Just a FYI- I recently got the Sony HDR-FX1 camcorder and measured the audio parameters using the RMAA (RightMark Audio Analyser) program, using the line-level input and feeding it from an audio CD player output. I got a measurement of 92 dB SNR. (that's referred to 0 dBfs, apparently some people quote SNR assuming 12 dB headroom in which case the figure would be 80 dB). This value is significantly better than the measurements of audio performance I made of my other two video cameras (Sony VX2000 and Panasonic DVX100). I put the details with spectrum plots here:

http://beale.best.vwh.net/measure/audio/FX1_RMAA.html

Subjectively my ears agree that the audio quality is very good on this camera. I have to say that the image quality is very nice as well!

I am not sure that all of that is good news... the Frequency Response is iffy. I have had great results with my Z1 and I am sure that the audio path is the same between models. But thank you for the testing.

My only thought is if you had to dial down the input to 3 is it possible that the input was too hot? There is an audio setting that inserts a "pad" or attentuation. If you heard a limiter kicking in, what did that sound like?

Thank you,
George

I forgot, the THD seems really high, but who is recording a symphony with this camera? Good for speech, but for the tough stuff you would want a stand alone recorder with cool German mics.

Yes, there is certainly low-frequency roll-off, and ideally it wouldn't be there. On the other hand, in my seven years of amateur and semi-pro shooting I have never had the opportunity to record in a venue that was clean of ambient rumble below 100 Hz, so I usually want more low-pass filtering anyway.

The limiter seems to turn down the gain on hot signals, from the limited testing I've done it looks like a soft limiter that starts around -15 dBfs and gets stronger as you go up towards 0 dBfs from there. I assume this is to prevent hard clipping. So if you suddenly turn on a fixed-level tone near full scale, it will have a sort of "spitting" attack as the limiter kicks in over a period of about 7 milliseconds. Again, this only affects the loud signals, quieter ones go untouched.

Another note, the THD is a strong function of the amplitude of the input signal, I wonder if it is related to the limiter. The ratio of signal to distortion is about 10 dB better with a -24 dBfs signal, than it is with a -14 dBfs.

John,

If it does a soft clip before 0dB, that is good, because as you know there is no headroom with digital. I still shoot BetaCam SP and you can push damn near everything into that without negative results. I also shoot with a DSR-500 and I set peaks at -20/ -12 just to be safe.

I am a child of analog, but am adapting to the new paradigm. I still listen to Vinyl records on a Michell Gyro-Dec.. sounds great to me. If you want to hack your CD player I have some tips.

Thanks for your hard work on the camera... good info for better or worse.

George

Problem is the audio is compressed to 384KB unlike the regular PCM uncompressed DV audio, so even if it is better from the camera, it will not last long if its compressed again anywhere in the editing path which is almost sure to happen if you want to edit your audio.

Note that I am using the camera in normal DV mode, which does not use any kind of compression on the audio. I haven't yet tried any significant recording of audio in HDV mode.

You might wonder, why buy a FX1 if you only are using it in standard-def? One reason is that the image quality of this camera is cleaner than any other SD camera I've used, by a significant margin. Clean images translate to better DVD encodes also. It's like finally getting a true broadcast camera for a semi-affordable price. Also the controls are nice, for example I like the FX1 ring zoom control better than any other "prosumer" camera.