Home |
Search |
Today's Posts |
#1
|
|||
|
|||
where to get RIAA test record / "RIAA NOISE"
Hi -
Well, I'm still in the midst of building & re-building *the* tube RIAA preamp. Got plenty of meaningless advice thus far (Pat's little post on 1st stage noise wasboth patronizing, incomplete,and *wrong*, but that's another story. I'm not looking for theory advice, but a few practical things: When I was a kid,. I remember getting a "HI FI test record" from the library, do such records still exist ( as in available *new*, hopefully with tone sweeps on both chans, noise which would look flat on a spectrum analyzer with proper RIAA eq, simulated "scratches" (one of my wonders works great on new records, but overloads on scratchy stuff - sounds *awful* on old records - not enough headroom in the first stage...). I'd love to have one so I can check the entire setup, not just the preamp. I'veinherited a pile of interesting carts, and one of my turntables has a standard headshell (~$5 for decent cheap ones), so it would be neat to see what they're doing on a spectrum analyzer. If anyone knows a source for decent test records (if it had a '78 section, that would be great - I know that 78's were all over the place as far as EQ was concerned, but a whole bunch of people ask me to dump their 78's on CD's, and, at this point, I do digital processing (sound forge, cool edit, pro tools, and all the assorted plugins) to get the 78's to sound "good" to my ear (some people ask me *not* to do it, so I'd like to know what 78's were supposed to sound like, the 'real" '78 curve. Once again, i know that 78's varied, but I've seen many good preamps with '78 eq settings. Also, it's a bit scary having a good turntable by my workbench (obvious reasons, including mess, clumsiness, and having to move big heavy things up & down the stairs). Did anyone create any software (like riaa - weighed noise, etc. for PC soundcards, for rough testing? Also, did anyone bother designing an attenuator for line-level signal, which is more than a voltage divider I use? something which will have -properties similar to various pickups? If one's not been done yet, I'd love to see as much data on various moving coil /moving magnet pickups, so I could try to hack one together (i'll post the schematics & layout if the results are useful, with settings for common carts). Finally, another *practical* question to people who *know*: It's been said, and, looking at different schematics, I believe it, that there *was no "gold standard' RIAA playback stage*. They all ranged from simple roloff filters to attempts at RIAA curves which "may" have worked with one particular pickup. The question is, how did the recordists EQ the records? did they slavishly follow the RIAA curve, or did they listen to them on themainstream gear of the day? I doubt that Motown recordings were expected to be played on transcription-grade gear - if I was doing the final EQ, i'd probably test it on some record-eater broadcast table transmitting to an AM car radio, and a portable record player with ~5g's of tracking weight & a nail for a stylus. Make it sound good on that stuff, and then see if I can satisfy the remaining few % of the listening /buying public. Anyone out there been there, or heard stories? Creating a perfect RIAA stage would seem pointless, if, in real world, the standard was not really adhered to. I listen to "quality" recordings ~5% of the time (by "quality" i mean the audiophool's delight acoustical recordings, jazz or "classical"), not the great "overprocessed' stuff that came out of pop studios in the 60's, 70's & 80's - I love the stuff. Club stuff was definitely mixed for clubs, and i know club setups well - used to do everything from setup / board to crewing for sound rental co & wiping puke off the cables while rolling them up. And getting bitched at by **** bands while running house boards at pro bono open mikes). So - any help? -dim |
#2
|
|||
|
|||
On Fri, 01 Apr 2005 16:04:20 GMT, "shiva" wrote:
Hi - Well, I'm still in the midst of building & re-building *the* tube RIAA preamp. Got plenty of meaningless advice thus far (Pat's little post on 1st stage noise wasboth patronizing, incomplete,and *wrong*, but that's another story. I'm not looking for theory advice, but a few practical things: When I was a kid,. I remember getting a "HI FI test record" from the library, do such records still exist ( as in available *new*, hopefully with tone sweeps on both chans, noise which would look flat on a spectrum analyzer with proper RIAA eq, simulated "scratches" (one of my wonders works great on new records, but overloads on scratchy stuff - sounds *awful* on old records - not enough headroom in the first stage...). Use passive rolloff for the 75usec break, and use it as early as you dare. Can compromise S/N if you're not real good with the first stage, but stops scratches in their tracks. See my page at http://www.lurcher.org/ukra/ for a possible topology. The gain blocks don't *have* to be op amps! :-) I'd love to have one so I can check the entire setup, not just the preamp. I'veinherited a pile of interesting carts, and one of my turntables has a standard headshell (~$5 for decent cheap ones), so it would be neat to see what they're doing on a spectrum analyzer. If anyone knows a source for decent test records (if it had a '78 section, that would be great - I know that 78's were all over the place as far as EQ was concerned, but a whole bunch of people ask me to dump their 78's on CD's, and, at this point, I do digital processing (sound forge, cool edit, pro tools, and all the assorted plugins) to get the 78's to sound "good" to my ear (some people ask me *not* to do it, so I'd like to know what 78's were supposed to sound like, the 'real" '78 curve. Once again, i know that 78's varied, but I've seen many good preamps with '78 eq settings. There is no 'real' 78 curve, over a dozen were used by different companies. The whole point of the RIAA curve, was that it was an agreed standard to bring order out of chaos. Also, it's a bit scary having a good turntable by my workbench (obvious reasons, including mess, clumsiness, and having to move big heavy things up & down the stairs). Did anyone create any software (like riaa - weighed noise, etc. for PC soundcards, for rough testing? No need for software, just use an inverse RIAA attenuator. Basically, the feedback section from a conventional 'all in one' RIAA preamp. Also, did anyone bother designing an attenuator for line-level signal, which is more than a voltage divider I use? something which will have -properties similar to various pickups? If one's not been done yet, I'd love to see as much data on various moving coil /moving magnet pickups, so I could try to hack one together (i'll post the schematics & layout if the results are useful, with settings for common carts). Imnpossible to do, as cart curves vary greatly with loading, hence all the varying R & C value options on 'universal' RIAA preamps. Finally, another *practical* question to people who *know*: It's been said, and, looking at different schematics, I believe it, that there *was no "gold standard' RIAA playback stage*. They all ranged from simple roloff filters to attempts at RIAA curves which "may" have worked with one particular pickup. The question is, how did the recordists EQ the records? did they slavishly follow the RIAA curve, or did they listen to them on themainstream gear of the day? The RIAA curve was the gold standard, and AFAIK, all major cutting facilities followed it exactly. Of course, cutting engineers would add their own special EQ to get the best out of any particular mixdown master tape, but the RIAA curve was always the base. I doubt that Motown recordings were expected to be played on transcription-grade gear - if I was doing the final EQ, i'd probably test it on some record-eater broadcast table transmitting to an AM car radio, and a portable record player with ~5g's of tracking weight & a nail for a stylus. The Yamaha NS-10 was the classic 'lowest common denominator' speaker used for final pop/rock mixdowns. Make it sound good on that stuff, and then see if I can satisfy the remaining few % of the listening /buying public. Anyone out there been there, or heard stories? Creating a perfect RIAA stage would seem pointless, if, in real world, the standard was not really adhered to. No, it is a viable target, but then you have to compensate your preferred cartridge to be as flat as possible to that curve. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#3
|
|||
|
|||
Behold, shiva scribed on tube chassis:
where to get ..... "RIAA NOISE" Any recording of their anti-filesharing huff-and-puff speech. ;-p -- Gregg "t3h g33k" http://geek.scorpiorising.ca *Ratings are for transistors, tubes have guidelines* |
#4
|
|||
|
|||
shiva wrote: Hi - Well, I'm still in the midst of building & re-building *the* tube RIAA preamp. Got plenty of meaningless advice thus far (Pat's little post on 1st stage noise wasboth patronizing, incomplete,and *wrong*, but that's another story. If my name is pat I am entitled to be patronizing, and lucky for you i didn't call my outfit Patronics.com, where patrons would have been given plenty of zing fo sure :-] Now there's an idea, Dimtronics.com. Just don't try selling TVs.... What was wrong with my post on stage noise? what was incomplete? I'm not looking for theory advice, but a few practical things: When I was a kid,. I remember getting a "HI FI test record" from the library, do such records still exist ( as in available *new*, hopefully with tone sweeps on both chans, noise which would look flat on a spectrum analyzer with proper RIAA eq, simulated "scratches" (one of my wonders works great on new records, but overloads on scratchy stuff - sounds *awful* on old records - not enough headroom in the first stage...). I'd love to have one so I can check the entire setup, not just the preamp. I'veinherited a pile of interesting carts, and one of my turntables has a standard headshell (~$5 for decent cheap ones), so it would be neat to see what they're doing on a spectrum analyzer. If anyone knows a source for decent test records (if it had a '78 section, that would be great - I know that 78's were all over the place as far as EQ was concerned, but a whole bunch of people ask me to dump their 78's on CD's, and, at this point, I do digital processing (sound forge, cool edit, pro tools, and all the assorted plugins) to get the 78's to sound "good" to my ear (some people ask me *not* to do it, so I'd like to know what 78's were supposed to sound like, the 'real" '78 curve. Once again, i know that 78's varied, but I've seen many good preamps with '78 eq settings. I have a couple of test tecords, one is the CBS 'ST100 Stereophonic Frequency Test Record' The test frequencies can be played through an amplifier to check the response of a system including the cartrige. The reference level is at 1 khz and designed to give the rated output for the cart, maybe 3 mV for an MM cart, or 0.4 mV for MC. F below 500 Hz are cut with constant amplitude, so the signal off the test record is one declining in amplitude at 6 dB/octave below 500 Hz. F above 500 are cut with constant velocity, thus giving a flat response with rising F. To plot the above responses mentioned requires an amp with no RIAA eq. So such a test record only tests the cart response. I found it difficult to examine the LF response sinse it declined so much with a flat amp, and I rigged a temporary eq network to allow LF to be plotted flat. 18 mths ago I compared several carts and TTs to see what differences I would find. Differences were not staggeringly different. Distortion was very visible on the CRO trace in the HF output above 6 kHz, and for Denon 103 MC I found 1k ohm plus 0.22 uF gave the flattest HF but with fairly well suppressed distortion in the signal. The distortion at 10kHz and above as like a really bad case of crossover distortion with a wriggle in the zero crossing point, so I suspected a mechanical whiplash effect. Now the problem with this damned test record is that you cannot just play a record into a preamp with a known correct RIAA eq, and simply expect to be able to plot a flat response easily. That because records are cut to the RIAA inverse filter shape, with boosted HF, to get around the noise problem on replay. So how to deal with amp overload? It is said to be most likely with passive eq preamps. Say you have a cart which makes 5 mV of signal at 1 khz from the preamp. Then to get the same amount of signal at 10 khz, the 10kHz signal is boosted at the cutting amp and comes off the record at around +12dB, so you have 20 mV going to the preamp at 10 kHz for a flat test signal. Bit say we allow the preamp to take 100 mV of signal. If its a 12AX7 mu follower, gain = 90, so there will be 9vrms output, which won't overload the stage. When 9 vrms at 20 khz is fed to the RIAA filter, it is reduced by around 20 dB, and we get 0.9 vrms out, and when that is amplified bt stage 2 of the phono amp with gain = 40, we'd get 36 vrms, which does overload the stage. In practice these abilities of a good MM amp allow for pretty darn good dynamic range ability, and such amps are rarely overloaded by noise spikes or music, which usally has a declining HF content above 1 kHz. A typical amp for MM is at http://www.turneraudio.com.au/htmlwe...tubepreamp.htm Now to increase the voltage ability of such an amp it is possible to place an attenuation network between the 12AX7 follower an the RIAA filter, but that network will require a change of all the R&C values used in the network. There were about a dozen different eq slopes used by record makers including all those used with 78, and you should know exactly what all the profiles and slopes were, and have a switched eq to suit all these records so that saving old records on CD isn't such a darned hit and miss affair. But that assumes record companies stuck to the eq they say was used on the lable, but many were "off tolerance" and exactly what was used is a secret. With tube preamps with NFB eq for RIAA, the overload capacity still is determined by the last stage, which will overload before the first stage. Since 0.4 vrms is a typical output from a vinyl preamp, having a 40 dB voltage overload is a decent margin. Its allowing for 40 vrms output. An SS opamp thinge would have only about 26 dB over load capacity, ( 8vrms ) Also in practice there is slew rate limiting of HF output with phono amps, but not until you have hugely excessive signals. So if your tube phono amps are overloading with records, the amp deign must be poor. Also, it's a bit scary having a good turntable by my workbench (obvious reasons, including mess, clumsiness, and having to move big heavy things up & down the stairs). Did anyone create any software (like riaa - weighed noise, etc. for PC soundcards, for rough testing? Sorry to sound patronizing, but clean that mess up and don't be clumsy, and don't be lazy. I have to carry players and amps all over the place. Its part of the deal of being a tech. Also, did anyone bother designing an attenuator for line-level signal, which is more than a voltage divider I use? something which will have -properties similar to various pickups? If one's not been done yet, I'd love to see as much data on various moving coil /moving magnet pickups, so I could try to hack one together (i'll post the schematics & layout if the results are useful, with settings for common carts). Just make your phono amp able to take 100mV of input at 20 kHz, and you'll never worry about overloading signals from records using magnetic pu. Ceramic pu with 100 mV average signal can be a problem, but not magnetic. Finally, another *practical* question to people who *know*: It's been said, and, looking at different schematics, I believe it, that there *was no "gold standard' RIAA playback stage*. They all ranged from simple roloff filters to attempts at RIAA curves which "may" have worked with one particular pickup. It is routine to get RIAA playback amps to +/- 0.5 dB compliance to RIAA. And not so uncommon to find carts +/- 1 dB accurate between 30Hz and 10kHz. The question is, how did the recordists EQ the records? did they slavishly follow the RIAA curve, or did they listen to them on themainstream gear of the day? There were 12 different curves used. After they finished cutting 78, and vinyl became king, they all agreed on the RIAA curve in the early 1950s I recall. I doubt that Motown recordings were expected to be played on transcription-grade gear - if I was doing the final EQ, i'd probably test it on some record-eater broadcast table transmitting to an AM car radio, and a portable record player with ~5g's of tracking weight & a nail for a stylus. Make it sound good on that stuff, and then see if I can satisfy the remaining few % of the listening /buying public. Anyone out there been there, or heard stories? Creating a perfect RIAA stage would seem pointless, if, in real world, the standard was not really adhered to. But so often the RIAA *was* adhered to in the cutting process. But studios eq'd the sound all over the joint for bass treble balance, or with graphic eq to make a pop star sound listenable, and therefore saleable. It all continues today, mauling of the signal. Nothing is real, all is unreal with recorded music. I listen to "quality" recordings ~5% of the time (by "quality" i mean the audiophool's delight acoustical recordings, jazz or "classical"), not the great "overprocessed' stuff that came out of pop studios in the 60's, 70's & 80's - I love the stuff. Club stuff was definitely mixed for clubs, and i know club setups well - used to do everything from setup / board to crewing for sound rental co & wiping puke off the cables while rolling them up. And getting bitched at by **** bands while running house boards at pro bono open mikes). So - any help? -dim Well, Have you tried placing the record on a tree stump? The best use of the nail you mentioned above is to nail that record down, and don't let it blow away. They put a hole in the record centre to make it easy to nail down. Good stumps have a bolt, and you can screw a nut tightly over the record. Then attach a stylus to a microphone, and hold it in the groove while running around the stump at 78 rpm. The slight wow and flutter is compensated by sound of the natural revolutions you are making, not to mention your fitness improvements. 33 rpm is easier to run, but you need to be more careful since the grooves are smaller. I saw a guy demonstrating the technique in a Monty Python show on british TV. He wore a flying helmet and goggles to cope with the breeze, and the chilly british weather. Patrick Turner. |
#5
|
|||
|
|||
On Fri, 1 Apr 2005 19:53:47 +0000 (UTC), Stewart Pinkerton
wrote: Use passive rolloff for the 75usec break, and use it as early as you dare. Can compromise S/N if you're not real good with the first stage, but stops scratches in their tracks. Excellent advice. One possibility for inductive sources (the usual "MM" cartridges are 1/3 to 1/2 Henry (yikes!) sources) is a resistive load to provide the 75uSec pole. This falls into the coupla-K-ohm range, so don't significantly reduce average level (source resistance is about a half a K-ohm). An interesting side effect is the elimination of the low-Q resonance otherwise needed to maintain response in the upper statosphere. And the two pole plumment above. No need for software, just use an inverse RIAA attenuator. Basically, the feedback section from a conventional 'all in one' RIAA preamp. Be sure to include a gain stop at the top. Chris Hornbeck 6x9=42 |
#6
|
|||
|
|||
On Sat, 02 Apr 2005 00:56:00 GMT, Patrick Turner
wrote: I have a couple of test tecords, one is the CBS 'ST100 Stereophonic Frequency Test Record' The test frequencies can be played through an amplifier to check the response of a system including the cartrige. The reference level is at 1 khz and designed to give the rated output for the cart, maybe 3 mV for an MM cart, or 0.4 mV for MC. F below 500 Hz are cut with constant amplitude, so the signal off the test record is one declining in amplitude at 6 dB/octave below 500 Hz. F above 500 are cut with constant velocity, thus giving a flat response with rising F. I'm sure everyone understands what you mean, but to be clear, you mean all this to be true when played back through a velocity responsive cartridge. Now the problem with this damned test record is that you cannot just play a record into a preamp with a known correct RIAA eq, and simply expect to be able to plot a flat response easily. That because records are cut to the RIAA inverse filter shape, with boosted HF, to get around the noise problem on replay. Perhaps a better way to describe the topic for first-time students is that the usual modern phono transducers are fundamentally dynamic, meaning velocity sensitive, meaning have their own 6dB per octave rising frequency response for constant amplitude signals. The "RIAA" curves we see everywhere take all that into account, but the cutters don't think that way. FWIW, non-velocity-sensitive playback cartridges are possible, and in the case of the MicroAcoustics, viable. They'd integrate with the dynamic world by differentiating with a loading R. Just babbling, sorry, Chris Hornbeck 6x9=42 |
#7
|
|||
|
|||
"Chris Hornbeck" Excellent advice. One possibility for inductive sources (the usual "MM" cartridges are 1/3 to 1/2 Henry (yikes!) sources) is a resistive load to provide the 75uSec pole. This falls into the coupla-K-ohm range, so don't significantly reduce average level (source resistance is about a half a K-ohm). ** This is just plain nuts. An L of 0.5H = 3141 ohms at 1 kHz so a load of 2 kohms will have a drastic effect. Output from the cartridge will be way down at high frequencies WHILE the pre-amp will have much more noise as the gain is maintained high and flat at all mid and high frequencies. An interesting side effect is the elimination of the low-Q resonance otherwise needed to maintain response in the upper statosphere. And the two pole plumment above. ** Another "interesting" side effect is a BIG increase in noise. ............ Phil |
#8
|
|||
|
|||
On Sat, 02 Apr 2005 01:18:43 GMT, Chris Hornbeck
wrote: On Fri, 1 Apr 2005 19:53:47 +0000 (UTC), Stewart Pinkerton wrote: Use passive rolloff for the 75usec break, and use it as early as you dare. Can compromise S/N if you're not real good with the first stage, but stops scratches in their tracks. Excellent advice. One possibility for inductive sources (the usual "MM" cartridges are 1/3 to 1/2 Henry (yikes!) sources) is a resistive load to provide the 75uSec pole. This falls into the coupla-K-ohm range, so don't significantly reduce average level (source resistance is about a half a K-ohm). I've almost always used MC carts, but that's quite a neat idea if you match the loading well to the cart - certainly kills any possibility of transient overloads! An interesting side effect is the elimination of the low-Q resonance otherwise needed to maintain response in the upper statosphere. And the two pole plumment above. No need for software, just use an inverse RIAA attenuator. Basically, the feedback section from a conventional 'all in one' RIAA preamp. Be sure to include a gain stop at the top. Um, It's kinda built into an attenuator, no? :-) -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#9
|
|||
|
|||
On Sat, 2 Apr 2005 06:41:20 +0000 (UTC), Stewart Pinkerton
wrote: No need for software, just use an inverse RIAA attenuator. Basically, the feedback section from a conventional 'all in one' RIAA preamp. Be sure to include a gain stop at the top. Um, It's kinda built into an attenuator, no? :-) Yes, but I didn't 'splain myself very well. An RIAA attenuator with 40dB midband loss (typical) will have a continuously rising response from 2122 Hz for two decades before leveling off, unless a gain stop is added. Chris Hornbeck 6x9=42 |
#10
|
|||
|
|||
On Sat, 02 Apr 2005 15:45:29 GMT, Chris Hornbeck
wrote: On Sat, 2 Apr 2005 06:41:20 +0000 (UTC), Stewart Pinkerton wrote: No need for software, just use an inverse RIAA attenuator. Basically, the feedback section from a conventional 'all in one' RIAA preamp. Be sure to include a gain stop at the top. Um, It's kinda built into an attenuator, no? :-) Yes, but I didn't 'splain myself very well. An RIAA attenuator with 40dB midband loss (typical) will have a continuously rising response from 2122 Hz for two decades before leveling off, unless a gain stop is added. Yes, but that's a correct response, if you include the later-added 50kHz pole which is the natural endpoint. -- Stewart Pinkerton | Music is Art - Audio is Engineering |
#11
|
|||
|
|||
"shiva" wrote in message news:8ie3e.662$7b.15@trndny06... So - any help? Sure.... feed'n scoop! |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
on topic: we need a rec.audio.pro.ot newsgroup! | Pro Audio | |||
Artists cut out the record biz | Pro Audio |