Often a smooth extended response gets characterized as soft where a sharply rolled response can sound bright. This is counterintuitive and seems wrong but I have seen it identified by observers consistently for many years. It may be a sensitivity in hearing to the envelope of the HF bursts and how that has been altered. We do not really hear tones above 8 KHz and are more sensitive to the envelope around the tones (part of the magic used by lossy compression schemes).
My experience too. In two different tests, listeners (including me) preferred the sound that had gone through a 15kHz brickwall FM filter to the unfiltered version. Reporting that the filtered version sounded 'clearer' and 'brighter'.
Jan
Jan
I believe it was Mooly''s experience that cheating was almost the norm. People would load tracks in Audacity or whatever, analyse them and then reported hearing differences.
The circulating boxes could be measured in all kinds of states, with all kinds of signals, and differences established.The only way to avoid that is to witness the listening tests which is not practical.
Jan
For an 'ultra clean' 30dB stage with 10mV input I would rather think of a composite than a single op amp. Graeme style, 2nd variety, would apply perfectly with a suitable op amp. Available loop gain at 10kHz would be at least twice the dB number for a single op amp at the same closed-loop gain. Something like 30dB vs 60dB++, and way more bandwidth and slew-rate.
For those that have perceived filtered HF as paradoxically brighter in some way, was the filter active or passive? In other words, if active, could the filter have added some THD in the passband at the same time it was providing filtering?
Last edited:
Switching noise from digital stages, Reverse recovery from rectifiers, Slew induced distortion from 20v/us opamps. Of course a 15k filter sounds better.
Right. Something similar was suggested in some posts little earlier in this thread. It was noted that if 30dB of gain was needed, it might take more than one opamp in order to keep HF distortion inaudibly low.
I don't know about that. I did look at the files in Reaper. That has nothing to do with whether I could hear differences. Assuming that cheating occurred would be jumping to conclusions. I assume that differences observable by examining a file visually or analytically in some way may not necessarily be audible. In fact, I'm sure that some small differences in files could not be other than inaudible.
But we can invite those with remarkable results to redo the test in controlled conditions before accepting them. This might be some sort of deterrence.
So long as the playback system used for verification is as good or better than the one used by those reporting hearing a difference at the time they heard the difference. I was using a Benchmark DAC-1 for Mooly's experiment, and I found two of the files nearly identical sounding despite a lot of effort strongly focusing attention on minute (to me) HF details.
And besides a near-state-of-the-art DAC-1, it takes a lot of practice to develop skilled listening to be able to do that. Similar to learning to recognize the subtleties of some new spoken language. Some people never fully learn to hear such subtleties as well as someone who learned the language starting at a very young age.
Last edited:
No it doesn't perpetuating myths, avoiding proper blind testing does exactly the opposite... Proper testing, proper measurements and telling the truth is what would create progress... A load of middle aged men claiming golden ears and promoting all the silly myths is STAGNATION.
I was with you until the "golden ears" and "silly myths" part. More science, yes. More denigrating language, no. If you want people to participate in the testing you would like to see, throwing insults around is probably not helpful.
In my case there was no additional distortion to speak of, but boatloads of phase rotation. That may have been a factor.
Jan
But examination could answer the question whether two files are the same or different, without even having to listen to them.
Jan
A long thread and I apologize for not reading all of it but I picked up a Crown D45 and was asked to compare the OEM installed OnSemi MC339078 to the TI TL084BCN. These are quad packages with a pair of op-amps acting as front end for each of the two channels. The first performs balanced to unbalanced, the second is high gain error signal.
Some results here:
http://www.northreadingeng.com/Forums/viewtopic.php?f=8&t=11
Looking to add to the list of packages to test in this configuration. Any suggestions? Used D45 are relative inexpensive and I would think this is a great DIYer project amp for folks to experiment with.
Some results here:
http://www.northreadingeng.com/Forums/viewtopic.php?f=8&t=11
Looking to add to the list of packages to test in this configuration. Any suggestions? Used D45 are relative inexpensive and I would think this is a great DIYer project amp for folks to experiment with.
I am not throwing insults just referring to those who do believe they have golden ears. To be perfectly honest are we getting anywhere, on this or many other threads... I ma angry because I want truths not silly myths of which there are far to many thrown around, just recently on this thread, regarding digital vs analogue; How is this progression, how can we move forward, when people cant even agree about the source...
As to silly myths, the hobby is full of them (BQPs', cable directivity etc.), all distract from the truth. The subjectives always get on their high horse regarding measurements (even though all those that support measurement also support proper listening testes and consider listening a part of the design cycle), engineering is often a dirty word, yet engineering is a BIG part of sound reproduction. So I stand by my rant, because I am getting angry with the lack of progress, 4000+ posts and we haven't even reached a consensus on op-amps...🙁
Whatever proof is put up there will be those that will never accept op-amps in audio designs, like some wont accept ferrites, digital etc. etc.
As to silly myths, the hobby is full of them (BQPs', cable directivity etc.), all distract from the truth. The subjectives always get on their high horse regarding measurements (even though all those that support measurement also support proper listening testes and consider listening a part of the design cycle), engineering is often a dirty word, yet engineering is a BIG part of sound reproduction. So I stand by my rant, because I am getting angry with the lack of progress, 4000+ posts and we haven't even reached a consensus on op-amps...🙁
Whatever proof is put up there will be those that will never accept op-amps in audio designs, like some wont accept ferrites, digital etc. etc.
So Marce, you want me to say a digital source is okay to really judge op-amp performance to make everyone here feel better? Op amps are really, really good and most here don't know how good they can be.
Yes a good digital source with a decent recording is better than an LP, but you could use both if you want to cover all bases and cater for those that believe LP to be a superior source🙂
Many have said it is the implementation of the circuit that is important...
Many have said it is the implementation of the circuit that is important...
Ears can be trained. When you design effects, synthesizers, you see shapes of waves, their change, and hear the same time. It calibrates your perception, teaches to recognize and imagine waveforms', specters', frequency response changes.
- Status
- Not open for further replies.