|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
173.66.105.23
Hey,
Does anyone know any specifics on the differences between the RIAA and NAB playback curves?
I got that Denon last week and earlier this week my Korg MR-1 came in. The MR-1 is a portable hard-disk based "field recorder" that does up to 24/192 in PCM and 2.8Mhz 1-bit DSD. In my years of doing (or trying) to do vinyl transfers to digital I realized the PC was one of the weaker points and rather than spend hundreds of dollars on a high-quality ADC and still have this noisy PC in the mix; a portable recorder that has a stable clock would probably simplify it. Having heard transfers someone else did with an MR-1 before I got ahold of one...and just doing my own testing with various analog sources...it's a fantastic device.
But that's not why I'm here. I'm here to ask about curves. See, the new version of the audio software you get for owning one of these devices has a built in phono EQ, in fact it's an actual DSD equalizer. Apparently they make a stand-alone ADC this is used in conjunction with...it has a flat no-EQ preamp stage and uses the software to do the equalization. Of course, even though I don't own that exact device, it will still happily apply deemphasis on my imported DSD recordings. The MR-1's microphone inputs handle the phono cart with zero issues. So I just hook the cart directly in to the unit and record flat.
This is where it gets interesting.
I did a test transfer of a Disco Tex LP (because it's a horrible album that I enjoy from time to time). This thing should be RIAA up and down all day long. But the NAB curve sounds better, in fact it sounds closer to the horrible quality CD Collectables put out in 1993.
Are the NAB and RIAA just close enough they could be interchangeable? Am I actually picking up differences in how the original recording was mastered?
I'm very tempted to just export this using NAB playback curve...except some wood-glue got missed in the outer groove and it makes horrible noises. So hopefully the second attempt at glue cleaning will get it out...or i'll wait for the sealed copy to come off eBay when I have my RCM.
Follow Ups:
Here is a link to an old page from the original Cartridge Database...see below. The same info has been ported over on the VE.
We don't shush around here!
Life is analog...digital is just samples thereof
Thanks Ed ! That really is a very nice chart. While there are many playback EQ charts available online, very few charts include db amounts at certain frequencies, plus the actual time constants related to the turnover and rolloff asymptotes. Also of interest, and rarely included is what actually happens below the low bass resonance frequency. Still, I have found no single chart/catalog which includes all the available EQ preemphasis and/or deemphasis curves, plus db/asymptotic frequencies (+ or - 3 db points) and time constants on any single site. Yes, that is a lot of info to ask for on only a few pages.
So, this chart from the original cartridgedatabase/VE does not cite the sources. I would love to know where the references for the db chart were derived. There are some small numerical errors, according to the "initially published to the trade" periodical, Audio Engineering.
The "AES" EQ curve and info, very important for early discs, appears in Audio Engnrg, V35 #1, Jan. 1951, pg. 24. The New Orthophonic EQ curve and info, standardized as the RIAA EQ in the early mid '50s, first appears in Audio Engnrg V37 #7, July 1953, even though it became the RCA 33 1/3 EQ curve in 1952.
Both of these curves' articles boasted that + or - 2 db from their published charts was "acceptable." Considering that the AES curve has about 1.5 to 2 db difference throughout the curve, compared with the RIAA curve, their acceptable deviation of 2 db is quite liberal and a bit too tolerant. For anyone concerned, when you hear a simple demo of an RIAA EQ recorded disc played back using the AES EQ, or vice versa, you will definitely hear that 1 to 2 db difference over the published curves. Before the RIAA EQ became the actual standard in '54 or '55, even though there were still record companies which continued to digress into the later '50s, most early stateside LPs of 10 inch and 12 inch size could be thoroughly enjoyed using either the "AES" or NAB/Columbia "LP" (later known as NARTB) EQ playback curves.
In fact, many of the earlier 78s benefit from using the AES playback EQ curve, as well. The Audio Engineering Society made a sincere attempt in 1950-1951 to integrate and average all of the then utilized recording and playback EQs, in order to standardize the industry, recommending to all record companies that their products should be produced and mastered to sound good when we consumers would use the AES playback EQ.
Researching preamplifier phono EQ evolution, it was a surprise to me that the earliest "cult" (later known as audiophile oriented) audio preamp equipment by Brook and Marantz actually considered or predicted the AES EQ would become the standard. The New Orthophonic EQ position on the original Marantz (1)1953 Consolette actually has a turnover of 400 Hz; not the "later" standard 500 Hz. While that 100 Hz difference does not seem like much, with the full frequency response curve being 1 to 2 db different than the real New Orthophonic curve, plus the Marantz 1 initial New Orthophonic High Freq Rolloff adhering just as close to the NAB Rolloff, it might now be clear how unstandard the standards really were.
Compounding this dilemna with the engineers' and producers' "house sound" plus the wide tolerance of frequency response deviation available, real standardization slowly evolved. If we wish to hear these pre-1955 discs under the best circumstances, having numerous playback EQ curves available for quick switching becomes desirable.
Please note that the record equalization chart Ed furnished includes some European standard curves. While "needed" stateside curves for early discs could number around five, there were almost that many European disc playback curves needed for the European discs of yesteryears.
Finally, here is a link to many audio related periodicals of yesteryear. Be patient when "streaming" and "scrolling" between title choices. When you have read more than a few issues, it might be a good idea to "clear history" often, as the streaming can hold up your PC. You should be able to link to the cited record EQ curves above, plus some great reading awaits...ENJOY !
They are different - hence Korg provided a separate option. The network for NAB equalisation only has two corner frequencies at 50Hz (same as RIAA) and 3150Hz. Since it shares the same low frequency corner with RIAA, but has a higher upper frequency corner the attenuation is higher above 1kHz so you will be getting a reduced glare or brightness that may have been in the original mastering.
As far as your preference goes, you can choose whatever you want!
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
Well, I knew they were different; I just wasn't sure *how* much different. IT mostly sounds like the NAB will roll off the HF a bit, which as someone mentioned in a different reply...might not actually be a horrible idea.
That depends on which NAB curve you are referring with. The Columbia "LP" curve is often referred as the NAB (circa 1951) EQ curve.
In one of your posts you seemed uncertain about preemphasis. To simplify, in cutting discs, the low freqs. are cut and highs are boosted (pre-emphasized). In playback, bass is boosted and highs are cut.
Since the recording curves are the inverse of our playback curves, keep in mind that the actual "curves" reflect a 6 db/octave slope above or below the transition frequencies. This is very important because bass and treble tone controls, as well as graphic and parametric equalizers are usually not setup for this 6db/octave slope. Fiddling with equalizer controls is not an instant tweak, whereas switching in proper EQ curves can be quick and easy; if we have them on our preamplifiers.
DewDude, you can have the best of both analog and digital playback methodologies, if you can build your own. If you can switch in or out an EQ curve, like the feedback RIAA EQ you mentioned, you can transcribe digitally as needed. More online research will reveal there were/are programs which include a phono EQ preamp, or simply use the high gain (mic) stage and digitally EQ and "process" the transcription.
Unfortunately, the degree of digital processing, scratch repairing and dynamic range enhancing is not usually publicized or "known" enough about, requiring practice and accepting the results when transcribing. While digitizing can be good, even outstandingly good, often the results do not sound as pleasing as we hoped to achieve. This fact makes full analog not only simpler, but even more "natural" sounding.
Keep On Groovin' ........................
First, let me say that this is a superb topic. Since our hobby really revolves (punny) around spinning discs, what we each hear can supersede what the engineers/producers had in mind. They, of course, each had different ears like we each have. So, if your RIAA disc playback curve does not sound as nice as the NAB (or is it the Columbia/NAB "LP" EQ ?), then it is your enjoyment that is most important.
The question really is "why do they sound different ?" Before this can become a digital vs. analog discussion, since the OP's initial transfer is done all digitally, that processing cannot be "predicted" as simply as spinning an early 1950s mono LP through a very good analog system which happens to have the various phono EQ curve settings available for quick switching during playback.
So, all of the available phono EQ curves relate to time constants and frequencies of turnover, boost and rolloff. Please let me chart a few, in an effort to simplify. Then, I will try to enlighten more via further discussion.
RIAA: 3180uS(50 Hz), 318uS(500 Hz), 75uS(2122 Hz); -13.7db at 10KHz
Col.LP:1592 uS(100 Hz), 318 uS(500 Hz), 100 uS(1592 Hz); -16db at 10KHz
AES: Optional Bass, 398uS(400 Hz), 63.5uS(2500 Hz); -12db at 10KHz
pre-1949 NAB: 3180-2653uS(50-60 Hz) 318uS(500 Hz), 100uS(1592 Hz); -16 db
It might be very apparent that RIAA "encoded" discs might sound a bit thinner in the bass and slightly duller in the highs when "decoded" via LP type EQ playback. Conversely, an NAB LP encoded disc could sound a bit boomier in the bass and a bit brighter in the highs when decoded via RIAA EQ playback. The reason for my "might" and "could" revolves around the original disc recording, mastering and production techniques, a la "house sound" artistic conceptions and musical creations.
Concerning the actual midrange Turnover, which is the common 500 Hz (3180 uS) for NAB, Columbia LP and RIAA EQ curves, any more or less prominent mids heard "should" be minimal, since the EQ slope starting from +3 db at 500 Hz slants upward toward the bass peak and/or resonance. Checking the chart, there are definitely differences in that bass resonance region. That "could" be an answer for increased or decreased perceived "definition" of notes, instrument "action" or musical passages.
Post WWII, the NAB curve went through some changes. Further complication of the EQ dilemna exists as the NAB became known as the NARTB with later NAB curves meaning Tape Head equalization. Any of the curves mentioned here have three time constants involved during recording/cutting of the disc:
1) Bass time constant, indicating bass peak or resonance or bass cutoff
2) Turnover, indicating where the mids begin their upward slope toward the bass.
3) Rolloff, indicating where the mids/highs begin their downward slope, rolling off the highs. The standard "0" level is 1000 Hz. The usual specified distinction is the -db level at 10KHz.
The NAB/NARTB Tape Head EQ turnover is near or above 1KHz, depending on era. That simply means the upward slope toward the bass peak began higher in frequency than the usual LP/45/78 disc playback EQ curves. Regarding playback of early discs, often times the usage of a higher than expected Turnover results in "better" sound, since the mids become more prominent, sometimes creating a more natural sound from the very early incarnations.
My favorite example of this is one of my acoustically recorded 1917 Columbia 78s. Imagine all of the musicians crowded around a large "phonograph" style horn used as the microphone. As simple as that seems, those were difficult recording sessions, to be sure. Certainly, today, their best sounding playback is indeed difficult. Quickly switchable phono EQ curves helps the "spinning early discs" sessions. Graphic and Parametric Equalizers can be helpful; but; they take time to implement correctly. Typically, that prewar 78 used a 300 Hz or 500 Hz Turnover. Blue label, early lighter blue label Columbias are said to use a 300 Hz turnover. However, even using the standard RIAA/LP 500 Hz turnover, those early discs often sound dull and almost lifeless. Lifting the high frequency rolloff only adds hiss and noise. But, if you happen to have a 629 Hz, or better a 700-800 Hz Turnover curve available, wow, that will wake up the musicality of these early discs.
I guess, there will and should be follow-up discussions concerning disc playback and digital transferring of analog. I highly recommend researching online. The audacity team and midimagic sites have compiled an amazing amount of EQ info. Today, the audacity team is the forerunner of digital transferring. Audacity also use "flat analog" recording(encoding) and digital EQ/recreation (decoding), but their EQ charting sites are a great source of info for us spinning disc devotees...
One thing that always bothered is is that while the curves we're supposed to use are standardized...how much deviation is there due to all the elements involved. I don't know much about cutting an LP..I don't know where de-emphasis is applied. But I always felt, especially with older LPs...this was all done on analog equipment that will have some drift, and maybe even the way the LP was cut could vary this....and that's before I took in to consideration variances in curves just from out-of-spec/loose values.
I just never got a setup serious enough to want to investigate flat recording...mostly because back then I didn't possess the skills to look at a schematic and make one out of a pile of parts. The MR1 stock seems to do it ok....I haven't actually checked the mic inputs to see if they're loading the cartridge properly; I may get some improvement by just building a preamp and taking out the RIAA feedback stage. I may just leave it alone because the test recording I made came out sounding relatively distortion free and the NAB curve gives it more punch. I do have a CD transfer to compare to...but the CD is done so horribly I don't know if I trust it.
Thanks for the information! It is an interesting topic because it starts to separate Technically Correct from Subjectively Preferable.
The iPhono phono stage has 2 alternative curves - Decca and Columbia. The instructions suggest the use of the Columbia curve even for RIAA discs through to the 80s. Technically incorrect. However the suggestion is made to ameliorate the brightness that many Pop discs from that era suffer from so he is advocating this curve as a form of Tone Control!
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
Anthony, I always enjoy reading your posts. Your technical knowledge, especially of phono carts is exceptional ! We share the joy of Pickering/Stanton. On a certain British site, your technical experience is notable and always appreciated.
Tone controls cannot be avoided. I believe our choices of equipment are actually analogous to a tone control. While we strive for linear, flat frequency response from our phono carts, preamplification and amplification, do any of us really have flat response from our speakers or listening rooms ?
While I can enjoy a Stanton 381, 581, 681, 881 Calibration Standard phono pickup, I seem to prefer more euphonic sounding phono carts. Still, linear frequency response and low distortion in preamps and amps is of utmost importance. How else would we be able to discern phono EQ differences when playback curves can differ by only 1 to 2 db along the curve ?
The iPhono2 seems like a good product for stereo only listeners. Why don't these design engineers recognize mono discs ? What I find most distressing in modern phono stages is the lack of "AES" EQ and no inclusion of a 700-800 Hz Turnover. Even early Elektra LP's 629 Hz turnover would be helpful for dull sounding early discs. Graham Slee's Revelation and Jazz phono stages also neglect some stateside disc needs. I actually wonder where the iPhono was designed ? They include some Euro curves, at least we think they include them, but switchability is not as versatile as some would like...
If you check the TempoElectric phono EQ site, those guys also believe Euro and UK Columbia disc's EQ differed from RIAA into the late '70s. If true, that would mean many of the Euro discs we crave actually have a turnover at 450 Hz, not 500 Hz. This is a problem for me. As I am planning some preamp builds, more switch settings means more work and more parts...
I have seen the insides of the first gen iPhono. Replacing parts or adding switches to tweak EQ would not be an easy task. At the current state of our art, I guess $500. is not too expensive a price for a switchable EQ phono stage. I just know DIY can be better; especially if using tubes.
Thank you for the kind words :) It is nice to know that my posts are enjoyable to some people - I tend to write as I am thinking so I've often been worried that I write too much!
I have nothing against using a different curve as a tone control and I'm no longer black and white regarding "accuracy" and neutrality. To me it is akin to remastering. The exception is if I am transcribing a record in which case I want an accurate facsimile of the recording...after which I can do what I want with the sound. This way I can always go back to the original if I change my mind!
My thinking on what I want to "hear" has evolved over the years as I learnt more about mastering. In fact it wasn't until I started doing my own digital recordings and realised that I was getting an adequately close facsimile of my vinyl that it became clear that unless one was present at the recording session, there is no way to know what the original sound balance was. At that point I stopped fussing over upgrading my equipment - although to be fair I had taken it about as far as I was prepared to pay!!
Remastered records are a case in point - it is yet another flavour of the original master. Purely someone's opinion of how it should sound.
WIth respect to phono stages and vintage curves, I think Kevin is the only one doing something universal in his Souvenir phono stage.
"I am planning some preamp builds, more switch settings means more work and more parts..."
Me too! Except I have been stifled by the Paradox of Choice - I couldn't decide which way to go with the topology. After I recently successfully tested out applying software RIAA using the Korg AUdiogate Export function (I bought a DAC-10R) on a wav file recorded on my preferred rig, I have shifted to simply designing a flat gain amplifier with adjustable loading.
I tried taking apart my iPhono, but couldn't slide the board out because of the DIP switches. I googled the images and realised there wasn't much I could do either. My Musical Fidelity X-LPS was quite another story!
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: