Honestly guys, I'm pretty proud of myself for going through this. I was just reading about another guy that was very sad to find his two H723's both had this dreaded radio cancer. He junked both of them! Just didn't have the heart or courage to tackle it. He thought all he could do was look for NOS ...

Q: What did the maggot do at the baseball game? A: Wormed the bench. So guys, I went ahead and did T3, the 3rd FM IF can. Radio sounds better on FM but I still need to do full alignment. I'm probably not going to bother with T3A, the 3rd AM IF can, because AM sounds fine and I plan to use this for F...

It ran all night with no issues. I'm gonna hold off on doing any more cans. A person that knows much more than me told me that aligning the FM limiter/discriminator is tricky. The FM IF cans I re-built can be re-aligned very easily but you get into a can of worms when you mess with the limiter/discr...

Things are looking good. No hisses or pops so far after T2. Maybe we should do T3, T3A, and T4 just for futureproofing. This thing can outlive all of us. My notes to share because I care: It's not so much about silver migration as it is about deterioration of the interfaces between the mica wafer si...

The Salvation Army trademarked the phrase in 1950. To me, it's too trite to be used by high IQ individuals.

I completed T1A and it improved things but it's still not fully fixed. Now, instead of the crashing thunder I get just a little hiss and occasional pop after several hours. I tackled T2 this morning and luckily had the needed micas on hand. Finished up T2 and the radio still works. Waiting now to se...

Makes sense. I'll do that.

Makes sense, thank you Albert . Guys, the caps arrived and I completed the T1A AM front-end can. Everything still works so I have to wait 5 hours to see if it starts thunderstorming again. If so, we move on to T2. One interesting thing is that one of the resistors associated with T1A was 350 ohms wh...

French toast Digi-Key shipping is like molasses. Never forget, patience is a virtue.

Curly asked me what is the point of virtuousness. Do you know?

I'm pretty sure Frankaboy knows the answer.

Or you can reduce your swap file size to 1GB and close chrome, etc. But yeah, if you're going to be serious about UHD work on linux you'll need more system RAM.

I'm using an app called nvitop.

You just need more memory. Can you install a second 8GB stick? 8GB is very small these days, especially for UHD workflows.

You have 8MB and half is gone for swap. Open a couple chrome windows and there's another 1-2GB used.

How much system RAM do you have? I'm showing usage fluctuating 8-10 GB. That could be the problem.

Maybe the DPB 16->20 change pushed things over the edge for you. I'll look into that.

EDIT: DPB 16->20 does not seem to affect system memory usage. You may just need more RAM on your machine.

Using

watch -d -n 0.5 nvidia-smi

while previewing shows GPU memory never passing 1GB.

Also, how do you get memory usage stats for the GPU etc.

First test is fine for me. Does it fail straightaway or after some time?

I have a 2080Ti which has only 3 GB more than your GPU.

My vsedit is r20.

Welcome back, Frankaboy. I hope your vacation was stellar.

Did that work before the recent changes?

I will try to duplicate this right now.

DGDecNV for linux build 502:

* Added log file generation to dgindexnv.
* Bumped DPB 16->20.

https://rationalqm.us/dgdecnv/binaries

Released build 502 with log file generation.

For sure. Started on the 2nd IF can (AM front end). Got the wafer removed and re-installed the can with just the coils. Then I measured the wafer capacitors to the best of my ability. Then I ordered equivalent mica caps from Digi-Key. So a week to do each can is forced by having to measure and order...

There is a 'wart' in the windows version. All the parsing is done in our code EXCEPT for the HEVC parsing, which instead uses NVDec. The problem is that we don't have NVDec for DGIndexNV linux. So what I did is replace the NVDec parsing in the windows version with our own parsing. It works just fine...

I did the first one, the front-end FM IF can. Radio still works and I thought I was gonna get lucky but it started thunderstorming again after 5 hours. So now I'm gonna do the front end AM IF can. Note that these two IF cans have windings in series, so if one is bad it will affect both FM and AM. I ...

OK, Mr Balti.

Rebuild 6 IF cans?

Sadness is discovering your radio has silver mica disease.

Hey folks, got delayed a bit on the linux logging. As I was implementing it I found issues in the windows version logging. I'm working with Rocky to fix those up. Then I'll get back to the linux version. I think Rocky will make a slipstream tomorrow for the windows version.

Uh huh and next Winkie is gonna teach me to refinish bakelite cases. Cuz I just put my allowance advance on a 1952 Zenith H723 AM/FM radio. DG said he only wants AM/FM now. Figure it out. It's our first plastic tube radio, marking the end of the wood tube radio era.