First try at ultrasonic cleaning

[yewneek]

1954 pressing, 1923 recording, new(ish) Technics SL1200 MK7, brand-new unused properly aligned AudioTechnica VM95ML. Silicone mats iterated to give least static with prior (new) VM95e. RCA Cables measured for lowest capacitance, chosen from a selection of 10+.
Assume the TT is performant as good as it can be for the money I have to spend.
This upgrade form the Rega is topping out at £1000, before recouping costs from sales.

TT Input into Cambridge Alva Solo, then into NAD C372, then into Tascam DR-05x, recorded at 96kHz 24bit, manual levels. Cables are iterated good, connections are iterated good, practice is optimal.
Cleaning solution proportions suggested from Monsieur ChatGPT: DDI, IPA and Fotospeed.

Before:

Dropbox

www.dropbox.com

After:

Dropbox

www.dropbox.com

That was really an experiment, before I get onto the pricey rare stuff. Proof of concept and equipment, satisfactory.

Thoughts welcome etc.

GB

 

[krabapple]

Is your jpg supposed to be showing before/after? Surely not?

 

[DVinylfan]

There's clearly a reduction in surface noise and a perceived volume increase in the after but the remaining noise level (damaged disc) and speed/pitch instability (I guess from the original recording?) renders this unlistenable for me.

 

[krabapple]

The top jpg looks like it was 'selected' (highlighted) so it's much brighter overall than the bottom.

 

[Timcognito]

My experience with Vevor ultrasonic cleaner and similar solution with Photoflow. My unit lets me do 5 at a time. Every record improved none where damaged. Improvement varied, guessing because of what they were contaminated with. 5 each, three sets, first bath, rinse, dry, 5 each, three sets second bath, rinse, dry, new sleeve >15 lps.

 

[yewneek]

The top jpg looks like it was 'selected' (highlighted) so it's much brighter overall than the bottom.

I'm a professional journalist, and such an accusation against a professional would render any chance of employment, zero. It's a slur. I hope it wsan't intended as such. This was a first try, it's in the title. It's empirical. I haven't touched anything.

 

[MatrixS2000]

I digitize all my records, what I have found is there is minimal, if any surface noise reduction but there can be some click/pop reduction. I had a quick listen to your files, ultrasonic cleaning will not correct that. You need to look into software to help you.

I downloaded your after track and ran the first song through my restoration process, you can hear it here. I used Izotope RX and Ozone for this example. It may not be perfect, but I did it in about 5 minutes.

Is your jpg supposed to be showing before/after? Surely not?

Agreed, that is too much of a change considering only an US cleaning.

Here is the spectrogram after my changes:

 

[MatrixS2000]

oh and FYI - excellent record cleaning document.

 

[yewneek]

My experience with Vevor ultrasonic cleaner and similar solution with Photoflow. My unit lets me do 5 at a time. Every record improved none where damaged. Improvement varied, guessing because of what they were contaminated with. 5 each, three sets, first bath, rinse, dry, 5 each, three sets second bath, rinse, dry, new sleeve >15 lps.

I've got some assignments to do but when I'm back from them I'll have some more mounts on the way so I can industrialise it to your level of multiples. Usually, on a tryout, the experience oscillates somewhat, whereas, this was a little more successful than I'd hoped. Just by clicking the spectrogram setting in Audacity which I'd downloaded only a minute before using, it appeared quite immediately with NO post production. I recorded the WAV file, I dragged it to the laptop via USB cable, and then dragged it into Audacity.
Journalistically, this is regarded as primary source. It's what I posted to dropbox.

 

[yewneek]

I digitize all my records, what I have found is there is minimal, if any surface noise reduction but there can be some click/pop reduction. I had a quick listen to your files, ultrasonic cleaning will not correct that. You need to look into software to help you.

I downloaded your after track and ran the first song through my restoration process, you can hear it here. I used Izotope RX and Ozone for this example. It may not be perfect, but I did it in about 5 minutes.


Agreed, that is too much of a change considering only an US cleaning.

Here is the spectrogram after my changes:

View attachment 536869

Recorded digitally, then imported, untouched, where a spectrogram was generated. I have never used that facility before, I found it by clicking in about ten seconds. If there's any forensic file investigation done, it will read with full CRC hash as identical.

I'll take the one you did as an education, thanks!

 

[yewneek]

There's clearly a reduction in surface noise and a perceived volume increase in the after but the remaining noise level (damaged disc) and speed/pitch instability (I guess from the original recording?) renders this unlistenable for me.

It's a 1954 pressing of a 1923 recording, right at the bleeding edge of when 33rpm was new technology(early '50s), recorded on technology that was before electronic recording (1925 advent).
Sorry if your ears are bleeding, I was hoping for it to be extreme, by intent.

 

[Timcognito]

Full disclosure, I stopped buying and playing records and all analysis was listening after cleaning four years ago. Subjectively improvements were very obvious on the majority of lps targeting thrift store finds and old favorites. For those who want stay under a $250 VEVOR sells systems that have all the fixtures and drying rack to do mutiple along with Ultrasound bath and their claim of 8-10 per batch was never met but 5 worked great.

 

[yewneek]

Full disclosure, I stopped buying and playing records and all analysis was listening after cleaning four years ago. Subjectively improvements were very obvious on the majority of lps targeting thrift store finds and old favorites. For those who want stay under a $250 VEVOR sells systems that have all the fixtures and drying rack to do mutiple along with Ultrasound bath and their claim of 8-10 per batch was never met but 5 worked great.

It's rare for me to buy new, I tend to hit used record emporiums in antique centres, and there's one up the road from me in a rural village. I live in a cow shed.

The idea is to do the 500 or so albums I have, then after that madness, to do the occasional ones when I get them from trawling bins of Jazz.

I like the idea of buying the packaged ultrasonic thing from Vevor, but I really like the idea of buying a handmade labour of love from a dude who's retired, who hand-cuts extra label covers to order.
The result is pretty much identical, yeah it's more polished, has a lifty thing, but I've already purchased!

https://www.ebay.co.uk/itm/306982314177

 

[MatrixS2000]

Recorded digitally, then imported, untouched, where a spectrogram was generated. I have never used that facility before, I found it by clicking in about ten seconds.

Understood, you most likely had the entire track selected in the first capture and unselected in the second capture. That is why there is a sharp contrast in the spectrograms.

Here are the waveform stats for your “before" capture:

And here they are for your “after" capture:

If you compare the intergrated loudness measurements for each, they are basically the same, meaning the loudness is basically the same before and after the US cleaning. If you look at the min. RMS levels, you can also see that the there is really no reduction in the noise floor that results from the US cleaning.

Now if I add the waveform stats from my single song restoration. You will see that the noise floor has dropped, which is expected due to the de-noise process I put it through.

PS - you should also try to record a little lower, you came close to clipping your ADC in the “before", and you did in the “after", when that happens its unrecoverable. You can always increase the volume in the digital domain.

 

[yewneek]

Thank you! The digital recording wasn't the purpose, it was a vehicle to document and demonstrate.

However, I have the workflow now with your helpful assistance and analysis, which is much appreciated.

The intent is to merely clean, learn how to clean, industrialise the cleaning and then enjoy the result of the industry.

That guide is goign to be invaluable reading too.

Cheers

GB

 

[krabapple]

I'm a professional journalist, and such an accusation against a professional would render any chance of employment, zero. It's a slur. I hope it wsan't intended as such. This was a first try, it's in the title. It's empirical. I haven't touched anything.

I'm a scientist, and I've messed around with recording software for decades now.

Your jpg looks like the top stereo track was 'selected' , as in , made currently active -- which highlights the track , and could easily have been accidental -- and the bottom one not.

And I listened to to your two samples. The difference simply doesn't sound as radical as your jpg.

Here is a random track on my laptop that I loaded into Audacity. Twice. In other words, it's the same exact audio, shown twice. There is literally no difference between the stereo track on top, and the one on bottom.
I tap the left side of the top one, and it lights up.
Look familiar?

 

[yewneek]

I'm a scientist, and I've messed around with recording software for decades now.

Your jpg looks like the top stereo track was 'selected' , as in , made currently active -- which highlilghts the track , and could easily have been accidental -- and the bottom one not.

And I listned to to your two samples. The difference simply doesn 't sound as radical as your jpg.

Here is a random track on my laptiop that I loaded into Audacity. Twice. In other words, it's the same exact audio, shown twice. There is literally no difference between the stereo track on top, and the one on bottom.
I tap the left side of the top one, and it lights up.
Look familiar?

View attachment 536928

My undergrad studies were in lab-based science too, then a couple of postgrads, including teaching and journalism. Rigor, empirically, and historically and legally, and visually, are my bag. So, we both understand epistemological discipline, as is with most of the users on this site!

Thanks for highlighting the highlighting! I made a similar rookie error using a GCMS back in the day, on a sample which was used as a control for a double-blind study. Oops! It was repeated with oversight, and error corrected!

 

[krabapple]

Yet your first response was to label it a 'slur'.

 

[JP]

 

[krabapple]

Bottom is before, top is after?