[MatW]

The same thing happens with my Dave but didn’t happen with my TT2.

Thanks. What was different with the TT2?

 

[JohnA1]

The Scaler bypass button is fully volume-matched. A/B testing couldn't be easier.

QuTest also pauses for 2 seconds between full scaling and bypass.
Dave I think gets the channels confused for a couple of seconds before deciding 'this is it!'
Probably glitch in the Dave code that got fixed subsequently for TT2 and Qutest

 

[MatW]

The Scaler bypass button is fully volume-matched. A/B testing couldn't be easier.

QuTest also pauses for 2 seconds between full scaling and bypass.
Dave I think gets the channels confused for a couple of seconds before deciding 'this is it!'
Probably glitch in the Dave code that got fixed subsequently for TT2 and Qutest

Thanks for the clarification!

 

[griff500]

Thanks. What was different with the TT2?

Exactly what we were talking about… It didn’t go to the right and then the left

 

[HughF]

Hi all!
Excuse me for not haven’t read every 899 pages… I’m curious about musical improvement by adding the MScaler to my Hugo TT (not TT2)

Thanks in advance!

I had owned a TT 1 for maybe a year when I bought the M Scaler. I moved to the TT 2 a few months after buying the M Scaler. If you like the tonal balance of the TT1 then losing money (50% ?) p/x'ing your TT 1 for a TT 2 is probably an expensive move : it's a useful jump in quality BUT will not guarantee more enjoyable music if you like the TT 1's sound as they do not have the same balance.

If you don't like the TT 1's tonal balance or its wacky ergonomics then it is much more worthwhile to move to the TT 2.
In my particular system, I'd rather swap my TT 2 and go back to TT 1 than ditch the M Scaler if I ever had to sell something. I would then hope to go to a (used?) DAVE later with the M Scaler rather than a TT 2 if my circumstances improved.

I found the M Scaler (with TT 1) made longer listening sessions MUCH more enjoyable (involving yet relaxed) than they were when I only had the TT 1. I graduated to listening to whole albums rather than jumping from track to track all the time in Roon.

Since the M Scaler can be used with all the current DACs, money spent on it is not wasted even if more benefit can be gained later by using the full 1M taps at 705/768kHz into a later DAC, not the 500K taps the TT 1 can use at 384kHz (single BNC cable).

HTH,

Hugh

 

[Reactcore]

Thanks for fast reply!
I’m not surprised but it would be interesting reading about your experiences…

I also have red post of users stating they like HMS's blue OP-SR over white (half M taps).. that uses single coax out which is compatible with TT1. Also users stated TT1 sounds warmer than TT2. U will eventually want HMS anyway.. and TT1 can be upgraded later too.

Even my Questyle DAC truly benefits from it and it can only do 192/24 on Spdif

 

[mammal]

So getting HMS tomorrow as a loaner for a week. To pair with my HTT2 and AB-1266, curious how much improvement will it make for EDM. Have been asking around for a while, so happy I can finally try for myself and report back to you all.

 

[Icenine2]

Not sure W/EDM but listened to some Eno yesterday w/lots o synth. It was spectacular. Better than I’ve ever heard

 

[Reactcore]

Yes MP3 loses critical information. Also, the HF is lower data rate and noisier than low frequencies; this means transients are not encoded with great accuracy. I think that's why MP3 sounds soft compared to redbook.

As to mic waveform to DAC waveform - for sure you would see big changes, but if it's phase or frequency response errors it's not subjectively damaging. The tiny errors that upsets the brains processing of the music are much more important, and you won't see that with waveform overlays - just the gross changes.

One of the reasons I am working on ADC's is so that the change of 768kHz>decimating to 48k>reconstructing back to 768kHz can be evaluated, and I will be able to see how much of an audible loss occurs. Ideally, we will not be able to hear a difference from a bandwidth limited 768k file, or a decimated then WTA reconstructed file - that is my goal.

Hi Rob,
Did you ever thought of just comparing the reconstructed digtal sound to non digitalized sound?
Why record or encoding and decoding to figure out if theres any improvement possible?

See my simple sketch below:
Musicians playing live while the sound being captured and straight being transferred through single driver speakers or headphones.
This way there are no timing issues since theres nothing digital to reconstruct.

The players hall can be set up much larger.
I wonder how well depth can be percieved with this setup..

 

[Rob Watts]

Hi Rob,
Did you ever thought of just comparing the reconstructed digtal sound to non digitalized sound?
Why record or encoding and decoding to figure out if theres any improvement possible?

See my simple sketch below:
Musicians playing live while the sound being captured and straight being transferred through single driver speakers or headphones.
This way there are no timing issues since theres nothing digital to reconstruct.


The players hall can be set up much larger.
I wonder how well depth can be percieved with this setup..

Yes I do indeed plan to do this, but with headphones, and have depth tests as well as impact tests (being able to perceive the sharp crack of a wood block being struck) and of course timbre variation tests. So these tests will tell me how well the ADC and mic and headphones work. The test I mentioned (768kHz>decimating to 48k>reconstructing back to 768kHz) is to test how well the ADC decimation and DAC WTA interpolation filters work.

 

[Reactcore]

The test I mentioned (768kHz>decimating to 48k>reconstructing back to 768kHz) is to test how well the ADC decimation and DAC WTA interpolation filters work.

Will you be just skipping samples of that 768khz file to get a 48khz one?

I wonder if the record industry will start bringing out 768khz music for download, even if those files will be huge in size. I've seen DAC chip specifications stating a supported rate of 1024khz already, but that will probably never be utilized.

 

[Triode User]

Will you be just skipping samples of that 768khz file to get a 48khz one?

I wonder if the record industry will start bringing out 768khz music for download, even if those files will be huge in size. It seems they kind of stopped at 192khz for the consumer market. I've seen DAC chip specifications stating a supported rate of 1024khz already, but that will probably never be utilized.

I have some 768kHz albums. The file size is unwieldy to say the least. Work on about 20GB per album and you are not far off. Also individual track sizes are large and mean that one is using wav files with long tracks being split into segments which are 'reassembled' using gapless play back. The faff factor is enormous. People used to request Rob to supply Mscaler processed files so that they could hear the effect without having to loan or buy and Mscaler and at the time I remember Rob saying the files would be huge.

So it is no wonder that it never happened and that if one wishes to hear the Mscaler effect the best advice was to just to plug in an Mscaler!

 

[Rob Watts]

Will you be just skipping samples of that 768khz file to get a 48khz one?

I wonder if the record industry will start bringing out 768khz music for download, even if those files will be huge in size. I've seen DAC chip specifications stating a supported rate of 1024khz already, but that will probably never be utilized.

No that would cause huge aliasing distortions! So to ensure no aliasing, you have to filter out any content above half FS and you use decimation filters to do this. Actually, I now have another headache - I have designed the decimation filters, but that was -240dB filtering, and work on the ADC sample rate converter indicates that aliasing at -350dB is audible - so I now have to re-do those filters...

 

[Reactcore]

I have some 768kHz albums. The file size is unwieldy to say the least. Work on about 20GB per album and you are not far off.

Very large memory size and fast transfer rates are not uncommon with today's computer resources.. so that shouldnt be the issue to supply it, as long one is not streaming it over the air or web.

I can imagine modern studio master files being recorded in this sample rate but released in downsampled format.

 

[Reactcore]

No that would cause huge aliasing distortions! So to ensure no aliasing, you have to filter out any content above half FS and you use decimation filters to do this. Actually, I now have another headache - I have designed the decimation filters, but that was -240dB filtering, and work on the ADC sample rate converter indicates that aliasing at -350dB is audible - so I now have to re-do those filters...

I thought creating a 768khz file is just a matter of 'regular' PCM encoding with only taking more samples in a time compared to 48khz.

The filtering becoming important only on the decoding side.. i guess i'm missing side effects caused by the finer structure.

How many 768khz ADC's are available in the market anyway? Or are they bad at doing their jobs?