1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

Large scale lp digitization Help needed

Discussion in 'Recording' started by lumendog, Nov 14, 2009.

  1. lumendog

    lumendog Guest

    Hello, everyone.
    I am a first timer at this forum, but have logged many hours at other forums under the name name.

    I am in the process of organizing a very large scale lp digitization effort.
    Eventually using up to 50 Technics 1200 tables running a good portion of the day.
    My initial thought was a 1200 with separate phono pre that has usb into a computer running audacity. This would mean 50 separate preamps as well as the Tables.
    An initial problem has occured that audacity can only record from one table at a time and will crash if two programs are running simultaneously.

    I need some professional help. I need hardware and software recommendations. I do not need state of the art mixing capabilities.
    Can I run several tables into a mixer or something that will record the files simultaneously? With what software?
    Thanks for any help in advance as this scale is different then my usual setup.
    I do not want to by 50 netbooks. :?
     
  2. BRH

    BRH Active Member

    Even if you did do it your way, how will you separate and name those files from the different LPs?

    I no longer give free advice.
     
  3. Kapt.Krunch

    Kapt.Krunch Well-Known Member

    Let's see...

    50 turntables. Stereo. 100 A/D converters to get them into a computer.

    Do you have access to 100 channels of interface?

    Do you have a computer that can handle 100 channels of interface?

    50 turntables. Spread out over a large area, obviously. A record on each one. Create and record-enable as many recording tracks in the software as it will allow that you have inputs for. Manually go through and check and set the recording levels for each LP's tracks, since they are all likely to be different. Start the recording in the software on stereo track one, manually drop the tonearm. Move to the next turntable, drop the tonearm. That recording starts later in the overall project. Move to the next one, drop the tonearm, it starts even later. And on, and on, for as many as you have. By the time you get them all going, you'll have many tracks of staggered start and end times, and recording a lot of nothing at the beginning and end of subsequent tracks. (Recording silence is still recording, and eats disk space and resources.) Also, by the time you go around to the 50th turntable, the first side of LP one should be nearly finished, or maybe even long finished. You can turn that one over, and start it. Just hope that that LP 3 isn't shorter in time than LP 2, LP 8 isn't shorter than LP 7, etc., because it'll mess up your carefully-planned sequence. Or, you can just let it sit there and record nothing for those periods to continue to tax your computer and take up disk space.

    If (when) the program eventually locks up or crashes, no problem. Just do some computer/software tweaks, and try again. Then, try again. And, again. Eventually, you'll learn how many you can actually do at once, and stack the other 35-40 turntables in the corner, out of the way. Now, you can start for real!

    After all that, and you finally get around to getting all those tracks recorded, all you have to do is separate them all into their own stereo tracks to work on individually. Hope you kept track of all the tracks, and named them first.

    You can start by chopping off the extra recorded nothing at the beginning, end, and in the middle (where you flipped the LP) of all of them, but leave a BIT at the beginning so you can create an FFT, or similar file, in case you need it for denoising purposes. Denoising takes a lot of computer power, so you'll really only want to do one stereo track at a time. And, you may have to experiment a bit with each album's track to set just the right amount of denoising parameters, since they'll all be different.

    Uh...where was I? Oh yeah, uh..don't forget that you'll eventually need to monitor each and every entire LP side to ensure that no skips or anything took place, since you really can't rely on a waveform view to catch everything. So, plan to take some time to listen to each, individually. You want to do that BEFORE you do any further processing. Otherwise, you'll have wasted time doing processing for a file you can't use, and you apparently are in a big hurry to get everything done as quickly as possible.

    Wait..it just hit me! Uhhhh...D'OH! If it were me, I might think about just doing the normal thing, and monitoring each LP while recording it one at a time! That way, I'll KNOW I have enough computer horsepower to not crap out...I'll KNOW there were no skips, I'll be able to keep track of things much easier without running around dropping tonearms, flipping albums, and hoping that everything plays all the way through, and keeping track of all of it.

    Hmmmm......I wish I had thought of that before going to all that other trouble and confusion.

    Kapt.Krunch :?
     
  4. Thomas W. Bethel

    Thomas W. Bethel Well-Known Member

    Dare I ask what this is for????

    This is NOT the way to do what you are asking about.

    Transferring records to computers is not done en masse because you cannot listen to all of the records at the same time and if one starts skipping or repeating how are you going to know about it? Also if you are serious about doing this you won't be doing this with a freebie program, a bunch of turntables and some audio to USB converters.

    We do a lot of vinyl to CD transfers and restoration. We start with a $1500 turntable, a $500 arm, a $200 cartridge and a $600 preamp. This goes into an $1800 ADC and is recorded at 88.1 at 24 bits. We record this in Wavelab and do any noise reduction and cleanup using RX Pro from Izotope and then burn it onto T-Y CDs with a Plextor burner.

    IMO I don't think you have the background or the knowledge to do what you are suggesting you want to do. It may be better to let some other company who has the equipment and the knowledge to do this correctly do the transfers for you or your client. There are lots of places on the WWW that do this type of thing daily and have the background to do it well and with professional results.

    IMHO this is NOT something that a person should be attempting without the knowledge or equipment to do it properly.

    This is not a flame but a simple statement of fact!

    Best of luck but do some serious thinking about what you are about to do!
     
  5. JoeH

    JoeH Well-Known Member

    My post was going to be what Tom wrote, almost verbatim.

    It sounds almost like a render-farm for movie animation. ;-)

    Do you have that many LPs that you need to go at it 50 copies at a time? Can't you hire someone (an apprentice or two or three?) to just come in, spend the day(s) transferring things a side at a time?

    All of the above posts are correct; LPs are not things you can just drop the needle down and go - unless it's completely useless crap - stuff that you wouldn't be bothering to transfer to digital anyway.

    The BEST you could hope for would be an 8, 16 or 24 track recorder running tracks in stereo pairs; you'd have to name them ahead of time, cue them up together, end them together, perhaps in 20 minute chunks (per side running time). You'd have a lot of wasted data, though, if some of them run short. Plus, the time you'd spend going back over each one to cleanup, trim, etc., would have been just as well spent on the front end, doing them correctly from the start.

    There's no free lunch, no shortcuts, and one way or another, the biggest cost in a job like this is TIME. Assuming you DO have 50 turntables and pre's, converters, etc. do you really have that many LPs that would justify this kind of expense?

    Not only that, what will you do with them all after the project is over?

    I'd say spend the $$ on a couple of good workstations (even laptops), get a half dozen converters, pre's and turntables, and hire a handful of temps. Spend a day or two teaching them your methods and quality control, and have it.

    Once the project is over, you won't have so many leftover TT's to flood Ebay with. :-?
     
  6. lumendog

    lumendog Guest

    Thanks for all of the replies.
    We will have plenty of people, so manpower is not the concern. Time is.
    Hence the multiple Turntables. We will be running them 8 or so hours a day so we have chosen the Technics 1200 and are settling on cartridges and preamps. This gear will be bought from a nationwide storefront and sold back to them at the end, so no need for ebay on this.
    We are a non profit, and so therefore do not have funds to hire this out.
    The albums are a gift and comprise over 100,000 lp's.
    Right now we are running the tables we have into audacity via usb phono preamps. The problem is that audacity crashes if two recordings are simultaneous. So either we need 50 computers or run virtual desktops. We have tried windows, ubuntu, and os operating systems with audacity and all fail when two windows are in use.
    My question is; Is there a better software that can handle the multiple recording on one computer ( we were thinking 6 Tables per computer). Also is there better hardware then going a single preamp to usb to do this multiple recording.

    I will admit to all that I am not a recording pro. That is why I am asking for your help. I am not trying to anger anyone, but really need some direction. But please, comments like BGH gave do not help anyone.

    :cry: :cry: :cry: :-?
     
  7. BRH

    BRH Active Member

    Haste makes waste!

    OK, I have a constructive crit. but will not explain the reasoning.

    Forget about Audacity!
    Forget about USB phono preamps!
    Forget multiple turnables into an audio recording program!

    Use one turntable per computer with quality phono preamp and hardware interface.
    Get a few volunteers that can work as interns for school credit over a semester or two.
    Have them monitor/name files/make logs as the audio as it's being recorded.

    If any of the discs are availble on CD, through them in the trash.
     
  8. RedStache

    RedStache Active Member

    Wow! Come on, folks; never say never!

    This does sound like a daunting task to say the least. If I understand you correctly, you are a non-profit organization with 100,000 LPs to digitize in a timely fashion. While I can't say I really understand what you are working on, I do understand very well the idea of trying to do something fantastic with resources that are, shall we say, less than fantastic. I get it that you are non-profit and trying to make the most from the least.

    Let me throw this out there and see what comes of it. Bear in mind that I really don't know anything except that I will not stop until I accomplish whatever wild idea comes to mind and I am more than willing to go unconventional, which is what you're thinking about already.

    By the way, what sort of quality level do you require? Do you need everything to be perfect? It doesn't sound like it, but I'm sure you need things to sound good enough. What sort of computers do you plan on using? These answers can help others answer your question for real. Putting your requirements in the terms of, "I need to record 100 tracks at once; how do I do that?", will also help get you some better answers. (That probably sounds mean and I really don't mean to. I really want to help! :) )

    So next, it's all about the converters-you must have one for each analog track and each stereo turntable has two, and most recording programs that I know about will only accept one interface set-up coming in at a time. It sounds like you are asking your PC to run two recording programs at a time, which I don't think any PC can do. That comes down to the core audio required out of the OS to handle the incoming audio, regardless of the recording program or number of 'windows' you want to run at once. One core audio per one PC equals one audio input set. I don't think you can do the 50 USB converters under those requirements, but there are options.

    Again, let me stress that I don't know much here, and an area of special weakness for me is the turntable/preamp thing. I do, however, wish to make a suggestion.

    I think you will probably need more of a program, and I think you will need a unified interface. The program is probably not that big of a deal cost-wise, but the interface, well that's a different story. By unified interface, I mean one or a multiple of one that will come into the computer as one. One option is to ADAT/lightpipe from one to another, but I know you are adding cost with ADAT.

    Here's a clip from Presonus:
    "The FireStudio Project can be daisy-chained with one or more FireStudio Projects or FireStudio Tubes to expand your number of recording inputs. Two and three units daisy-chained in any combination has been tested at PreSonus, however, the total number of units that can be daisy-chained depends on your computer’s processor speed, amount of RAM and computer configuration."
    Found at: http://www.presonus.com/ That is saying to me that presonus thinks you can add tracks 'til the cows come home, but I know there are limits. For one thing, computer hardware and software capabilities will come into play. I think that it is reasonable to think 24 tracks per recording computer/software combination, I wonder if anyone on the forum has done more. So with this idea you're looking at each Firestudio costs at least $400. You need one for every 4 turntables. That is a total of 13 interfaces in order to be able to record 50 turntables at once. 24 tracks per computer covers 12 turntables per computer. That adds up to 5 computers to cover 50 tracks at once. Add the software licensing and the personnel, and you probably have the most realistic requirements list so far. I don't know what sort of resources you actually have, but I think you are looking at this kind of requirement list if you want to get it done quickly and without paying for the service. You must check to make sure the software will take all those tracks at once, and somehow I don't think Audacity will do it. I don't know about that. So 24 tracks per computer via 3 Firestudios gives you 12 turntables in at once. You can name the tracks before recording and maybe figure out how to get volunteers to listen for major hickups during the recording. Of course we are skipping the USB converters for each and every pre and feeding line inputs into the Firestudios instead. And, of course you will still face the incredibly awesome task of doing something with the tracks once you have it all digital. Wow! You ARE talking about a major task here! I don't know, but maybe this is the kind of course you need to think of. Key points are: 1. One program won't handle multiple interfaces at one time. 2. One PC will only run one recording program instance at a time. 3. 'Daisy-chained' interfaces with multi-track capabilities are needed. 4. A program capable of handling 24 tracks or more is needed. 5. To do 50 turntables at once will require 100 tracks (almost certainly more than one computer can handle but not as much as 50 interfaces will require). Hopefully this helps you on your way to get this done. I wish I could help, this sounds like quite a feat your attempting. Let me know how it turns out. Rock On!!! Robert
     
  9. Thomas W. Bethel

    Thomas W. Bethel Well-Known Member

    I guess the major question is what are you going to do with all the materials after you have transferred them? We are talking about 100,000 records X 40 minutes average playing time for both sides or 66666.67 hours or 2777.8 days or 7.6 years worth of music. Depending on how you do this that is a lot of hard disk space. That is also a lot of room for errors to creep in especially keeping all those recordings documented with the correct track name and what record they came off. If you use volunteer help the problem is going to be magnified.

    I think Joe's idea is the best but if you did it his way it is a project that could literally take years to complete. At 8 hours per day you could transfer about 12 records per work station per day barring unforeseen problems. So to do 100,000 would take you about 8333 days and with ten work station would take you about 833 days which is more than 2 years working 8 hours per day 365 day per year which is probably not realistic. Also the wear and tear on the equipment is going to be a BIG factor. Styli only last so long and even the 1200s and the cartridges are going to wear out eventually probably well before they can be resold. Computer hard disks are good but even the best used that heavily will probably start to fail sooner than you think and so you should have a backup of all the hard disks.

    I think you have one king sized problem that is only going to be solved with some money for the equipment and its maintenance and lots of GOOD volunteer labor.

    Is it possible to determine what are the most important recordings and only do those and not try and digitize all of them en masse?

    Could you approach some large company like Microsoft with your problem and ask for them to supply the equipment and some money for replacement parts? At least that would take the money part out of the equation.

    Again best of luck!
     
  10. BRH

    BRH Active Member

    [/quote]

    Is it possible to determine what are the most important recordings and only do those and not try and digitize all of them en masse?
    [/quote]

    Now you are talking like a professional archivist.
     
  11. Cucco

    Cucco Distinguished Member

    I'm sorry...
    I have to cry foul here.

    This can't be legit or even real. It sounds like a school project of some sorts.

    First, I don't think there are 100,000 LPs that exist that are public domain. And if you're archiving 100,000 LPs, I have to ask - What are they of? My initial thought was perhaps some church that archived all of its sermons or similar, but it makes my head hurt thinking how many Sundays that would be!

    On top of that - I know of NO store that stocks 50 Technics turntables and certainly NONE that would buy them back after that much work has been done on them!

    Another thing - you say you can't afford to hire this out. You can't afford to do this yourself! Trust me. Do the math.

    For 50 stereo inputs, you're going to need a LOT more than Audacity, some USB preamps and a single computer.

    If I absolutely had to tackle your problem, I would probably do the following:

    Different turntable (x80 - the extras are to account for downtime during breakages) - one that has a built-in phono preamp (pioneer and sony both make inexpensive ones. Let's face it, if 100,000 copies are the order of the day, you can't be overly concerned with that last bit of quality - these turntables will do fine).

    SSL A/D Converters (x4) (24 channels per)

    RME Fireface (x4)

    Computer (x4)

    Any DAW capable of recording 12 stereo tracks at a time (most).

    The problem is, you're now talking about an initial investment of over $30,000 with a cost-per-hour of the duplication itself. It's ridiculous.

    Reproducing 1,000 LPs can a be a single person's life-long endeavor. Increasing that by 2 orders of magnitude is simply ridiculous.

    Please provide more details about what this is for and if this is legit.

    Until then, I consider this to be a hoax and encourage others to treat it the same.

    Cheers-
    J
     
  12. pmolsonmus

    pmolsonmus Well-Known Member

    Cucco stole my thunder. A real understanding of copyright law is probably in order here.
    Who and what are these recordings of?

    Non-profit doesn't mean its ok to skate around the law.

    From the Public Domain Website:

    "There are no recordings in the public domain in the USA"


    http://www.pdinfo.com/index.php
     
  13. Thomas W. Bethel

    Thomas W. Bethel Well-Known Member

    This forum is not the only one that lumendog posts on.

    http://www.vinylengine.com/phpBB2/viewtopic.php?p=162187

    I cannot imagine a project like this ever being started or completed. If it was done by a pro sound company and done correctly it would probably cost somewhere around 3 million dollars just for the time (@$45.00 per hour) not to mention all the equipment and spares and take years to complete.

    Also the amount of hard disk space would be enormous especially if these were done using AIFFs or WAV Files. Even if they were all converted to MP3s the hard disk space would still be in the multiple terabyte range. All these transfers would have to be cataloged and maintaining the data base would be a full time job for a group of people.

    I too question where there are 100,000 records that "need" to be transferred and but unless the OP tells us more we will never know what these records are.

    The OP is from the Bay Area according to his posting on the vinylengine website.

    To the OP. I think all of us are more than willing to give you some good advice but this whole project is just a bit hard to grasp and or understand and so can you tell us what this is really about??? and PLEASE go into more details.

    Thanks!
     
  14. Codemonkey

    Codemonkey Well-Known Member

    I bet it's only 100 records; somebody forgot to carry a 1.
     
  15. Davedog

    Davedog Distinguished Member

    There are trolls amoungst us. Do the math.


    BTW. I have a new project I'm going to start any day now. I'm borrowing 100,000 monkeys from various animal service organizations and I need the name and model number of the typewriters that will properly input into windows (in English) the assuredly complete works of Shakespere.
     
  16. Boswell

    Boswell Moderator Distinguished Member

    Nah, it's a game of Chinese Whispers. There's actually a database with 100,000 records that needs to be transferred to another computer.
     
  17. Codemonkey

    Codemonkey Well-Known Member

    Pfft, Davedog... monkeys ain't that smart. You'd need 100,000 years and that means a lot of dead monkeys => more monkeys. More monkeys, always more!

    Boswell, LOL. Could copy that amount of information in about three seconds (six if it's an MS Access file).
     
  18. Thomas W. Bethel

    Thomas W. Bethel Well-Known Member

    Until the OP tells us what this is all about it is all conjecture.
     

Share This Page