Audio & FPS

B

boothba

Well-known member
Any word from Red about synch speeds and audio recording? I'm assuming all 24fps and 30fps (50fps for PAL) frame rates are on the drawing board (4K & 2K 24p & 30p, 1080p, 1080i, etc.)

Thinking outside the box here, would it still be possible to record standard audio when shooting offspeed? Even if it's just a primative "wild audio grab" of the event that we have to re-synch later? You know - for those great 60fps shots of people screaming "Nooooooooo...." that are a must for every serious film.
 
I too would love to shoot 60p slow motion of Waves and have the audio too that I can then slow down in post and pitch correct.
 
I think all we can do for audio in off-speed modes is record it to a seperate file and let you do with it what you will. Does that make sense?

Graeme
 
yes, because to me ( note : to me ) anything else is pure fx-audio and unnecessary in 99,999999... % of all projects
 
Graeme_Nattress said:
I think all we can do for audio in off-speed modes is record it to a seperate file and let you do with it what you will. Does that make sense?

Graeme
Click to expand...

Yes that would totally rock Graeme. Right now I'd love to have some kind of audio record on the HVX in VFR but nothin'. PS- what's the audio specs?

-Noah
 
Graeme_Nattress said:
I think all we can do for audio in off-speed modes is record it to a seperate file and let you do with it what you will. Does that make sense?
Click to expand...
What speeds would be considered non-off-speed? That is aren't all speeds treeded equal?
 
Well, if you have the camera set to 24, then any speed used other than 24 is off speed. If you've got it set to 29.97, any speed other than that is off speed, etc.

Graeme
 
But what does it really mean to set the camera to 24 fps? I mean, if you're shooting at 43 fps (to pick an arbitrary non-standard frame rate), you're shooting at 43 fps. The camera doesn't need to care too much about what speed you intend to play that footage back at. If anything, this would just be a piece of metadata.

Now, obviously if you're shooting sync sound at 43 fps and you intend to play back at 24 fps, the sound might be of rather marginal utility... but it seems to me the right approach for a professional camera is to let the operator worry about that. That is to say, just record sync sound as if the footage was going to be played back at 43 fps, and let the operator do whatever he wants with the data in post... slow it down in editing to match the length of the clip when played at 24 fps, use it as wild sound in mix, whatever.

The technical problem of syncing up sound with video at arbitrary frame rates can't be that hard, right? QuickTime and other desktop video frameworks don't seem to have any trouble with it, anyway.
 
Ah, but you can tell the camera what your target fps is. Say you're making a film, you set it to 24. Now when you shoot 60fps, it knows you want that as slowmo, or 10fps as fastmo.

Say you then tell the camera you're doing a 30p production, if you then shoot 24, it's fastmo, and 50fps would be slowmo etc.

Graeme
 
Wow, that's very useful! Good stuff.

I do note that you think the audio is 4 channels of 24bit/96kHz, but I have read elsewhere that it will be 4 channels of 24bit/48kHz.

I know all specs are subject to change, plus the audio side isn't really your department, so would it be possible to clarify the latest thoughts at Red HQ as to which is the correct one? Personally I think 48kHz is enough for a camera, but 96kHz would be sweet as it basically futureproofs the camera for the upcoming HD audio standards which all specify 24/96 type audio.

No probs if you can't lock it down, I understand everything is in a state of flux right now :)
 
Cool stuff Graeme. So if I understand this right, the REDCINÉ will take eg materal shot at 60p, but flagged as for a 25p programme, and output a Quicktime Movie that is 25fps slow-mo. That is very useful, since as far as I know there is no easy way in FCP to put a 60p clip in a 25p time-line and make it fit "frame-for-frame". Or does one of the Nattress plugins do it? If not, consider it for the next revision.

It would also be useful if REDCINÉ could do this same frame-rate conversion "after the fact" in case the camera was set wrongly when the footage was shot. But then you've probably thought of that already. You do seem to be very good at thinking of everything!

Nick
 
G Map Frames in the standards converter can do this, or people also use the Cinema Tools conform feature.

And no, you're not tied to the metadata recorded as you shoot. You can alter that to decode things differently.

Graeme
 
Audio specs, off-speed playback

Audio specs, off-speed playback

Usual caveat about specs subject to change ertc.... but the audio specs are are published in the on-line spec sheet -

4 channel uncompressed, 16 / 24 bit, 48KHz

Speed up / down isn't necessarliy performed in the REDCINE, although it could also be done there. What happens is you specify a PROJECT resolution and frame rate - for example 1080p 24fps. This defines which timecode to use, HD-SDI output formats etc...

So if you capture frames at a faster or slower rate than that, we just write the frames to disk (or flash) faster or slower. Then when we play the clip back (on camera via the REDCINE) the clip is played at the PROJECT frame rate - in this case 24fps; and you automatically have your speed up or slow down.
 
Cheers Graeme. I should read the instructions in your plugins properly before asking stupid questions!
 
Stuart English said:
Usual caveat about specs subject to change ertc.... but the audio specs are are published in the on-line spec sheet -

4 channel uncompressed, 16 / 24 bit, 48KHz

Speed up / down isn't necessarliy performed in the REDCINE, although it could also be done there. What happens is you specify a PROJECT resolution and frame rate - for example 1080p 24fps. This defines which timecode to use, HD-SDI output formats etc...

So if you capture frames at a faster or slower rate than that, we just write the frames to disk (or flash) faster or slower. Then when we play the clip back (on camera via the REDCINE) the clip is played at the PROJECT frame rate - in this case 24fps; and you automatically have your speed up or slow down.
Click to expand...
Thanks Stuart, that's an amazingly simple yet effective way of capturing variable framerate footage!

My only additional question to add to extrapolate from that would be - have you considered speed ramped footage as well? Fixed, variable framerates are one thing, but having the ability to actually ramp the speed while shooting using an intervalometer of some sort would be absolute heaven and the final big hurdle to truly make this the ultimate high def camera!
 
Offspeed capture

Offspeed capture

Maintaining our analogy to a film camera, the above process would work irrespective of whether your off-speed acquisition was at a single fixed frame rate, continuously varying frame rate, or in time lapse mode - for example when driven by an external intervalometer or motion capture rig controller.
 
In regards to the audio specs. It has come up on these boards before about the RED offering too few channels of audio. Personally, I think four is just right. And I advocated that in most cine situations, the sound recordist is going to e recording to a high res audio recorder and feeding the camera a mixdown or guide track. So in those situations, does it matter that the audio is any higher than 48K, probably not. The 24bit 48K spec is certainly acceptable. However, I do wonder if the 24bit 96K might be a better approach, if you have the bandwith to do it. Here is my reasoning:

Let's assume two things. This is an independent film and they do not have a dedicated hard drive recorder. The sound recordist will be feeding audio directly to the camera. This audio will be what ends up in the film. This to me warrants the 24bit 96K recording format and it also makes me curious about the mic pres that will be used. If the above mentioned film did have a mixing board that was feeding line level into the camera, then as long as the ciruitry was clean on the camera inputs, there would be a nice 24bit 96K audio signal (there is a siginificant difference between 48 and 96K, especially later on when you go to process the signals). But let's say they decide to hook a nice scheops into the mic pre on the camera and those mic pres are not that good. You will be recording at a higher quality, but you'll have some inherent problems with the poor quality of the mic pres.

So, I guess where I'm going with this is that if you were going to do the onboard audio at a high enough level to make it work well in a Cine situation or in ENG, the specs would look like this:

4 XLR (or mini XLR) inputs
Mic/Line switchable
Low noise, high gain mic pres
24 bit 48/96K switchable
Good limiters
Low cut filters (for dealing with wind, low end sound problems)
Good volume adjusters, either knob or data(If data, make sure they are in small enough increments for fine adjustments)

Now, I have gone on record before as saying I think the visuals should come first with this camera, and I do believe that, as I think a double system works much better when doing movies. The sound recordist should have full control over their domain and not be tethered to the camera. But I also see the need for situations where budget will not allow for a hard disc recorder or even a DAT/DVD-RAM recorder. So, having the option to record high qaulity onboard audio would be a good thing.

Also, perhaps there could be something in RedCine that would allow audio to be converted if necessary from 24bit/96K to say 16bit/48K if that is the spec for offline editing. Just a thought.

Steve
 
You must log in or register to reply here.
Back
Top