Nagra IV help

[Anna Harmon]

Hey all,

Anyone here know where I can access an instruction guide or a how-to on the Nagra IV?

Never used one before (I know I'm lame).

Edit: Originally had a deadline but now I don't. Still curious though. Seems simple enough but I don't have one to tinker with so any leads would be great.

I know it's not digital but I'm sure some of you seasoned guys can hook a sister up.

[Brooks Harrington]

Nagra IV which one? I'm assumimg it's the IV ST.
Here's a 4.2 mono version.

http://www.temple.edu/fma/referencel...gra4point2.pdf

Uh, ? Why would you record to Nagra with HVX200 which is probably going to lay down 4 tracks of good audio wheither you have anything hooked up to it or not?

[Anna Harmon]

Thanks Brooks.

I know. It's really dumb. Turns out the producer I was talking to had no idea what he was talking about. It sounded pretty damn strange to me as well.

The HVX-200 is an interesting camera. I might be wrong, but I've only been able to assign 2 channels. The other channels use the internal mic. And it only works in 4 channel mode when using the P2 card which makes sense.

[Brooks Harrington]

I believe you are correct. 2 channels for internal mics and 2 for external XLR.
From the limited time I played with the camera, it seemed to always bring in all 4 channel on import using P2. I don't know about the SD tape mod. When in doubt, read manual.

[Anna Harmon]

Ha ha. I always carry the HVX manual with me. You never know when you need it or when you're working with a cameraman who's using it for the first time.

Ah, the life of freelancers.

I'd like to grab the HVX-500 manual at some point as well. I don't think the audio portion is that different. Worked with them before. Very audio friendly.

[Wayne Brissette]

Originally Posted by Anna Harmon

The HVX-200 is an interesting camera. I might be wrong, but I've only been able to assign 2 channels. The other channels use the internal mic. And it only works in 4 channel mode when using the P2 card which makes sense.

No, that's how it works. I've gotten a few calls after the fact asking me about this since I usually feed it two channels of audio only. I'm still not sure what value the on-board mics provide in when you're feeding it two clean channels of audio, but you can't turn the channels off, you simply have the option of reassigning them. Although in FCP and other tools you can delete those audio channels.

Wayne

[Bob Hart]

Sentimental attachment to a sweeter sound I guess.

To someone with good ears that haven't been belted to perdition by heavymetal out of the headphones, digital audio for video still sounds a trace like fingernails on a blackboard and there is a big dead spot on the high end you find yourself straining to hear.

Double system analogue sound of course has its own bag of problems. I don't miss manually syncing up one little bit. I don't miss the high noise floor either.

[Jim Andrada]

Interesting about the "big dead spot on the high end" with digital.

Could you elaborate? I know something seems to be missing with CD sound and I was wondering if it might be related to the 44.1k sample rate.

Our newest CD player has a vacuum tube final stage and it sounds lots better. It may not be accurate (whatever that means) but it definitely sounds better

Anyhow, just wondering because we often notice the "something missing" feeling with digital sound.

[Bob Hart]

I noticed it more on the theatre sound when digital first came in for movie soundtracks.

To me, analogue dolby pro-logic sounded sweeter.

Early broadcast TV digital audio altered speech, an example perhaps inaccurate, -- "ssssleep" becomes "sshleep" or "ffleep".

Audio CDs always had me straining to hear them. Turning them up louder was no solution. They do have greater dynamic range and no more groove jumps from heavy kick drums like vinyl did in its end days.

At the end of it all, it comes down to trade-offs.

Apparently it is something to do with the sampling rate of digital recording.

CD audio is 44.1K. The highest frequency that actually comes out in the form of sound is apparently 22K approximately. It is said that adult human hearing rolls off above 12K.

It has also been published that human hearing may pick up frequencies much higher in the harmonic relationships between these we supposedly cannot and the lower frequencies we can hear. - the so-called colours of sound.

Also puiblished is an account of a mix desk which was reported by the sound engineer as defective in two channels. No one else could hear it. The story goes that a veteran engineer was called back in by the manufacturer.

Eventually it was discovered that in two channels, some capacitors had been inadvertently omitted. The audio was cutting off at 48K. The sound engineer was picking this up, obviously with very well trained hearing.

This may have been exceptional and possibly what subliminally brought the sound engineer into his craft in the first place.

I know in my childhood, I could not be in the same room as an old black and white tube television while it was warming up. The whistle off that thing was intense. My parents never heard a thing.

[Jim Andrada]

I've noticed that lately they're rating speakers as capable of 40kHz and I've seen that Schoeps has a version of their mic amp that they claim has pretty fflat response up to over 40k.

There is definitely "something" there.

Have you experimented with a digital recorder by setting the sample rate to 44.1(CD), 48 (DAT) 96(DVD) and 192 (Bat mating calls?????) and seeing (or rather hearing) if there's any perceived difference?

Of course it would all depend on the playback system as well, but I know there is something missing woth CD's. My classical pianist wife also notices it. Can't quite put our sonic fingers on it, but there is something that's not there.

[Abe Dolinger]

The big dead spot you hear on audio for video might come from compression - the DVX/HVX class of cameras almost all record to 380kbps mp3, which doesn't reproduce above (I believe) 16khz.

I know what you mean about CDs . . I think it's mostly the limitations of 16 bit. Jim, a good test for this would be to make a nice deep sine wave with a bunch of overtones. At 44/16 it will sound wrong somehow . . grating, metallic, forced. There are huge sensory areas missing in the English language and sound is no exception. But at 48/24 it'll sound a lot more like music.

One Sanken non-measurement (recording-oriented) mic claims to be flat up to 100k. Interesting stuff.

In my own non-scientific tests I can hear a sine up to about mid-15k, and yet, if I roll off everything above 18k in music I'm working on, I can hear it instantly.

I'm sure you know this already, but for illustrative purposes - 16 bit means you only have ~65k possible voltages to send to your speaker (at maximum volume). Well, the theory goes that all sounds can be reproduced by combining sine waves of different frequency and intensity. Digital is great for reproducing stuff like square waves, or things that naturally would occur in the digital domain. With a sine wave, it's going to be comprised of a series of tiny voltage steps, and as the original signal decreases in volume, the number of steps it can use to express itself goes way down. With 24 bit you have 16.7m possible "steps" and thus much greater resolution. With analogue you have potentially limitless resolution. It's my belief that this roundabout, "lots-of-little-right-angles" way of reproducing a curving line is what leads to digital's notoriously somewhat cold sound.

Sorry for rambling! Anna, if you need mixer contacts in NYC, shoot me an email. I don't know many but we all stick together.

[Jim Andrada]

Everybody blithely tosses about comments re the Nyquist frequency and samplig rates etc etc.

Just for the heck of it I've tracked down the papers by Nyquist and Shannon on the topic.

http://www.loe.ee.upatras.gr/Comes/Notes/Nyquist.pdf

http://www.stanford.edu/class/ee104/shannonpaper.pdf

The other point that I think gets forgottenis that the Nyquist frequency applies to bandwidth limited signals.

Of course, the signal can be more or less imperfectly bandwidth limited by filtering, but I think what's at issue is whether the part of the signal that gets filtered out may in fact contain information that shouldn't be filtered out.

I think the assumption about human hearing only going to 16kHz or 20kHz etc is based on having people listen to a sine wave and telling someone when they can or can't "hear" it. Lately there seems to be a growing ralization that whether one can hear the higher frequencies in this fashion or not is not the whole story and that in fact higher frequencies can be perceived in some fashion or other, and can constribute to the perception of sound.

Of course, bit depth is also a big piece of the puzzle

Anyhow, I have some reading to do. It's been almost 50 years since college, but maybe I can slog my way through this!

[Petri Kaipiainen]

I talked to a AES engineer about this and he said there are NO valid test showing people could hear anything above 20kHz. Only because it is there and modern digital equipment can record it does not mean it is worth doing (and loosing sleep over). There are zillions of other bigger problems in recording audio than trying to satisfy bats.

Another thing is the relation of s/n ratio and high frequences. It is impossible to get high (24 bit) S/N ratios at above 20kHz, because only small diaphram microphones can record those frequences and they are inherently noisy. Thus a dream of 24/96 audio is an impossible one. It is either 16/96 or 24/44.1 (in quality) and there is no way around this.

Relating to this; the above mentioned enigineer and a couple of his hi-fi journalist friends reviewed the $50000 JBL Everest loudspeaker system which goes clean to 40kHz. While they all agreed that it is The Best loudspeakers in the world by a sizable margin, they also could block the superhigh tweeter (above 20kHz) and did not hear any difference in sound. Draw your own conclusions...

[Jim Andrada]

Petri,

Good points!

But why does CD audio sound - "cold" to so many people? Much colder than the performance itself. Even when heard in the venue in which it is recorded.

Why do a lot of people think that vacuum tube systems sound closer to the way the instrument sounds.

I play brass and my wife plays piano and my cousin played cello for years in the Philadelphia Orchestra and every one of us has had the same feeling of a lot being somehow missing from CD recordings. In spite of the noise and static and other problems, we didn't feel this so much with vinyl records.

Do you think it's a consequence of our sound concept being formed by being closely coupled to the instruments and feeling the sound as well as hearing it?

I doubt that we're all fooling ourselves.

(I also know that a lot of people can't tell that anything is "missing")

OK, we could say that the sound has to be different because a loudspeaker isn't a real acoustic instrument so all sound reproduction is a rough approximation or delusion, I guess

I'd also be interested in why you think that 24 bit/44.1kHz or 16 bit/96kHz have some kind of equivalence. Why would bit depth and sampling frequency be so symmetrically related?

[Petri Kaipiainen]

Analog audio and tube equipment add a small amount of harmonic distortion to the signal. This makes the signal "warmer" and smoother, which to some listeners equals "better". It might be so, but it is not accurate. I think the recording system should record the signal as faithfully as possible, if people want to modify the signal to make it sound "better" it is up to them. But we need to be honest about the "better" signal not being more accurate.

It is no possible to mic the instrument and get the same sound and feeling as the artist playing it. The best we can try is to get the sound the audience is getting. Certainly the violin sounds different to the fidder than to the listener, but what do we want to record. And should the listener hold a loudspeaker like a violin? :-)

16 bit recording is bashed because it is now possible to record with better dynamic range with 24 bits (bit depth ONLY affects the maximum dynamic range, or how far the system noise floor is, NOTHING to do with more "detail" whatever people mean by that, etc.), Still, good 16 bits is capable of over 94 dB of dynamic range which is considerably more than what people can reproduce at home (listening room noise floor is typically around 40 dBSLP and good stereo played real loud can output 110 dBSPL peaks = 70 dB usable dynamic range).

Real 24 bit or about 140 dB dynamic range does not exist in real life, or at least nobody can reproduce and listen jet take-offs and whale grunts at home at realistic levels (more than once at least...). Another thing to consider is the S/N ratio of the analog preamps and A/D conversion. Even the best recorders like Sound Devices 7xx series can muster only 110 dB S/N ratio, and best mics about 90 dB. So there is no automatic quality advance of 24 bits over well done 16 bits in real life. With measuring equipment and special listening arrangements maybe yes, and off course the headroom safety in recording at 24 bits is a good thing for engineers.

What I was trying to say with that 16/96 and 24/44.1 is that it is a vain thing to demand 24/96 recordings and claim them to be superior because of their extended frequency range and high dynamic range. This is because it is not possible to record above 20kHz signals with better than about 75 dB S/N ratio, which is about 14 bits in bit depth. So, if you claim 96 kHz sample rate is needed to get those above 20 kHz frequences you have to admit that 24 bit bit depth is not needed, 16 is enough, and if we want 24 bits for better dynamic range and lower noice floor, 96 kHz sampling is an overkill as we have to use microphones whic roll off at 20 kHz and can not get high frequences. Can not have both at the same time.

I still maintain there is no reliable scientific proof that recording at above 20 kHz gives any benefit, or that more than 16 bits are needed for real life applications. And it is not because there have been no tests, just that in controlled tests people have not been able to hear any difference.

Addendum: there were some problems with early digital recordings for two reasons: distortion added by analog systems hid some mic placement related problems which became audible with more accurate digital systems. For example DG had to remaster all their older multitrak recordings for CDs, because the time differences between mics were not compensated for LPs. Another problem was the lowpass filters used in A/D and after D/A in CD players to block signals above 20kHz. Those early filters had some phase altering ripple effects which extended to below 20kHz. Systems have gotten better since then and it is true oversampling helps in this matter, which might be the reason some people claim to hear the difference between 44.1 and 96; not the sampling rate itself, but the filter anomalies which in 96 systems are well outside the hearing range.