wrote in message
oups.com


Actually the original post comes from a conversation I had
with Mr. Krick one day as I lamented my current 20 bit
Layla
based set up. Wasn't bad for it's day. In fact was kind
of
fun.

I still have a 20 bit Layla, but it ties my DAW in knots
after about 30 minutes of recording with 20 channels.

But I've done so much to improve sound quality in other
areas
in the interim. Got some great pre's (including the
Mercenary
River NV). Got some great mic's including a U87. It's
time
to do something in the area of conversion. Want the best
converters I can find in the $3K-5K range.

I think several posters have mentioned Lynx Studio. If
you're spending that kind of money Lynx can absorb it
getting to 12 channels, and you'll get great performance for
your money.

Resolution sort of became the focus of this discussion.

Audible or measured?

Probably my fault. Should have made more clear that my
top
consideration is best quality converters for the price
range
indicated, end of story. Am looking into Aurora 16.
Thank
you for that suggestion.

So then you'd need a Lynx AES16 card on top of the Aurora?

All of this said, I'm not sure resolution isn't a
consideration.

There's a controversy over what is enough resolution.

Some here have said that there is no
difference, you'll only dither to CD quality anyway.

Worse than that, look at the dynamic range (IOW resolution)
of the real-world input signals. The real world is a noisy
place compared to modern converters.

Maybe
so. I couldn't claim to know for certain. But after
recording a track I tend to apply multiple filters (high
pass,
compression, limiting, etc). It makes no sense to me from
a
computational standpoint that greater resolution during
filter
applications does not minimize error at the final
dithering
stage.

Of course not. However, if you have a signal with 80 dB
dynamic range (outlandishly high in the real world) and pass
it through a converter with just 90 dB dynamic range, the
resulting signal's dynamic range is degraded to only 79.6
dB.

Have any of you out there who've said no difference
recorded
large scale projects at extremely high resolution and
A/B'd
them with lower resolution versions?

Not large scale projects. I did some work with really
stripped-back projects to get a project that started out
with super high dynamic range.

Do you know for certain
that artifacting isn't minimized overall?

Listen for yourself:

http://www.pcabx.com/technical/sample_rates/index.htm

They are billed as sample rate tests, but I also did some
dithering down.

Here's some simpler tests:

http://www.ethanwiner.com/BitsTest.html

and

http://www.pcavtech.com/test_data/