Hifi Forum TDA-1541A DAC Review

[filo97s]

And you would not hear a difference between this one and the Topping D90 in a proper double blink test.

Yes, I think so. Human hearing is not a spectrometer. I don't think that anyone can distinguish something that goes -90dB SINAD from something -120 even if the difference is obvious for the instrumentation.

 

[PeteL]

Use a elder 'quad switch' (AD550) R2R Design - the voodoo people will love it ...

Or better - DATRAC, a tube DAC from 50's - must sound like Beethoven live on stage

For the Philips side - i recommend a DAC08 ( https://pdf1.alldatasheet.com/datasheet-pdf/view/17629/PHILIPS/DAC08.html ).
That thing kills them all - a nice tube buffer at the end.... and sell it for 10k bugs ... i have an idea



Yes. Modern DAC pull out more information. The reason is, they can reconstruct tiny little fragments and bring them out to the analog side.
Funny thing is, ADC was not such a problem at all - in the old PCM is more information as u thing.
The difference is the precision and the shape of information. In the old files it lock like a mess, but no problem for modern designs. The elder DAC Design have more trouble, special with the tiny fragments.

A modern DAC is like a wizard, its pure magic what these Chips can pull out from a PCM file. And still highly impressive.

I have no idea, why they use these elder DAC Chips - technically it's like a Intel 66Mhz 486 against modern Intel i7 in a game Benchmark ...

Yes, I fully agree but that was not exactly my point. Here’s an analogy. If you dig out an old VHS tape, and try to play it on your SOTA 4k big screen, it looks extremely poor, If you instead connect this vcr to an old CRT TV. Well, it’s somehow more acceptable, watchable. wondering if there is some of that in this love for older DAC tech. Like if you are a big fan of late eighties CDs, could it be possible that it does in fact sound more acceptable with DACs of that era, since that’s what was use at mastering?

 

[Veri]

Yes, I fully agree but that was not exactly my point. Here’s an analogy. If you dig out an old VHS tape, and try to play it on your SOTA 4k big screen, it looks extremely poor, If you instead connect this vcr to an old CRT TV. Well, it’s somehow more acceptable, watchable. wondering if there is some of that in this love for older DAC tech. Like if you are a big fan of late eighties CDs, could it be possible that it does in fact sound more acceptable with DACs of that era, since that’s what was use at mastering?

Would kinda doubt that since there's not a big enough difference, at 90+ dB resolution one would be hard pressed to A/B this DAC with a modern performer. Because performance is fine, really.

 

[Wuzel]

Yes, I fully agree but that was not exactly my point. Here’s an analogy. If you dig out an old VHS tape, and try to play it on your SOTA 4k big screen, it looks extremely poor, If you instead connect this vcr to an old CRT TV. Well, it’s somehow more acceptable, watchable. wondering if there is some of that in this love for older DAC tech. Like if you are a big fan of late eighties CDs, could it be possible that it does in fact sound more acceptable with DACs of that era, since that’s what was use at mastering?

U never ever master for a special Device, DAC or something. U master for a (transfer) medium and the standard(s) behind it.
For CD, u go to Red Book Standard.
In most cases the press shop do the job (they get the (pre) master for it) - but today often the DAW itself or a VST do the job automatically.

When u mean, somebody wish the same 'hearing' experience, then buy a 10$ china DAC from Ali, the sound will be the same

The old stuff where hard technically limited, there is no need to use them in these days. We have way better options.
Same as Vinyl, Cassette and the other esoterish ... (put the word here )

 

[restorer-john]

How do you account for the fact that there is no research at all to back that and the entire industry as moved on to delta sigma DACs?

Nobody needed research back when delta sigma D/A converters hit the market the first time around. They were terrible. R2R outperformed "one bit" by miles. It took years for one bit/PWM/delta sigma to catch up.

Philips invented the "bitstream" (PDM) converter for one reason and one reason only- it was cheap to manufacture. It was designed specifically for cheap, portable or car audio- never home high fidelity. It required no grading or laser trimming. They'd hit on the concept after struggling to make a proper 16 bit converter in the first place and resorted to O/S digital filtering with their 14 bit D/A (TDA-1540) and had the lightbulb moment- "what if we took oversampling to the extreme and just used one bit?". Imagine that.

So all the development of delta sigma over the last 30+ years has resulted in D/As that only outperform R2Rs simply due to strapping on more bits (24 etc). This TDA-1541 is not an R2R. It is 10bit (passive using emitter scaling) and a 3x2 bit active (DEM) give 16 bits total. Its default utilization was 4 or 8 x O/S. Its default THD including noise was typically -95dB (0.0018%) in standard selection (not S or S1).

If you tested a good 16bit R2R D/A converter implementation with 16 bit content, not 24 bit content and compared it to the best D/A converter you can find, also with 16 bit content, then you have a valid comparison. Otherwise, lumping it into the 'poor' category is silly. You are, after all, feeding 24bit AP test signals to a 16 bit converter and then proclaiming it in the "poor" category. The THD of this D/A converter is over twice (2x) what the TDA-1541S (or S1) regularly achieved in commercial CD players at the time.

Trouble is, people reading this will forget that most of their content is 16 bit and they have absolutely no idea how their so-called SOTA D/A converter performs on 16 bit content because nobody is testing it.

 

[KSTR]

If you tested a good 16bit R2R D/A converter impementation with 16 bit content, not 24 bit content and compared it to the best D/A converter you can find, also with 16 bit content, then you have a valid comparison.

For a reference this is the spectrum of a 1kHz @ -130dBFS sine dithered to 16 bit, using a 1M FFT at 44.1kHz (for high FFT noise reduction). We note the RMS noise level of -95dBFS, and I've put the vertical cursor there to indicate the equivalent level for a single singe.
While this -130dBFS tone is still measurable it's actually 35dB down in the (white) noise.
If this DAC would measure similar (with same FFT settings), with the -130dBFS peak still sticking out of the noise it would be stellar and Amir's 1kHz @ 0dBFS spectrum does indeed suggest this.

 

[KSTR]

And now a real-world 24 bit DAC.
-150dBFS sine is rendered and the noise floor (20..20k) is -120dBFS which means with a good DAC we have at least 15dB of possible digital volume reduction headroom before the output signal gets noise limited more than already by the dithering to 16bit in CD playback (and other) data.
When I applied dither to that signal at lesser LSB numbers I could go down to 21bit dither until the noise floor in the FFT started to rise, and from that I would infer that DAC is good for at least 20bits of resolution as far as limited by noise.
(Note: dBV actually represents dBFS in the plot, a tribute to the way I applied tha 20dB gain of the measuring ADC).

 

[PeteL]

Would kinda doubt that since there's not a big enough difference, at 90+ dB resolution one would be hard pressed to A/B this DAC with a modern performer. Because performance is fine, really.

I agree, the reasoning is based on he assumption that the difference is audible but I doubt too.

 

[MeLikeyTacos]

I have one of the original Philips / Magnavox CDPs with this chip inside. Love the pop-up top, never have to worry about a CD getting stuck inside.


View attachment 175561

That's just lovely. Machines such as this were part of the reason I liked the CD format so much in the first place, and still do.

I have an early 90's Rotel CDP with one of these 1541 DACs in it, given to me by a friend because it just outputs a static-like sound when playing. I need to dig it out one of these days and attempt a repair. I'd like to hear what it's supposed to sound like...

 

[aj625]

Yes, I think so. Human hearing is not a spectrometer. I don't think that anyone can distinguish something that goes -90dB SINAD from something -120 even if the difference is obvious for the instrumentation.

But how come these people can differentiate between r2r and DS dac ?

 

[voodooless]

But how come these people can differentiate between r2r and DS dac ?

Who can? What r2r DAC?

 

[aj625]

Who can? What r2r DAC?

Audiophiles like R2R more than DS dacs but same people can not (or rather don't want to ) differentiate between 120db and 80db sinad saying the difference as inaudible.

 

[amirm]

But how come these people can differentiate between r2r and DS dac ?

They use their eyes in addition to ears when they do that....

 

[voodooless]

Audiophiles like R2R more than DS dacs but same people can not (or rather don't want to ) differentiate between 120db and 80db sinad saying the difference as inaudible.

I’m pretty sure those are not the same groups of people

 

[amirm]

Otherwise, lumping it into the 'poor' category is silly.

I didn't give it poor rating. It falls in the second bucket of "Fair." Today we have recordings with true dynamic range of about 18 bits so we need better DACs. 16 bit is not the only digital format anymore.

 

[aj625]

I’m pretty sure those are not the same groups of people

Ask any r2r lover ! Even if you don't ask some will appear automatically in this thread.

 

[aj625]

They use their eyes in addition to ears when they do that....

 

[filo97s]

But how come these people can differentiate between r2r and DS dac ?

It's called suggestion or psychoacoustics.

 

[PeteL]

Audiophiles like R2R more than DS dacs but same people can not (or rather don't want to ) differentiate between 120db and 80db sinad saying the difference as inaudible.

If this is true, that Dacs with more than 80 dB SINAD are indistinguishable from the top ones then all should get a golfing panther. Any engineering ressources that have no benefits in real world usage have absolutely zero value. In fact they should have points taken off for wasting ressources that can be useful elsewhere. Interestingly, “objectivists” purchasing a product that the only thing of value added is to be able to brag about numbers, are not in any way more “objective” than subjectivists that claim hearing unmeasurable stuff. Now, I am not the one saying they are undistinguishable.

 

[PeteL]

U never ever master for a special Device, DAC or something. U master for a (transfer) medium and the standard(s) behind it.
For CD, u go to Red Book Standard.
In most cases the press shop do the job (they get the (pre) master for it) - but today often the DAW itself or a VST do the job automatically.

When u mean, somebody wish the same 'hearing' experience, then buy a 10$ china DAC from Ali, the sound will be the same

The old stuff where hard technically limited, there is no need to use them in these days. We have way better options.
Same as Vinyl, Cassette and the other esoterish ... (put the word here )

I agree, but the question is, do they sound different? You don’t master for specific device, but you do master with your ears, with reference class equipment for the state of the art at the time. You will never hear the same as what was heard in the mastering suite, but it can still can be argued that if the output of new DAC is proven not the same, meaning the reconstructed analog waveform is not the same and does not sound the same, which one is more “right”?