Re: Stumped, baffled, and perplexed.......

"Frank" <frank@xxxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message news:loufc5docj42gutnld9mk5ve6esivgmc3r@xxxxxxxxxx
On Sat, 3 Oct 2009 11:52:05 -0400, in 'rec.video.production',
in article <Re: Stumped, baffled, and perplexed.......>,
"Smarty" <nobody@xxxxxxxxxx> wrote:

"Frank" <frank@xxxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
I fully understand what you're saying about reading the Theory of
Operation. Unfortunately, the Sony service manuals don't cover such
territory. It would be nice if they did.

Of all of the service manuals that I have there, the only one that
comes to mind that discusses theory, or at least includes some sort of
narrative of how it works is, believe it or not, the service manual
for the Sanyo VPC-FH1. I found that one to be interesting reading.
Take a look at it if you have the time.

The VPC-FH1, in case you're not familiar with it, is a consumer-grade
AVCHD camcorder that claims to do 1080p59.94 video - a world first,
along with the VPC-HD2000.


I will check out your website. Been awhile since I last visited, and I am looking forward to seeing some of the AVCHD camcorder details in particular.

Not to be my usual picky self or anything, but you wrote "sampled at
64 kbits/sec". 64 kbps is the data rate, not the sampling rate.



Indeed, and the spectral analysis of the voice content I see here when equalizing this audio track looks like it is band-limited to around 100-3300 Hz. I am guessing it was actually sampled at maybe an 8 KHz rate but have not looked at the header to see what was actually used.


Six 2 GB DDR3 sticks totaling 12 GB. Must be ECC (else why not just
get/build a fast gamer-like desktop?), so that means slower 1066
memory rather than faster 1333 memory, or so I understand. I *hate*
I think that Apple uses 1066 MHz Registered ECC DDR 3 memory in all
current Mac Pros. They don't offer ordinary DDR 3 memory in any Mac
Pro, whether 1066 MHz or 1333 MHz, as far as I know.

Puget only offers regular 1333 MHz DDR 3 memory or 1066 MHz Registered
DDR 3 ECC memory, although I do also see regular 1600 MHz DDR 3 memory
listed on their Web site. Now that should make a difference for
someone like me who mostly edits uncompressed and therefore is moving
a lot of data.


Having paid the considerable RAM premium once for the Mac Pro RAM, I would personally opt for faster non-ECC, and would love to own a 1600 MHz DDR3 motherboard and RAM. I have to believe that good thermal design with enough heat sink / spreader / blower / cooler hardware would give you a stable and extremely high performance solution for less total cost and higher throughput. I'm not a big proponent of liquid cooling, nor do I appreciate fan noise, so I guess I would have to take a hard look at the trade-offs.

By the way, as an old guy, I come from the days when as far as memory
is concerned, there are (were) nine bits in a byte - eight data bits
and a parity bit.


Some of the not-too old machines with SIMMs still support parity checking, but apparently DIMMs no longer offer this option other than the full ECC approach. I would be surprised if soft memory errors occur often enough to justify the extra cost for video work. If I were building a number cruncher for science or finance apps, then I would probably spend the extra $$ for ECC.

I believe that the fastest chip that Apple is currently offering in
its Mac Pro systems is a quad-core Xeon X5570 at 2.93 GHz whereas
Puget offers not only that chip but also a qual-core Xeon W5590 at
3.33 GHz. It's raw processor speed that I'm looking for so I would go
for the 3.33 MHz chip.


MaximumPC magazine and elsewhere have been doing some reliable overclocking above 4 GHz, but I would personally never go that route. The NLE software and hardware is flakey enough that the last thing I would want to add was a marginally stable CPU and bus.

Audio will be either Lynx Studio Technology (probably an Aurora 8
since Apogee has given up on the Windows market) or maybe a pair of
Benchmark Media Systems A/D and D/A converters in the highly unlikely
event that I can talk myself into just two channels in and out (I've
owned a DAC1 for years now and love it) and video will be from
Blackmagic Design (don't know which card(s) yet, but must be able to
support a variety of formats as a lot of the work that I do is with
legacy formats).



I would want at least 5.1, possibly 7.1 editing and authoring, but it depends a lot on what your customers demands are and what content you will be handling. Then again, with Japan Incorporated ready to launch 3D television into the world market, it will be interesting to see what if any new formats arise which demand alternative video and audio processing.


Besides, I want a 25 GHz chip (even if it's liquid cooled), not a slow
chip with more cores.


It is taking a long time for the migration of software to parallel processing architectures. Many if not most of the apps out there do not exploit the true parallelism which the cores provide, and Vista does a poor job of spreading the work and adjusting workflow dynamically. Multiprocessor OS designs have been around for decades, with IBM's Deep Blue and boxes from MassPar, Cray, Amdahl, and others, and even the minicomputer vendors of the 1970s and 1980s like DEC were doing reasonable distribution of processing with VAX clusters and DataGeneral Eclipses. It will be a very long time, IMHO, before the PC world makes this migration, and thus the uber-fast single CPU (25 GHz!!) solution wins my vote for sure. The biggest issue is the liquid nitrogen cryogenic hardware, which is still a bit pricey. Not sure what temperature Gallium Arsenide melts at.......................................... And if Microsoft still "owns" the PC market for operating systems, they have yet to figure out how to get single threaded, single core code to run reliably, so I would speculate that our great-grandchildren may be among the first to see truly reliable and fast multicore solutions such as now used in the supercomputers with hundreds of cores. (Ironically, the Playstation 3 is a good example to the contrary.)

.

Relevant Pages

  • Re: Verbose functional languages?
    ... Speaking of multiple cores: when I look at what's Intel talking about, ... Even memory allocation would create all kinds of mutual wait ... I'm less convinced about general use of semi-space moving collectors ...
    (comp.lang.functional)
  • Re: Target market for Intellasys.
    ... I was wrong about that Ambarella chip, it's average power requirements are more than I thought. ... With the 1 transistor dram, the substrate acts as a capacitor, so theoretically you get many times more memory density, good speed etc. ... I for one would be dropping in 10+DACS, extra processors, extra memory, and if available 36bit processor cores and full external SRAM memory buss mapped to one core. ... But such a scheme would allow customers to easily order a module populated with a desired amount of memory cores, and it would cost intellasys a lot less than putting memory on the processor. ...
    (comp.lang.forth)
  • Re: Stumped, baffled, and perplexed.......
    ... I just recently added a number of service manuals for Sony AVCHD ... memory rather than faster 1333 memory, ... I believe that the fastest chip that Apple is currently offering in ...
    (rec.video.production)
  • Re: Straight-Up Replacement For P4P800-E?
    ... work with a couple SATA drives, and use it's memory. ... The reason I did it the way I did, is Win2K only supports 2 cores, ... Expected System Bus Frequency: 800 MHz ... there is a "hidden warranty" for ICH5/ICH5R failures. ...
    (alt.comp.periphs.mainboard.asus)
  • Re: Suggestion on computer for synthesis and simulation of FPGA
    ... Will DDR3 memory have an advantage over DDR2 memory for FPGA ... FPGA placing and routing (I don't know about simulation) is *mostly* single-threaded. ... So unless your simulator is fully multi-threaded, anything more than 2 cores will not speed up FPGA work. ...
    (comp.arch.fpga)
Loading