|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
60.227.51.246
In Reply to: Re: the Lynx but... posted by drwkng on January 30, 2007 at 21:25:05:
I find the E-MU 1212M to have a slight edge sonically over the Lynx L22. Technically: slightly better DACs, audio stage isolated from the PCI bus, and native rate clocks rather than synthesized.The breakout cable for the L22 tends to pick up noise - I noticed some harmonic multiples of A/C frequency on a friend's L22.
Follow Ups:
As I pointed out before, this may be due to band width limiting on the 1212. Use a top quality play back system, and the result may be different.
.
There are serious issues in imposing band limiting at 20k on a 96/192k card.First it makes the measurements better and EMU/Creative are good at this game.
Second, why introduce phase issues by having an extra hf filter when it is the design and operation of the digital filter that plays a key part in making hirez digital audio sound good.
Yes, we all know you hate Creative, but why do you have to lie and make up untruths to justify your bias?There is no "extra filter", and at 96kHz the E-MU 1212M has a usable frequency response (+/-3dB) up to 48kHz. I have no idea what you mean by "band limiting".
I suspect you have never owned this card and have never measured it, and all you are doing is spreading misinformation based on prejudice.
Hey Fangirl - see EMU specAnalog Line Outputs (2) Type: Balanced, low-noise, 3-pole low-pass differential filter
D/A converter: CS4398
Level (software selectable):
- Professional: +4dBu nominal, 20dBu max (balanced)
- Consumer: -10dBV nominal, 6dBV max (unbalanced)
Frequency Response (20Hz - 20kHz): + 0.0/-.35dB,
Dynamic Range (1kHz, A-weighted): 120dB
Signal-to-Noise Ratio (A-weighted): 120dB
THD+N (1kHz at -1dBFS): -105dB (.0006%)
Stereo Crosstalk (1kHz at -1dBFS): < -115dB
There is a low pass filter other than the brickwall filter. I read this in a review before I decided not to buy the 1212M.Just because you own one and seem to think that you are the only authority, there is no need to accuse people of lying.
Actually the EMU manual says it is a two-pole filter. Why don't you eat your own words and accuse EMU of lying?
What's so strange about the specs? Doesn't say anything about "extra" filter to me. You do realise that every DAC needs a low pass analog filter post DAC, don't you, and that E-MU is simply describing the characteristics of this filter?*** There is a low pass filter other than the brickwall filter. I read this in a review before I decided not to buy the 1212M. ***
If you are referring to the low pass analog reconstruction filter, then yes every DAC needs one (and it is separate from the digital oversampling brickwall filter). This is also the case for the RME and Lynx and in fact the majority of DAC designs, so you shouldn't buy them either.
The only DACs that don't have two filters are non-oversampling or non-sigma delta modulated designs (which are in a tiny, and disappearing, minority).
*** Just because you own one and seem to think that you are the only authority, there is no need to accuse people of lying. ***
No, I don't think I am the "only" authority, but I do speak from experience. If you truly believe there is some sort of extra "evil" filter in the 1212M (over and above the digital filter and the required analog reconstruction filter), why don't you show us exactly where it is located on the circuit board?
Incidentally, since you are so hot against "extra" filters and "band width limiting", you do realise that this is exactly what upsampling is? Upsampling is a fancy term for an extra, redundant, "band limiting" digital filter.
If you are as knowledgeable as you make it out to be, you will know that a 3-pole active filter is not necessary from 20k. There are manufacturers who use much slower slopes for better sonics. In this sense, the EMU is band limiting. I would summise that this is for spec reasons, to cut out hf artefacts.As for digital filters, there are filters and filters. See dcsltd.co.uk for white papers on how these affect the sound.
.
nt
If you can't, then you are either lying, or don't know what you are talking about, or both.PS - I hope you realise that the number of poles does not indicate the steepness of the filter. You are confusing between poles and orders I think. In any case, measured data shows that the frequency response is +/-3dB to 48kHz for a 96kHz sampling rate I fail to understand exactly where the bandwidth is limited, except at Nyquist.
Why do you think that people lie. Perhaps this tells more about yourself.
Let's see ...1. You claim there is an "extra" filter, but cannot or will not show us where this filter is located.
2. You say the output is "bandwidth limited", yet measured results say otherwise.
Why do I think you are lying? Gee, I don't know ...
Furthermore, you confuse between filter slope and the number of poles in a filter - two unrelated concepts.
Can you give us one reason why we should even pay any attention to anything you say in the future? Otherwise, it's time to move on.
as i said. it is my personal preference.
i'm aware that 1212M has better dacs but i dont' think it is the whole story. i've not made any loop back test on the l22 yet. 1212m still suffers power supplies effect. the effects of 60hz power supply noise is still clearly visible on the loop back test.regarding the breakout cable, i made my own so i don't know whether the one came with the card is any good.
*** the effects of 60hz power supply noise is still clearly visible on the loop back test. ***Not on mine. Not in the loop back test, and not during recording. I can send you some graphs if you want. And I've tested it in several PCs.
Perhaps you have a defective unit?
-----------------------------------that's interesting. i've seen not just on mine but those from the interent as well. it is on the RMAA intermodulation distortion spectrum. anyway if you managed to get rid of it, good. here is the link of one of the results
I don't really see any issues with power supply noise. There is a weird spike around 1800Hz in the 1212M results, but other than that the results are pretty clean.So I'm not sure what you mean when you say you have seen it in other results on the Internet.
I also have an E-MU 1820M, and the results are virtually identical with that unit. So i don't think my results are atypical.
I could show you the other graphs, but they are pretty clean ...
But I run a dedicated 20A circuit for audio, and apply power supply filtering (not just between the equipment and mains, but also between equipment connected on the same circuit).
i didn't go to that extreme but i've good power supplies for a dedicated pc for measurement.
what's the intermodulation distortion spectrum look like? do you see the 50/60 hz?
There's some harmonic distortion, but they are multiples of the fundamental frequencies, not the power supply (no multiples of 50Hz)
what do make of the signal near the 50/60 hz region? lynx is the only card that does not show that signal whatever that is. here is the link for lynx
The normal IMD test signal consists of 70 Hz and 7000Hz, which is what is displayed on my graph and also the results you linked to.Your results are weird because they are not showing these fundamentals.
You must have changed Rightmark options to measure IMD using 11kHz and 12kHz signals. This is not the standard way to measure IMD as per EIAJ.
Incidentally, I noticed that you results show a significantly higher noise floor between 100-500 Hz, indicating that you do have some issues with power supply noise.
Try measuring again after changing the settings in Rightmark.
is because any harmonics of 11 and 12kHz are higher than Nyquist (assuming 44.1kHz sampling rate) hence the test will give artificially good results for any soundcard.
they are not at 44k. they are 96k. in fact lynx performs better at 96k than 44k
and they appear to be significantly better than yours at 96k (3dB lower noise floor, which means the noise level is half that of the lynx, plus lower THD and IMD). The only area where the L22 wins is frequency response (since the 1212M is slightly aggressive in it's low pass filter).This is not surprising, because the 1212M has a number of design advances over the L22 (slightly better DACs, noise insulation from the PCI bus, and better clock). To my ears, this translates to a slightly better sound as well.
The reason the 1212M is cheaper is not because of quality, it's economies of scale. Lynx is a small company, and the L22 uses expensive FPGAs. E-MU on the other hand is using the 10K2, the design cost has been fully amortised through the high volume sales of Creative's consumer cards. E-MU is probably leveraging the production efficiencies of Creative as well.
o.k. i'll try and see.
and it seems that Rightmark has been configured to generate test signals at 19 and 20kHz for IMD.This is a silly way to measure IMD, because it's so close to Nyquist it's unlikely to result in any measurable harmonic distortion.
This post is made possible by the generous support of people like you and our sponsors: