New Sound Devices 8 Channel Recorder Announced
 

Just saw this announcement on the SD website, thought I'd share it.

http://sounddevices.com/products/788t.htm

DROOL!

Dear Steve,

I found a price, from one dealer, for the 788t: $5,995.

I really like that the 788t has 8 microphone preamps.

I would like to learn more about the circular LEDs around the gain controls. I assume that these are tri-color leds that indicate signal level.

Comparing this to the 744t, you gain six more mic preamps and 4 more tracks, but you lose being able to record over 48,048 samples per second, so no 96K or 192K recording on any channel.

I wonder if a 744t can be linked with the 788t in the same way that multiple 744t's can be linked together.

UPDATE: 788T Has Video Sync!
 

FYI - was just reading up on the new SD 788T recorder announced the other day and according to SD it now has the ability to slave its sample clock to any flavour of incoming NTSC and PAL video/blackburst and HD Tri-Level sync signals. The older SD recorders would only slave to wordclock.

Dan - I think they are status indicators (still reading up). For example, when slaved to external sync, if the sync signal is lost the ring LEDs flash yellow.

I'm still not a fan of the LEDs as the VU meter. I didn't like them on the Wendt X2, and I didn't like them when I was using the 722 on a project. I wish there was another option for these, but I'm sure it's a space thing.

Wayne

48Khz sampling rate limitations?
 

Some who record live classical music are complaining about the lack of a 96Khz sampling rate in the new SD 788T. Would one consider that to be a negligible limitation when it comes to recording location dialogue? As far as recording live music, what form of instrumentation (acoustic, etc.) do you find benefits most from using a 96Khz sampling rate? As the new SD unit does boast an increase in A/D dynamic range, would that help compensate for the 48Khz sampling limit to some degree (at least where location dialogue is concerned)?
Thanks.

At risk of criticism, I'll say that 24 bit 48 KHz sampling rate has proven to be the sweet spot in acquisition. I don't know of pros who are recording at 24/96, everybody's working at 24/48.

Meaning half the disk space - twice the recording time for no perceivable difference in the recording.

OTOH, I'm not hanging out with people who are recording classical music, and don't know what those folks consider to be best practice or what may have value as bragging rights in that market.

I'm perfectly happy recording acoustic instruments at 24/48, and I do a lot of it.

48 kHz is only a problem if you're putting out an SACD. It doesn't really matter if your release is on CD.

Regards,

Ty Ford

According to the manual these do light up in colors and with intensity that tell you what is afoot with the associated signal. Green proportional to input signal turning to red as the limiter level is approached and then yellow in the non linear part with intensity proportional to the amount of limiting.

The C.Link connectors are present (and apparently a cable is part of the kit) but the manual says that these connectors are "not functional". This suggests to me that linking will be available with a later firmware release i.e. it looks as if they put linking on the back burner to get the thing on the market sooner rather than later.

We could certainly empty a few beers over this one. SACD is a 1 bit Delta Sigma modulation at a very high sampling rate with an effective bandwidth of about 100 kHz and a noise floor equivalent to about 20 bits in straight PCM. The A/D converters in this new machine are doubtless sigma-delta converters (but not one bit) with a bandwidth of 22 kHz and a spec dynamic range of 120 dB (123 dB A weighted) [Note: Sony and Phillips have claimed commenusrate SNR in the 20 - 20kHz part of the band]. Thus they have a nominal advantage over SACD of about 6x4 = 24 db but then the extra quantizing noise in the SACD is in a part of the spectrum where you can't hear it (22 - 100 KHz) so the PCM's advantage should be reduced by 7 dB corresponding to 10 log(22/100) (approximately). So a SACD recording would record sounds we can't hear (up to 100 kHz) but do so with less dynamic range. Where the application is audio that people are going to listen to for entertainment I'm convinced 48k (as implemented by a modern sigma-delta converter) is adequate (more than adequate if you are over 20). If you are doing research on bats' ultrasonic chirps or something of that sort then a higher sampling rate would be required.

Dear A. J.,

I would like to learn the difference between a delta-sigma and a sigma-delta Analog to Digital converter (ADC).

Could you please explain or point me to a location that explains the differences?

The 788t looks like a great product. Much cheaper than I expected, too!

Dan,

AFAIK Sigma-Delta and Delta-Sigma are equivalent names for the same architecture. Try the Wikipedia page for a reasonably good description. The article "AN-283: Sigma-Delta ADCs and DACs (pdf, 1,740,324 bytes)
Overview of Sigma-Delta Concepts: Oversampling, Noise Shaping Using the Sigma-Delta Modulator, Digital Filtering and Decimation" at http://www.analog.com/en/cList/0,288...255F43,00.html (Analog Devices is one manufacturer) goes into a little more detail. You may also find the data sheets for some of their A/D converters helpful (the article talks about 1 bit quantization - it is now common to quantize the differences (deltas) to several bits. Warning: this is pretty geeky stuff.

Dear A.J.,

Thank you for the information. I will check this out soon.