Jump to content

Audio Responses


Recommended Posts

3 hours ago, Reg Wilson said:

The best processing happens between your ears. It's the difference from being able to drive formula one and an auto family sedan.

Let's see, that probably means I'm driving a pretty good Formula Atlantic car.
All the best Reg.

Link to comment
Share on other sites


The whole point of this idea is to give your brain and ears more raw information to process. A good analogy would be seeing with two eyes instead of one. The Zed does give you a warble or wobble over certain targets, and I think that kind of descriptive audio or even more descriptive audio can be applied to any detector that gives you two sets of data (signal strength and signal phase or x and y for ZVT/pulse) in all-metal mode.

Machine learning still has a long ways to go before it can rival a human, provided whiskey hasn't killed off too many brain cells. There is more raw audio information to be presented to a user that isn't currently being explored.

  • Like 2
Link to comment
Share on other sites

In my view, any threshold variation can be indicative of a gold target. Too many factors determine what sounds it would make under what circumstances. Sometimes you can guess, but you should never rely on it. I don't know of any serious operator of a high end gold machine who would reliability discriminate by sounds. Got to dig them all my friends.

  • Like 3
Link to comment
Share on other sites

4 minutes ago, phrunt said:

I am betting a few people have missed their find of a lifetime by ignoring a screamer.  Junk fatigue sets in.

Yup. Steve was also pointing that out in his great youtube clip. 

Link to comment
Share on other sites

On 8/6/2020 at 2:21 PM, klunker said:

So how bout a trainable detector where each time you find a target it stores all the audio information for each target and  you push  button that that says "good target" (gold for nugget hunting / non ferrous for coins an relics) or" bad target" (hot rocks / trash) so that after a couple of hundred target digs the detector would be "trained".

I think this is a significant part about what a multi-frequency detector is all about.  There are certain algorithms programmed in that give an audio response in addition to a screen.  The latest update to the Equinox has proven to make the point.  They have taken a lot of the 'scratchiness' out of corroded pennies.  It is a penny after all so why not take off some of the 'edges' and make it sound a little bit harder on the ears.  I'm hoping the same has happened to nugget sounds.

When you have digital processing a good imagination or a proper understanding of conveying a good target to a 'dig me' audio response could be synthesized from the data.  I can't do it and I know there are limitations but you can basically make any response have any tone you want.  You just say 'make all F flats sound like C sharps' and I'll be ok as a programming choice for instance.

The trainable part is where the engineers take all of the possible targets and plot them, which they already do.  The accept/reject if simple enough should isolate good targets or mask bad ones.  When this happens is there still enough audio data to determine to dig or not.  As Simon mentioned there is brain fatigue with all our detectors if you listen to pure raw data.

To filter or not to filter is the choice of all detectorists.  As Jasong said, swing speed and direction are audio filters.  It allows for audio interrogation of the target.

Technique and coil control separate a lot of us now.  We are also separated as Jasong said about not being completely familiar with the ground we are detecting.  I know my beaches intimately but don't know my nugget ground.  I have to get 'in the ball park' but often times I'm not really a player.

Mitchel

  • Like 2
Link to comment
Share on other sites

4 hours ago, tboykin said:

The whole point of this idea is to give your brain and ears more raw information to process. A good analogy would be seeing with two eyes instead of one. The Zed does give you a warble or wobble over certain targets, and I think that kind of descriptive audio or even more descriptive audio can be applied to any detector that gives you two sets of data (signal strength and signal phase or x and y for ZVT/pulse) in all-metal mode.

I like the mental exercise you are giving us to consider ALL of the data to help us find nuggets.  Sometimes I am a contrarian willing to express a thought.

You reminded me of a mentor I have in another hobby.  He was 'handicapped' in a way because he was legally blind in one eye.  

I have one of his books and he lived about 100 years ago.  Snooker was his sport.  Joe Davis was his name.  He wrote a book called 'How I Play Snooker.'  There are several pages in this book that describes his technique (stance, swing, cue control, levelness, etc.) but it was all based upon the fact that he could only aim with one eye.  The other eye didn't distract him.  His aiming was not a matter of dominant eye but the ONLY eye aiming.  Joe was the Snooker World Champion for 18 years.

When I try to use his technique I have to try and forget about my other eye.

So what does this have to do with detecting?  MAYBE we don't need that much raw data to find nuggets.  MAYBE one set of data (rather than two) with the proper information is enough.

Reg and others found lots of gold with 'less' information.  Can we still do it?

  • Like 2
Link to comment
Share on other sites

On 8/7/2020 at 12:26 PM, jasong said:

Now you know at least one. 😀

The nice thing about prospecting is that everybody has their own technique and whatever works the best should be adopted for individual needs. So, all good. I can just say that for me making sound decisions is not a productive way. So often I have been wrong when guessing what the GPZ is signaling, so I just don't take any chances. In certain cases, when I know the area really well, I disregard surface screamers if I know the gold is deep, and if my boot can just kick it away. However, for me the sound discrimination is just not a good practice, especially for a detector like the GPZ. 

  • Like 2
Link to comment
Share on other sites

When I get a target I lift the coil till I get a soft volume lets say 10% of full volume before I let the brain decide should I dig, then I dig and see if I was wrong.😉   

  • Like 1
Link to comment
Share on other sites

On 8/7/2020 at 11:32 AM, 2Valen said:

When processing audio the audio takes up so much memory that you would not have enough storage room to handle that much information. When you use a phone to video anything the thing that uses the most memory is the audio not the video. The ratio is close to 10 to 1 when talking audio versus video.

That is probably the main reason it is not done, there would not be enough room for the detector to store all the information.

Huh? Audio bitrates are significantly lower than video. We have dedicated GPUs to handle video processing with their own RAM. Audio processors and DSPs have far fewer resources than video for a reason - audio is far simpler than video. Maybe I’m not understanding correctly but audio does not need more memory than video. 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...