Kir
Max Output Level: -90 dBFS
- Total Posts : 18
- Joined: 2006/01/19 14:15:49
- Status: offline
RE: Wanting 192hz...what's best
2007/08/25 17:34:56
(permalink)
Mike...very cool...thanks.......yeah i have heard the bowels do some amazing things..............and Jinga........I have no followers...and no philosophy........but I was a professional peace worker for the UN in Bosnia.......so yeah....I'm looking for answers any place i can and anyway i can to help out................
Well thanks all......I got what i needed from this.........I think we have kicked this dead horse to mush...wouldn't you say........
I just wanted to hear some opinions before i jump furthur.............Like i didn't know that lots of people had 192 capacity and still choose 96 or 88......that is helpful to know.....I have only been computer recording for a year............and there is lots of experience in this forum.............Thanks again...........
|
John
Forum Host
- Total Posts : 30467
- Joined: 2003/11/06 11:53:17
- Status: offline
RE: Wanting 192hz...what's best
2007/08/25 17:57:40
(permalink)
Well thanks Kir I had no idea I was so useful to you.
Just so you know I am John not Mike.
Best
John
|
Jose7822
Max Output Level: 0 dBFS
- Total Posts : 10031
- Joined: 2005/11/07 18:59:54
- Location: United States
- Status: offline
RE: Wanting 192hz...what's best
2007/08/25 21:30:50
(permalink)
ORIGINAL: John
Well thanks Kir I had no idea I was so useful to you.
Just so you know I am John not Mike.
Best
John 
|
vintagevibe
Max Output Level: -51 dBFS
- Total Posts : 2446
- Joined: 2003/12/15 21:45:06
- Location: Atlanta, Ga
- Status: offline
RE: Wanting 192hz...what's best
2007/08/25 23:16:18
(permalink)
ORIGINAL: Jose7822
Recording at 192KHz is like making music for aliens or something. As you know, the human ear can only go as high as ~20KHz (and that's a young and well preserved ear by the way). When you record at 192KHz you are capturing sounds at up to 96KHz--Do you see what we mean now? This is an old argument that has long been debunked. What you are missing is that the frequencies that are captured above the level of human hearing have a harmonic effect on the frequencies below them. The sound will be different if those frequencies are not captured.
|
Jose7822
Max Output Level: 0 dBFS
- Total Posts : 10031
- Joined: 2005/11/07 18:59:54
- Location: United States
- Status: offline
RE: Wanting 192hz...what's best
2007/08/25 23:23:24
(permalink)
ORIGINAL: vintagevibe
ORIGINAL: Jose7822
Recording at 192KHz is like making music for aliens or something. As you know, the human ear can only go as high as ~20KHz (and that's a young and well preserved ear by the way). When you record at 192KHz you are capturing sounds at up to 96KHz--Do you see what we mean now?
This is an old argument that has long been debunked. What you are missing is that the frequencies that are captured above the level of human hearing have a harmonic effect on the frequencies below them. The sound will be different if those frequencies are not captured. Nyquist Theorem--read on it. That's all there is to it  .
|
Geokauf
Max Output Level: -72 dBFS
- Total Posts : 912
- Joined: 2003/12/01 20:59:45
- Location: Port Chester, NY, USA
- Status: offline
RE: Wanting 192hz...what's best
2007/08/25 23:32:04
(permalink)
Hello,
The S/N ratio of a Studer 24 track was about 50dB. One of the greatest acoustic guitar records of all time ("Beggars Banquet") was recorded on analog equipment with inferior specs compared to the most menial home DAW. Today I was listening to John Lee Hooker (singing "Crawling King Snake"). Just John Lee his 335 and his foot (stomping to keep time). It was mesmerizing! The performance and the material is everything.
GK
|
Kir
Max Output Level: -90 dBFS
- Total Posts : 18
- Joined: 2006/01/19 14:15:49
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 01:34:29
(permalink)
the frequencies that are captured above the level of human hearing have a harmonic effect on the frequencies below them. The sound will be different if those frequencies are not captured. Ah there we go...that is what I was trying to say..... Nyquist Theorem.........it is just what it says it is........a theorem....that's all
|
Jose7822
Max Output Level: 0 dBFS
- Total Posts : 10031
- Joined: 2005/11/07 18:59:54
- Location: United States
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 01:48:11
(permalink)
Ok, this is from the MAN himself. If this doesn't do it for ya then I don't know what will: Hello Bob! I spent all weekend reading your articles, they're great! But I'm an amateur engineer and lots of questions came to my mind.
You favor the new 24/96khz a lot and you mentioned on the "Liftoff!" article that you're recording at 24/96khz. Now, the mics you used aren't capable recording more than 20hkz. So, can you explain your position on that?
Sincerely
Sun
Hello, Sun. Thanks for your comments.
Actually, the B&K Mikes are quite linear beyond 20K, but that's not what's important to your question. I have written an article that appeared in Audiomedia magazine which explained an experiment I performed... Basically, the superiority of 96 kHz sampling is in all probability not due to its extended bandwidth, but rather to all the other improvements which can be measured in the 20-20 kHz band, such as improved linearity of frequency response with no ripple in the passband, less phase shift at the frequency extremes, and other irregularities which can be caused by inferior 20 kHz filters required for use at lower sampling rates.
Hope this helps,
Bob Katz This can also be found at: http://www.digido.com/index.php?option=com_kb&Itemid=43&page=articles&articleid=114
post edited by Jose7822 - 2007/08/26 01:57:07
|
Kir
Max Output Level: -90 dBFS
- Total Posts : 18
- Joined: 2006/01/19 14:15:49
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 02:19:34
(permalink)
such as improved linearity of frequency response with no ripple in the passband, less phase shift at the frequency extremes, and other irregularities which can be caused by inferior 20 kHz filters required for use at lower sampling rates.
Ok I'm listneing....but then what is he using to get all those improvements? Do all 96k capable rigs have these better filters that he is talking about?
|
Jose7822
Max Output Level: 0 dBFS
- Total Posts : 10031
- Joined: 2005/11/07 18:59:54
- Location: United States
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 03:37:49
(permalink)
Original: Kir
Ok I'm listneing....but then what is he using to get all those improvements? Do all 96k capable rigs have these better filters that he is talking about? Well, I couldn't answer that question because I haven't scientifically tested all the converters in the world. But, as he (Bob) has explained in that quote, inferior filters (from inferior converters) cause irregularities at lower sampling rates that negatively impact the audible range. For example, recording at 44.1 KHz would cause the filters to cutoff at 22.05 KHz, a.k.a. Nyquist Frequency. That means that on bad converters these irregularities will be placed near the audible range. So, to avoid these irregularities, a solution would be to raise the sampling rate which will also raise the cutoff frequency of such filters at around 48 KHz. Remember though, that this is not the case with good converters as they posses better quality filters. But that's not all, there are also issues with jitter, aliasing and quatization errors as well. A brief explanation of each would be: - Jitter: Is caused by timing errors in the AD process. This has to do with the converter's clock not being able to accurately take evenly spaced snapshots of the waveform. - Aliasing: Is the distortion caused by the overlap of frequencies during sampling. - Quantization Errors: Has more to do with bit rate and takes place during the A/D process while the signal is being converted to digits. If the resolution is not high enough it introduces noise in the audio. Anyways, it would take a long time to explain these things. Just go to the Bob Katz website if you wanna learn more about how converters work at www.digido.com HTH. P.S. Please feel free to correct me if I have not given accurate information. Thanks!
|
Jose7822
Max Output Level: 0 dBFS
- Total Posts : 10031
- Joined: 2005/11/07 18:59:54
- Location: United States
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 11:39:38
(permalink)
Hey guys! I have found an article from 2002 that compares DSD (Direct Stream Digital) to PCM (Pulse Code Modulation). In this article Craig Anderson from Craigman Digital (not to be confused with Craig Anderton - the guitar player) have made some experiments using a 10 KHz square-wave generator as the basis of his test. The results were pretty interesting (even though they contradict what I've been saying here). If you go to the next page (Page 10), you'll read about the audible differences of various sampling rates and bit depths. Here's the article: http://www.smr-home-theatre.org/surround2002/technology/page_07.shtml So what do you guys think (read UnderTow, Bitflipper, you know who you are )?
|
DaveClark
Max Output Level: -71 dBFS
- Total Posts : 956
- Joined: 2006/10/21 17:02:58
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 11:40:48
(permalink)
Hi All, As Jose7822 quoted, pushing the Nyquist frequency out to 48 KHz has advantages. The fundamental problem is that 22.05 KHz is an *absolute* cutoff. This means that rolloff should start much earlier, say at 11.025 KHz. This latter frequency is below the highest frequencies we can hear, as we all agree. Too many folks think that if we can hear at best only to about 20 KHz, then 22.05 KHz is plenty. It's *not* because as I said, this is an absolute cutoff. This is too much of a restriction on filter design. Being able to go all the way out to 48 KHz as an abolute cutoff is much better for filter design if you want fidelity out to 20 KHz. It really is that "simple." On Edit: I should also have re-iterated my previous comment that in addition, higher sample rates are also advantageous for numerical processing, which I tend to utilize a lot --- plugins, offline processing such as computational acoustics, and so on. Regards, Dave Clark
post edited by DaveClark - 2007/08/26 11:51:52
|
Kir
Max Output Level: -90 dBFS
- Total Posts : 18
- Joined: 2006/01/19 14:15:49
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 12:07:32
(permalink)
Thanks Jose........I'll go into both sites you gave and look around........from an alien perspective of course 
|
bitflipper
01100010 01101001 01110100 01100110 01101100 01101
- Total Posts : 26036
- Joined: 2006/09/17 11:23:23
- Location: Everett, WA USA
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 13:58:26
(permalink)
Many of the widely-held misconceptions regarding A/D converters stem from converter concepts that are decades out of date. I first learned about converters in the 70's when I attended electronics school, and carried around an inaccurate mental model for a long time based on what I knew about converters of that era. Consequently, for many years I, too, held many intuitively sensible but inaccurate ideas about converters. It wasn't until I revisited the subject in earnest about a year ago that I realized that many of the old concepts no longer apply.
Nyquist's elegant theorem only holds true in a band-limited system. Limiting bandwidth is absolutely essential to the process; not doing so - or doing it poorly - results in aliasing, the introduction of new frequencies that were not in the original source material at all. And it sounds really, really bad. Avoiding it requires severely limiting everything over 20KHz.
In early converters, this was a big problem. You've only got the space between 20KHz and 22.05KHz for your rolloff, requiring a very steep filter. However, such steep filters introduce their own kind of distortion. The alternative was to begin the rolloff at a lower frequency, allowing a gentler curve but sacrificing some high frequency content in the process.
Fortunately, a third option became available as the cost of CPUs, RAM and disk space came down: raising the sample rate. Raising the sample rate raises the Nyquist frequency to a point well above the hearing range, allowing a gentle filter that did not begin its rolloff within the range audible frequencies. The result was a higher-fidelity recording, at the cost of higher CPU overhead and much greater disk storage requirements.
That was the state of affairs 10-15 years ago. Since then, oversampling has become the norm. With oversampling, the converter actually samples at a much higher rate, allowing the filter to be designed with a cutoff frequency way, way above the audible range. Now you have an intermediate digital image at a ridiculously high SR, which is then downsampled to the desired "real" sample rate. The oversampled signal is already digital at this point, allowing the final filter to be a digital filter, which can be as steep as we want without the phase issues inherent in analog filters.
The consequence of all this is that whether we ultimately end up with 44.1, 48, 88.2, 96 or 192 is just a matter of how many oversamples will be discarded by the converter. The historical benefits of filtering at higher Nyquist frequencies no longer apply.
Once you've eliminated the filter problems, and can band-limit to 20KHz without compromising fidelity, Harold Nyquist's math works exactly as advertised. This is the basis of my claim that if you can hear a difference between 44.1KHz and higher rates, it's due to inadequacies of your converter, not in any inherent advantage of higher sample rates.
Regarding that test using 10KHz square waves, keep in mind that you are NOT going to ever record a square 10KHz wave at any sample rate, nor even analog. Nor would you ever need to. Remember that the squareness of the wave is an indication of the presence of odd harmonics. As you filter away the harmonics, the waveform loses is squareness and ultimately becomes a sine wave if you filter out all harmonics.
And here's the thing: you cannot hear any difference between a 10KHz square wave and a 10KHz sine wave! Really. This is not about golden ears. Try it with a laboratory signal generator (your synth won't do), a simple RC filter, an oscilloscope and some high-quality monitors. I've done exactly that, and trust me, the ability to record a 10KHz square wave is a total nonissue.
 All else is in doubt, so this is the truth I cling to. My Stuff
|
syrath
Max Output Level: -34.5 dBFS
- Total Posts : 4075
- Joined: 2005/08/11 05:40:08
- Location: Ayrshire, Scotland
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 14:19:42
(permalink)
ORIGINAL: Jose7822
Kir,
Most of the people here have the gear to record at 192KHz and choose not to. That's what they've been trying to tell you all along. Recording at 192KHz is like making music for aliens or something. As you know, the human ear can only go as high as ~20KHz (and that's a young and well preserved ear by the way). When you record at 192KHz you are capturing sounds at up to 96KHz--Do you see what we mean now? On top of this, you'll end up burning your songs to CD which has a 44.1KHz/16bit quality or worse to MP3. It's like recording in HD quality to end up in Youtube quality. Now does it make sense? If you're doing acoustic material then 96KHz is way more than you or anyone for that matter needs. Take care!
ORIGINAL: CJaysMusic
Thats was put beautifuly and its very true..
Cj
For what its worth HD quality is still a compressed picture. Its equivalent in music is an MP3. A better quality than what you may be used to be seeing, but its still an mp3. Uncompressed video is measured in gigabytes per minute.
|
UnderTow
Max Output Level: -37 dBFS
- Total Posts : 3848
- Joined: 2004/01/06 12:13:49
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 14:31:13
(permalink)
ORIGINAL: bitflipper
And here's the thing: you cannot hear any difference between a 10KHz square wave and a 10KHz sine wave!
This pretty much sums it up. All this stuff is just marketing departments doing their best: Formatting information to give a misleading view to try and con unsuspecting consumers. DSD was marketed because the patent on CDs ran out. No other reason. It is in fact inferior to good old PCM in many ways. Don't be fooled by the pretty folders. Btw, that article has the very annoying and misleading habit of calling higher sample rates "high resolution". Increasing sample rates does not increase resolution in any way or form whatsoever. It really is very simple: Good converters will sound great at 44.1Khz or 48Khz. Lesser converters might sound better at 88.1Khz or 96Khz than the lower sampling rates. 192Khz and anything above is total and utter bollocks. UnderTow
|
The Maillard Reaction
Max Output Level: 0 dBFS
- Total Posts : 31918
- Joined: 2004/07/09 20:02:20
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 14:33:01
(permalink)
My favorite thing to do with video editors who tell me they edit HD uncompressed is to ask them what codec they use... tricks em everytime :-)
post edited by mike_mccue - 2007/08/26 14:41:40
|
Boogie
Max Output Level: -54 dBFS
- Total Posts : 2106
- Joined: 2003/11/19 15:45:21
- Location: CALIFORNIA
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 14:36:55
(permalink)
ORIGINAL: Clydewinder
you need a pretty impressive amp & monitor system to hear if there even is a difference at all.
Not to mention golden ears. I even view the opinions of those who supposedly have golden ears with skepticism. Is the difference really that major, or are they just trying to sell you expensive gear? If you can't hear a difference in a blind taste test, don't waste your money or computing resources.
|
The Maillard Reaction
Max Output Level: 0 dBFS
- Total Posts : 31918
- Joined: 2004/07/09 20:02:20
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 14:42:07
(permalink)
How about if a clinically deaf person told you they heard something beyond YOUR range of hearing? Would you quote your physics textbook... or just call the person a fool??? Or maybe just think what if?
best,
mike
|
wicked
Max Output Level: -90 dBFS
- Total Posts : 5
- Joined: 2007/07/17 10:14:11
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 16:58:53
(permalink)
The nyquist theoremis only an absolute or the low end. Meaning you MUST sample at thatr frequency to accurately reproduce the waveform. But if you want more content of the signal, you must sample at greater than that.
If you have a sin wave at 1 hz and you sample at 2 hz anything in between the two sample points will be averaged to the two points, like conect the dots so what if there are small spikes in the wave between the points, they will not be there. What is the highest absolute sample freq to use , Don't know but the lowest is 2fc
Thanks, not a DSP eng but did have some course work
|
DaveClark
Max Output Level: -71 dBFS
- Total Posts : 956
- Joined: 2006/10/21 17:02:58
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 17:15:35
(permalink)
Greetings All,
Regarding Dave's (bitflipper's) recent post about converters:
I agree with what Dave posted. However, just as it is easy to be misled about the current status of converters, it's easy to get misled into thinking that what applies to converters applies to the entire audio chain. Not true! Many folks use plugins, parametric filters in a mixer, and other audio processing without oversampling.
One way to look at oversampling is that higher rates are used temporarily and automatically where it matters so you don't have to. If you use processes that do not oversample, and most people do, then you should achieve higher fidelity with 88.2 or 96 KSamples/sec than with the lower rates. Now it is true that not all software takes advantage of the higher rates, leaving cutoffs at really low frequencies, so you may experimentally find no difference such as no difference between 10 KHz square and sin waves. But ideally you can achieve higher fidelity. (This is one really good reason to write your own software!!)
This problem with "not all software takes advantage..." is the reason Bob Katz is advocating and has been advocating an industry standard of 96 KHz. If *everyone* did that --- get everybody on the same page --- then all hardware and software could be designed with this in mind. The "RC filter" that Dave mentioned could be redesigned with a higher cutoff. Then you would hear a difference.
Again, I'm just saying, be careful also about applying ideas from converters to the entire audio chain. You'll be drawing some very wrong conclusions about what should be done in general about sampling rates.
Regards,
Dave Clark
|
Jose7822
Max Output Level: 0 dBFS
- Total Posts : 10031
- Joined: 2005/11/07 18:59:54
- Location: United States
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 17:31:05
(permalink)
Bitflipper and UnderTow, Thanks so much for this info. It really messes with your head to see stuff like this floating around the net. Now I have an even better grasp on the subject and can enjoy a normal life again  . Take care!
|
Jose7822
Max Output Level: 0 dBFS
- Total Posts : 10031
- Joined: 2005/11/07 18:59:54
- Location: United States
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 17:46:49
(permalink)
ORIGINAL: syrath
ORIGINAL: Jose7822
Kir,
Most of the people here have the gear to record at 192KHz and choose not to. That's what they've been trying to tell you all along. Recording at 192KHz is like making music for aliens or something. As you know, the human ear can only go as high as ~20KHz (and that's a young and well preserved ear by the way). When you record at 192KHz you are capturing sounds at up to 96KHz--Do you see what we mean now? On top of this, you'll end up burning your songs to CD which has a 44.1KHz/16bit quality or worse to MP3. It's like recording in HD quality to end up in Youtube quality. Now does it make sense? If you're doing acoustic material then 96KHz is way more than you or anyone for that matter needs. Take care!
ORIGINAL: CJaysMusic
Thats was put beautifuly and its very true..
Cj
For what its worth HD quality is still a compressed picture. Its equivalent in music is an MP3. A better quality than what you may be used to be seeing, but its still an mp3. Uncompressed video is measured in gigabytes per minute.
I think HD quality would be more akin to CD quality and that SD or lower would be like MP3. I say this because we ourselves record sounds at higher rates and depths only to deliver the final product to a lesser quality like CD.
|
The Maillard Reaction
Max Output Level: 0 dBFS
- Total Posts : 31918
- Joined: 2004/07/09 20:02:20
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 17:55:22
(permalink)
Bitflipper, "Since then, oversampling has become the norm." Is it not also true that oversampling has recently become unfashionable? I have a friend who makes a hobby/side business replacing DACs in CD players. The hot trend is to eschew oversampling and embrace the latest much more accurate clocks. He replaces the DAC and clocks with new parts. Question: Was Nyquist's Theory exclusive (and in anticipation of) to digital (quantized) audio or did he also apply it to bandwidth issues in the analog domain... things like analog sound recordings and or radio transmission? Now here's something to consider. I found the following info last month when I was purchasing 2 Portico 5012 units from Rupert Neve Designs. Anyone care to call Mr. Neve crazy? Maybe someone would like to accuse Mr. Emerick of being a liar?: "Which brings us to the next interesting design point: The 5012 has a frequency range of roughly 10Hz to 160kHz. Yes, that’s right, forget the Nyquist theorem, this is at least eight times the ceiling of average human hearing! I had read an interview with Neve some time ago and tracked it down so I could re-tell the story (or at least give you the gist of it). OK, here goes: The year is 1977, the location is George Martin’s AIR studios, the concerned engineer is Geoff Emerick, the console is a recently delivered Neve and the problem, which only Emerick’s clued in to, is anyone’s guess. Ultimately, testing revealed a 3dB peak at 54kHz on three channels due to some mis-wired transformers.
The point of this story is that while humans may not hear much above 20kHz, there is some sort of perception at work. Anyone familiar with Pauline Oliveros’ work with tones will know that in addition to this perception, one can also hear the interaction of frequencies above the range of human hearing if combined properly. What I’m getting at is this though: The 5012s stated frequency range not only looks good on paper, but makes for good science and even better listening. RND set out to make an affordable, compact, transparent preamp with a minimum of crossover distortion. They have achieved this and our ears will thank them for it. And if anyone cares to check out the interview with Mr. Neve in which he tells the Geoff Emerick story it can be found in greater detail right around here: prosoundweb.com/chat psw/transcripts/ rupert.php." edit: the actual link is: http://www.prosoundweb.com/chat_psw/transcripts/rupert.shtml
post edited by mike_mccue - 2007/08/26 18:10:22
|
Jose7822
Max Output Level: 0 dBFS
- Total Posts : 10031
- Joined: 2005/11/07 18:59:54
- Location: United States
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 18:04:43
(permalink)
Hey Mike!
I have to apologize to others first because what I'm about to ask is not related to the subject at hand but I can't resist. How are the 5012's sounding? I'm thinking of getting one myself hopefully sometime in the near future. Thanks!
|
The Maillard Reaction
Max Output Level: 0 dBFS
- Total Posts : 31918
- Joined: 2004/07/09 20:02:20
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 18:11:05
(permalink)
Jose,
A 720P Panasonic Varicam HD records at 100Mbs. A 1080P Sony Cine Alta HD records at 150Mbs.
Please consider that these are highly, but effectively, compressed compared to true full frame 1920x1080P image sequencing.
miniDV, DVCAM, DVCPro25, and HDV are 25Mbs.
BetaSP is usually digitized at 50Mbs but the quality is dependant on the codec and conversion hardware.
Distribution of "HD" via cable or satellite is apprx 12Mbs. FWIW most of it is still part of the mpeg2 definition.
You can use these numbers visualize your own metaphor between video and music production.
best regards,
mike
|
The Maillard Reaction
Max Output Level: 0 dBFS
- Total Posts : 31918
- Joined: 2004/07/09 20:02:20
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 18:13:45
(permalink)
Jose,
I very much like the 5012's. They are a huge upgrade for me. I don't have any other pres of that quality to compare but I would recomend them without reservation.
I intend to order a API 3124 very soon. It's time for me to collect preamps.
best,
mike
|
bitflipper
01100010 01101001 01110100 01100110 01101100 01101
- Total Posts : 26036
- Joined: 2006/09/17 11:23:23
- Location: Everett, WA USA
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 18:14:58
(permalink)
The nyquist theoremis only an absolute or the low end. Meaning you MUST sample at thatr frequency to accurately reproduce the waveform. But if you want more content of the signal, you must sample at greater than that.
If you have a sin wave at 1 hz and you sample at 2 hz anything in between the two sample points will be averaged to the two points, like conect the dots so what if there are small spikes in the wave between the points, they will not be there. What is the highest absolute sample freq to use , Don't know but the lowest is 2fc
Thanks, not a DSP eng but did have some course work
This is an excellent example of what I was referring to in my previous post. Yours is a perfectly reasonable, logical assumption that makes sense intuitively -- but unfortunately, it's also quite incorrect. Don't feel bad, I shared the same opinion for many years. The Nyquist Theorem does not just apply to the low end, unless by that we mean the audible frequencies below 20KHz. Believe it or not, you actually CAN reconstruct an EXACT replica of the original waveform with as few as two samples -- the only prerequisite is that those two samples must describe an audible frequency. Forget the "connect the dots" image, that's what's tripping you up. I know this is a difficult concept to wrap your head around. At least, it was for me. And I had the advantage of decades of experience in audio electronics and analog synthesis to draw upon and still it was not intuitively obvious to me. The key to understanding it, for me, anyway, was to remember that any waveform other than a sine wave is comprised of multiple overtones (most of them harmonically related to the fundamental, but not necessarily harmonics per se). That means you can't talk about a 1KHz square wave, or a 10KHz square wave, or a 1Hz square wave without treating it as an amalgam of frequencies, a compound sound if you will. Take away the harmonics and it is no longer a square wave, or, conversely, take away the squareness and the harmonics are gone. Now, the reason a square wave sounds like it does (and makes it distinguishable from a triangle or sawtooth) is its particular constituent harmonic content. It will never be perfectly square, because that would require the algebraic sum of an infinite number of harmonics. Consequently, truly square waveforms only exist on the printed pages of textbooks and engineering timing specs. They do not exist in real life. But that doesn't prevent us from recognizing the sound of a square wave when we hear it, as long as enough of those harmonics are in there. So how square does a square wave need to be? Square enough so that no AUDIBLE components are lost. For a 1KHz square wave, that would be every odd harmonic between 3KHz and 19KHz. Beyond that we don't care, inaudible components are irrelevant. A 1KHz wave will be represented not by 2 samples, but by 44 samples, just enough to accurately represent a 1KHz "square" wave. Now consider a 10KHz square wave. Guess what? NONE of its harmonics fall into the audible range. In the world of bandwidth-constrained audio, there is no such thing as a 10KHz square wave. This is the basis of my earlier statement that you can't hear the difference between a 10KHz sine wave and a 10KHz square wave. If you have a sin wave at 1 hz and you sample at 2 hz anything in between the two sample points will be averaged to the two points, like conect the dots so what if there are small spikes in the wave between the points, they will not be there. Ah, but if there were any "small spikes" in the wave between the points, it would not be a sine wave. If the sampling frequency were 44.1KHz, the sine wave in question would have to be 22.05KHz and any nonsinusoidal components ("spikes", if you will) would constitute inaudible harmonics that by definition are disallowed within the band-limited spectrum described by Nyquist. In fact, any blip, bump, dip or wrinkle in the waveform that's too brief to be represented by at least two samples represents an inaudible component (above 20KHz) and is therefore irrelevant.
post edited by bitflipper - 2007/08/26 18:29:39
All else is in doubt, so this is the truth I cling to. My Stuff
|
Jose7822
Max Output Level: 0 dBFS
- Total Posts : 10031
- Joined: 2005/11/07 18:59:54
- Location: United States
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 18:17:04
(permalink)
ORIGINAL: mike_mccue
Jose,
A 720P Panasonic Varicam HD records at 100Mbs. A 1080P Sony Cine Alta HD records at 150Mbs.
Please consider that these are highly, but effectively, compressed compared to true full frame 1920x1080P image sequencing.
miniDV, DVCAM, DVCPro25, and HDV are 25Mbs.
BetaSP is usually digitized at 50Mbs but the quality is dependant on the codec and conversion hardware.
Distribution of "HD" via cable or satellite is apprx 12Mbs. FWIW most of it is still part of the mpeg2 definition.
You can use these numbers visualize your own metaphor between video and music production.
best regards,
mike Well, now you're getting too picky . Thanks for commenting on the 5012's. I only hear good things about it so it makes it an easy buy. Take care!
post edited by Jose7822 - 2007/08/26 18:25:52
|
Jose7822
Max Output Level: 0 dBFS
- Total Posts : 10031
- Joined: 2005/11/07 18:59:54
- Location: United States
- Status: offline
RE: Wanting 192hz...what's best
2007/08/26 18:43:02
(permalink)
Original: bitflipper
Now consider a 10KHz square wave. Guess what? NONE of its harmonics fall into the audible range. In the world of bandwidth-constrained audio, there is no such thing as a 10KHz square wave. This is the basis of my earlier statement that you can't hear the difference between a 10KHz sine wave and a 10KHz square wave. OK, I can understand that because the first harmonic of a 10 KHz square wave will be located at 20 KHz (its octave) and rest of it's harmonics will be at higher frequencies than that placing them in the inaudible range but, is it really be true that they do not matter? What I mean is that, are we not able to sense them somehow even if we cannot hear them? I think this is what's kept the argument about higher sampling rates alive.
|