Hi Dan,
I find that most of the "high-end" expensive stuff is greatly inferior to some "Mid-Fi" stuff. This is especially true of power supply design and layout. One reason might be that larger companies can actually afford to hire and keep real engineers, while the "high-End" companies are smaller and typically have butt one person who does the designs. Often those people aren't engineers or even good technicians. It's mind blowing to see what they cook up without any counterbalance to what they are doing. Recently I worked on some amplifiers put out by a US person who worked at Hafler for a while (maybe still does) as a technician. He came out with his own products, but those designs and layouts are wrong. By reimagining the ground layouts and some other changes, I was able to drop the noise floor by 30 dB! That's huge! The distortion was dropped by nearly 20 dB. THe owner of those products had the equipment back to the manufacturer and various service people over a 20 year period. We got him inot some Marantz gear that I improved and he was going to sell the preamp that kept blowing up and crackling, and the matching amps. Once I was done, he had a quiet preamp that was reliable (we also went from two external power supplies to one), and his amplifiers blew him away. The Marantz was still a lot better, but he said that had his stuff sounded as good as they did now, he wouldn't have changed gear. Now he is selling that old system with a clear conscience.
This kind of thing isn't an isolated incident Dan. I had a similar experience with a nut-bar from Oz. Forgot his name, but he made amps in a chrome chassis, one model was Icarus and the pair brought to me had to be totally redesigned to be reliable. The fact that the sound quality improved a lot just shows how important it is for the parts to be happy (not overheating and in a linear circuit). If you mention his name I would remember it. The customer in Canada sent them back for warranty three times and they kept blowing. Most techs refused to work on them. Why I did I can't say. He was a customer of mine with another product. So two amps redesigned and one was simply unrepairable (reliably) and didn't have speaker protection at all. There are other memorable situations like these. Some even included companies that were large enough to have no excuse.
I find that corrected design errors and component matching improves things more than component changes. Certainly more than fancy wire. Once wire is better than some minimal quality, more expensive cable doesn't buy you much. Same for AC power conditioning. If you need a power line conditioner, either you have horrible noise on teh line, or poorly designed power supplies (normally the case).
So I believe you to a point, beyond which I think is confirmation bias.
-Chris
I find that most of the "high-end" expensive stuff is greatly inferior to some "Mid-Fi" stuff. This is especially true of power supply design and layout. One reason might be that larger companies can actually afford to hire and keep real engineers, while the "high-End" companies are smaller and typically have butt one person who does the designs. Often those people aren't engineers or even good technicians. It's mind blowing to see what they cook up without any counterbalance to what they are doing. Recently I worked on some amplifiers put out by a US person who worked at Hafler for a while (maybe still does) as a technician. He came out with his own products, but those designs and layouts are wrong. By reimagining the ground layouts and some other changes, I was able to drop the noise floor by 30 dB! That's huge! The distortion was dropped by nearly 20 dB. THe owner of those products had the equipment back to the manufacturer and various service people over a 20 year period. We got him inot some Marantz gear that I improved and he was going to sell the preamp that kept blowing up and crackling, and the matching amps. Once I was done, he had a quiet preamp that was reliable (we also went from two external power supplies to one), and his amplifiers blew him away. The Marantz was still a lot better, but he said that had his stuff sounded as good as they did now, he wouldn't have changed gear. Now he is selling that old system with a clear conscience.
This kind of thing isn't an isolated incident Dan. I had a similar experience with a nut-bar from Oz. Forgot his name, but he made amps in a chrome chassis, one model was Icarus and the pair brought to me had to be totally redesigned to be reliable. The fact that the sound quality improved a lot just shows how important it is for the parts to be happy (not overheating and in a linear circuit). If you mention his name I would remember it. The customer in Canada sent them back for warranty three times and they kept blowing. Most techs refused to work on them. Why I did I can't say. He was a customer of mine with another product. So two amps redesigned and one was simply unrepairable (reliably) and didn't have speaker protection at all. There are other memorable situations like these. Some even included companies that were large enough to have no excuse.
I find that corrected design errors and component matching improves things more than component changes. Certainly more than fancy wire. Once wire is better than some minimal quality, more expensive cable doesn't buy you much. Same for AC power conditioning. If you need a power line conditioner, either you have horrible noise on teh line, or poorly designed power supplies (normally the case).
So I believe you to a point, beyond which I think is confirmation bias.
-Chris
I ran an experiment last night as follows -
Playback of mono wav file into amplifier, 7 ohms loading on each channel, amplifier output active terminals connected to soundcard balanced input.
I recorded three times with both IC's in normal direction...A1, A2, A3.
I reversed one interconnect and repeated....B1, B2, B3.
I then inverted A2 and B2 and played each of them individually against the other files individually.
Confounder was that files did not perfectly synchronise causing tube sound effects despite being loopback recordings (variable latency ?).
Perhaps I need to run ASIO or something to cure this mission critical problem, advice needed.
Dan.
How did you align the files in time so as to null them, by eye, not at all, something else?
Hi Chris.
The Hi-End I mean is stuff that is designed competently, not the silly boutique stuff.
30dB too high noise floor and 20dB too much THD must have sounded like a 70's battery portable record player...what was it and what were the before/after measurement numbers ?.
Perhaps you mean metaxas.com, if so yeah that stuff is full on tug fest, 'art' in the right (well heeled) setting.
Metaxas gear has always been known for extreme bandwidths, and for blowing up just by staring at it for as little as a millisecond too long.
I agree by experience that matching and fixing stuff like dumb *** earthing routing is the very first thing to do and removes a bunch of dependencies in one swoop.
Once these fundamental errors are cured, then filtering AC power (energy feed) reduces a bunch more dependencies (of typical gear) and then lower level stuff starts to dominate.
Tweaking by cables etc is then an easy/practical way for owners of setting system final 'tone', however what sounds good on one system may not be optimal for another.
It's all about recreating fun music.
Dan.
Last edited:
I didn't try, I was depending on the timing of each recording event to be the identical...they weren't and it was time for bed.
I might try installing ASIO and see what I get.
Any advice on fixing this latency problem will be most appreciated.
Dan.
I would suggest loading all the files into Reaper at once, one file per track, expanding the view out in time until you can see individual samples, then sliding each track individually in the time (horizontal) axis into the best alignment you can (using the left mouse button and dragging), as compared with the other tracks. While still in Reaper you can mute all but two of the tracks, invert one of them, and play them back to see how well they null.
Note: To prevent "snapping to time grid" behavior when sliding the tracks into time alignment, you can either turn off the "snap to grid function" or hold down the ctrl or alt key while sliding (whichever works) to temporarily override snapping.
Sorry for the learning curve, but use is extremely easy once understood. All multitrack editing/recording programs do the same things and all the main ones are very powerful. Probably good to start getting used to one of them. Reaper is cheap and good, free for 60 days, and only gentle nag reminders thereafter.
Note: To prevent "snapping to time grid" behavior when sliding the tracks into time alignment, you can either turn off the "snap to grid function" or hold down the ctrl or alt key while sliding (whichever works) to temporarily override snapping.
Sorry for the learning curve, but use is extremely easy once understood. All multitrack editing/recording programs do the same things and all the main ones are very powerful. Probably good to start getting used to one of them. Reaper is cheap and good, free for 60 days, and only gentle nag reminders thereafter.
Last edited:
True.
Shape/dimensions are critical at RF of course, at audio ?.
Yes.
Contact pressure cures this.
Assuming reasonably ideal contacts, ime it remains that each individual material has unique signature, next question.
The very last connection is the most important, previous connections scale.
I am soon to compare SE twisted pair in this system.
You reminded me about Reaper, thanks.
I used Cool Edit Pro, I opened a wave file into multitrack track 1, rec armed track 2, hit home, hit record, hit stop, unarm and mute track 2, arm track 3, home, record, stop...and repeat through to track 7.
I found one pretty close time match, but other combinations were way out and a couple sounding like coming down a long pipe, cool sound effect but no use to me right now, I want tracks that are already synchronised in order to prohibit any false variance.
I thought any DAW would do 'bit perfect' loopback starts, maybe not so.
Any experts ?.
Dan.
Downloading Reaper now.....I used it heaps ages ago, I like it.
My eyes need eyeglass -4 and +1, but I can solder 0603 SMD without eyeglass. My eyes only can focus at certain distance (about 15 - 20 cm).
I tend to put marker pulses in the sample so re-alignment is easier. And yes, I think ASIO is pretty much mandatory both for perfect controlled latency as for guaranteed bit-transparency. Cool Edit and earlier Audition can't do ASIO so I use Audition 3.0. Further, I wouldn't use speaker outputs if any possible, rather use the cable in a simple loopback fashion directly connected to the soundcard at both ends, this allows for much lower noise floor. With a power amp the residual often gets dominated by mains hum and buzz.
As for procedure, please revert cable direction after each recording, so you have a sequence A1, B1, A2, B2, A3, B3. This is because we want to space any effect of unplugging/re-plugging/re-routing the cable evenly accross the test. If you don't, that potential source of subtle changes could be mistaken for directionality effect. We need to control all variables as much as we can, and we need to know our measurement resolution / noise floor and we need enough trials (2x10 would be much better than 2x3, btw) to discard outliers and to have some certainity that the result isn't likely just a false positive by random chance.
As for procedure, please revert cable direction after each recording, so you have a sequence A1, B1, A2, B2, A3, B3. This is because we want to space any effect of unplugging/re-plugging/re-routing the cable evenly accross the test. If you don't, that potential source of subtle changes could be mistaken for directionality effect. We need to control all variables as much as we can, and we need to know our measurement resolution / noise floor and we need enough trials (2x10 would be much better than 2x3, btw) to discard outliers and to have some certainity that the result isn't likely just a false positive by random chance.
I have installed Reaper/ASIO and quick run gives good cancellation.
My point is I don't want to do manual alignment for this test, other tests yes.
Getting ready for dinner party tonight, I might run your tests suggestions after.
Dan.
My point is I don't want to do manual alignment for this test, other tests yes.
Getting ready for dinner party tonight, I might run your tests suggestions after.
Dan.
All the digital chains have some delay and if you are not using asio may have some Windows sound manipulations in them.
What are you using for adda?
What are you using for adda?
Old school Tascam US-122 Mk1 44k/48k 16bit/24bit, pretty noisy but it might be good enough....my good soundcard is on loan.
Dan.
I specifically wrote a caveat that the rain forest thing might not be of any importance, as in maybe it's all compression. You then got all weird about it, as if I were making the claim without caveat. So I insinuated maybe you only understand the inside of IC's.
You also stated it in a factual manner, BTW I don't how wire can compress either.
perhaps another one of those conditions that are 'sounds like' compression. Or dynamic range is less?
Added 'stuff' (harmonics?) would also seem to limit the dynamic range or could say compressed. Not compressed as in the way of electronic compression. Bottom up rather than top down. Sonic result is similar.
Poor contacts most often produce odds...3H. and who knows what air born pollution coats the connector and what distortion that produces. Oxides too.
The non-porous, thick gold plated RCA made by Monster are best connection.... air tight and damn hard to remove - very tight.
THx-RNMarsh
Got bass? Them Changes -- Buddy Miles.
Added 'stuff' (harmonics?) would also seem to limit the dynamic range or could say compressed. Not compressed as in the way of electronic compression. Bottom up rather than top down. Sonic result is similar.
Poor contacts most often produce odds...3H. and who knows what air born pollution coats the connector and what distortion that produces. Oxides too.
The non-porous, thick gold plated RCA made by Monster are best connection.... air tight and damn hard to remove - very tight.
THx-RNMarsh
Got bass? Them Changes -- Buddy Miles.
Last edited:
I thought we were talking sound. You guys just keep saying/making up this stuff like it's a fact.
Mr. Marsh did your guys at LLNL run around worrying about the directions of all their cables? I've never been in a physics lab or lab of a major instrument company where any of this stuff was ever discussed.
- Status
- Not open for further replies.