XL1 12-bit vs. MiniDisc

[Ola Tuvesson]

Hi All,

If I wanted to record four audio tracks (2x lavs, 2x booms) during a shoot which would give me better sound quality: 4 channel 12-bit on the XL1 or 2 channel 16 bit on the camera and an extra two tracks on a MiniDisc?

I'm tempted to go 12-bit/4 channel as it would save a lot of hassle, both during and after shooting. Would good quality external pre-amps and compressors help counteract the poorer resolution?

Any input much appreciated!


Ola

[Glenn Chan]

I'm tempted to go 12-bit/4 channel as it would save a lot of hassle, both during and after shooting.

Some (all?) NLEs like Final Cut take two passes to capture the other 2 channels.

Double system sound would take more effort though.

12-bit audio isn't very good... the bit depth is 12 bit as opposed to 16bit, and the sampling rate is 32khz instead of 48khz.

Do you have a laptop by the way? You can add a USB or Firewire input device and turn your laptop into a good field recorder.

2x booms

2?!

[Ola Tuvesson]

Thanks for your reply Glenn, I would be using Scenalyzer for capture which can render tracks 3 & 4 as a wav file during capture. The wav gets the same name as the avi so it's easy to locate even with many clips.

I have heard a lot of comments about 12-bit/32khz sound not being very good but I wanted to know how it compares with MD recordings. MD uses a lossy compression called ATRAC (similar to MP3) whist 12bit/32khz DV sound remains uncompressed. So, is the compromise in terms of compression (MiniDisc) or sample rate (12-bit) preferable?

The laptop option is an excellent suggestion, I will immediately look into what FireWire A/D converters are out there that would let you capture 4 channels simultaneously.

Btw - what are your opinions on using a pre-record compressor/limiter during shoots? I got it in my head that this could not only help prevent digital clipping (which is an absolute no-no) but also improve general sound quality. legibility and noise floor. Wrong?

[Adrian Douglas]

Ola why not look into buying/renting a portable mixer. That way you could have youur 4 channels at 48k/16bit. The only draw back is that you won't have as much control during post. I'm not sure on which is the best mixer but Mackie's seem to be quite popular.

[Ola Tuvesson]

Hi Adrian, thanks for your reply. I've looked into the mixer route but it sort of negates the benefits of having four mics in the first place; security and flexibility. Having four independent tracks in post is just too tempting.

[Douglas Spotted Eagle]

Avoid 12 bit. If there is noise, it's harder to deal with in post due to the lower resolution. Canon and anyone else making a 12bit/32K camera should be hanged by their lack of interest in good sound. Even 16/48 sucks by true audio standards, 32K is inexusable. Hmmm....gee.....why doesn't someone build an NLE with 6bit color correction? Because the resolution sucks. Same with 12 bit/32k. Maybe 12 bit 192K would work. 12 bit still sounds worse than an MD with ATRAC. Compression and resolution are related indeed, but with compression, you can have various resolutions that are compressed. With uncompressed low resolution, you can't decode it and have a decent sound, because it was never encoded. Consider how a zip file works, maybe that's a better simple explanation.

[Ola Tuvesson]

Well the XL1 and other sub $100k cameras may well suck but they remain truly amazing tools for what they are. I have to say I'm hard pressed to tell the difference between dialogue recorded at 12-bit/32khz and 16-bit/48khz. But I also collect MP3s and I'm fairly familiar with the side effects of compression which can clearly be heard even at 160kbps on some tracks. ATRAC compression is far more similar to MP3 compression than ZIP in that it is a psychoaural algorithm that removes things our ears/brains hopefully won't notice as opposed to ZIP which is a lossless compression (simliar to GIF) where no data is lost.

As you say, noise would be a bigger problem at 12-bits, but what about using a good compressor/pre-amp to hot up the signal before going to the camera. Wouldn't this negate much of this lost resolution?

[Ola Tuvesson]

I just read through this thread and realised it could potentially be totally disorientating... I am of course talking about to completely different kinds of compression here; one being the digital data compression used to squish as much possible storage space onto a MiniDisc (ATRAC) - the other being the use of an analogue audio compressor outside the camera to even out the difference between weak and loud sounds (in the pre-amp).

[Glenn Chan]

If you aren't doing run and gun then you can set levels properly to avoid digital clipping. On better equipment with greater S/N ratio/dynamic range you will have more headroom. A compressor or automatic limiter can help avoid digital clipping but isn't necessary if you set levels right.

http://www.dv.com/features/features_...se_feature1102 <-- that article has some information on how to record good 2-channel sound on the XL1 by using the RCA inputs (you'll likely need a field mixer to do that).

Not sure if 12bit/32khz is ok. It might work for dialogue, where you don't need high frequency response. The best thing to do would be to test it out in real world conditions (actually recording 4 channels on camera, not sitting around our computers at home and downsampling stuff). If you can't tell the difference... then practically speaking is there really a difference? Maybe try processing the sound a little to see if 12-bit becomes an issue (quantitization error should be higher and higher noise?). If you read the dv.com article linked above, you can see specs on MA-something adapter that allows 4 channels. Using the RCA inputs is a bit better. I actually have no idea if 12bit/32khz is good enough so take what I said with a grain of salt.

There may be alternate approaches with your mic setup that may get around your problem of 2 channels on the XL1. Maybe you just need a boom and a lav?

[Douglas Spotted Eagle]

I didn't say the camera sucked, I said the 12 bit audio sucked. We've used XL and GL's to great successes.
The problem is, due to the low resolution and potential for headroom, it's difficult to edit anything with the audio without artifacting. I'm the last guy to argue that multiple tracks is better. In fact, I highly recommend it. I simply don't recommend using a camera to do it, that's all.
I fyou can't hear the difference, then it's likely not to matter. Most folks can hear it cleanly. Using a good pre will help, but it surely doesn't negate anything. Worse, if you hit your AGC and it's 12 bit audio, you're effectively screwed. You asked if the 12 bit 4 channel will give you better sound quality. The answer is a loud, resounding "no." Will it give you more flexible sound quality? Yes. Enough that it warrants the significant loss? That becomes a matter of opinion, and in my opinion, 2 mono mics at higher resolution beats 4 mono channels of low resolution any day. Not to mention the additional capture hassles.

[Ola Tuvesson]

Thanks for the link Glenn, unfortunately dv.com seems to be down at the moment but I'll check it out once it's back up.

It is of course possible to set levels so that you don't get distortion but in a feature the dialogue can sometimes be quite dynamic, suddenly going from gentle whispers to loud shouting. If the soundman is familiar with the script I suppose it would be possible to ride the gain but setting a fixed level would mean low voices are dangerously close to the noise floor. This is probably more true on "prosumer" class cameras like the XL1 which has pretty poor s/n ratio and this is precisely why I thought of using external pre-amps with gate/compressor/limiter. I would be very interested to hear what people think of the use of compression for on-set recording, I've used it a lot in the studio and think it could prove quite useful?

I will do some tests later on a/b comparing 12-bit & 16-bit on my XL1, if there is interest I can upload the files so you can access them.

But the real question remains; how do 12-bit 32khz recordings compare with MiniDisc? It seems the use of MDs is quite poular here?

And if I were to use only one lav (omni) and one boom (hypercardioid), how would I ensure getting good sound from all actors? Or do you just go with what you get and rubberband the levels in post?

[Douglas Spotted Eagle]

This is where things can get really silly....
If you are willing to carry additional hardware to the shoot, such as a rack of 4 compressors, why not just carry a laptop or a 4 track HD recorder, and use that to capture PCM files? It seems to make more sense. You'll have greater dynamic range, full resolution, more flexibility in terms of mixing and operating the mix, and less headache overall.
The bigger question that I've not heard asked yet, is "what are you shooting?" This is the critical component in all this. If it's a band, or set of a number of actors, then you're best off indeed, going multitrack. It's rare on video shoots, but it's the way to fly if you want maximum control in the post process. You could do very well with 2 boom operators, or wireless lavs on all actors, tying into the aux sends of a house mix console, etc. A lot of the question boils down to whether or not you're willing to carry a tremendous amount of extra gear, and for what purpose? What kinds of shoots are you doing?

That said, if you are a musician, and you appear to be based on your knowledge of audio, you'll DEFINITELY hear the difference between a 12/32k recording and a 16/48k recording. It's no small difference, unless it's well mic'd dialog, and even then, listen to the background noise more than you listen to the dialog. You'll hear what's happening, and due to the lower sample rate, it makes editing more difficult.

[Ola Tuvesson]

Ok Douglas, sorry if I misunderstood your post, I just get a bit frustrated when people say "oh, there's no way you'll be able to get good sound/picture with that camera/recorder, you should get yourself professional gear instead" For most of us this is simply not an option and compromise is the name of the game. It's not that I think you can do as well with semi-pro gear, of course you can't, but it is what I have to work with and I have to make the best of it. I'm sure many are familiar with the situation...

Besides, compared to the S-VHS & Hi-8 cameras I started off with, MiniDV provides far superior sound and picture and it is possible to come surprisingly close to "pro" quality.

I'm also aware that 12-bit sound can never sound better than 16-bit, my question was how it compares to MiniDisc which imposes its own set of compromises but still seems to be widely used.

I have a bit of money to spend for an upcoming feature shoot and wanted to check opinions before rushing off to the shops. I have two EVO lavs (for talent), one Senn shotgun (as backup) and an AT hypercard (for ambience) and I want to record all four mics individually and mix-down to mono in post. As far as I can see I have three options:

1) Use the camera's four tracks and spend the money on good quality preamps (possibly poor sound quality)

2) Use two camera tracks and record the other two on an external device, i.e. MiniDisc (extra set/post work, sound quality?)

3) Record all four tracks onto a laptop via a FireWire A/D box (fiddly, no money left for preamps)

The reason I like option 1 so much is its simplicity. No aftersync, no clapboard, no extra extra boxes, no extra "rec" buttons to forget pressing, no additional capture and all stored on a single tape. But if the consensus here is that sound quality will be poor I will take that advice and re-think.

[Douglas Spotted Eagle]

If you had a device such as the Echo Mona to plug directly into your Mac or PC laptop, you don't need pre's, cuz they're already built in. You can also get something like the M-Audio or PreSonus box, which also have preamps built in. They aren't Hardy or Avalon pre's, but they are decent. Certainly better than any Beachtek or other device in that vein. If you have a PC, you can get your hands on Mackie's new Tracktion recording software. It's pretty amazing for what it is is, and it's currently free, soon to become $80.00. It's got decent compression, EQ, and other functions built right in.

[Ola Tuvesson]

Thanks Douglas, you do have a point there! My plan was to get four JoeMeek MicroMeek MQ1 and power them off a 12v lead-acid battery pack. The whole kit would easily fit in a shoulder bag and would be highly portable.

A four track battery powered HD recorder sounds like the ideal solution but I have yet to find one! Sure, there are a lot of porta style mixer/recorder combos but they all seem flawed in one way or another, does anyone know of a box with four balanced inputs and a record button?

And yes, you're quite right, my sound recording knowledge is purely from the musical world - although I wouldn't consider myself a pro. I've done a fair bit of studio work and know my way around consoles and outboard gear pretty well. This is precisely why I'm so keen on multitracking and processing, something I'm aware is quite rare in the video world. It just makes so much more sense to me!

Btw - I will be doing sound on a $100k feature with scenes of up to four actors. It will be mostly shot indoors and I will have access to AC power but I want the kit I buy to be flexible enough for any future projects too. I normally do documentary/interview style shoots so I want it to be highly portable, battery powered and quick to work with both during and after shoots.