing system. Wouldn't it be so much nicer if these very same editors could appreciate the company that makes THEIR editing system for not screwing up the sync for files coming from all the other recorders in use? I remember when the Avid wouldn't even recognize a poly file that had an odd number of tracks and the editor on the movie I was working on "appreciated" that I recorded blank tracks (un-needed) so that his Avid could actually use the tracks but that didn't stop him from vehemently appealing to Avid to fix their faulty software (which fortunately they did, eventually).

[macrecorder]

Well, we or the editors can shout at avid a lot if that is what you think best. Don't know if it will work, though.

[John Blankenship]

At one time some Avids would distort captured audio files.

Solution (as seems to be promoted by some here):  Record silent audio.

  

At one time some Avids wouldn't lock to the time code of several industry-standard machines.

Solution:  Take the time to manually capture everything.

 

At one time some Avids wouldn't handle even semi-long timelines well.

Solution:  Break every project into smaller sections.

 

Meanwhile... Avid as a company has been steadily shrinking year by year.  The question appears to be, "Will Avid start listening to the industry rather than thinking they can still dictate to it?"  So far, they've not demonstrated the foregoing to be a direction they've considered.

 

Disclosure:  In my studio we have Avid, Final Cut Pro, Premiere, Pro Tools, SADiE, and whatever else we choose to use.  I'm not married to any one of them.  I have watched, however, how Avid, year by year, has only taken token steps to overcome their "we rule the non-linear world" avid-tude. 

 

It will be their downfall in the end.

 

[Jeff Wexler]

This reminds me of the Fostex DV-40 problem and the birth of the infamous "F-mode" setting that we had to use if DV-40s were in place at the facility doing the sound transfers. The "F" stood for either Fostex or Fu**% up depending on how you felt about THAT company and their recorders.