For interviews, I'd personally do something like:

1. Mic(s) into stand alone recorder, just let it roll for the entire interview. Recording sound separately like this might have another benefit in that if the 5D2 cuts off, at least you still have continuous sound that you could drop some B-roll over if need be.

2. Outputs of recorder fed into inputs of 5d (as mentioned in above post), probably will need adapters, and will also need to experiment with levels to make sure the 5D2 can handle what you're giving it (experiment before the interviews obviously since you can't actually monitor the 5d2).

3. Optional, but recommended. Somehow slate each take on the 5d2. The most simple form of this would be to have someone use a slate/clapper board, or to just have the talent clap once once the camera is rolling. This will make syncing later much easier, but as long as you have the same audio on the camera, you should be able to manually sync regardless (just WAY FASTER if the clap is there). Another low-budget take on slating that I use sometimes is to record the display of the recorder at the begging of the take. Although it won't be "true" timecode, it still can help locate the "approximate" location of the clip in post. In fact, with some recorders, like the one I use (an R44), both the record time AND the filename is show on the display, so if I record that on the video, I get both what audio file to sync to and an approximate timecode in that file. Although during an interview doing that might be disruptive.

4. Even more optional would be to slate the tail of the 12-minute video clip. This would only be needed if the video and your audio get significant drift over a 12-minute clip (e.g., you sync the head, but the tail is out of sync still). Sometimes you can sync the middle of the clip, say 6 minutes in, and then spread the drift out across the whole clip (slightly off at beginning, good in middle, slightly off by the end). But for 12 minutes, I bet drift isn't that bad. I record a lot of music, and over the course of an hour or two (obviously using a video cam without the 12-min limit) drift is a real issue that almost always needs to be addressed if you aren't completely synced in real-time using high end gear. But since this cam can't do more than 12-minutes, drift probably won't be an issue.

Personally, I think the 5D2 would be tough in a long interview situation if you are going solo mode (just yourself, no crew). It seems to me that there's a real chance of disrupting the flow of the interview with the start/stop of the camera (especially if swapping out CF cards, maybe not so much if just hitting record again). However, if rolling audio separately, maybe that's mitigated by the fact that if needed, you can miss some video but can depend on solid audio (think b-roll again). On the other hand, the interview footage will look so freakin sweet with the 5D2, and the ability to do it in low light settings, it might just be worth it...

The camera records 30fps, while NTSC is 29.97fps. How does this 0.1% difference affect syncing? How long in a take until it's noticeable?

Regarding latency, when playing a keyboard, 10ms is the threshold. When watching video, the threshold for audio delay is closer to 50ms. We're more sensitive if the audio is early, than if the audio is late.

With a 0.001 error, the delay will be noticeable (50ms) by about 50 seconds. The bottom line is that if you squeeze the video, you also need to squeeze the audio. The other option is to drop the occasional frame. The maximum error is then about 17ms.

Thanks Bill !!– did a homemade clapper test run last night.
Suggestion # 1 is great; having additional audio might be a lifesaver in the future!

#2 When using a standalone audio recorder > running audio from the audio recorder into the camera is great because it made syncing the audio so much easier. Without the camera’s audio waveform I was doing a video(clapper) to sound sync which was a little bit tedious – When I had the camera’s audio waveform as well as the audio recorder’s waveform I could easily and quickly identify the sync positions.

#4 is good practice for longer takes with equipment that doesn’t have some form of time code sync. I come the audio world and having a master word(time)clock is imperative!

Actually, I’ll be doing one-man interviews – lots of them in fact! Each interview will last at least an hour; I’m a little worried about the stopping the interview 5 or so times! But I’m pretty good with interviewees and my interview topic is quite light hearted so hopefully it won’t be troublesome, and you’re right the visual outcome will more than make up for it!

This is confusing Salvador, and if my limited knowledge of video time code is correct the difference between 30fps and 29.97fps is in how the frames are counted and not actual frames

30fps is referred to as “non-drop frame”
And
29.97 is also referred to as “drop frame”, but in actuality it does not drop a frame but merely drops a frame number: Simple example: 1,2,3,4,6,7,8,9. The five is dropped, there are only 8 frames but the counter would display “9”

3000(or any amount of) frames in both 30fps and 29.97fps will last the same amount of time – is it only the time code that is calculated differently. These two time codes only count video frames and they count differently.

I don’t want to do the math but a quick a google search will give you the differences in time code running time display for real world duration/frame amount.

However, and more importantly – Broadcasters only accept Drop Frame - AND what makes it so confusing… and this seems crazy… is that 29.97 is in fact the actual frame rate for NTSC! You can edit your 5D footage in drop frame or non-drop frame – doesn’t matter – but for broadcast your final output should be in 29.97 drop frame!

Drop Frame vs Non Drop Frame for details or to confuse yourself more.

If i was wrong, someone PLEASE correct me:)

The add-on grip holds two batteries, which should easily get you through an hour with a 32gb card. If starting and stopping video with a remote works. the interviewee doesn't even need to be aware of the interruption. The video turns on and off quickly.

A manual or disconnected EF lens seems necessary to ensure the camera doesn't do anything crazy with aperture.

Hi Mike!

NTSC is 29.97 and can have either NDF or DF Timecode associated to it. The camera actually records 30fps. In order to conform it to NTSC standards you have to play the video back 0.1% slower (30fps->29.97fps). To keep audio sync you'd also have to slow your audio by the same amount. However when slowing or speeding audio your pitch changes as well. I haven't played with this much, but I guess that with today's DAWs a 0.1% speed change produces a negligible pitch change. This is also necessary when converting 24p (real 24p no t 23.976p) to 29.97.

After reading that link, I think your example is incorrect.

If you have 2 videos, one shot at 30fps, another at 29.97fps
the number of frames between the 2 will not be the same for 30 minutes worth of video.
But they're both 30 minutes long -- the last frame on the 30fps video will represent the same 'moment' as the last frame in the 29.97fps video.

And this would be the same for audio (whether it's recorder in-camera or in a separate recorder).


You can also re-render a 30fps video to 29.97 without slowing it down (and it would still be 30 minutes and it would match the audio), but that could introduce ghosting/frame-regeneration/etc.

It's better to slow down the 30fps video to 29.97, but it won't be 30 minutes anymore (as Salvador mentioned).

Correct me if I'm wrong, too !

It's best to slow down the 30p video to 29.97, dropping one frame in 1,000. This solution keeps your video at the correct duration, and is still at real time, more or less. No need to stretch your audio either.

Post removed! Sorry:)

When i read or hear 30fps i automatically think 29.97 non-drop frame... but the 5d is in fact true 30fps --- VERY VERY SORRY! All this time i thought it was just NTSC non drop frame.

Sorry - Sorry - Sorry

What tool/workflow do you use/recommend ?
I'm using Windows.

Also, even without slowing down the video, I find that I still need to stretch/shorten the separate audio track (5 minutes long) to synch with the video (I'm using Vegas and Sound Forge to stretch/shorten the audio track). That was with an MP3 file though. Need to test this with WAV files.

After Effects is excellent at handling multiple frame rates and providing good control over the result. It lets you specify everything, including selection of speed change vs. frame drop/double vs. frame blend vs. motion interpolation.

One caveat: motion interpolation is slow and flawed. There are plugins available, but I have yet to see anything I like for non-computer generated material.

For the 5D, going from 30p to 29.97, either slow down the video/audio or drop every 1000th frame. And for 30p to 24p, I'd generally drop every 5th frame.

On slow shutter material, like a smoothed out waterfall shot from a tripod, go with frame blending.

If you jump over to the BeachTek site, it looks like they're going to be offering an XLR adapter with phantom power and headphone monitoring specifically for the Canon EOS 5D Mark II. BeachTek website

I've been searching and can't really find a good answer to this, but, what's the point of bringing your pro audio into this camera as long as it has auto-gain so integrated?

I've heard/seen a few examples on Vimeo where they use a nice shotgun plugged straight in to the 5dM2 and it sounds good until there's quiet. Then you can really clearly hear the audio levels creep up during those quiet times.
And worse, the levels drop for a loud noise and take a while to recover - just like any auto-level camera/setup.

I don't specifically remember the video, but they were showing a guy who was talking; he tapped something, producing a light mettle CLANG and the levels dropped to compensate, taking his voice with it.

Even taking audio in via pro mics into the 5DM2 via XLRs into a Beachtek wouldn't fix the auto levels, right?

Near as I can tell, the only thing the camera's audio is good for is reference.

I must be missing something obvious.

Agreed, auto gain in audio is just bad! When audio is even slightly important to your shot/production you need a separate audio recorder!

maybe some one can make an audio recorder (or any device) that has a universal record button. you press "record" and all your devices start recording at same time. Although i guess each device will have different stop to record times, it would still make things a lot easier!

First, to Salvador, Jon, Bernard and anyone else, my apologies for being confused and formerly posting wrong info regarding time code and the 5D. I’m reading too many threads on too many forums!

As for converting 30p to 29.97 do you complete your edit at 30fps first then do your conversion OR do you covert your clips to 29.97 and then edit?

BTW having my eyes opened to the fact that 5D is in fact true 30fps and not 29.97 non-drop, is to me more of a pain in the ass than all the other problems I’ve been reading about and trying to find workflow workarounds for! Sooner or later the ratio of pain-in-ass:beautiful images is going to be too high to ignore :(

It's probably easiest to do at the end, but if you're going to intermix 29.97 content, then definitely convert up front. Then again, if you have 10,000 30p clips and one 29.97 clip, convert the odd clip to 30p, edit, and then go to 29.97.

It definitely and certainly all depends. :)

I agree with you, and today I discovered a clever solution for synchronizing an external audio recorder to the 5D Mark II, using the 5D Mark II's audio track as a time code reference track.

Microphone for 5D mkII - FM Forums

A little lower in the thread he has posted a diagram of his solution. He proposes using a Zoom H4 with a SMPTE LTC time code track already recorded on its memory card. Because the H4 can playback at the same time as it records, he suggests sending the LTC via a wireless link to the 5D Mark II. Then the video and audio will have the same LTC tracks attached and can be easily synchronized in an editor. Credit to "jray" there for this idea.

Has anyone tried this? Or what about attaching a portable time code generator with LTC output to the audio input of the 5D Mark II and jamming it to the time code generator in an external audio recorder? Frame accuracy isn't essential, just good lip synchronization. The less manual work needed to synchronize the video and the audio recordings, the better it is for all of us who record sound too.

get a fostex fr2-le
 

I dont know if this has been mentioned, but i'd simply get a fostex FR2-LE.

There's no way to get decent audio into the 5D2. Period. The only way to go is with an external recorder, and sync later in post (use a slate), it takes seconds. That on the other hand brings new possibilities, as the mics can be put anywhere, like far away from the camera and close to the source, if you're shooting with a tele lens, and you won't need radio transmiters for the mics.

Unfortunately, all the cheap audio recorders are full of compromises, either dont have 48v, or don't allow proper confident monitoring and have weak and hissy phone outs, or their preamps are very poor, or the AD is subpaar, etc, etc. Think that of mic as the lens, and the preamp+A/D as the CCD. Quality costs.

I mean, you spend many thousands in the camera, lenses, a video editing software etc, and then pretend to skip on the the audio, which is 50% of videography? Cheap solutions will be like using the kit lens from an entry level DSLR, you wouldn't do that.

This is not photography. Consider that audio gives continuity to images. Keep a good audio source and you can mix all kinds of video footage, regardless of picture quality, colour correction etc. The audio keeps it consistent. Now do the opposite, get perfectly edited and colour corrected footage and pair it with poor audio: it falls apart.

There's no other way to it (if you care about it) than to get a pro recorder, which are all quite expensive, starting from $1,000 up to $6,000. Under the $1,000 mark, the best deal (i've done comaparisons) is with the LE version of the Fostex FR-2, which can be found for $500 or less. It doens't have many features, but what it has, it's up to professional specs, and it's an unit you can use for any kind of proaudio work, even including classical music (for which I'd use at least hi-end external pres though, but im an audio engineer in another life).



Anyway, chek it out. It's not the cheapest option, but one to consider.

how about acceptable audio?

are those shotgun mics that can mount on the flash shoe on the camera good enough for something like an interview during an event? will it still catch the annoying IS motor sound?

What about the Zoom H4? That actually does provide 48 volts as far as I know. I hope it works well, since I just bought one.

Anyway, I'm buying up all the little pieces I need to hopefully start working with the 5d Mark II.

The options for powered mics are the Zoom H4, the Zoom H4n (not shipping just yet), and the M-Audio Microtrack II. The Microtrack has 1/4" TRS inputs, so you need an adapter cable, but given that, it works fine with dynamic mics.

I'm a few hundred miles from our production right now, but I heard a short bit from our Microtrack capture last night. It's a bit darker than I expected with our mic/environment, but I was able to boost the highs and they came through without any appreciable noise.

I don't know about the Zoom, but given the Microtrack results I heard last night, I'd upgrade to a top mic first, and a top recorder/mixer second.

The microtrack ii works decently well with dynamic mics but if you are trying to pick up detailed low level sounds I found that pre-amp can be a little bit noisy. I poked around the M-audio site and found that M-Audio even says it's not optimal for dynamic mics. From their FAQ page:

A workaround I plan to try is to use my wireless sennheiser setup to see if I can get cleaner signals using the sennheiser's pre-amp.

I would think that most of us would be using condenser, rather than dynamic mics, so this shouldn't be an issue.

The pre-production BeachTek DXA-5D is specifically engineered to compensate for Auto Gain by generating a high pitched tone which is inaudible to the human ear. The high pitch tone keeps the audio at a consistent level.

Guy, that's an awesome solution for the auto gain problem. Very clever idea by Beachtek. Can't wait for the dxa5d to finally be released.

My only concern, though, will the high pitch tone be harmful to household pets or Superman? :)

Well that's fairly ingenious and amazing.
Does that solve all problems from silent times (I can see how it would) to loud hits?
How does that all fair in Post?

Even simpler - every camera has at least onboard audio. Looking at the waveform, you should be able to see a spike on the VIDEO's waveform as well as the seperately recorded audio. You can literally line the two spikes up and get perfect matching.

I have thought about this for a while now. My original idea was to use something like this for other camcorders with mic inputs and auto-gain (then I bought a HV-30).

I think the idea is sound. You could go with either a ultra-low frequency (20-50 Hz) or ultra-high (15Khz to 20Khz). I think an ultra-low frequency would work better since I would expect better frequency response from the recording circuitry here. But I'm splitting hairs - experimentation could show which is better if there is a difference anyways.

The key design parameters are the ratio from the pseudo subsonic signal to the mic signal has to be well matched. You don't want any audio spikes overloading the input (over the subsonic signal) or the agc will kick in. On the other end you don't want the subsonic signal too high in relation to the mic signal or it will be more diffiicult to filter out in post.

The overall gain of both has to be well matched to the auto-gain circuit as well to get the best dynamic range without overloading - quite similar to adjusting manual gain for manual gain capable camcorders - except there's no indicators to help you here.

Could be a fun project someday . . .

On another note - regarding seperate audio recording - there is a big issue I have experienced myself. I record audio for my band and use seperate audio recorders to do it. It just so happens I have both a Maudio Microtrack _and_ a Zoom H4. I keep one recorder on stage to record the band instruments via two PZM mics. The other recorder goes to the house PA to pick up the Vocals and any PA amplified instruments straight off the board. I have also recorded video of the band with these recorders capturing the audio in the same fashion.

Both of these recorders have issues with the internal clock sources. They are no where near each other and neither ever matches the camcorders. The funny thing is when I use multiple camcorders they are never off by more than one frame (audio+video). I guess camcorders have significantly better clock sources then either of these recorders do.

Anyways, when I sync all the audio in post the difference between either recorder and the camcorder is startling. It is very easy to find the opening notes of a song/clip to begin sync but very quickly you can start hearing the sound phasing (caused by one signal lagging the other) and then it starts to sound like a slapback echo. BY the end of the song it sounds like avant garde music - LOL.

The fix is to time stretch the audio (I find an opening cue point and an ending cue point. I calculate the delta in time and then apply the stretch accordingly - using the camcorder audio time frame as the "golden reference"). I use Cubase 4 and it has the ability to do this. It works very well but it is a pain in the butt.

YMMV but something to consider with either the Microtrack or Zoom H4 since they both have this issue.

Do you have the original Microtrack, or the Microtrack II? The second generation had a number of important improvements - better preamps, improved handling of unbalanced mics in the 1/4" inputs, full 48VDC phantom, rather than ~30VDC...

Also, any preference in sound between the two?

I had read that the H4 had some noise from the blinking LED, but I'd imagine that was on the unbalanced inputs or onboard mics, rather than the XLRs.

Depends on on where your bar is. I am, among many other things, a professional audio engineer, and to me the differences between cheap and pro gear are easily noticeable. Also, these days ALL manufacturers lie with specs. For instance having 48v doesnt mean the mic is continually sourced with this current without any voltage dropout, or how long the recorder's batteries will last when giving 48v away (maybe minutes instead of the 8 hrs from pro units). Same when they present frq responses, or A/D/A, and they say nothing about the analog circuits feeding those converters, etc etc.

Im not saying the zoom might a bad unit as i havent used, but i tried other zoom products and i found them products for the average consumer.


Some people might be satisfied with a rode mic into the canon. I wouldn't set for anything
less than a sennheiser MKH or Schoeps mic along with a pro external audio recorder.
And later in post, I would pay as much attention to audio as to video, building a soundtrack in layers, with all the classical elements if required: (production) dialog, ADR, foley, walla, fx and music (and all at 48khz/24bit ).

I have the first Microtrack. I am quite happy with the sound quality from both. Since I am doing live recording from PZM's and PA feed I am not using the phantom function (or onboard mics) and I am sure I am not feeding in the best signal possible to really test the units. My biggest beef with both is the clock source issue.

My second beef is battery life for both are awful. The Microtrack has a non-replaceable internal battery that goes dead whether it's powered up or not. And for being a LiON battery it's stay alive time when charged is pathetic. The Zoom H4 also eats batteries (2 AA's) like they are going out of style.

Luckily for me I bought both of them them for convenience, price and size - not battery power. [Note - I was hoodwinked into the Zoom H4 thinking I could do 4 simultaneous track recording with the 2 onboard mics + 2 offboard mics - nope can't be done] You could always use battery packs and power cord adapters to drive them for better battery life.

One more note - they are targeted to practicing musicians so yeah, they both probably are not what one would consider pro-quality "field" recorders. If you are looking for "studio" grade recording in a field recorder there are much better (and more expensive) options out there.

Heres what I am stongly considering doing for my 5Dmk2. It is important to me to be able to record audio in cameras since syncing in the edit is not the best option for every format of shooting (reality, doc, event, web).

I've been looking at the Azden FMX-32 field mixer which is less than half the price of a sound devices mixpre with similar capabilities (phantom power, vu, etc but no built in tone or slate mic). The azden mixer seems to have better battery life, headphone amp/monitor built in and is less expensive compared to a beachtek (DXA-6vu is what i was considering before).

according to john saunder's tests (Canon 5d Mk2 Audio Test on Vimeo) and also it sounds like a similar route to what beachtek is taking, applying a consistent tone to the channel will prevent agc from activating, and the agc is linked for both channels, so worst case you apply a tone to R channel to control agc and record actual audio to L channel. And a 3 channel mixer would allow for one xlr tone generating dongle (like this Goldline GL1K*::*Test Equipment*::*Toolbox*::*Installations*::*Blue Dog Audio) on channel 3 mapped to R output and leave 2 xlr inputs that can be mixed into the L output channel. Of course if we could do what beachtek does and use a high-freq tone that does not interfere with other recording on the same channel (and doesnt need to be removed in post??), I might actually be able to get two recordable agc-free channels. anyone know what kind of tone generator might be appropriate for this? might be tough to monitor the channel with the tone and mic input on it since the VU meter would probably just read the peak of the tone... maybe there is a way to apply the inaudible tone after the mixer before the camera?

anyone see any holes in this plan? any suggestions?

if it works it seems pretty reasonable and flexible compared to beachtek's options which seem more expensive in general. maybe if beachtek released modular products it would be more helpful, like making a agc-defeating in-line inaudible tone generator or something.

As Noah said I did some audio tests with my mixpre and tone to see how the ACGs react. Here they are if anyone is interested.

Test with 1khz tone
Canon 5d Mk2 Audio Test on Vimeo

Test with ultra high and low tones
Audio test for 5DM2 extreme low and high frequencies on Vimeo

Hope that helps.

Zoom H4 - Michael Schoenfeld
 

Hey everyone,

Michael Schoenfeld here,

I use the Zoom H4 - it's pretty good based upon what mic you attach.

I have two options - Beer, and champagne.

Sennheiser G2 with a Countryman B6 (Beer)

Schoeps CMIT5 (Champagne)

The tricky part is not to forget that the "normal" file will be 30P, NOT 29.97 in the US of A, so it needs to be conformed with some tool so the movie AND the audio file will sync perfectly.

I conform all my files with MPEG Streamclip to 29.97 and the audio sync works that way - the timeline wants to make the clip "out of sync" with the H4 file because of this stupid Canon thing - why oh why Canon, are you not making this camera pro friendly (I "know" the answer (as good as anyone can) so no flames, please).

The nikon and zeiss glass is the only way to fly with this beast right now - thanks Canon,


Michael Schoenfeld

Just to clarify, are you having synch problems when putting 5d footage on a 29.97 timeline or are you having problems even on a 30.00 fps timeline? I ask because I ran some tests of several long clips (using MP3s captured simultaneously with a microtrack II) and with Vegas I'm seeing something on the order of perhaps 10ms of drift across a 13+ minute segment. It doesn't seem to matter whether or not the project is set at 29.97 or 30.00 fps.

People, this is not that difficult to understand. No one should be expecting perfect sync for long periods, it really is that simple.

If you are: (1) using a separate device to record audio and (2) not fully timecode jammed/synced, then drift will -- in all likelihood -- occur to some extent or another. That's true for even the most pro recorders in existence and may occur regardless of 30 or 29.97 fps or 44.1 vs. 48 kHz. Even two SD744s will potentially drift from one another over extended times if they aren't jammed, and that's two of the exact same high end units.

So, if you don't have drift problems over extended periods, you should consider yourself lucky, and if you do have drift issues, you should consider that to be pretty normal.

To me, the dual-system single system comes down to one thing: Redundancy. One little blip of static can ruin a shot unless you have a second audio source.

(In one particularly telling moment, my acting-DoP, who was really my PA, who was filling in for my REAL DoP who was arrested for terrorism - long story - actually forgot to turn on the microphone for a few minutes while recording. I'm damn glad the backup audio was running then!)

I record dual system when I can, but I always make sure to use the really good stuff on the camera if I have to choose between the two. I'd rather have a good audio track on the camera than have to sync in post.

I've recently become a convert to the XLR workflow - get a Juicedlink or Beachtek - I prefer the former for the preamp - and you should be fine.

Although Apogee has always made topnotch products, I still would recommend a Sony PCM D-50 for this kind of usage that we're talking about here.

-- peer

does anyone know of a way to monitor mic levels BEFORE they get into the camera? I was thinking some type of small adapter that the mic could plug into before going to the mic input on the MK2?