|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
Hi inmates, yes i'm probably crazy. Ok i've got four DBX 700 digital audio processors, remember these? They were used in mid 80's. The manual states they convert analog audio into digital bits and then codes the bits into the standard video NTSC (National Televison System Committee)format, at a sampling rate of 644Khz and 644Kb/sec. The Anti-aliasing filters are -3db@37Khz. Instead of using the intended VCR/VTR U-matic (limited life tape media) machine, what kind of results could I expect using a DVD+RW to record the video signal? Am I correct in that the DBX digital signal would be converted into Mpeg 2 video code to disc? And of course the signal would need to flow/read from disc back to the 700 for D/A for listening. The audio tracks (PCM type processing) of the DVD+RW would not be used, unless I want them for A/B comparisons or something else. I understand that the recorded signal will not be compatible with other systems, (unless they also have a 700 processor), but do you guys have any thoughts on this? Good and Bad. Why am I looking to do this you may ask? I've never like CD's, have a room full of gear and still don't own a player, and oh yes my local dealer has loaned me units costing up to 3.5K, I don't like what I hear, they are very fatiguing and harsh to my ears. You've all heard the criticisms before so no need to cover it again here. The DSD system sounds great, but I don't have the 25k or more to get into that, and I don't see any indications that we will be offered a prosumer DSD recording system in the near future, nuff said there. If you guys don't see any reason why this idea might not work out ok, then I want to archive my analog vinyl collection and studio reel tapes of bands I recorded years ago to disc, (main tracks DBX/video) and maybe use the high res PCM audio tracks of the DVD+RW as well, as a back up or what ever. I recall the sound of the DBX 700 compander as being much better than red book cd,(no 20Khz brick wall), and the PCM systems still don't sound natural to me. (Plez don't be cruel). The ultimate goal is to have a library (juke box)with [no physical contact with media], there by no wear. so I would appreciate any thoughts you Pros can share, no straight jackets please....;) Thanks,Rick
Follow Ups:
Denon offers a record player which has a useful feature, a variable digital out. it may be worth looking at the DENON DP-DJ151
Hi Ben,
Thank you for your input, very interesting turntable, however I think that my Sota Nova and Shappire tables would bet it in a sonic A/B comparisons. The built in A/D converter uses the standard 16 bit 44.1kHz 'brick wall' format, I've never care for it much, chops off the upper octaves.
Thanks again for pointing out some new tech stuff though!Rick
Hello Rick. May have missed it, but it's not clear to me whether you primarily want to archive master recordings having more than two tracks. If not -- that is, if you want to archive mainly two-track masters and LPs, etc., did you notice that Tascam recently started selling a version of their high-density 8-track prosumer deck (a version of the Model 98-HD, I think) that records in the Sony DSD format (1-bit, 2.8 MHz): 2 tracks DSD capacity on the machine (which does 8 tracks at 24-bit/96kHz) -- I think the list price is under six-thousand dollars. I also wonder if you've considered going direct to stereo FM analogue audio on videotape. I've heard complaints about the VHS Hi Fi system for this purpose (a reported ripple around 40 Hz, as I recall, from the switching between the heads on the drum). I've used a Sony Hi-8 analogue videocam with FM stereo analogue audio (TRV series) for archiving, and it sounds pretty good to me. It seems that Sony has stopped making FM audio videocams since they've switched their sonsumer line to mostly digital (the few low-priced analogue models they now list apparently aren't FM analogue audio), but there are probably still some of them around for sale. I'd be curious to know what expereinces others on this list have had with using VCRs or videocams with FM analogue stereo audio for recording sound. I remember the reviews and test-reports when the first models (VHS) came on the market in the mid-80s -- there was a lot of excitement about the quality (signal-to-noise is something like 90 dB, and very extended frequency response because of effective tape-speed of something like 400 ips thanks to rotating heads. In those early reports didn't hear any complaints about noise from the head-switching. I wonder if any of the higher-end video-studio type VHS VCRs (with audio meters, etc.) are made to good enough tolerances to avoid that problem.
Hi Mahc'l,
Thank you for your input on this 'Mad' project. Yes I've looked at the Tascam machines, very nice, also still very pricey.The main objective here is to archive to 'Non-Contact' optical disc, with out using the PCM format if possiable. The newer DDR type recorders look very interesting, I'm still checking into their use for this project, may be too pricey al well though.
Thanks again for your thoughts.
Rick
...I think I remember those... 1984 APRS show in London I saw one, some ridiculous dynamic range... -The first delta-error prediction system, if I rcall correctly. The sample rate was through the roof, but the data rate was comparable to PCM, since it only recorded the differences from the previous sample (which considering the sampling rate, thus the super-short time elapsed between consecutive samples) should only be very small!The only trouble this thing had was with square waves... -If I remember correctly, it used the logic that if it had got the last few guesses close -or correct- than it used to 'zoom-in' to the next level of resolution. each error larger than the current 'target window' used to cause it to 'zoom out'. -Problem was that with square waves, it used to keep zooming in on the flat part of the waveform, until the sudden, massive leading/trailing edge used to cause colossal overload. -For this reason, it's transient response was always lamentable, but the steady-state sine wave response was out-of-this-world!
Well, if you attempt to compress the video data, (MPEG-2)You're gonna end up with absolute unlistenable crap. -I don't mean bad-sounding audio, I mean it just plain old isn't going to work.
Unfortunately, the compression systems for picture result changes that have been developed to not be too offensive to the eye... unfortunately, the effect on a data stream would almost certainly be catastrophic!
Most of the video-based digital encode/decode units used to make a 'picture' comprising a set of black vertical bars on a white background, and the width of each black bar changed at each of the horizontal line intersections, to represent a one or a zero in the data stream recorded to the VCR. This means that vertically adjacent pixels on a complete image could easily -and frequently did- alternate black/white/black... this is probably going to be re-written by an MPEG encoder as a block of grey... losing all vertical distinction, and presenting the decoder with an interesting puzzle to take a guess at!!!
Take a look at the NTSC video stream coming out of an encoder... I've never looked at one from the DBX series, but the lower-resolution (16-bit, 44.1/44.056/48/47.whatever) Sony PCM-F1 and PCM-501/701, PCM-1610/1630 etc. all worked as described above. The front end was radically different on the DBX's, but it would stand to reason that the video back end would use similar methods.
Seems to me to be a few steps too many to convert audio to bits, then convert the bits to a picture, then convert the picture to bits again for storage... Naturally a manufacturer wouldn't do this, but back in the mid-eighties, video tape was cheap (still is) but affordable large-volume archive data storage just wasn't available.. hence the various video-evolved systems. -The precursor to DAT (and all the problems that that brought!) but not even as robust, since DAT at least uses saturation recording, so stands a better chance of survival over time.
Keith
Hi Keith,
Thank you for your very informative reply, your memory and knowledge is impeccable. Yes the manual claims a dynamic range of 110db, and a sampling rate of 644Khz and a bit rate of 644Kb/sec.
Your correct, in the the Technical background notes by Robert Adams state that the 700 uses Companded Predictive Delta Modulation.I agree with you in that there would be to many conversions in the chain, I was concerned with this as well. Can you or anyone suggest a media format that would work, other than the VCR/VTR tape media? Is there a burner/format that would do the job of recording the video signal without the compression/conversion problem?
As I mentioned in the orginal post, VHS/tape in not a good archiving media, I'm just thinking there may be a way to place the signal on to disc, I don't know.
Anyone else is welcome to chime in on this please.
Rick
I'm sure that Keith is right here. DVD uses mpeg data compression of the order 5:1. Results for video can be stunning, I have a Spruce system, and comparisons between the original and the encoded stream are very good. But as to wether it would work with your DBX, I would be most suprised if it did. In this age of high resolution encoders and sensible AD converters I would seriously look again at the current crop of digital 96khz encoders. I know many hardened analogue fans who are converts. Having used both in my recording career, I have to say I can't see any benefit in using analogue any more. I would suggest that if you are hearing harsh sounds from digital kit, either the recording or the kit thats being used is at fault. Smooth top end, is perfectly achievable by good recording practices!Regards
Roland
Hi Roland,
Thank you for your reply as well. I've enjoyed and learn a lot from your numerous posts, and respect your opinion. I did not invest a lot into the 700 processors and don't mind adding them to a collection of audio relics, if that be their fate. Many fine albums were mastered using them. I have yet to hook them up to a VCR/VTR and experiment/play with them, but will soon and enjoy seeing just what the little beasty can do.Your right in that I should give digital another listen to and try to leave my bias' at the door. The 'harshness' of CD's mentioned in the earlier post was while listening to some jazz with horns, it may have been cheap mics, and/or bad levels at the mix, or just a poor pressing, however after listening to about 6 CD's at the local Hi-Fi shop, I still didn't like what was coming out of the speakers.
I have yet to hear a digital source reproduce proper phase relations in the room, (many recordings sound inside out), stage imaging has allways been a criticism of mine as well, the sound stage sounds like a almost 'flat' two dimensional affair, pinpointing individual instruments has improved over the years but not like listening to some old, yet clean analog vinyl. Solos can sound very clean to the point of sounding sterile, lacking the harmonic rich 'roundness' of pure sine waves, the 'live presents in the room' is still a few steps behind a good analog recording IMHO, but without the 'noise' that's always present with the analog. (I hear the guys with the straight jackets coming down the hall, better stop here on these points)...:-)
Ok, so I've been researching what's available without breaking the bank, can any of you give a opinion on Alesis' MasterLink using up to 24bit, 96Khz vs. Euphonix Model Two HDCD processor using up to 24 bit and 192Khz? (I hate format wars!!!) Any other systems/formats that I've missed? As I mentioned in the first post the DSD system sounds great, but way too expensive for me, a rack of SADiE outboards would be great for christmas, but then I'll be eating rice & beans for the next three years too..;) I can probably arrange a loan of a Alesis ML from the guys downtown to try out, but the Euphonix may require a license...don't know yet. Opinions please!
Lastly, (still beating the dead horse) there is a 25 pin "D" type expansion jack accessory connector on the back of the DBX 700 encoder/decoder, all the manual says is that, 'this is were the digital 'inputs' and 'outputs' are available for equipment such as the dbx 700D Disc-Mastering Delay', no further information is given except that any wrong connection will damage the unit. I 'suspect' that this may be were one could access the encoder/decoder processor BEFORE it's converted to the NTSC video signal, but this is only a quess at this point, carefull tests may follow. If any of you high powered brain types (Keith..;-) have any thoughts on how one might utilize this port directly to disc recording please let me know!
Ok, the very last, I promise, restating the original question another way, excluding DVD-RW or DVD+RW / MPEG type encoding due to previous explanations, is there not some type of CD-burner that would record bit for bit data without errors that could replace the VCR/VTR???....(now I hear the Doctor coming down the hall with the syringe...I'll be back later to see what you guys have to say when I wake up...;-)
Rick
The disc mastering delay is indeed a digital output of the encoded signal prior to video encoding... I'm not sure of the format however. -You may well know the idea behind mastering delays, but just in case you're not familiar, here's a quick rundown:Back in the later all-Analog days, disc-cutting rooms had progressed to computerised disc-cutting lathes that could maximise the amount of time you could wring out of a vinyl disc pressing. -Louder signals make the groove deviate from side to side more, so louder signals in the days before computers meant shorter recordings, since you had to space the spirals out to give more room between adjacent grooves to stop them banging into each other. (...yes, I know...there's only one long groove on each side, but most folks still think of them as 'adjacent grooves'!) Louder cuts meant shorter recordings, longer recordings meant that the signal to noise ratio went down.
The advent of decent computerised lathes (The Neumann VMS-70 series was a great favourite!) meant that the computer could slow down the lead screw (tightening up the grooves) in quieter passages, -allowing the groove spacing to close up when it deviated less- and speed up the lead screw to space out the groove spacing when the signal got louder. -The only thing that complicated the process was that it had to be able to predict the future... it had to know how loud a signal was going to get and when, so that it could give the groove enough space to get loud before the signal actually got loud...
The analog tape machine solution was to provide two sets of playback electronics and two playback heads, with a fairly long tape path between them. -The first head (the 'preview' output) would allow the lathe to see what was going to happen in 1 revolution's time, and let the computer figure out how much space to give the next groove for signal level. -the second playback output -one full revolution later- would be the signal that was actually sent to the cutting head, now that the computer has spaced it as wide as it needs, but as tight as it could.
The length of the tape path between the two playback heads would need to be different for different turntable speeds, -45RPM and 33+1/3 for example- and this was normally accomplished by having the heads fixed, but running the tape through different 'pulley-paths' by routing the tape over different guide rollers for each cutting speed.
Some mastering facilities used single head playback machine and used stereo digital delays to generate the delayed signal, the advantages included only one stereo playback alignment instead of two, and the delay line could have setting memories to quickly switch between the different delay times for different cutting speeds. -The big disadvantage was that the signal was degraded by the A/D & D/A conversion & filtering, especially given the state of the art of converter design at that time!
Now.... back to the DBX units... Using the boxes you have, the two signals ('preview' and 'cut') should be identical, although you'd need two decode devices. -The first playback converter would decode the video and convert it to raw digital information, which would be sent both to the D/A converter (sent to the lathe as it's 'preview' signal) and to the mastering delay output. This would bedelayed -still in it's digital form- by the required amount (one lathe revolution) and its output would then be fed to a second playback decoder, where the identical raw digital data (just later in time) would be re-constructed into an analog waveform by an identical converter to the first (preview) converter's output.
The delay line and second converter would all have to run from a common sample clock, so either the 25-pin port would contain the timing clock along with the data, or there would be an additional clock source -similar to a BNC wordclock connector used with SDIF, MADI or SDIF-2 nowadays... -Hmmm.... I don't think that it could actualy be SDIF-2 though, because of the high (nonstandard) sampling rate... -It might use similar line drive hardware though...
Unfortunately, the great bulk of my expertise is in Analog (long may it reign!) -if you can get a computer data storage geek to figure out a way to lock onto the clock, strip the data and send it to a common storage system, you might have a nice little setup there, but I'm afraid I don;t know anything about that end... as soon as music stobs being a waveform and starts being a stream of numbers I just lose interest... (-It seems more like accounting to me than signal treatment!!!)
I suppose Roland's shorter answer is a better one... Simply put, there's no way unless you know -or are- someone who's really good at data storage!
I'd love to know if you can get it to work, though...
Keith A.
Hi Keith,
I again thank you for a voluminous and educational reply. I vaguely remembered the reasons for the mastering delay, and appreciate your detailed explainations very much. The DBX 700's do have external BNC type 'input' & 'output' jacks labeled 'SYNC' on the rear panels. The manual states: 'These are used for broadcastiing and recording applications where the 700 must be synchronous with video sources coming either before or after it (e.g., satellite transmission and multiple-700 applications). If a standard video-synchronization signal is the input-it must be 4-v, negative-going, and ac-coupled as well -- then the 700 will generate video in synch with it. In other words, the time rate of the 700 adjusts to that of the source, with the Sync output a buffered version of the input. If there is no such input, the ouput is at the 700's free-running rate (within 0.1 percent of the NTSC standard)'.
However, the Addendum to the maunal states: 'Model 700s without our dealer-installed Copying-function modification require a synch signal that is negative-going, dc-couple (not ac-coupled, as the text says), and with levels that are 4 v and ground. (8v peak-to-peak loaded by 75 ohms, as is the case here, drops to 4 v.) These units have serial numbers below 794 (with the exception of no.774). If your particular unit is in this group, you can have your dealer perform the Copy modification. Units no.794 and above, as well as pre-794 unis that have received the dealer Copy modification, accept the industry-standard ac-coupled synch inputs'.
Unfortunately, two of my four units are below this serial number, and two units are above, without any indication if the earier units have received this modification, and I don't want to damage something without determining this first.The main point here is that the 700s do have the SYNC clock capability, this is good news, if I can lock the clocks to a PC and then as you have indicated strip/rip the data to a hard drive / burner, then I might have something niffty to archive with. However I'm not there yet..;)
Keith, I've been racking my brain all day trying to understand paragraph seven of your post though...'Now back to the DBX units'.
It seems as though your saying here that I would need to record video to a VCR/VTR unit then decode to a second decoder to send the raw digital signal to the PC. If the digital signal, Encode & Decode are available at the port, without the NTSC video format process applied, would this not be the signal to 'Burn' to a disc? I understand that the 700s are not simultaneous Encoders/Decoders, but with four units, I was hoping that utilizing the port to a PC, that I might be able to have 8 channels one way or 4 channels simultanously. I think that I've miss understood what your saying, the signal flow you were discribing, can you please give a flow chart description?I do know a computer geek that just might have an idea of how to record the raw data, I'll give him a call. And of course I'll let you know what results I get. If you think a pdf file of the manual would help I could post on a web page for your viewing as well. Thanks again for your input on the 'MAD' idea....;)
Rick
Rick,
In paragraph seven, I was continuing on the mastering theme, specifically with the need for two units. *-If-* you could somehow strip and re-construct the data, you'd still need only the one unit to record each pair of two channels, and the same one unit to read the data back. -Thinking about it, it's by no means certain that there's a "read-back" input on the mastering delay connector... -It's possible that the delay device might also contain the second (identical) converter circuit... -this would remove the need to buy a second encode/decoder, with all it's 'wasted' redundant input circuitry...To know the answer to that question, you'd need to know more about the delay option, or have information on the pin-out of the interface connector.
I would be curious to know what the nearest integer multiple of the NTSC line frequency (15.xxxxkHz, instead of the much lower frame rate of 29.97Hz) to the sample rate/bit rate would be... this might give a clue as to how they're synchronising the whole affair...
About this point any knowledge I might have is already shared... from here on, it would be detective work, conjecture and guesswork!
I know that what you're looking into would be beyond my abilities, but even failed efforts usually turn out to be quite educational! -If you're not the type of person to get too frustrated with a lack of progress (and you have plenty of time on your hands for a 'project'!) then by all means investigate away! -If you're not known for patience and have a calendar with more ink than white spaces on it, then perhaps it's time to clear a spot on a shelf in the pro-audio museum for the DBX's ;-)
Good luck... -if by some stroke of genius you get anything to store & retrieve, I'd be delighted to know about it!!!
Keith
Keith,
Ok, on paragraph seven, got it, that's what I thought, but wanted clarification, thanks. I agree with you that there may not be a 'read back' on the port, but I've been refreshing my self on the use of the break-out box to see what sex this little beasty is and determine the hand-shaking involed. I'll approach it as a puzzle, it should be fun.
Using the 700s with the video ins & outs is the way it was intended to be used and that maybe the best way to utilize them.
I'm now investigating the use of DDRs, (Digital Disk Recorders) which claim a frame by frame lossless accuracy. I'm just learning about them, seem to have different forms of compression and even no compression formats. Depending on their cost it might be a way to go, but dunno yet.Thanks for all your input, you are a asset to this forum and appreciate your involement. I will post results in the future.
Rick
Just figured out that 29.97 (frames per sec.) x 525 (lines per frame) = 15.734.25 kHz... -NTSC line frequency -unless someone can correct me on that- I hope that's correct. (Never was keen on video, and coming from Britain, I was just getting the hang of PAL...!!!)Keith
Simple answer no. You still need a video recorder.Regards
Roland
This post is made possible by the generous support of people like you and our sponsors: