The farmer and his plow - When I set out to build a record cutter

[markrob]

Hi Thomas,

Great work! Did you write your code using floating point calculations? If so, you can really speed things up if you use integer math. Takes a bit of work, but in a time critical situation like you have here, its worth the effort. The Arduino has no floating point hardware, so using 32 bit integers is much faster. If you need help with the coding, let me know and I'll see if I can give you some tips.

If you really need floats, then maybe look into the Teensy. Uses the same Arduino IDE so its easy to get up to speed with. Its a 600 Mhz ARM processor with an FPU.

https://www.sparkfun.com/teensy

Mark

[farmersplow]

great job....fyi. it's easier to just route the main signal into another channel and apply delay there digitally, then there is no need for playing back 2x2 channels and when you change the main signal level the pre-view will follow.

Thank you for the suggestion! Originally I wanted to go exactly this way, but I didn't know how. My plan was also to make it easier, but my knowledge with the DAW is limited. I don't know how to route a signal to another channel and then delay it digitally. I work with Cakewalk - maybe this DAW can't do that or I haven't found anything about it. Maybe you can give me a tip on how to do this. I should be able to use this for several song titles (LP with 6 songs) and set different peak levels for each one. If it's easier to do this via loop and delay, that would be great.

Thomas

[farmersplow]

Wow. I thought the speed changes would be very slight and difficult to discern but no. There's really quite a difference in speeds.

Quite amazing, isn't it? How strongly the groove spacing control reacts depends of course on the system and the effort you make to get the maximum out of it. I think I have exaggerated a lot here.

Thomas

[farmersplow]

Hi Thomas,

Great work! Did you write your code using floating point calculations? If so, you can really speed things up if you use integer math. Takes a bit of work, but in a time critical situation like you have here, its worth the effort. The Arduino has no floating point hardware, so using 32 bit integers is much faster. If you need help with the coding, let me know and I'll see if I can give you some tips.

If you really need floats, then maybe look into the Teensy. Uses the same Arduino IDE so its easy to get up to speed with. Its a 600 Mhz ARM processor with an FPU.

https://www.sparkfun.com/teensy

Mark

[farmersplow]

Hi Thomas,

Great work! Did you write your code using floating point calculations? If so, you can really speed things up if you use integer math. Takes a bit of work, but in a time critical situation like you have here, its worth the effort. The Arduino has no floating point hardware, so using 32 bit integers is much faster. If you need help with the coding, let me know and I'll see if I can give you some tips.

If you really need floats, then maybe look into the Teensy. Uses the same Arduino IDE so its easy to get up to speed with. Its a 600 Mhz ARM processor with an FPU.

https://www.sparkfun.com/teensy

Mark

Thank you Mark! You're right, you won't get far with floating point calculations. I already had problems with the “normal” groove control, especially when I did some tests with 78 rpm and wide groove spacing. At that time, I didn't use floating point calculations. I then had to make extensive changes to the software. If I hadn't done these tests back then, I would have had problems now at the latest and your tip would have saved me. I didn't want to switch to the Teensy yet because it has too few connections and I used them all up (a small excuse, because I could have avoided that too). In truth, I was too lazy to do that.

But now I had to go back to a Teensy 4.0 because I'm working on the turntable drive control. For the necessary FOC control, I not only need an encoder with 1024 steps, but also a very fast magnetic field measurement. The Arduino is completely overwhelmed by this.
I have a short video where I do the encoder measurement with the Teensy 4.0.
.
.

Video.png

20250528_190017.mp4

[/size]
.
.
In fact, I still have problems with the Teensy because not everything is the same with the Arduino. If I need some help with this, I would be happy to get in touch with you.
Thomas

[misjah]

great job....fyi. it's easier to just route the main signal into another channel and apply delay there digitally, then there is no need for playing back 2x2 channels and when you change the main signal level the pre-view will follow.

Thank you for the suggestion! Originally I wanted to go exactly this way, but I didn't know how. My plan was also to make it easier, but my knowledge with the DAW is limited. I don't know how to route a signal to another channel and then delay it digitally. I work with Cakewalk - maybe this DAW can't do that or I haven't found anything about it. Maybe you can give me a tip on how to do this. I should be able to use this for several song titles (LP with 6 songs) and set different peak levels for each one. If it's easier to do this via loop and delay, that would be great.

Thomas

i never used cakewalk so can't help you there but most DAWs can simply route 1 output to another input internally...if not, you can always go physically out (digital and back in with AES cable or whatnot)

[mratx]

In Cakewalk you want to send the channel with your recording to a bus, and then apply the delay to the bus. In console view, you need to create a new bus, then, in the send section in the middle of the console channel for your track, set the send to go to that bus. Change the output for the channel (located at the bottom of its console channel) to one set of your audio interface outputs. Then set up a single delay in the effects section of the bus (middle of the bus console channel), and set the output of the bus to another set of the audio interface channels. The delay has to be 100% wet. You'll then get the original signal in one set, and the delayed signal in another.

This is assuming you're starting with a stereo recording on one channel. If you have two mono channels, you have to do the process for each channel, and each channel's output would be a single audio interface output.


Mark

[farmersplow]

Thank you very much for your help! I know my weaknesses and one of them is using the DWA and mastering. An area in which you are very good.
Two keys to improving the processes are “Routing” and “BUS”. I will try as Mark suggested and let you know how it goes. Many thanks in the meantime

Thomas

[farmersplow]

In Cakewalk you want to send the channel with your recording to a bus, and then apply the delay to the bus. In console view, you need to create a new bus, then, in the send section in the middle of the console channel for your track, set the send to go to that bus. Change the output for the channel (located at the bottom of its console channel) to one set of your audio interface outputs. Then set up a single delay in the effects section of the bus (middle of the bus console channel), and set the output of the bus to another set of the audio interface channels. The delay has to be 100% wet. You'll then get the original signal in one set, and the delayed signal in another.

This is assuming you're starting with a stereo recording on one channel. If you have two mono channels, you have to do the process for each channel, and each channel's output would be a single audio interface output.
Mark

In Cakewalk I have now managed to send the channel to a bus. I then set the delay there and the EQ for the output signal. I then sent both signals (input signal and bus signal) to the two audio interface channels. Everything has worked so far.
There was only one thing I couldn't do. If I increase the volume at the bus output, I can't establish a connection, so the volume level at the input also increases. I also tried using two bus-to-bus connections, but it didn't work. It is easily possible that if I increase the input level, the output level will also increase, but vice versa? There is also no tutorial describing how to change the volume level from the bus back to the input level. It only ever mentions that there is a signal flow chart, from top to bottom and then from left to right. But that would be important, because otherwise I have to leave it as it was. Mark, maybe you can tell me something about this?

Thanks in any case!

[farmersplow]

Automatic, variable groove spacing control - Part 4 (last part):

I would like to show some more information about my progress.

The advantages of variable groove control can be seen in the following pictures. Here I have cut the song “I will always love you” by Whitney Houston. The song lasts 4min 34sec. With a fixed groove spacing I can only cut 4min 20sec (if I record at slightly above 0dB). The last 14sec of the song no longer fit on the single (The song cuts right into the automatic run-out groove). It looks like this:
.
.

A10.jpg

.
.
With variable groove control, a lot of space is saved on this song. Everyone knows the song and knows that there are quiet parts, especially at the beginning and end, where there is only voice without music. You save a lot of space there in particular. In this case, the same recording looks like this. The whole song fits easily on the single and there is still space left over:
.
.

A20.jpg

.
.
For your information: With extreme techno music, where there is a lot of bass and volume from start to finish, no space is saved.

Greetings from Austria
Thomas

[farmersplow]

And this is what the groove looks like under the microscope:

Fixed groove spacing:
.
.

A30.JPG

.
.

A40.JPG

.
.
Variable groove spacing:
.
.

A50.JPG

.
.

A60.JPG

.
.
Video showing the variable groove spacing:
.
.
VIDEO 1:

A80_Video_compr.mp4

.
.
VIDEO2:

A70_video_compr.mp4

.
.

Greetings from Austria
Thomas

[farmersplow]

Cutting music again at last

I still had to fix a few signal problems, which I will write about in the next reports. After I had finally fixed the interference, I was able to make a music cut again.
An old friend from school visited me in Austria again. He now lives near San Francisco (California) and is currently studying music. So he was able to help me with the mastering. Thank you Bernhard.

Here is an excerpt from the Whitney Houston cutting. In my opinion, very difficult to cut because there are many very quiet spots that show every mistake of the setting and the material.
I think I managed to get a good shot.

TESTCUT: Whitney Houston - I will always love You

45rpm 7" +1dB (5cm/sec) no post-processing!
.
.

cutted whitney houston.mp3

.
.
Greetings from Austria
Thomas

[spinnertownblanks]

We are in the studio mastering this afternoon, look forward to hearing it!

[zdenek]

Cutting music again at last

I still had to fix a few signal problems, which I will write about in the next reports. After I had finally fixed the interference, I was able to make a music cut again.
An old friend from school visited me in Austria again. He now lives near San Francisco (California) and is currently studying music. So he was able to help me with the mastering. Thank you Bernhard.

Here is an excerpt from the Whitney Houston cutting. In my opinion, very difficult to cut because there are many very quiet spots that show every mistake of the setting and the material.
I think I managed to get a good shot.

TESTCUT: Whitney Houston - I will always love You

45rpm 7" +1dB (5cm/sec) no post-processing!
.
.
cutted whitney houston.mp3
.
.
Greetings from Austria
Thomas

Hello, I listened to your recording and analyzed it using electronic devices.
A professionally recorded vinyl record sounds slightly different because the professional heads use feedback.
The difference between a digital recording and a professional recording is that you hear the SPECIFIC CRACK of a PROFESSIONALLY RECORDED VINYL RECORD, but what you actually hear is the FEEDBACK RESPONSE. That is, you hear an echo superimposed, mixed feedback signal on the line signal, with the line signal coming from the signal source.
This is what absolutely distinguishes a digital recording from a professional vinyl recording. The most interesting and amusing thing is that no correction device, studio equipment, or computer program today can imitate this echo coming from the cutting head. I think you understand.

[farmersplow]

Hello zdenek, to avoid misunderstandings, I would like to correct some of your statements.

A professionally recorded vinyl record sounds slightly different because the professional heads use feedback.

That has nothing to do with it. My cutting head works so well with feedback that an additional EQ to tame the resonances is almost unnecessary.
The main reason why professional (store-bought vinyl records) sound better is the material being cut into (lacquer) instead of PET-G plastic. I like to compare it to cutting into butter instead of brick.

The difference between a digital recording and a professional recording is that you hear the SPECIFIC CRACK of a PROFESSIONALLY RECORDED VINYL RECORD, but what you actually hear is the FEEDBACK RESPONSE. That is, you hear an echo superimposed, mixed feedback signal on the line signal, with the line signal coming from the signal source. This is what absolutely distinguishes a digital recording from a professional vinyl recording.

The ECHO referred to has nothing to do with professional cutting. It occurs when the spacing of the groove becomes too narrow. An average groove width is 40-70µm (20-35µm deep). If the groove spacing is 80µm, there is no problem because a “land” of 10-40µm remains. If music is recorded loudly, the groove needs more space and if the groove spacing is not increased sufficiently, the recorded sound “pushes” some of the “land” outwards onto the previous groove. As a result, some of the music is always shifted to the previous groove. During playback, you can already hear a record rotation beforehand, which will only occur after 1.8s (at 33.3rpm).
This is called ECHO!

If you choose a large, fixed groove spacing from the outset, the problem will not occur. Professional record cutters usually use variable groove control to put more music on the record. This can lead to this problem.
Even Neuman lathes had this problem. However, there is also a solution. Normally you don't hear the problem because it is hardly relevant compared to the loud music. But if it is very quiet and then suddenly gets louder, it can happen. Especially if the pre-signal control does not recognize all changes correctly.
This is why the Neuman turntables had an “ECHO” function. This could be used to prevent this problem. Just before it gets loud (1/2 plate revolution before) the ECHO function was activated, which forced a larger groove spacing.
I also recorded with variable groove control and also built an “ECHO” function into my lathe (via MIDI). It was a long road to get it all working like that. At the beginning of my attempts there were always problems until I finally fixed it.
Here is an old example when the variable groove control didn't work so well. The ECHO is very audible here.
.
.
FAIL-CUT: too little groove spacing (with bad variable groove spacing control – without “ECHO”-function)

fail cut WN.mp3

.
.
As I said, it has nothing to do with a lack of feedback (feedback cutting head). (Because it is actively in use).
Professionalism also has nothing to do with audible echo - on the contrary. Echo should be avoided and can only be heard in cheap productions.
But one thing is true and was clear to me from the start. If you cut in brick instead of butter, the cut won't be as clean. That's just the way it is. My aim was always to do it as well as possible.
(I could also cut in lacquer, but then I wouldn't listen to the record very often).

Greetings from Austria
Thomas

[displacedsnail]

Hello Thomas, outstanding work! That is an excellent digital rip, hard to believe you've been building this machine. Seriously, I have such a hard time conceptualizing machinery and electronics, to see your creativity - and that of the other trolls - is very impressive and inspiring. Bravo!

You've inspired me to take on the "Whitney Houston Challenge" (is this Lathe Trolls first meme challenge?!) myself and try the same section of song that you cut, both on my mid-fi Presto 6D and on my lo-fi Rek-O-Kut Challenger. The records are embossed into PC and are in mono. I gave them the standard "starter plug-in pack" I give all my cuts, but didn't work on them beyond that. I'm also attaching a picture of my ROK, for no reason other than I like the way it looks, ha!

Very much looking forward to hearing more of your records and seeing more of your wunderbare maschine,

Joe

[farmersplow]

Hello Thomas, outstanding work! That is an excellent digital rip, hard to believe you've been building this machine. Seriously, I have such a hard time conceptualizing machinery and electronics, to see your creativity - and that of the other trolls - is very impressive and inspiring. Bravo!

You've inspired me to take on the "Whitney Houston Challenge" (is this Lathe Trolls first meme challenge?!) myself and try the same section of song that you cut, both on my mid-fi Presto 6D and on my lo-fi Rek-O-Kut Challenger. The records are embossed into PC and are in mono. I gave them the standard "starter plug-in pack" I give all my cuts, but didn't work on them beyond that. I'm also attaching a picture of my ROK, for no reason other than I like the way it looks, ha!

Very much looking forward to hearing more of your records and seeing more of your wunderbare maschine,

Joe

Wow Joe! I think that's cool! Inspiring someone to record the same song. I really like the idea of the meme challenge. Stereo always has a small advantage over mono for our ears because it seems more spatial and fuller. (Just like louder sounds better than quieter).
But your mono recordings are also really good. Presto in particular is great. With the ROC you notice that something is missing from 4kHz onwards, but it's still very accurate up to that point. These old boxes don't just have charm, they're also really good at it!
I'll show you more of my recordings and my lathe soon - I promise.

[zdenek]

Hello zdenek, to avoid misunderstandings, I would like to correct some of your statements.

A professionally recorded vinyl record sounds slightly different because the professional heads use feedback.

That has nothing to do with it. My cutting head works so well with feedback that an additional EQ to tame the resonances is almost unnecessary.
The main reason why professional (store-bought vinyl records) sound better is the material being cut into (lacquer) instead of PET-G plastic. I like to compare it to cutting into butter instead of brick.

The difference between a digital recording and a professional recording is that you hear the SPECIFIC CRACK of a PROFESSIONALLY RECORDED VINYL RECORD, but what you actually hear is the FEEDBACK RESPONSE. That is, you hear an echo superimposed, mixed feedback signal on the line signal, with the line signal coming from the signal source. This is what absolutely distinguishes a digital recording from a professional vinyl recording.

The ECHO referred to has nothing to do with professional cutting. It occurs when the spacing of the groove becomes too narrow. An average groove width is 40-70µm (20-35µm deep). If the groove spacing is 80µm, there is no problem because a “land” of 10-40µm remains. If music is recorded loudly, the groove needs more space and if the groove spacing is not increased sufficiently, the recorded sound “pushes” some of the “land” outwards onto the previous groove. As a result, some of the music is always shifted to the previous groove. During playback, you can already hear a record rotation beforehand, which will only occur after 1.8s (at 33.3rpm).
This is called ECHO!

If you choose a large, fixed groove spacing from the outset, the problem will not occur. Professional record cutters usually use variable groove control to put more music on the record. This can lead to this problem.
Even Neuman lathes had this problem. However, there is also a solution. Normally you don't hear the problem because it is hardly relevant compared to the loud music. But if it is very quiet and then suddenly gets louder, it can happen. Especially if the pre-signal control does not recognize all changes correctly.
This is why the Neuman turntables had an “ECHO” function. This could be used to prevent this problem. Just before it gets loud (1/2 plate revolution before) the ECHO function was activated, which forced a larger groove spacing.
I also recorded with variable groove control and also built an “ECHO” function into my lathe (via MIDI). It was a long road to get it all working like that. At the beginning of my attempts there were always problems until I finally fixed it.
Here is an old example when the variable groove control didn't work so well. The ECHO is very audible here.
.
.
FAIL-CUT: too little groove spacing (with bad variable groove spacing control – without “ECHO”-function)

fail cut WN.mp3
.
.
As I said, it has nothing to do with a lack of feedback (feedback cutting head). (Because it is actively in use).
Professionalism also has nothing to do with audible echo - on the contrary. Echo should be avoided and can only be heard in cheap productions.
But one thing is true and was clear to me from the start. If you cut in brick instead of butter, the cut won't be as clean. That's just the way it is. My aim was always to do it as well as possible.
(I could also cut in lacquer, but then I wouldn't listen to the record very often).

Greetings from Austria
Thomas

I'm not talking about echo, the crossover of sound from a track to an adjacent track in the case of variable pitch with groove spacing correction and variable signal amplitude.
I'm talking about something completely different.
A head with feedback enabled sounds completely different from the same head with feedback disabled. That's what I'm talking about. I don't know if you understand this. The echo of the feedback signal from the feedback coils, the NEGATIVE, REVERSED PHASE feedback signal, penetrates the source signal, giving a KICK IN THE ASS to the power amplifiers, which then deliver a KICK IN THE ASS to the head, i.e., the drive coil. There's no way to achieve professional-quality recordings with a head without feedback, which has no feedback, but everyone should consider why this is so.

If I have time, I will make a recording for YouTube directly with the same head with feedback turned on, closed loop and I will turn off the feedback during recording and then you will hear a DIAMETRAL difference.
I was afraid to turn it on and close the feedback loop, but I took the risk and I can still hear STEREO with the clone head with the feedback turned on. If I turned the feedback off, the recording would be of poor quality, very weak, believe me, I checked it.
https://www.youtube.com/watch?v=8BG8Y26mu_k

[farmersplow]

I'm not talking about echo, the crossover of sound from a track to an adjacent track in the case of variable pitch with groove spacing correction and variable signal amplitude.
I'm talking about something completely different.
A head with feedback enabled sounds completely different from the same head with feedback disabled. That's what I'm talking about. I don't know if you understand this. The echo of the feedback signal from the feedback coils, the NEGATIVE, REVERSED PHASE feedback signal, penetrates the source signal, giving a KICK IN THE ASS to the power amplifiers, which then deliver a KICK IN THE ASS to the head, i.e., the drive coil. There's no way to achieve professional-quality recordings with a head without feedback, which has no feedback, but everyone should consider why this is so.

If I have time, I will make a recording for YouTube directly with the same head with feedback turned on, closed loop and I will turn off the feedback during recording and then you will hear a DIAMETRAL difference.
I was afraid to turn it on and close the feedback loop, but I took the risk and I can still hear STEREO with the clone head with the feedback turned on. If I turned the feedback off, the recording would be of poor quality, very weak, believe me, I checked it.
https://www.youtube.com/watch?v=8BG8Y26mu_k

Fortunately, I built a cutting head with feedback and also closed the loop.

If you simply turn off the feedback during recording, then of course there will be a lot of degradation because the resonance correction will turn off! You can of course also run a feedback head in open loop (without feedback compensation). But then you have to perform a very complex EQ correction, which should also be adjusted depending on the cutting depth, recording volume and music track. I myself have been using my head in open loop mode for a long time and have been able to achieve good results, albeit always with very complex EQ characteristics.
And FB has nothing to do with stereo. If your head can cut stereo, then it can do it with and without FB.
So either you work with a closed loop, or you create a complex EQ curve. But simply switching off the FB (without EQ adjustment) makes no sense. (Believe me)

[zdenek]

I'm not talking about echo, the crossover of sound from a track to an adjacent track in the case of variable pitch with groove spacing correction and variable signal amplitude.
I'm talking about something completely different.
A head with feedback enabled sounds completely different from the same head with feedback disabled. That's what I'm talking about. I don't know if you understand this. The echo of the feedback signal from the feedback coils, the NEGATIVE, REVERSED PHASE feedback signal, penetrates the source signal, giving a KICK IN THE ASS to the power amplifiers, which then deliver a KICK IN THE ASS to the head, i.e., the drive coil. There's no way to achieve professional-quality recordings with a head without feedback, which has no feedback, but everyone should consider why this is so.

If I have time, I will make a recording for YouTube directly with the same head with feedback turned on, closed loop and I will turn off the feedback during recording and then you will hear a DIAMETRAL difference.
I was afraid to turn it on and close the feedback loop, but I took the risk and I can still hear STEREO with the clone head with the feedback turned on. If I turned the feedback off, the recording would be of poor quality, very weak, believe me, I checked it.
https://www.youtube.com/watch?v=8BG8Y26mu_k

Fortunately, I built a cutting head with feedback and also closed the loop.

If you simply turn off the feedback during recording, then of course there will be a lot of degradation because the resonance correction will turn off! You can of course also run a feedback head in open loop (without feedback compensation). But then you have to perform a very complex EQ correction, which should also be adjusted depending on the cutting depth, recording volume and music track. I myself have been using my head in open loop mode for a long time and have been able to achieve good results, albeit always with very complex EQ characteristics.
And FB has nothing to do with stereo. If your head can cut stereo, then it can do it with and without FB.
So either you work with a closed loop, or you create a complex EQ curve. But simply switching off the FB (without EQ adjustment) makes no sense. (Believe me)

I've tested it and I know this for sure. It's impossible to accurately record the difference between the left and right channel signals (i.e., the stereo effect, depth of space) without precise feedback from the feedback coils. Just as it's impossible to record a vinyl record correctly at the appropriate level and with the full acoustic range without feedback, it's impossible, and that's all there is to it. I've read a lot of nonsense and thought it was possible, but unfortunately, it's not. Then I had to have it explained to me by a SPECIALIST, a MASTER who has "cut his teeth" on these devices and proved to me that for PROFESSIONAL VINYL RECORDING, the head MUST HAVE PRECISE AND VERY STRONG FEEDBACK. Ask Master Flo from Switzerland, because I can't explain it to you. Best regards. And it costs a lot of money [at least for me it's $20,000 for a professional head]