Deafening feedback loop inside the GT-1000CORE

Started by szszoke, August 10, 2021, 09:32:12 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

szszoke

Hello!

Today one of my worst audio-related fears came to life. Audio feedback at full volume while wearing headphones.

The source was my GT-1000CORE so I started to investigate.

I was able to narrow the issue down to the FX loop, which creates an audio feedback loop under certain circumstances. I already wrote to Roland support, but I want to do some more investigation. I only have a sample size of one after all.

Below are the steps that I am using to reproduce the issue on my end. There are two things connected to my unit: power supply and a pair of headphones to the MAIN L output.

Firmware version: 1.02.

1. Turn on the unit
2. Switch to the P001 patch
3. Go to MENU > HARDWARE SETTINGS > OTHER and set SEND1 SETTING and SEND2 SETTING to SUB OUT
4. Switch to the P002 patch
5. Enable the 4th block from the right, which is the FX loop in my case.

The feedback loop can be hard at full volume after the 5th step.

Changing the patch after step 3 is necessary, otherwise the bug won't manifest.

Changing the setting at step 3 to SEND or SUB OUT & AUX MUTE will prevent the bug from manifesting.

BE AWARE THAT THE MAIN VOLUME KNOB WILL NOT ACTUALLY AFFECT THE VOLUME OF THE FEEDBACK. IT WILL BE AT FULL VOLUME REGARDLESS OF THE POSITION OF THE POTENTIOMETER.

The only way to affect the volume of the feedback is to change the return volume of the FX loop.

I would like to ask for the help of other GT-1000CORE to try to reproduce this issue on their own unit. Obviously make sure that noting expensive is connected to the outputs, or if there is, then the volume is limited after the GT-1000CORE as the built-in volume knob will not affect the volume of the feedback.

I want to find out if I have a faulty unit or we have a widespread bug that can cause hearing damage when encountered.

szszoke

Here is a YouTube video illustrating the issue:


Brak(E)man

Quote from: szszoke on August 12, 2021, 01:45:08 AM
Here is a YouTube video illustrating the issue:



If your unit is not faulty then Boss might have a huge lawsuit on their hands.
(Even if it's only your unit this is horrible)
This bug can cause irreparable hearing problems and a bring music career to a definite end.
I don't have that particular unit , so I can't test this.
Did you send this to Roland/Boss?
swimming with a hole in my body

I play Country music too, I'm just not sure which country it's from...

"The only thing worse than a guitar is a guitarist!"
- Lydia Lunch

JiveTurkey

I no longer have the Core. But I do recall trying to wrap my head around how the fx loops worked and the usual convoluted Boss implementation of them and had a similar experience. Good luck getting them to acknowledge and fix it.

Chumly

#4
It seems to me the equivalent can be done with many effects processor in which you either effectively or actually take the output and connect it to the input without sufficient attenuation.

Years ago I sold a power amp I was no longer using, and the new owner subsequently connected the outputs to the inputs (why I'll never know) and then called me up to complain I sold him a non-working amp.
I think it's much more interesting to live not knowing than to have answers which might be wrong. - Richard P. Feynman

szszoke

Quote from: JiveTurkey on August 12, 2021, 04:27:01 AM
I no longer have the Core. But I do recall trying to wrap my head around how the fx loops worked and the usual convoluted Boss implementation of them and had a similar experience. Good luck getting them to acknowledge and fix it.

I already had a discussion with customer support and they confirmed this issue during their own testing.

I was told that the issue will be forwarded to the office in Japan.

szszoke

Quote from: Brak(E)man on August 12, 2021, 03:47:05 AM
If your unit is not faulty then Boss might have a huge lawsuit on their hands.
(Even if it's only your unit this is horrible)
This bug can cause irreparable hearing problems and a bring music career to a definite end.
I don't have that particular unit , so I can't test this.
Did you send this to Roland/Boss?

I got a blast of this and I was exposed to it for about 3 seconds while wearing headphones.

I'm no doctor so I don't really know if my hearing could have been damaged in that time.

I'd much rather have my hearing unaffected by this than trying to sue..

I talked to the EU based customer support and I have an email saying they were able to reproduce the same issue during their own testing.

szszoke

Quote from: Chumly on August 12, 2021, 10:21:31 AM
It seems to me the equivalent can be done with many effects processor in which you either effectively or actually take the output and connect it to the input without sufficient attenuation.

Years ago I sold a power amp I was no longer using, and the new owner subsequently connected the outputs to the inputs (why I'll never know) and then called me up to complain I sold him a non-working amp.

If you do that then I'd call that an user error. In this case a seemingly unrelated configuration caused this behavior to manifest.

I suppose much of the firmware is shared between the GT-1000 and GT-1000CORE.

Since the GT-1000 has dedicated sub output and send/return sockets, they can always be used simultaneously.

On the GT-1000CORE you have to chose if you want the send sockets to behave as FX send or sub output.

They are likely using the same code for the Send/Return block in both devices and forgot to account for this scenario.

Of course, this is just speculation.

Brak(E)man

#8
Quote from: szszoke on August 12, 2021, 10:35:50 AM
I got a blast of this and I was exposed to it for about 3 seconds while wearing headphones.

I'm no doctor so I don't really know if my hearing could have been damaged in that time.

I'd much rather have my hearing unaffected by this than trying to sue..

I talked to the EU based customer support and I have an email saying they were able to reproduce the same issue during their own testing.

Depending on the actual Db you where exposed to it's more than enough time to cause permanent hearing problems. At levels above or near 120 DB it can be instant, also somewhat depending on the frequency, it can be at a lot lower level.
You're unfortunately already exposed , so there's nothing to do ,except rest your hearing for a while ,but unless you experience ringing afterwards, you might be lucky.
If I where you I'd rest my ears for a week.
It's been established that one loud exposure like this can be made worse even at low volumes if not "left to heal".  You might not be affected though, it's very unclear to the medical profession why some people don't get hearing loss or/and tinnitus, hyperacussis under the same circumstances.


My point being that this bug is a extreme hazard and that Boss has to fix it immediately.


swimming with a hole in my body

I play Country music too, I'm just not sure which country it's from...

"The only thing worse than a guitar is a guitarist!"
- Lydia Lunch

szszoke

Quote from: Brak(E)man on August 12, 2021, 01:47:32 PM
Depending on the actual Db you where exposed to it's more than enough time to cause permanent hearing problems. At levels above or near 120 DB it can be instant, also somewhat depending on the frequency, it can be at a lot lower level.
You're unfortunately already exposed , so there's nothing to do ,except rest your hearing for a while ,but unless you experience ringing afterwards, you might be lucky.
If I where you I'd rest my ears for a week.
It's been established that one loud exposure like this can be made worse even at low volumes if not "left to heal".  You might not be affected though, it's very unclear to the medical profession why some people don't get hearing loss or/and tinnitus, hyperacussis under the same circumstances.


My point being that this bug is a extreme hazard and that Boss has to fix it immediately.

I did some testing with the limited equipment that I have at home.

I put a sound level meter in the cup of my headphones and triggered a bug again.

I got a reading of about 100dB.

Then I tried to make a better seal with my palm. That way I got a reading of about 113dB.

I suppose the seal was not as good as when the headphones are worn on the head and the "dose" I received was actually higher.

I didn't notice anything weird afterwards with my ear but I also read that hearing damage is cumulative so it cannot be good.

Chumly

Wow, sounds like it's a real concern! I hope you'll be OK, I sometime used my FMR Audio Really Nice Compressor as a brick wall limiter when I'm running IEM's or headphones and there's a chance of excessive transients.  I suppose I should make a habit of it all the time.
I think it's much more interesting to live not knowing than to have answers which might be wrong. - Richard P. Feynman

PAH

Definitely not good for your ears, but I wouldn't worry too much about it. Going to concert and forgetting ear plugs would causing far more hearing damage than a momentary exposure to this.  Even many movie theaters in recent years have the volume so loud, you need ear protection.  Similarly, if you play in a band with a drummer (unless they play very softly i.e. jazz drummer playing with brushes), the cymbals are destroying your ears unless you're wearing plugs or IEMs.

gumtown

Quote from: PAH on August 13, 2021, 04:13:48 PMthe cymbals are destroying your ears unless you're wearing plugs or IEMs.
Yes, I can still hear drummers cymbals 24/7.
Free "GR-55 FloorBoard" editor software from https://sourceforge.net/projects/grfloorboard/

gumbo

Quote from: gumtown on August 13, 2021, 05:11:15 PM
Yes, I can still hear drummers cymbals 24/7.

Some decades later, the hi-hat destroyed the hearing in my left ear....   ::)

...unfortunately, the perpetrator of the "Honey-Do" list has learned to stand on my right side when issuing instructions.. :'(
Read slower!!!   ....I'm typing as fast as I can...

Brak(E)man

#14
Quote from: PAH on August 13, 2021, 04:13:48 PM
Definitely not good for your ears, but I wouldn't worry too much about it. Going to concert and forgetting ear plugs would causing far more hearing damage than a momentary exposure to this.  Even many movie theaters in recent years have the volume so loud, you need ear protection.  Similarly, if you play in a band with a drummer (unless they play very softly i.e. jazz drummer playing with brushes), the cymbals are destroying your ears unless you're wearing plugs or IEMs.

This is a dangerous misconception.
Momentary and/or long term exposure are as dangerous.
I learned the hard way.

https://www.vguitarforums.com/smf/index.php?topic=14274.msg112139#msg112139

Ear protection is vital in both cases.
IEM might actually work the other way but that's a topic for another thread.

This bug is extremely hazardous!


A one time momentary exposure can cause irreparable damage as well as long time exposure.
It all comes down to several factors that even the medical profession within this particular field have no explanation to why.
Some people seem insensitive and others sensitive . Nicknamed: Stone-ears Vs Glass-ears.

(There seems to be one advice that's medically proven.
After a sound exposure like this if you get another one in proximity (like the next couple of days) to the first one, the level of sound which can cause damage is way lower Db-wise. How much lower no one knows but the general advice is to rest the ears from sound a week or so. The theory seems to be that the follicles in the cochlea that might be damaged are bent but not broken off , the rest makes it possible for them to "unbend" and mend , if exposed again they might brake permanently. This is not a proven theory but resting after a sound exposure is proven to be regenerative.)
swimming with a hole in my body

I play Country music too, I'm just not sure which country it's from...

"The only thing worse than a guitar is a guitarist!"
- Lydia Lunch

aliensporebomb

I used to work at a hearing aid place years ago.  One of the audiologists there told me about the most profoundly deaf individual he'd ever met.  The guy had been an artillery gunner during world war 2.  He also had been a polka band drummer for 40 years - full kit with cymbals.   Hearing aids did not really help him.   He said it wasn't so much long duration high volume that did the most damage, it was short, sharp reports of cymbals, drums, artillery gunning that did the damage.   Without hearing protection you're kind of sunk.  At that point, mid 1980s I started wearing earplugs to both band practices and concerts I'd attend.   Made a huge difference: that persistent ringing after a show or practice was eliminated and I felt a lot better mentally and physically after.   So is this feedback loop happening merely because the CORE doesn't have as many outputs as the 1000 unit?  Or is it endemic to both units?
My music projects online at http://www.aliensporebomb.com/

GK Devices:  Roland VG-99, Boss GP-10, Boss SY-1000.

admin

#16
Quote from: aliensporebomb on August 16, 2021, 07:57:12 AM
So is this feedback loop happening merely because the CORE doesn't have as many outputs as the 1000 unit?  Or is it endemic to both units?

Never occurs on my GT-1000 - but I always SAVE patch edits on my MFX gear  - Before changing to a different patch in the middle of editing ( apparently this is not common knowledge?) 

Why am I the only one who caught the Pilot error here?





re watch his video - he starts on one patch, sets its FX loop controls,
but midway during his FX Loop level edits, he changes to a different patch ! (why?) (*** WARNING this WILL corrupt the FX loop settings for  the original patch being edited, and may require a SYSTEM Restore to get things back working again.

I would classify this as  pilot error 

so do this instead


1. Turn on the unit

2. Switch to the P001 patch

3. Go to MENU > HARDWARE SETTINGS > OTHER and set SEND1 SETTING and SEND2 SETTING to SUB OUT

3A -save the current patch being edited 

4. Switch to the P002 patch

5. Enable the 4th block from the right, which is the FX loop in my case.

5A P002 Patch FX Loop should now work correctly , without the high pitch chirps which got accidentally configured due to NOT saving the original P001 patch being edited , before changing to a different patch (P002) in the middle of editing the patch P001 FX Loop.



Brak(E)man

Quote from: admin on August 16, 2021, 08:38:40 AM
Never occurs on my GT-1000 - but I always SAVE patch edits on my MFX gear  - Before changing to a different patch in the middle of editing ( apparently this is not common knowledge?) 




re watch his video - he starts on one patch, sets its FX loop controls,
but midway during his FX Loop level edits, he changes to a different patch ! (why?) (*** WARNING this WILL corrupt the FX loop settings for  the original patch being edited, and may require a SYSTEM Restore to get things back working again.

I would classify this as  pilot error 

so do this instead


1. Turn on the unit

2. Switch to the P001 patch

3. Go to MENU > HARDWARE SETTINGS > OTHER and set SEND1 SETTING and SEND2 SETTING to SUB OUT

3A -save the current patch being edited 

4. Switch to the P002 patch

5. Enable the 4th block from the right, which is the FX loop in my case.

5A P002 Patch FX Loop should now work correctly , without the high pitch chirps which got accidentally configured due to NOT saving the original P001 patch being edited , before changing to a different patch (P002) in the middle of editing the patch P001 FX Loop.




Even put down as a pilot error it's something that should not be possible to happen, there's no cables attached , so in other words built in design , hard- or firmware .
Clearly imho a extremely bad and hazardous design flaw by Boss and should be treated as such.
swimming with a hole in my body

I play Country music too, I'm just not sure which country it's from...

"The only thing worse than a guitar is a guitarist!"
- Lydia Lunch

admin

At least its understood this occurs only when changing patches in the middle of an FX Loop configuration edit.  - I never do that   

szszoke

Quote from: admin on August 16, 2021, 08:38:40 AM
Never occurs on my GT-1000 - but I always SAVE patch edits on my MFX gear  - Before changing to a different patch in the middle of editing ( apparently this is not common knowledge?) 

Why am I the only one who caught the Pilot error here?





re watch his video - he starts on one patch, sets its FX loop controls,
but midway during his FX Loop level edits, he changes to a different patch ! (why?) (*** WARNING this WILL corrupt the FX loop settings for  the original patch being edited, and may require a SYSTEM Restore to get things back working again.

I would classify this as  pilot error 

so do this instead


1. Turn on the unit

2. Switch to the P001 patch

3. Go to MENU > HARDWARE SETTINGS > OTHER and set SEND1 SETTING and SEND2 SETTING to SUB OUT

3A -save the current patch being edited 

4. Switch to the P002 patch

5. Enable the 4th block from the right, which is the FX loop in my case.

5A P002 Patch FX Loop should now work correctly , without the high pitch chirps which got accidentally configured due to NOT saving the original P001 patch being edited , before changing to a different patch (P002) in the middle of editing the patch P001 FX Loop.




Thanks for chiming in but this bug is likely not applicable to the GT-1000 Non-CORE due to hardware differences.

The original GT-1000 has dedicated ports for the FX loop and sub output. The CORE version is sharing them between SUB OUT/AUX and the FX loop.

Consider that my video is specifically made to show the easiest way to trigger the bug with the least amount of customization which includes connected hardware and patches.

The PXXX patches are read-only. I choose to trigger the bug with them to have a common starting point when troubleshooting is performed.

The bug happens with my own patches as well and I do not need to edit anything. It's enough to switch to any patch where the FX loop is enabled to trigger the bug.

szszoke

Quote from: admin on August 16, 2021, 09:15:57 AM
At least its understood this occurs only when changing patches in the middle of an FX Loop configuration edit.  - I never do that

That is not true. See my previous reply.

Initially the bug happened when I turned on my unit and started switching patches. There was no editing involved.

Only after further testing I came up with the steps that you see in the video.

admin

#21
I have Posted this thread link here, it should get seen by Boss Support 
https://www.facebook.com/BOSS







But always file a Backstage Pass support ticket to your local Boss Distributor when you have a problem like this
https://www.roland.com/backstage/#/login?redirect=%2Fsupport#service_request





szszoke

Quote from: admin on August 16, 2021, 09:20:24 AM
I have Posted this thread link here, it should get seen by Boss Support 
https://www.facebook.com/BOSS

But always file a Backstage Pass support ticket to your local Boss Distributor

Thanks!

I already connected Roland Customer service and we concluded that the phenomenon is an issue.

They were able to reproduce as well.

It should be resolved via a firmware update.

szszoke

After some discussion in a different thread regarding something seemingly unrelated, a hypothesis has been established about a potential cause.

The SEND and RETURN ports are internally connected in a way that when nothing is plugged in to the RETURN port, the signal from SEND appears on it.

This is good when the ports are used for the FX loop because if the signal chain is not broken regardless of what is or isn't plugged in to the RETURN port.

When the SEND1 and SEND2 ports are configured as SUB OUT, RETURN1 and RETURN2 are working as AUX IN, but the internal connection is still there, so now whatever the unit outputs is fed back into the unit via the return ports.

The theory is that the RETURN ports are sampled regardless of the settings and when the FX loop block is enabled, the output of the unit is reintroduced somewhere in the chain and then output again.

If something is plugged in to the return ports, then the issue goes away, because the physical link is broken. Similarly if the sub outputs are enabled with AUX IN muted, the issue goes away.

The behavior is partially documented, meaning the parameter guide mentions that the signal from SUB OUT will appear on the main output when SEND1 and SEND2 is configured as SUB and nothing is plugged into the RETURN ports.

I still think this is a bug though because the parameter guide also says that the FX loop block will not work unless the SEND SETTING is set to SEND, but this isn't actually the case. It seems the the FX loop block is very much working and causing feedback.