Topic: Vintage Cassette Deck Question

[Gutbucket]

No output that is not tied to the volume knobs?

The implication here is that if present, use of an alternate, fixed-level output which does not have level-control potentiometers in line with it is likely to offer the cleanest output with lowest distortion, making it a better option unless you actually needed to adjust output level to match the input of the ADC or whatever is downstream.

But if that's the only output option, it's a classic gain-staging question on how to set it - typically best wide open if its simply a stereo attenuation pot and the downstream gear can handle the hot output.  For best results clean the pots if they aren't sealed.  If they are, exercise them a bit and listen for scratchiness in the output, hopefully eliminated by the exercising/cleaning.  If some scratchiness remains, avoid the positions were it seems worst.

[Scooter123]

One does not know if setting it in the middle is a line output level, or if setting it at "10" (wide open) utilizes gain from the deck.  My Tascam Professional Cassette Deck has a similar knob and through experimentation, I generally set it at "5," straight up in the middle.  That sounds about right to my ears.  If the channels are different levels, I don't mess around with the cassette deck, nor the digital recorder (in my case a DR680) and set the digital recorder at "0" which in theory is line input, and save the tweeking for a DAW in post. 

[rocksuitcase]

I believe the OP's original question was answered with the post telling him manual says set it to 6. However, both scooter and GB mention important issues which I will take up to continue the basic premise of the thread:
what and how loud of a signal do we need to output from the cassette deck to a digital recording chain?
my answer in short is set it to maximum. Let it go out of the 40 year old vintage electronics as straight up as possible. (My earlier reply stated I use 9 typically.)
I would say for the OP's deck, since they found the manual reference saying 6 was 0dB or unity, then obviously this is the way to go.
when going into the digital chain, I have done it most of the options we have:
1] stereo analog input on a DR680 or PMD661 etc. setting the input gain per tape.
2] analog input on a ADC such as my Benchmark AD2k then that digital output goes into the digi in on  a bit bucket such as a DR680 or PMD661. with that method, the ADC sets the level and you do not need to set it on the bit bucket.
3] Analog input into a computer soundcard such as a Echo Gina, or more modern card. Generally you don't set the gain on the soundcard or PC, so max output on the Cassette deck.

again to the OP- amazing deck, refurbed as yours is, it is widely considered (on tapeheads.net) to be a top of the line deck for both playback and recording.

[Gutbucket]

Most potentiallty influential will be the condition of the tape transport (cleaned, demagnatized, aligned) and after that, if that output pot or the outpot jacks are dirty. The digital side is pretty acommodating, choose what you like. Remember, a perfect pro-cassette deck with a brand new tape is capable of about 9 bits of dynamic range. Most will ne closer to 6 bits!

[rocksuitcase]

Remember, a perfect pro-cassette deck with a brand new tape is capable of about 9 bits of dynamic range. Most will be closer to 6 bits!

While mathematically accurate, the reason to take off analog recordings into a 24/48 or 24/96 file is for editing. There are a LOT of scientific reports about each side of the discussion.

What follows is entirely MY opinion- also IMO, there is no right or wrong answer here, it is purely preferential for the archivist.
Intuitively, I go for the visual explanation. When editing on a DAW, the more bit depth you can actually see the details or artifacts you may be trying to reduce/remove much better.

https://www.tnt-audio.com/sorgenti/rip_it_2_e.html         for just one man's opinion ( I respect this dude for his work)

What sample rate to choose? While psycho-acoustics suggests that 44.1kHz is almost enough, and while there seldom is real signal content above 20kHz on LPs, it still is wise to record at a higher rate.

How high? Quadruple rates (i.e. 4 x 44.1 = 176.4kHz and 4 x 48 = 192kHz) may appeal to the numbers brigade, but are often quite pointless. One reason is the lack of content in the source signal, the other main reason is that many quad-capable ADC chips in reality aren't.  snip

 So back to 88.2kHz and 96kHz for recording LPs. One benefit of using 2x rate, even when targetting a CD, is that this way the ADC's anti-alias filter cuts at 44.1kHz or 48kHz, and not at 22 or 24kHz. ADC filters are part of the silicon chip, and thus are often made for economy rather than performance. Having them out of harm's way enables one to use a superior software-based resampler while later converting the data to 44.1kHz for release on disc.

Another objective reason for choosing 2x is that most digital post-processing just works better in a wider frequency space. Minimum-phase digital equalizer software, as used in Audition, operates as a transform of its analogue counterpart. This yields something similar as the equivalent analogue filter, but with a subtle difference: in the discrete sampled domain half the sample frequency takes on the same properties that an infinitely high frequency has in the continuous domain. As a result the digital filter response deviates progressively with frequency from its analogue target function. This deviation can be compensated for in software, a method called filter warping, but this is not always done. Hence it is safe to assume that filters may work better at an elevated sample rate.

Now of course, he is talking about LP recording to digital, but the math is the same. My archiving guru insists that 24.96 is the way to go for tapes. He buys into the math of 24-48-96 being the proper amount for post processing, especially because we are rarely these days downsampling the final output to disc. So a processed 24.96 file can easily downsample to 16.48 for storage.

[Gutbucket]

Agreed that for some processing, doing so at a higher sample rate is helpful.  However, I'm not convinced the best way to achieve that is to sample at a high rate to begin with.  I'm of the opinion that for audio that did not have useful info up there to begin with, sampling at a reasonable rate of, say 48kHz (16bits is plenty, 24 is harmless if you prefer, but no better), and up-sampling in the DAW for the processing - typically via the oversampling option built in to most processing tools and plugins is actually the better option.

Here's a youtube video from FabFilter which makes a compelling argument for this approach- https://youtu.be/-jCwIsT0X8M

It concludes:

27:36
So, lets try to form a conclusion.
28:04
The original proposition: “samplerates, the higher the better, right?” is clearly
28:09
wrong.
28:10
Higher samplerates do nothing to improve the quality of audible content below 20KHz, and
28:16
can result in lower quality due to increased intermodulation.
28:20
Higher samplerates only provide any benefit when using non linear processes that can add
28:24
extra harmonics above Nyquist.
28:26
But in this case its usually better to oversample each non linear stage individually, rather
28:31
than just increase the project samplerate.

[Gutbucket]

Excellent primer on digital audio here: https://youtu.be/cIQ9IXSUzuM

Posting that here (have posted it previously at TS) as it explains why 16bits is plenty for this type of application and higher bit depths do not increase "resolution" within the pass-band (pass-band is determined by sample rate, or more specifically, the ADC cutoff filter uses for the sample rate in question).  And, similar to the reasoning for up-sampling within each process as mentioned in my previous post, modern DAWs do all internal processing calculations in a 32bit floating point calculation space, regardless of the bit-depth of the original file.

BTW, this video is where I plucked that comment about the dynamic range of cassette being "9 bits at best, 6 more typical".

[rocksuitcase]

Clearly this is an "Ear of the beholder" dealio as there is no right or wrong.
However, this dude (excerpted below) thinks like I do. I feel no need to take off cassettes at 24/96. I use 24/48 as my preference.
I suppose it is possible the Vinyl has more "detail" than cassette, but I do not think it is that much more.

The original proposition: “samplerates, the higher the better, right?” is clearly
28:09
wrong.
28:10
Higher samplerates do nothing to improve the quality of audible content below 20KHz, and
28:16
can result in lower quality due to increased intermodulation.

From Werner O's cite I made above:

Yet another argument for using higher sampling rates is that any non-linear operation executed in the digital domain insiduously violates the sampling theorem: a non-linearity generates harmonics and intermodulation products, and some of these products exceed half the sampling rate and thus alias irrevocably back into the audible band. Avoiding this requires alias-aware processing software, which is rare. And so the second line of defense is, again, working at a higher sample rate. This is even advised when doing only seemingly-linear operations, such as gain change and equalization. Indeed, in the quantised domain no operation is truly linear: rounding and truncation due to limited numerical accuracy are non-linear in nature and the errors so injected will be multiplied by aliasing. This phenomenon is often not known or understood by designers and users of signal processing software alike, but in my opinion it may well be one of the main reasons why modern digital productions are often thought to sound bad, and also why high-res recordings subjective are preferred over standard resolution!

But you addressed this non-linearity in your cited quote:

Higher samplerates only provide any benefit when using non linear processes that can add
28:24
extra harmonics above Nyquist.

I'm betting live audience concert recordings have a certain non-linearity acoustically given location, mic choice, pattern etc. That linearity may not present itself electronically but it is there in the 3 dimensional cues (time of arrival and frequency dependent stuff).

[Gutbucket]

Yeah, I think we are agreeing.  I'm just advocating for letting each program or plugin that is actually doing the non-linear effect processing do its own up-sampling, rather than recording at a high sample rate.  ..And accomplishing that via the now common (2x, 4x, 8x) "quality" up-sampling selector typically available in modern plugins and stand alone software, which is made user selectable so as to allow one to manage processor load in bigger projects.  For 2-channel projects with relatively minimal processing going on simultaneously, one can probably select the highest up-sampling option, just to be safe, without overly bogging down the machine.

Also agreed that in "small" projects like this there is likely to be no audible problem transferring at 96kHz and processing at that rate, yet I see no benefits in doing so, while I do see several drawbacks: All the files will be twice as large, the processing can actually be done at a much higher sample-rate than it would be natively at 96kHz (by choosing the 4x or higher up-sampling option), and the inter-modulation issue goes away - even if that doesn't represent much of a problem, better to avoid it.

That's my stance and option.  Not saying other ways are wrong, but I do buy into the advantages of doing it this way.

Sorry to drag the discussion into technical details, but if anywhere, this is where its actually relevant. Thanks for it and to each his own. Preserve and long live great music!

[edit..  Rocksuitcase wrote-
I'm betting live audience concert recordings have a certain non-linearity acoustically given location, mic choice, pattern etc. That linearity may not present itself electronically but it is there in the 3 dimensional cues (time of arrival and frequency dependent stuff).

You know those live audience recorded 3d cues are near and dear to my ear.  But they are linear and captured linearly by quality high digital recording.  Any non-linearity in capture is technically a distortion.  If an argument can be made for recording at 96kHz, I think it applies more to recording of live performances, or studio recording, rather than to transfers, especially of cassettes.  Personally I'm satisfied with making new digital recordings at 48kHz, and doing the processing using the highest up sampling my machine can support.]

[Gutbucket]

To clarify.. and flog the dead horse,

The reasoning I lay out above applies to the question of sample-rate in regards to processing - which is all happening fully within the digital domain.  It does't apply to the somewhat messier job of converting from analog to digital and back to analog again, which involves analog circuitry.  I think it is there were stronger arguments can be made for higher sampling rates, due to the converters being able to use easier to build less steep stop-band/anti-alias filtering.  But converters, and the filtering using in them, have gotten better and better and less and less costly for really good performance, and I don't think the general understanding of what used to be a more significant issue has kept pace.

More from Monty on this= https://youtu.be/FG9jemV1T7I?t=678

[rocksuitcase]

This is a fun topic! GB- we do agree for the most part. In fact, it is probably just the difference in source material and what we output into which is more challenging.
I guess if this were to go further OT, then it should be a new thread.
To pull us back OT a bit:
I am about to become responsible for the repair/refurbish of a Nakamichi Dragon and a Nak 550 portable recorder. I own neither, but have become part of this project due to the death of long time NYC, Long Island taper best friend of one of my good buddies.
I will put up pictures once I have them.