PSA: PC "Line" inputs inadequate for 1v RMS line-level signals. Simple test.

[10PSPMemSticksForMiku2007]

Hello everyone. I have a pile of a relative's cassettes, videotapes and vinyls that need to be digitized. I am recording from the RCA output of my Yamaha KX-390 deck to the line-in of my motherboard, ASRock Z370 Extreme4. This motherboard has good sound output (to my ear) and the line-in seems fine unless it clips.
I thought I heard clipping on some of these cassette transfers. There was no digital clipping in Audacity as recording levels were set sanely, but it occurred to me the signal from my deck's RCA jacks may be too hot for the line-in. The transfer was much better with an attenuated signal coming from the deck's headphone jack. Easy fix for now, but I may have problems when I need to record from a VCR.
This strikes me as a problem I shouldn't have. Isn't this blue jack on my PC designed for recording from line-level sources? It kind of fails at its one job. I decided to measure how much it can take, and also to see if my old laptop could do better.
Computer USB - > ODAC B 90% volume -> O2 headphone amp -> Y-cable -> Computer Line-in
Y cable -> male to male 3.5mm cable -> multimeter
Using RightMark's pre-test screen, I adjusted the amp knob and PC input volume until I got a clean signal as loud as possible. If I turned the amp knob any higher it would result in a "highly distorted" signal, but without clipping digitally.
At the same level, I then played a maximum volume sine wave in Audacity to measure the maximum voltage that corresponds to these volume settings. This voltage was much higher than what I read during the pre-test screen. The test did complete successfully at these settings.
The two computers performed very differently. If you want to record stuff using your pc's built in codec, it seems you have no choice but to measure what each computer is capable of and make sure your sources are not too hot for it.
Any comments on improving my testing would be welcome. I am not sure I did this properly. I mean, I'm sure that I can't feed a genuine 2v source into my pc and expect good results, but I'm not sure I found out the maximum levels I CAN record properly. (Edit - RightMark's test tones are not full scale, so the actual voltage accepted by these soundcards is even worse than I thought).
I really don't want to buy an interface for this project. I think I'll use my O2 to attenuate the source and give myself a good measure of safety. Its price would be too much for a volume knob but hey, I already have it.
Also, recommendations for cheap USBs that can record line level properly would be appreciated. While I probably won't buy it, I'm sure many people have digitization projects they would rather not waste their time with. I have one of those cheap composite to usb things coming. When it arrives I may test that for the hell of it.

 

[JSmith]

Yamaha KX-390 deck to the line-in of my motherboard, ASRock Z370 Extreme4

The RCA outputs on the KX-390 are fixed-level outputs;

Output: 0.57V (line)

https://www.hifiengine.com/manual_library/yamaha/kx-390.shtml

JSmith

 

[restorer-john]

The RCA outputs on the KX-390 are fixed-level outputs;

No, not fixed level outputs. Where di you get that idea from?

It depends on the recorded level on the tape itself.

 

[Blumlein 88]

Set Audacity recording level to 100%. If you don't get digital clipping it should be fine. There is some voltage level which will be 0 db FS in Audacity set this way. If you find you are getting clipping, you will need to reduce the voltage level going into the sound card input jack.

The output of that Yamaha is specified as .57 volts on the line out. Could be the distortion was on the tape.

 

[10PSPMemSticksForMiku2007]

Set Audacity recording level to 100%. If you don't get digital clipping it should be fine.

I definitely get digital clipping if I set Audacity's recording to 100% or even 65% when recording from the RCA jacks. In fact I will have to redo some tapes I tried transferring at 65%. I wish it was better explained what the computer line-in jacks do. It would seem the record level boosts the signal beyond its original digital value, at least on some motherboards. On the macbook, the best SNR I got was with the level at 0%. That one is in line with the mic preamp.

The output of that Yamaha is specified as .57 volts on the line out. Could be the distortion was on the tape.

Thank you for this information. I'm going to do another thorough cleaning of my cassette deck to try and improve it further. Maybe I will try to measure the voltage coming out with the loudest section of tape I could find.

I think the main thrust of this thread still stands, especially if you have a laptop. The Mac clips at .6 volts, making it unsuitable to record even the .57 nominal line level of this tape deck, and definitely no good for something that actually puts out 2 volts.

It depends on the recorded level on the tape itself.

I think some of my tapes are putting out double the voltage of others.

 

[kemmler3D]

Isn't this blue jack on my PC designed for recording from line-level sources?

Maybe. What are the specs they offer for the input and ADC? If any?

If you want to record stuff using your pc's built in codec, it seems you have no choice but to measure what each computer is capable of and make sure your sources are not too hot for it.

Agreed. On ASR it's commonly repeated that DACs are a solved problem and ADCs are also generally pretty good. This is true. However, I think this community underestimates how crappy the analog portions of these devices can be in practice. I've experienced some really audibly awful performance from built-in audio on laptops and motherboards.

My advice is get the cheapest "real" (read: outboard USB) audio interface you can, e.g. a secondhand focusrite scarlett. This will give you a lot more headroom on the input and probably better quality (lower noise, anyway and probably lower distortion) overall.

 

[wwenze]

Most well-built soundcards should have 2V inputs from personal experience, but anything other than well-built soundcards is whatever. That said it is usually closer to 1V.

I use my Fiio headphone amp as a preamp, this helps to ensure full THD+N performance on the soundcard anyway.

DACs are a solved problem and ADCs are also generally pretty good

When you are specifically hunting for and buying one that solved the problem. The rest of the world including pro-audio home-theater seem to be convinced that 80dB SINAD is good enough and that's the target they are going with for computers, phones, AVR etc. When their budget is $0.35 I can't blame them. The USB-C w/DSP cable of the Tanchijm costs $2 more than the 3.5mm cable, like wat.

 

[Blumlein 88]

I see a couple options. I like the one kemmler3D suggested. A Focusrite Scarlett 2i2 or 2i4 2nd hand. Even a 1st or 2nd generation is fine. You see them on ebay for like $75.

Or use a passive volume controller to reduce the levels going into your PC soundcard. A Schiit SYS would do this for $49. Amir has even reviewed and tested one. Pretty simple device.

Review and Measurements of Schiit Sys Passive Pre-amp/Switcher

This is a measurement and review of the Schiit Sys switcher and passive pre-amp. It retails for just $49 plus shipping which is reasonable. The unit is on loan from a member so not new. I will do a teardown later but for now, there are just two inputs, one output, and a volume control in...

www.audiosciencereview.com

Schiit Audio: Audio Products Made in USA

www.schiit.com

The benefit of buying a Scarlett on ebay or Facebook marketplace is you'll have a better device for recording and when done you can sell it to get most or all of your money back. Here is one just listed for $70 and a 3rd generation model. Some cosmetic issues, but free shipping.

Focusrite Scarlett 2i2 2x2 Interface - 3rd Gen. for sale online | eBay

Find many great new & used options and get the best deals for Focusrite Scarlett 2i2 2x2 Interface - 3rd Gen. at the best online prices at eBay! Free shipping for many products!

www.ebay.com

 

[kemmler3D]

The benefit of buying a Scarlett on ebay or Facebook marketplace is you'll have a better device for recording and when done you can sell it to get most or all of your money back. Here is one just listed for $70 and a 3rd generation model. Some cosmetic issues, but free shipping.

QFT. I'm a big proponent of secondhand stuff and this is one of the big reasons. Sometimes I buy speakers or amps, use them for a while, and make a few bucks when I sell them again. Sometimes I lose a few bucks. Either way, whatever depreciation you get between new and used is already eaten by the original owner.

What you need to think about is total cost of ownership, not just sticker price, as long as you can afford to tie up some money in gear for a a while. Anyway, @10PSPMemSticksForMiku2007 you will not regret using proper recording equipment to do these transfers, I am sure.

 

[restorer-john]

Here is one just listed for $70 and a 3rd generation model.

That one is actually a 2nd gen. You can't trust ebay sellers...

 

[10PSPMemSticksForMiku2007]

Specifications sometimes are accurate but don't tell the whole story. Looks like the tape deck's specifications measured output at 0db. However with tapes you can get signal levels much higher and the VU meters show as much. With commercial tapes, especially classical music that needs dynamic range, you often get much stronger peaks.
To see how much the deck can output, I recorded a sine wave onto a blank tape at +8 and +0 on the VU. This time I used an RCA splitter between the right RCA jack of the tape deck and the PC.

I noticed clipping on this Mozart tape. Using the multimeter you can not measure music volume. The reported level is much lower. all you can do is match the level of a given sine wave to a segment on the VU. During the loudest moments I saw +6 several times. This roughly matches what I figured out by ear. Well, I'm glad I was not "hearing things."
It's good by the way. CD version.

Also, thank you for the gear suggestions Kemmler3D and Blumlein 88. I normally buy used equiment as well. If I ever need an interface I'll keep an eye on those. For now, I don't need one for the tape deck, as I can measure and mark exactly where I need the headphone out to be. I probably also won't need one for the record player. I plan on using the DJ Art preamp that Amir reviewed. If .306 volts is really as loud as it gets, I won't need to attenuate it. That solves most of my problem already. The sound quality of my line-in jack at the correct level will do no harm to these sources, I think.

I mainly want to make people aware of this issue that surely affects millions of computers all over the world.

 

[kemmler3D]

Specifications sometimes are accurate but don't tell the whole story. Looks like the tape deck's specifications measured output at 0db. However with tapes you can get signal levels much higher and the VU meters show as much. With commercial tapes, especially classical music that needs dynamic range, you often get much stronger peaks.
To see how much the deck can output, I recorded a sine wave onto a blank tape at +8 and +0 on the VU. This time I used an RCA splitter between the right RCA jack of the tape deck and the PC.

I noticed clipping on this Mozart tape. Using the multimeter you can not measure music volume. The reported level is much lower. all you can do is match the level of a given sine wave to a segment on the VU. During the loudest moments I saw +6 several times. This roughly matches what I figured out by ear. Well, I'm glad I was not "hearing things."
It's good by the way. CD version.

View attachment 386395

Also, thank you for the gear suggestions Kemmler3D and Blumlein 88. I normally buy used equiment as well. If I ever need an interface I'll keep an eye on those. For now, I don't need one for the tape deck, as I can measure and mark exactly where I need the headphone out to be. I probably also won't need one for the record player. I plan on using the DJ Art preamp that Amir reviewed. If .306 volts is really as loud as it gets, I won't need to attenuate it. That solves most of my problem already. The sound quality of my line-in jack at the correct level will do no harm to these sources, I think.

I mainly want to make people aware of this issue that surely affects millions of computers all over the world.

This is great info and it's very helpful that you are sharing it here! Thanks for the contribution.

I would say that if you are keen to get these tapes captured very cleanly, you should still consider upgrading the interface... I see you're prepared to get the most from this gear, but I might expect to see high-ish THD even if you're not hard-clipping the input. If you just need a decent but not perfect copy this is probably fine.

Either way nice work running down the voltages.

 

[10PSPMemSticksForMiku2007]

Here's a small but useful piece of information I for those recording with the line input of their computer. Sorry I forgot to mention it earlier, but many people may not know. In the windows sound control panel applet, there is a slider for your input level with a percentage display. Percentage of what? Who knows. It varies by each audio device and possibly driver version. This percentage is useless and meaningless unless you have a db number to compare to. There is no visual indication of this, but you can right click the slider to select between percentage and db. You must set it to 0db. Any other value can cause problems with your recording.
On my Realtek ALC1220 codec, the +0db is equivalent to 62%. Not 0, 100, or 50, but 62. Who knows how they came up with this. The default value of 100% is +12db!
For my macbook 1,1, I do not feel like booting it to take a screenshot, but I hope you'll trust me that +0db is the leftmost value or 0%.

By the way, I'm really not worried about adding noise or distortion to these recordings. The quality of the Realtek ADC may be mediocre in digital terms but I have no doubt it's 20db better in every specification than my analog sources. The tapes and records themselves are pretty worn out as well, or should I say they are well-loved. Maybe I will post a few samples later.