I'm a DJ. I've been collecting vinyl for over 20 years and I'm ready to convert since the vast majority of my events nowadays don't require vinyl (or my hard-earned ability to scratch and cut).
Here's my dilemma: after a six month project of narrowing my collection and ridding myself of all but the records I either love to listen to or would actually use at an event, I'm still left with roughly 3500 records. No matter what hardware/software devices I've found, they all suffer from one fatal flaw: they convert in real time. That means for *every* record I want to convert, I'd spend between 10 minutes (assuming time spent labelling and editing the .mp3) and 2 hours depending on the length of the record.
I'm trying to find out if there is a vinyl scanner out there. I've seen Vertical record cleaners and Laser turntables. I've also found an article where a guy used a scanner (because of the size of the bed vs the size of the record, he had to scan in 4 parts and stitch the images together) and was able to use it to scan records.
What I haven't found is someone combining these two ideas: the vertical form factor of the record cleaner with a scanner (and perhaps a small motor to rotate the record?) Does that already exist or should I find a programmer and run to the patent office?
Here's my dilemma: after a six month project of narrowing my collection and ridding myself of all but the records I either love to listen to or would actually use at an event, I'm still left with roughly 3500 records. No matter what hardware/software devices I've found, they all suffer from one fatal flaw: they convert in real time. That means for *every* record I want to convert, I'd spend between 10 minutes (assuming time spent labelling and editing the .mp3) and 2 hours depending on the length of the record.
I'm trying to find out if there is a vinyl scanner out there. I've seen Vertical record cleaners and Laser turntables. I've also found an article where a guy used a scanner (because of the size of the bed vs the size of the record, he had to scan in 4 parts and stitch the images together) and was able to use it to scan records.
What I haven't found is someone combining these two ideas: the vertical form factor of the record cleaner with a scanner (and perhaps a small motor to rotate the record?) Does that already exist or should I find a programmer and run to the patent office?
ideal vs reality
Scanning the LP, even in the ideal way you describe, would still take time, just as it does for pages of text and photos. I imagine an application would be needed to convert the scan into an audio file. Then there would be the task of editing the files to break them up into individual tunes, to set the overall audio level, and to deal with defects such as ticks and pops and scratches.
I regularly copy LPs to CDs. I'm not an expert but I try to produce a clean recording. Procedure involves cleaning LP on cleaning machine, recording into computer, editing saved files, burning CD, scanning cover and making up cover art and play list for the CD and printing the cover. All this takes me three or four hours per LP, varying greatly with the amount of time devoted to fixing defects.
Minimum time for cleaning LP, scanning LP with your ideal scanner and editing sound file I think would still be an hour per LP even if you had software which automated such things as separating tunes according to breaks between cuts. Estimating an hour a day to scan or to listen to an LP you have at least ten years of scanning ahead of you no matter what. Myself I transfer LPs to CDs for other people but prefer listening to the LPs.
Scanning the LP, even in the ideal way you describe, would still take time, just as it does for pages of text and photos. I imagine an application would be needed to convert the scan into an audio file. Then there would be the task of editing the files to break them up into individual tunes, to set the overall audio level, and to deal with defects such as ticks and pops and scratches.
I regularly copy LPs to CDs. I'm not an expert but I try to produce a clean recording. Procedure involves cleaning LP on cleaning machine, recording into computer, editing saved files, burning CD, scanning cover and making up cover art and play list for the CD and printing the cover. All this takes me three or four hours per LP, varying greatly with the amount of time devoted to fixing defects.
Minimum time for cleaning LP, scanning LP with your ideal scanner and editing sound file I think would still be an hour per LP even if you had software which automated such things as separating tunes according to breaks between cuts. Estimating an hour a day to scan or to listen to an LP you have at least ten years of scanning ahead of you no matter what. Myself I transfer LPs to CDs for other people but prefer listening to the LPs.
You can't successfully scan an LP with any existing optical scanner. Some of the fine detail in the grooves has an amplitude less than the wavelengths of visible light. The ELPJ laser turntable does it by measuring the phase of the reflected beam from the groove wall relative to the transmitted beam, allowing sub-wavelength resolution.
In theory, because it doesn't physically touch the groove, the ELPJ turntable could run at several times "real speed", just like a CDROM drive reading an audio CD.
Or you could develop a scanner that scans a line from centre to edge of the record, as the record rotates slowly under the beam. I'll leave it to you to work out the wavelength of light (extreme ultraviolet?) and the resolution of the sensor required.
In theory, because it doesn't physically touch the groove, the ELPJ turntable could run at several times "real speed", just like a CDROM drive reading an audio CD.
Or you could develop a scanner that scans a line from centre to edge of the record, as the record rotates slowly under the beam. I'll leave it to you to work out the wavelength of light (extreme ultraviolet?) and the resolution of the sensor required.
He's also going to go through up to 10 styli for his cartridge in the process of playing 3500 LPs...
If one must record in real time there's the possibility of playing the record at higher speed. An LP could be recorded at 45 or 78 rpm to save time and the speed returned to normal by the software. I never have done this but just now I opened a wav file of a three minute song in Audacity and tried to change the speed from 33 to 78. The window said the processing time would take about 7 or 8 minutes, so I didn't bother to let it finish. My computer is new but not a high powered one so perhaps the processing time could be reduced. Still, there seemed little advantage in recording at higher speed.
Re-reading Nehesi's post I see that his goals may not require high quality or recording a whole LP, but the time required seems daunting all the same.
Quicker than a scan would be snapshots with a high resolution digital camera or maybe a video done of one revolution of an LP.
-Terry C
Re-reading Nehesi's post I see that his goals may not require high quality or recording a whole LP, but the time required seems daunting all the same.
Quicker than a scan would be snapshots with a high resolution digital camera or maybe a video done of one revolution of an LP.
-Terry C
In principle this works, but has practial problems arising: When youu transpose a recording made at 78rpm back to 33rpm, you also transpose the frequency spectrum by the same scaling factor. Suppose your phono cartridge (or A/D card) has a frequency response up to 20kHz... after speed correction of the high-speed recording, you will end up with a maximum frequency response of 20kHz * 33/78 = 8.46khz.
Additionally, playing a disc at 78rpm will wear the stylus considerably faster than at 33rpm.
What I Can suggest you is Continue tu use your Records, but when you use one in your work you send the signal from that turntable to a recording software on the PC. You play The Song Completely and save it in FLAC Format.
Each time you need to use an old record....Do the same.
While one is playing you can remove clics and pop and archive the final product at the same time! (on Pre-recorder Vinyl Song)
Regards
Marc
Each time you need to use an old record....Do the same.
While one is playing you can remove clics and pop and archive the final product at the same time! (on Pre-recorder Vinyl Song)
Regards
Marc
Also the factor of the EQ curve being totally wrong when played at the wrong speed.
I would think that trying to play the record at a higer speed would be problematical on a few points.
The cartridge would mistrack on the high frequencies , unless it was an exotic MC.
The frequency response would be incorrect due to the RIAA eq.
The Recorded audio would need to be recorded at an high sample rate.
One point I would like to make is when digitising Vinyl is that for best results avoid actually playing the audio through speakers, so that there is no accoustic feedback into the record deck & pre-amp.
The cartridge would mistrack on the high frequencies , unless it was an exotic MC.
The frequency response would be incorrect due to the RIAA eq.
The Recorded audio would need to be recorded at an high sample rate.
One point I would like to make is when digitising Vinyl is that for best results avoid actually playing the audio through speakers, so that there is no accoustic feedback into the record deck & pre-amp.
With no intent to be argumentative I point out that old surround sound quadrophonic systems employed carrier frequencies of 30khz - 45 khz on the LPs, so cartridges could track them even if sound cards can't handle them. Supposedly greater wear of stylus from higher playing speed is still proportional to greater number of LPs done in shorter time.
Then there were the half-speed mastered LPs by Mobile Fidelity. Regarded as superior quality, did they employ eq technics to accommodate speed change? Also, high speed tape duplication is a common practice, especially with audio cassette tapes. According to one web site typical speeds are 16x, doing a C90 cassette in 3 minutes.
Then there were the half-speed mastered LPs by Mobile Fidelity. Regarded as superior quality, did they employ eq technics to accommodate speed change? Also, high speed tape duplication is a common practice, especially with audio cassette tapes. According to one web site typical speeds are 16x, doing a C90 cassette in 3 minutes.
Quad (CD-4) systems did use a HF carrier. But even special CD-4 cartridges were not linear up to those frequencies. Assuming you had an LP with content up to 20 KHz, and played it at 2x speed, you would in theory capture frequencies up to 40 KHz. But after downsampling the result to 20 KHz again, you would hear significant distortion in the 10 to 20 KHz range, manifesting itself as unnatural sibilants, spits and pops(*). This distortion did not affect the playback of CD-4 LPs, because the encoding used (FM based) was insensitive to the level variations and resonances.
(*) Been there, tried that, back in the late 80's:
- CD-4 cartridge (JVC?)
- Lenco variable speed turntable set to 66.6 RPM
- Homebuilt phono preamp with modified RIAA time constants
- Ampex RtR tape recorder (record at 30 ips, playback at 15 ips)
Sound cards which support 24 bit / 192 KHz sampling rates are relatively cheap and common these days. You buy them in musical instrument stores, not computer stores.
So the problem is still the playback of the LP at high speed, not any subsequent stage.
(*) Been there, tried that, back in the late 80's:
- CD-4 cartridge (JVC?)
- Lenco variable speed turntable set to 66.6 RPM
- Homebuilt phono preamp with modified RIAA time constants
- Ampex RtR tape recorder (record at 30 ips, playback at 15 ips)
Sound cards which support 24 bit / 192 KHz sampling rates are relatively cheap and common these days. You buy them in musical instrument stores, not computer stores.
So the problem is still the playback of the LP at high speed, not any subsequent stage.
High speed tape dubbing is a totally different ball game as there is no complex EQ curve which gets screwed up with speed. Also high speed dubbing is one of the reasons why bought cassettes always sounded worse than recording your own on even a fairly basic hifi.
High quality actually *is* one of my goals: To make it worth the effort, it would need to equal or surpass the results currently obtained through USB mixers or other Analog to Digital converters. I like your camera idea (assuming the software could be created to convert the video) but wouldn't the camera be subject to the same issues with reflectivity of the record and angularity of the track that current scanner technology would have? I'm a DJ not a programmer nor an engineer, hence my original post here to bounce the idea off you guys.
The applications, with the exception of one that would convert the scan into audio, are already out there. I currently use Polderbits - www.polderbits.com - which already has the ability to automatically add breaks, clean up the audio, limit time recorded, etc.
I'm aware that I could simply buy CDs of the stuff that's commercially available. I guesstimate that 60 - 80% of the stuff I have I could find. For a few of them I'd have to convert anyway because the commercially available versions suck/are missing stuff (i.e "La-Di-Da-Di" by Slick Rick/Doug E Fresh is missing 9 measures in the middle of the song where Slick Rick is singing) or good luck on finding the long version of "Heartbeat" by Taana Gardner online anywhere other than Limewire, where quality isnt regulated.
Part of the reason I'm here, though, is that I know there are other DJs with the same logistics issue I'm running into: lots of records, not a lot of time and a desire to avoid paying reams of cash to replace songs they already purchased. Its not just DJs who have this issue, but prety much anybody with an extensive vinyl collection who isnt a hardcore collector (who would consider the ELP turntable at $12,000) and hasn't already replaced them all with CDs. I just wish I had had this idea 10 years ago when more people still had vinyl.
@AOC - I'm using Technics 1200s, which are fairly well isolated to begin with (when I was first starting out as a DJ I did several events with them right on top of the speakers) and they're physically removed from the speakers in my music room. I haven't had an issue with the files I've converted in the past, but I'll take that into account and use headphones to monitor the output.
I just tried a pitch convert in Audition 3 to convert from 78 to 33. My 'reference' file for process time is Don McLeans American Pie at 8:37. Process time was a little under 39 seconds for a 'resample'. But given the mechanical parameters of a phono stylus, I would never copy at higher speed to save time. For example, the Shure V15 type Y tuned the stylus to 33KHz and when combined with the electrical response of the cartridge coils yields flat response. The electrical response changes don't bother me at all but the mechanical response is totally out of your control. Trackability numbers go out the window when you more than double the play speed. I would expect the final output to be poor at best. The most effective way to speed up the process is a second or third turntable and 'multi process'.
Personally I'd skip MP3 and go to FLAC. Current price for a 2 T drive is $125 and I'd save the 'raw' unprocessed files in case you want to re-process later. Treat them like a photographic negative.
G²
I'm surprised the suggestion to play the LPs faster for recording generated such response, especially when it was apparent from the start that the hassles of coping with the speed change offset the time saved. I learned things from the comments and appreciate the information.
Yes, most likely. I got the impression you were interested in saving time doing the transfers. Instead of solving the issues for a slow scan why not solve the issues for a quick photo? I really don't know any more about it than from general information available. I'm not an engineer. You're already way more advanced in the business than I. Thanks for the link to polderbits, but I'm using linux.Nehesi said:
A few links to the US Library of Congress' IRENE project -
IRENE Sound Restoration
ScienceMatters @ Berkeley.
IRENE Sound Restoration
ScienceMatters @ Berkeley.
Always monitor with headphones, or the speaker level turned way down. I've demonstrated this party trick quite a few times:
Connect a turntable / preamp to a recorder.
Place the stylus on an LP on the turntable, but do not spin it.
Play music through your speakers from another source, while recording the signal from the turntable.
Now play back the recording.
The results can be surprisingly loud.
Did anyone suggest a (gasp!) record changer? Drop a stack on there, hit play, and start recording with Audacity. Ideally, arrange something to stop or pause when done. Say, a scrapped USB keyboard or gamepad plus a microswitch or optical sensor (light barrier), with Girder if necessary to turn a "fire" click into a "pause" command. That might be pretty slick. Play through a stack. Flip over, start playing, and Audacity would unpause again. Divide into albums and tracks later.
Or, accept that this is a long process, put together a convenient setup (dedicated deck and PC) and do a few every day.
Another thought: PCs capable of audio recording are essentially free. Turntables can still be bought pretty cheaply. So, use multiple turntables and PCs. Build phono preamps as required, or use USB interfaces that have phone preamps built-in.
Or, accept that this is a long process, put together a convenient setup (dedicated deck and PC) and do a few every day.
Another thought: PCs capable of audio recording are essentially free. Turntables can still be bought pretty cheaply. So, use multiple turntables and PCs. Build phono preamps as required, or use USB interfaces that have phone preamps built-in.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.